showvikdbz/code-reviews · Datasets at Hugging Face

text stringlengths 1 5.2k	labels listlengths 1 1
this change is just a bug flags help is a string and will never be equal to flags true if we just want help to return failure there are much better ways	[ 0 ]
removing silence means these will spam admins to hell and back	[ 1 ]
> changing charcoal to fix a problem here just load the fucker up on amatin instead you nerdlet	[ 1 ]
no otherwise the compiler will abort every time you hit one which is pretty much guaranteed	[ 0 ]
crap, I missed the other CRD.	[ 1 ]
it s perfectly normal for the exemplar timestamp to appear in the future relative to when the scrape started	[ 0 ]
yes yes i fucked this up but too im doing shit hold on	[ 1 ]
It's for consistent indentation. `void` and `auto` have the same width and by using trailing return types there is no need for whitespace fuckery (or "grooming").	[ 1 ]
@AndyAyersMS That is a case, but I have plenty of those things but at a higher level of abstraction where evicting the interface at the class level would have a deep architectural impact. My latest Generalized Allocators work has to do pretty nasty stuff to be able to do everything struct based, because of the usage pattern a class level dotnet/runtime#9682 would simplify it a lot. I dont care about code bloating in those cases because I am making an absurd effort to ensure that codepaths are created for each type of allocator in a custom way.	[ 1 ]
Oh, yuck. I think we had to use `dill` for some related stuff in DeepQA. But, as I said, I don't have a strong preference here, it's just what I would have done if I had done the work. Don't feel like you have to change it to using pickle.	[ 1 ]
I really can't get behind this line of thinking. It feels like we are treating ourselves as apes that are incompetent of noticing a `scoreManager.Import(beatmap)` line during review, and resorting to huge complexity overheads just to add this unnecessary safety. If this is still required, I'm going to rewrite `DownloadTrackingComposite` from scratch to avoid the hard-model requirement in the first place (have been considering this anyway, as there are still blocking factors surrounding its usages elsewhere).	[ 1 ]
pid	[ 0 ]
If you want relays you'll have to edit file sonoff_template.h and add them to your NodeMCU config. I won't support easy web based relay config as the interface will get to clumsy and this is about sonoff not espeasy.	[ 1 ]
that one should go away it is a default and it will be killed today	[ 0 ]
remove fanout	[ 0 ]
@codingCoffee sometimes the MoteCarlo approach cannot be avoided, Unless you find a ridiculously elegant solution like @rkurungadam did :-) About the only Night Mode theme I like is the One Dark theme of the Atom IDE. https://atom.io/packages/one-dark-ui	[ 1 ]
This one's missing the "requires CE access" line. Also holy shit, props for writing all these descriptions.	[ 1 ]
oh fuck, yeah. thanks.	[ 1 ]
silly nit take with a grain of salt perhaps s throw up raise throw up brings other things to mind	[ 0 ]
white space change should be reverted	[ 0 ]
why are not you using actual filters instead of writing them by hand	[ 0 ]
Closing. This is awful to use. The block was easier to use originally, and aside from an abstract sense of correctness hiding redundant interactors (which may in itself present challenges), it produced valid DOM and worked without issues that I could find. This just seems to introduce edge-cases and unexpected behaviour.	[ 1 ]
I know, not your fault. They just suck.	[ 1 ]
dome fix the order of calling dirname and realpath	[ 0 ]
If we're gonna make it 0 we need to make it buyable only once or else some chucklefuck will shit out tons of them.	[ 1 ]
i have a PR in the wings to make this not such a shitty singleline but this will do for now (just saying that in case you conflict later)	[ 1 ]
add an assertrequests here	[ 0 ]
hi hongbin can we add an ut this is so rat	[ 0 ]
runtimeenabled contactsmanager	[ 0 ]
should have been a single colon	[ 0 ]
Crap, really? I'll have to see why. It should just "work". :-(	[ 1 ]
is exact false	[ 0 ]
trust id metadata ref get trust id	[ 0 ]
these two are unnecessary for realloc so you can put them under ifdef guard blocks	[ 0 ]
Ugh, the pylint build just started failing with something wacky. Once https://github.com/openedx/edx-platform/pull/30533 is through, I'll rebase, which'll hopefully resolve pylint.	[ 1 ]
special value	[ 0 ]
dpid datapath self dp list items	[ 0 ]
this will never happen because of kill kill	[ 0 ]
i think you should kill this as we already have l same with the others like this	[ 0 ]
~~@AryaArunB the issue that you linked to is a pull request, not a bug so I'm not sure what this is fixing?~~ Sorry I'm talking rubbish!	[ 1 ]
wait a minute do we need to run this as root we do not and shouldn t start shellinaboxd as root so i am not sure why we need to kill it as root i was able to change this to false in my local environment can we change this to false	[ 0 ]
@Mothblocks We fucked up here, it should've been 0.01 SECONDS. Now the Biogen uses 300kw instead of 3kw. I will make a PR to fix that. I noticed this only after it got into master.	[ 1 ]
oh i was wondering why you use ifneq local force static executable true below	[ 0 ]
ipxe timeout should not be boolean better to use it as ipxe timeout	[ 0 ]
Hey wait a bloody minute. I multiplied the amount of juice pulp returned by the waterless recipe by 2 already, so this should be 1 juice plus 4 pulps. >.<	[ 1 ]
destroy	[ 0 ]
observe as they completely fucking obliterate you with no possibility of survival because all 6 of the cocksuckers are electrocuting you every other fucking server tick	[ 1 ]
this calls looks to be used by different thread so is seems we have race conditions i think need to add an mlock object to synchronize all the public methods	[ 0 ]
you are right i will remove this useless classes	[ 0 ]
No, the entire (dubiously funny) joke is that the faction name is AMF because the last thing they say to you is "adios, motherfucker."	[ 1 ]
This seems like something that could probably just be set in the field initializer instead. Hell, you could even make it static.	[ 1 ]
this section is weird, the original lines send overflows errors, but if we split this works fine...	[ 1 ]
nit delete orphan is redundant as the fk is also part of the pk in poolmonitorassociation so it should not be possible to detach it	[ 0 ]
i did not look at all closely at the native impl i plead incompetence there and it has already been heavily scrutinized but it looked like we would actually get a dcheck failure if name was null so perhaps the docs here should be updated or the code there should throw	[ 0 ]
>`as num` God damn it	[ 1 ]
Chawanmushi won't be getting buffed because it's fucking disgusting	[ 1 ]
use syslog log facility savanna cuz syslog log facility is undef	[ 0 ]
I haven't had time to fix this KG, I knew it broke the shit. It was just a matter of I didn't have time to fix it.	[ 1 ]
androidappsinfo	[ 0 ]
How the hell is this passing CI when we dont have this type? I am mystified.	[ 1 ]
can we elaborate on the comment something like intent for doing blah blah and let us call virtual call instead of virtualcall	[ 0 ]
have i mentioned today that it is stupid that we still have instance type id flavor id and flavor name	[ 0 ]
not a comment on your code, but how was this not spotted before holy hell.	[ 1 ]
i think we should change self health check into self check agents health	[ 0 ]
oh crap, I wanted to remove them, will fix soon	[ 1 ]
i think this should probably pull from slaves	[ 0 ]
multi line docstrings should end on a new line	[ 0 ]
valid set self gone set	[ 0 ]
yeah we have always listed nested git checkouts otherwise they dirty git status third party gitignore has a bunch more examples	[ 0 ]
the active looks for a pid file i did not see this generated maybe it is my bad	[ 0 ]
seems that this block will never be executed because in plugin nsx yaml you enumerate this puppet task only for groups primary controller controller	[ 0 ]
annoying extra whitespace if you would like to nuke it	[ 0 ]
holy shit, amazing catch	[ 1 ]
these two lines should be removed now	[ 0 ]
It'll certainly help the coders of the future if they ever have to go "WHO CODED THIS PIECE OF SHIT"	[ 1 ]
this also highlights a bug in the code i would removed identical block ranges in any operation not just move this would have caused big issues with say bsdiff how lame	[ 0 ]
I hate gays, cuz they are really gay	[ 1 ]
perhaps a stupid question given that the native library initialization is a single threaded process and well profiled into the start common sequence why do we need to run madvise infrequent text random so early doing it after finch has initialized would be less error prone and would yield the same memory savings no	[ 0 ]
WTF? Anyone know why Culture specific formatting us used here? Seems like overkill to me.	[ 1 ]
i think using is sorted would be a clever abuse since a reasonble alternative exists adjacent find i am using the alternative	[ 0 ]
add an assertrequests here	[ 0 ]
there are lots of changes downstairs on the indentations do you really mean to do it	[ 0 ]
my google fuck is weak today what is a span	[ 0 ]
s message callback message callbacks why is this a naked type while we have typedefs for eventtypestrings and subscribedevents	[ 0 ]
very small nit imagetypes should be image types	[ 0 ]
> Consistency. Ash storms, snow storms, and void storms do not use defines. You should be adding consistency by making code better with defines, not more shit with strings	[ 1 ]
should be using le instead of	[ 0 ]
into a human readable message method the method	[ 0 ]
> Putting your fucking name in the code.	[ 1 ]
remove ws here an below	[ 0 ]
this was a rename from the old zipfiletest that git did not detect	[ 0 ]
expected attrs is useless here i think	[ 0 ]
It's not a VALID dmca, if this were to become valid; you could use the exact same logic if you were a malware author; antivirus companies catch your shit, you DMCA the companies to have them remove you from their database. It's _INSANITY_	[ 1 ]
Here too, I just suck at clicking at line numbers.	[ 1 ]
nits while you fixing other nits would be nicer if you fix these indentations too	[ 0 ]
self key matrix should be private i e sel fuck	[ 0 ]
then zeros will be appended sorry i must be dumb	[ 0 ]
why do bronze arrows suck so much	[ 1 ]
yes because if ovs l pop l pop is true in ovs if ovs only l pop is false in ovs if l pop only does not make sense at all	[ 0 ]
Having both seems like an abuse of the platform tag, and wheel filenames are getting a bit ridiculous. If this was on purpose upstream, it'd be nice to reconsider.	[ 1 ]

this change is just a bug flags help is a string and will never be equal to flags true if we just want help to return failure there are much better ways

[ 0 ]

removing silence means these will spam admins to hell and back

[ 1 ]

> changing charcoal to fix a problem here just load the fucker up on amatin instead you nerdlet

[ 1 ]

no otherwise the compiler will abort every time you hit one which is pretty much guaranteed

[ 0 ]

crap, I missed the other CRD.

[ 1 ]

it s perfectly normal for the exemplar timestamp to appear in the future relative to when the scrape started

[ 0 ]

yes yes i fucked this up but too im doing shit hold on

[ 1 ]

It's for consistent indentation. `void` and `auto` have the same width and by using trailing return types there is no need for whitespace fuckery (or "grooming").

[ 1 ]

@AndyAyersMS That is a case, but I have plenty of those things but at a higher level of abstraction where evicting the interface at the class level would have a deep architectural impact. My latest Generalized Allocators work has to do pretty nasty stuff to be able to do everything struct based, because of the usage pattern a class level dotnet/runtime#9682 would simplify it a lot. I dont care about code bloating in those cases because I am making an absurd effort to ensure that codepaths are created for each type of allocator in a custom way.

[ 1 ]

Oh, yuck. I think we had to use `dill` for some related stuff in DeepQA. But, as I said, I don't have a strong preference here, it's just what I would have done if I had done the work. Don't feel like you have to change it to using pickle.

[ 1 ]

I really can't get behind this line of thinking. It feels like we are treating ourselves as apes that are incompetent of noticing a `scoreManager.Import(beatmap)` line during review, and resorting to huge complexity overheads just to add this unnecessary safety. If this is still required, I'm going to rewrite `DownloadTrackingComposite` from scratch to avoid the hard-model requirement in the first place (have been considering this anyway, as there are still blocking factors surrounding its usages elsewhere).

[ 1 ]

pid

[ 0 ]

If you want relays you'll have to edit file sonoff_template.h and add them to your NodeMCU config. I won't support easy web based relay config as the interface will get to clumsy and this is about sonoff not espeasy.

[ 1 ]

that one should go away it is a default and it will be killed today

[ 0 ]

remove fanout

[ 0 ]

@codingCoffee sometimes the MoteCarlo approach cannot be avoided, Unless you find a ridiculously elegant solution like @rkurungadam did :-) About the only Night Mode theme I like is the One Dark theme of the Atom IDE. https://atom.io/packages/one-dark-ui

[ 1 ]

This one's missing the "requires CE access" line. Also holy shit, props for writing all these descriptions.

[ 1 ]

oh fuck, yeah. thanks.

[ 1 ]

silly nit take with a grain of salt perhaps s throw up raise throw up brings other things to mind

[ 0 ]

white space change should be reverted

[ 0 ]

why are not you using actual filters instead of writing them by hand

[ 0 ]

Closing. This is awful to use. The block was easier to use originally, and aside from an abstract sense of correctness hiding redundant interactors (which may in itself present challenges), it produced valid DOM and worked without issues that I could find. This just seems to introduce edge-cases and unexpected behaviour.

[ 1 ]

I know, not your fault. They just suck.

[ 1 ]

dome fix the order of calling dirname and realpath

[ 0 ]

If we're gonna make it 0 we need to make it buyable only once or else some chucklefuck will shit out tons of them.

[ 1 ]

i have a PR in the wings to make this not such a shitty singleline but this will do for now (just saying that in case you conflict later)

[ 1 ]

add an assertrequests here

[ 0 ]

hi hongbin can we add an ut this is so rat

[ 0 ]

runtimeenabled contactsmanager

[ 0 ]

should have been a single colon

[ 0 ]

Crap, really? I'll have to see why. It should just "work". :-(

[ 1 ]

is exact false

[ 0 ]

trust id metadata ref get trust id

[ 0 ]

these two are unnecessary for realloc so you can put them under ifdef guard blocks

[ 0 ]

Ugh, the pylint build just started failing with something wacky. Once https://github.com/openedx/edx-platform/pull/30533 is through, I'll rebase, which'll hopefully resolve pylint.

[ 1 ]

special value

[ 0 ]

dpid datapath self dp list items

[ 0 ]

this will never happen because of kill kill

[ 0 ]

i think you should kill this as we already have l same with the others like this

[ 0 ]

~~@AryaArunB the issue that you linked to is a pull request, not a bug so I'm not sure what this is fixing?~~ Sorry I'm talking rubbish!

[ 1 ]

wait a minute do we need to run this as root we do not and shouldn t start shellinaboxd as root so i am not sure why we need to kill it as root i was able to change this to false in my local environment can we change this to false

[ 0 ]

@Mothblocks We fucked up here, it should've been 0.01 SECONDS. Now the Biogen uses 300kw instead of 3kw. I will make a PR to fix that. I noticed this only after it got into master.

[ 1 ]

oh i was wondering why you use ifneq local force static executable true below

[ 0 ]

ipxe timeout should not be boolean better to use it as ipxe timeout

[ 0 ]

Hey wait a bloody minute. I multiplied the amount of juice pulp returned by the waterless recipe by 2 already, so this should be 1 juice plus 4 pulps. >.<

[ 1 ]

destroy

[ 0 ]

observe as they completely fucking obliterate you with no possibility of survival because all 6 of the cocksuckers are electrocuting you every other fucking server tick

[ 1 ]

this calls looks to be used by different thread so is seems we have race conditions i think need to add an mlock object to synchronize all the public methods

[ 0 ]

you are right i will remove this useless classes

[ 0 ]

No, the entire (dubiously funny) joke is that the faction name is AMF because the last thing they say to you is "adios, motherfucker."

[ 1 ]

This seems like something that could probably just be set in the field initializer instead. Hell, you could even make it static.

[ 1 ]

this section is weird, the original lines send overflows errors, but if we split this works fine...

[ 1 ]

nit delete orphan is redundant as the fk is also part of the pk in poolmonitorassociation so it should not be possible to detach it

[ 0 ]

i did not look at all closely at the native impl i plead incompetence there and it has already been heavily scrutinized but it looked like we would actually get a dcheck failure if name was null so perhaps the docs here should be updated or the code there should throw

[ 0 ]

>`as num` God damn it

[ 1 ]

Chawanmushi won't be getting buffed because it's fucking disgusting

[ 1 ]

use syslog log facility savanna cuz syslog log facility is undef

[ 0 ]

I haven't had time to fix this KG, I knew it broke the shit. It was just a matter of I didn't have time to fix it.

[ 1 ]

androidappsinfo

[ 0 ]

How the hell is this passing CI when we dont have this type? I am mystified.

[ 1 ]

can we elaborate on the comment something like intent for doing blah blah and let us call virtual call instead of virtualcall

[ 0 ]

have i mentioned today that it is stupid that we still have instance type id flavor id and flavor name

[ 0 ]

not a comment on your code, but how was this not spotted before holy hell.

[ 1 ]

i think we should change self health check into self check agents health

[ 0 ]

oh crap, I wanted to remove them, will fix soon

[ 1 ]

i think this should probably pull from slaves

[ 0 ]

multi line docstrings should end on a new line

[ 0 ]

valid set self gone set

[ 0 ]

yeah we have always listed nested git checkouts otherwise they dirty git status third party gitignore has a bunch more examples

[ 0 ]

the active looks for a pid file i did not see this generated maybe it is my bad

[ 0 ]

seems that this block will never be executed because in plugin nsx yaml you enumerate this puppet task only for groups primary controller controller

[ 0 ]

annoying extra whitespace if you would like to nuke it

[ 0 ]

holy shit, amazing catch

[ 1 ]

these two lines should be removed now

[ 0 ]

It'll certainly help the coders of the future if they ever have to go "WHO CODED THIS PIECE OF SHIT"

[ 1 ]

this also highlights a bug in the code i would removed identical block ranges in any operation not just move this would have caused big issues with say bsdiff how lame

[ 0 ]

I hate gays, cuz they are really gay

[ 1 ]

perhaps a stupid question given that the native library initialization is a single threaded process and well profiled into the start common sequence why do we need to run madvise infrequent text random so early doing it after finch has initialized would be less error prone and would yield the same memory savings no

[ 0 ]

WTF? Anyone know why Culture specific formatting us used here? Seems like overkill to me.

[ 1 ]

i think using is sorted would be a clever abuse since a reasonble alternative exists adjacent find i am using the alternative

[ 0 ]

add an assertrequests here

[ 0 ]

there are lots of changes downstairs on the indentations do you really mean to do it

[ 0 ]

my google fuck is weak today what is a span

[ 0 ]

s message callback message callbacks why is this a naked type while we have typedefs for eventtypestrings and subscribedevents

[ 0 ]

very small nit imagetypes should be image types

[ 0 ]

> Consistency. Ash storms, snow storms, and void storms do not use defines. You should be adding consistency by making code better with defines, not more shit with strings

[ 1 ]

should be using le instead of

[ 0 ]

into a human readable message method the method

[ 0 ]

> Putting your fucking name in the code.

[ 1 ]

remove ws here an below

[ 0 ]

this was a rename from the old zipfiletest that git did not detect

[ 0 ]

expected attrs is useless here i think

[ 0 ]

It's not a VALID dmca, if this were to become valid; you could use the exact same logic if you were a malware author; antivirus companies catch your shit, you DMCA the companies to have them remove you from their database. It's _INSANITY_

[ 1 ]

Here too, I just suck at clicking at line numbers.

[ 1 ]

nits while you fixing other nits would be nicer if you fix these indentations too

[ 0 ]

self key matrix should be private i e sel fuck

[ 0 ]

then zeros will be appended sorry i must be dumb

[ 0 ]

why do bronze arrows suck so much

[ 1 ]

yes because if ovs l pop l pop is true in ovs if ovs only l pop is false in ovs if l pop only does not make sense at all

[ 0 ]

Having both seems like an abuse of the platform tag, and wheel filenames are getting a bit ridiculous. If this was on purpose upstream, it'd be nice to reconsider.

[ 1 ]