workspace
stringclasses 4
values | channel
stringclasses 4
values | text
stringlengths 1
3.93k
| ts
stringlengths 26
26
| user
stringlengths 2
11
|
|---|---|---|---|---|
clojurians | clojure | looks like the site is down today? (?!) | 2017-12-16T03:07:16.000020 | Kalyn |
clojurians | clojure | <@Berry> why don't you use comma for that ? | 2017-12-16T03:10:42.000033 | Jami |
clojurians | clojure | it seems that '|' is a vaid symol, like + or - | 2017-12-16T03:11:28.000040 | Jami |
clojurians | clojure | Your implementation looks flawless in also closing the file "the transducer way" | 2017-12-16T03:21:05.000018 | Kalyn |
clojurians | clojure | Is there any reason :clj: uses `2 + Runtime.getRuntime().availableProcessors()` for `Agent/pooledExecutor` (<https://github.com/clojure/clojure/blob/clojure-1.9.0/src/jvm/clojure/lang/Agent.java#L50>) and `pmap` (<https://github.com/clojure/clojure/blob/clojure-1.9.0/src/clj/clojure/core.clj#L6941>)? Why the `2 +`? | 2017-12-16T06:47:55.000018 | Andra |
clojurians | clojure | <@Daniell> that code seems to defy the normal order of compilation, in that it refers to `fib` in its own `def`inition, and it's not obvious how the compiler handles that! | 2017-12-16T07:37:38.000038 | Beulah |
clojurians | clojure | <@Beulah> symbols referring to vars are implicitly dereferenced by the compiler | 2017-12-16T08:10:35.000002 | Rosia |
clojurians | clojure | <@Beulah>which means ? | 2017-12-16T08:30:32.000126 | Beulah |
clojurians | clojure | <@Rosia> which means ? | 2017-12-16T08:30:52.000056 | Beulah |
clojurians | clojure | it means you are allowed to self-refer the var you're defining, as long as the access is deferred (what lazy-seq does) | 2017-12-16T08:32:29.000069 | Rosia |
clojurians | clojure | when the root value is evaluated, the var has already been created | 2017-12-16T08:33:17.000025 | Rosia |
clojurians | clojure | <@Rosia> what do you mean by 'root value' ? | 2017-12-16T08:38:14.000047 | Beulah |
clojurians | clojure | the value held by your var | 2017-12-16T08:38:51.000092 | Rosia |
clojurians | clojure | which in this case is `(concat [1 1] (lazy-seq (map + (rest fib) fib)))` | 2017-12-16T08:39:21.000082 | Beulah |
clojurians | clojure | yes, but during the evaluation of that, the root value is undefined, that's why you need to defer | 2017-12-16T08:40:50.000055 | Rosia |
clojurians | clojure | try e.g `(def a a)` | 2017-12-16T08:41:13.000102 | Rosia |
clojurians | clojure | ok, just to be clear, a var has a value, are you saying it also has another value called its 'root value' ? i.e. does a var have 2 values: its value and its root-value and if so what's the difference ? | 2017-12-16T08:42:27.000079 | Beulah |
clojurians | clojure | no, the difference is meaningful only for dynamic vars | 2017-12-16T08:43:12.000006 | Rosia |
clojurians | clojure | but this isn't a dynamic var | 2017-12-16T08:43:31.000076 | Beulah |
clojurians | clojure | static vars have always one value and this value can be a sentinel if this var has not yet been bound | 2017-12-16T08:44:40.000004 | Rosia |
clojurians | clojure | ok so we could just say 'value' in this case | 2017-12-16T08:44:58.000061 | Beulah |
clojurians | clojure | I guess so | 2017-12-16T08:46:34.000001 | Rosia |
clojurians | clojure | alright so the point is that `lazy-seq` defers evaluation (until realized by `take` or something) | 2017-12-16T08:47:11.000096 | Beulah |
clojurians | clojure | so what appears to be an intractable self-reference isn't really because at the time the `lazy-seq` is realized the var `fib` has a value, which is `(concat...)` | 2017-12-16T08:49:02.000039 | Beulah |
clojurians | clojure | and when the `def` is evaluated, the `lazy-seq` isn't actually realized, it's only created | 2017-12-16T08:49:54.000033 | Beulah |
clojurians | clojure | right | 2017-12-16T08:50:05.000094 | Rosia |
clojurians | clojure | it's confusing because it doesn't read like that | 2017-12-16T08:51:06.000070 | Beulah |
clojurians | clojure | it reads as an atomic definition | 2017-12-16T08:51:16.000010 | Beulah |
clojurians | clojure | there's no indication as to the 2-pass evaluation | 2017-12-16T08:51:36.000056 | Beulah |
clojurians | clojure | I think it's hard to reason about time in that code | 2017-12-16T08:52:42.000030 | Beulah |
clojurians | clojure | it's not that much common afaik | 2017-12-16T08:53:22.000004 | Rosia |
clojurians | clojure | no I've only ever seen that example - it's lovely though! | 2017-12-16T08:53:43.000038 | Beulah |
clojurians | clojure | The only way --in really most programming languages-- to write code that isnt' evaluated is to put it into a function. Naturally `(def x (foo))` is different to `(def x (fn[] (foo))` | 2017-12-16T08:54:12.000044 | Randee |
clojurians | clojure | So that's exactly what the `lazy-seq` does (it's a macro that wrapps the code in a function) | 2017-12-16T08:54:41.000091 | Randee |
clojurians | clojure | <@Randee> or `delay` | 2017-12-16T08:54:58.000021 | Beulah |
clojurians | clojure | It's just a function definition. Under the hood lazy seqs hide that. | 2017-12-16T08:55:03.000126 | Randee |
clojurians | clojure | <@Beulah> Again, delay is JUST a function under the hood | 2017-12-16T08:55:15.000090 | Randee |
clojurians | clojure | Try writing a lazy seq in Javascript. Totally possible since you have functions. You're API would just be a little different accessing that lazy seq. | 2017-12-16T08:56:02.000051 | Randee |
clojurians | clojure | that doesn't sound enjoyable | 2017-12-16T08:56:27.000033 | Beulah |
clojurians | clojure | Your mind probably woulnt' have a problem with that code: `(def foo [0 1 (fn [] [2 3 (foo)])])`, right? | 2017-12-16T08:58:56.000038 | Randee |
clojurians | clojure | it's pretty straightforward - you just need an object that holds a function and an optional value | 2017-12-16T08:59:01.000008 | Margaret |
clojurians | clojure | Hi. What lib you would recommend to use with GraphQL? I found several and don't know which I should to choose | 2017-12-16T08:59:12.000128 | Marcos |
clojurians | clojure | I understand right what I need to use GraphQL both in cljs and clj sides of my site? | 2017-12-16T08:59:18.000003 | Marcos |
clojurians | clojure | <@Marcos> lacinia | 2017-12-16T09:59:19.000029 | Marty |
clojurians | clojure | is maintained by walmart | 2017-12-16T10:00:05.000189 | Marty |
clojurians | clojure | <@Marty>, thanks | 2017-12-16T10:17:50.000082 | Marcos |
clojurians | clojure | there is also <#C4DLZKR9T|graphql> which could be interesting if you start using it | 2017-12-16T10:18:42.000104 | Daine |
clojurians | clojure | <@Daine>, joined, thanks | 2017-12-16T10:19:45.000043 | Marcos |
clojurians | clojure | Hi! Tomorrow, on Sunday at 12am UTC, I’m going to stream live Clojure coding solving some issues for my open source Etaoin library. Recorded version will be stored on Youtube automatically. Watch it here: <https://www.youtube.com/watch?v=cLL_5rETLWY> | 2017-12-16T11:45:25.000010 | Verna |
clojurians | clojure | Is there a de facto parser combinator library for clojure? | 2017-12-16T14:17:36.000049 | Johana |
clojurians | clojure | So there is/was fnparse, but the original is super stale the maintainer vanished and there’s no clear winner of the forks. | 2017-12-16T14:19:49.000061 | Charity |
clojurians | clojure | It lets you write pretty trivial recursive decent parsers easily. | 2017-12-16T14:20:06.000113 | Charity |
clojurians | clojure | Instaparse exists and I’d suggest you start there. Using the ordered choice operator you can optimize your grammars moderately well. | 2017-12-16T14:20:33.000020 | Charity |
clojurians | clojure | Antlr4 or one of its wrappers if you really need performance. | 2017-12-16T14:20:43.000071 | Charity |
clojurians | clojure | <@Charity> I wanted to check if I could get something faster than InstaParse and nicer than hand written (for some value of nice):
<https://github.com/borkdude/aoc2017/blob/master/src/day16.clj#L118> | 2017-12-16T14:21:24.000074 | Johana |
clojurians | clojure | <@Johana> try replacing unordered choice `|` with ordered choice `/` in your grammar and see how that does. | 2017-12-16T14:22:26.000010 | Charity |
clojurians | clojure | <@Johana> Instaparse tries really hard to support ambiguously parsing syntaxes with fully unordered choice. Looks like your grammar is actually LL(1) and not ambiguous so you can use ordered choice to make it clear that your grammar is less general. | 2017-12-16T14:23:18.000089 | Charity |
clojurians | clojure | It actually becomes a bit slower that way | 2017-12-16T14:24:49.000052 | Johana |
clojurians | clojure | Not significantly, probably doesn’t help nor hurt | 2017-12-16T14:25:10.000024 | Johana |
clojurians | clojure | One issue is that Instaparse has to generate nested items first which get resolved in the transform phase | 2017-12-16T14:25:40.000073 | Johana |
clojurians | clojure | :shrug: | 2017-12-16T14:25:43.000022 | Charity |
clojurians | clojure | And with a parser combinator lib I might get to do that during processing | 2017-12-16T14:25:58.000069 | Johana |
clojurians | clojure | You can. That’s how my first compiler written in fnparse worked. You can also use Antlr4 visitors/listeners to perform evaluation during parsing. | 2017-12-16T14:26:33.000101 | Charity |
clojurians | clojure | Instaparse doesn’t and won’t support doing this for good reason. I talked to <@Edelmira> about this at some point. | 2017-12-16T14:27:11.000078 | Charity |
clojurians | clojure | Besides you’re down in ms here, the measurement error is probably significant. | 2017-12-16T14:27:35.000032 | Charity |
clojurians | clojure | With | 582 ms, with / 591 ms, measured by criterium | 2017-12-16T14:34:01.000014 | Johana |
clojurians | clojure | As Alex explained, IP does a lot of bookkeeping, don’t know the exact details, but this can make it slower than simple hand rolled parsing. | 2017-12-16T14:34:38.000076 | Johana |
clojurians | clojure | Parser combinators don’t seem really popular in Clojure somehow. | 2017-12-16T14:35:02.000003 | Johana |
clojurians | clojure | I've been playing around a little with <https://github.com/blancas/kern>. I'm pretty new to parser/combinators as a tool though, so I can't comment on what's good or not so good. | 2017-12-16T14:41:10.000104 | Bibi |
clojurians | clojure | Thanks! | 2017-12-16T14:41:43.000044 | Johana |
clojurians | clojure | Funny, I'm using AoC to learn this, too :slightly_smiling_face: | 2017-12-16T15:06:42.000051 | Bibi |
clojurians | clojure | yeah - I'd never had that issue before, but it's up now | 2017-12-16T16:23:24.000067 | Margaret |
clojurians | clojure | Kern implementation:
<https://github.com/borkdude/aoc2017/blob/master/src/day16.clj#L115> | 2017-12-16T16:54:22.000049 | Johana |
clojurians | clojure | advanced clojure book written from the post 1.7 perspective? recommendations anyone? :slightly_smiling_face: | 2017-12-16T17:31:57.000059 | Kalyn |
clojurians | clojure | I see this one isn't out yet by the way, no idea what is its table of contents like
<https://www.amazon.com/Programming-Clojure-Alex-Miller/dp/1680502468/ref=sr_1_7?s=books&ie=UTF8&qid=1513463397&sr=1-7&keywords=clojure> | 2017-12-16T17:34:15.000035 | Kalyn |
clojurians | clojure | Is there anyone using Lacinia ?
Great job! But, I am so sceptic. I have some questions. | 2017-12-16T17:36:37.000036 | Olen |
clojurians | clojure | <@Kalyn> Programming Clojure 3rd Ed would be a good purchase -- assuming you like e-books and don't mind getting updates while it's in pre-release form? Otherwise I'd recommend Clojure Applied if you want something beyond intro/mid-level. | 2017-12-16T17:49:25.000018 | Daniell |
clojurians | clojure | <@Olen> at <#C4DLZKR9T|graphql> there are quit some questions, I also want to try it at somewhere soon | 2017-12-16T17:50:01.000056 | Daine |
clojurians | clojure | I have a probably dumb compojure-api question. (happy to take it somewhere else there's a better channel for this). I think I'm missing something very obvious, but if I return a core.async channel from a route in `compojure-api`, I get an error about channels not implementing `render`. I'm running 2.0-alpha16 and doing something super similar to the example in the readme. I'm sure this will be a facepalm moment, but what's the obvious thing I'm missing? | 2017-12-16T17:57:46.000037 | Michelina |
clojurians | clojure | <@Daine> thank you. :innocent: | 2017-12-16T18:00:18.000054 | Olen |
clojurians | clojure | you finally pushed me to start the advent of code. my solution to day 16: <https://github.com/wiseman/advent-of-code/blob/master/2017/day16/src/day16.clj#L39-L59> | 2017-12-16T18:14:18.000068 | Alline |
clojurians | clojure | Ah, nevermind. I just needed to show httpkit how to interface with ring-async | 2017-12-16T18:22:38.000029 | Michelina |
clojurians | clojure | Just saw this and thought it was really interesting...but I don't know what IK is. Seems like there was some part of this discussion I missed? | 2017-12-16T18:24:20.000058 | Jerold |
clojurians | clojure | <@Sandy> Was just revisiting this article: <http://blog.cognitect.com/blog/2016/9/15/works-on-my-machine-understanding-var-bindings-and-roots> and noticed that there was never a follow-up on `Var$TBox` :stuck_out_tongue:
> The rationale behind TBox-es will have to wait until a later time
What's the reason `Frame` bindings are maps of `Var -> TBox` instead of just `Var -> <value>`? Since `dvals` is thread-local, all threads a `Frame`'s `TBox`s are the same, which my experiments seem to confirm. | 2017-12-16T18:55:42.000028 | Andra |
clojurians | clojure | <@Andra> It's mostly to stop you from shooting yourself in the foot. There's some strange cases where you could cause cross-thread mutation with `set!` and bindings. I don' have a REPL infront of me at the moment, but I think this would cause the error:
```
(def ^:dynamic myval nil)
(binding [myval 32]
(set! myval 42)
(future (set! myval 45)))
``` | 2017-12-16T20:45:31.000067 | Sandy |
clojurians | clojure | So that's the gist of it, we bound a dynamic var, set! a value locally, then tried to set it again in another thread. I think that will cause an error and that error occurs when you set an already set val from a different thread than the first set!. The thread that set the value is tracked in TBox. | 2017-12-16T20:47:55.000059 | Sandy |
clojurians | clojure | Using TBox also improves performance a bit, since set! is then allocation free. Without TBox you'd have to do an assoc into the contents of the current Frame, and that would require an allocation. | 2017-12-16T20:48:51.000017 | Sandy |
clojurians | clojure | I see, so it's to handle binding conveyance? B/c AFAICT that's the only way frames are shared b/t threads.
Relevant code: <https://github.com/clojure/clojure/blob/clojure-1.9.0/src/jvm/clojure/lang/Var.java#L219> | 2017-12-16T20:55:51.000053 | Andra |
clojurians | clojure | But I'm actually not sure why the two `set!`s is a problem? Each thread gets its own frames, so they won't interfere w/ each other. | 2017-12-16T21:06:33.000069 | Andra |
clojurians | clojure | Does anyone else here see a need for something like: <https://github.com/aaronc/c-in-clj> but targetting both webassembly (to be called from cljs) and cuda (to be calledc from clojure/jcuda)?
I want to write high performance numerical code in something that is clojure-like, and be able to invoke it directly from clojure | 2017-12-17T00:35:54.000014 | Berry |
clojurians | clojure | <@Berry> Did you look at Ferret? <https://github.com/nakkaya/ferret> | 2017-12-17T00:51:33.000020 | Silas |
clojurians | clojure | <https://ferret-lang.org/> looks pretty amazing | 2017-12-17T01:41:44.000017 | Berry |
clojurians | clojure | Is there a way, either using transient! vector or a native java array, to get CONSTANT TIME, not log_32 N time for following op:
set n-th element of array/vector/object to :foo | 2017-12-17T02:51:14.000105 | Berry |
clojurians | clojure | <@Michelina> You might not be running the server on async-mode? Adding `{:async? true}` to jetty opts enables the async chain. | 2017-12-17T03:56:59.000033 | Alla |
clojurians | clojure | there is also <#C0A5GSC6T|ring> for ring async and <#C06GSN6R2|ring-swagger> with c-api things too | 2017-12-17T03:57:28.000043 | Alla |
clojurians | clojure | <@Daniell> oh is it already available as an e-book? | 2017-12-17T05:46:40.000070 | Kalyn |
clojurians | clojure | I wish there was a TOC of it to look at, but on Amazon there’s none yet | 2017-12-17T05:48:38.000041 | Kalyn |
clojurians | clojure | going to start in 10-15 minutes
<https://www.youtube.com/watch?v=cLL_5rETLWY> | 2017-12-17T06:48:35.000014 | Verna |
clojurians | clojure | <@Verna>, we are waiting! | 2017-12-17T06:51:16.000011 | Olen |
clojurians | clojure | what is this? are you live casting using a library to control a web browser via clojure ? | 2017-12-17T06:52:52.000098 | Berry |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.