<< 18-11-2015 >>
00:00:55*strcmp2 quit (Ping timeout: 240 seconds)
00:04:02desophosi suspect that any port of a large project is very difficult to maintain
00:05:59BlueProtomanECSes don't have to be large
00:06:01desophosbecause in addition to writing the port, you have to keep up with development on the original project
00:06:12desophosand that includes maintaining all dependencies
00:06:47desophosi was under the impression that you wanted to port a large one
00:08:06BlueProtomanArtemis and Ash (two most likely candidates) aren't all that large
00:10:21desophosAsh is in AS3??? lol
00:11:05BlueProtomanOf course, a port would be mindful of AS3/Java/something else tricks that aren't relevant to Nim
00:11:31desophosis Artemis abandoned? looks like artemis-odb is the actively maintained fork?
00:11:44desophospseudo-fork
00:12:12BlueProtomanRaw artemis is abandoned, yes, but the design is still solid. artemis-odb is massive, and a lot of it seems to be Java-specific optimizations
00:13:04desophosthe site is gone
00:13:12desophosthe artemis site
00:14:06BlueProtomanThere's a fork here https://github.com/tescott/artemis
00:17:51desophosit would be interesting to "port" a java framework
00:18:20desophosbecause java's very strong opinions conflict with nim's
00:18:32desophosi mean java's concept of OOP
00:18:47BlueProtomanAnd a port to Nim would pay close attention to those opinions so they don't accidentally pollute it
00:18:53desophosyes
00:18:59*pregressive joined #nim
00:20:20*applejugz joined #nim
00:20:26desophosdo you know AS3 or java/
00:20:27desophos*?
00:20:35applejugzhow can I make an executable in nim?
00:20:37desophosor both, i suppose?
00:20:52BlueProtomanJava, yes. AS3 is easy to read
00:20:58desophosapplejugz: something wrong with `nim c myfile.nim`?
00:21:26applejugzoh, that can be ran across platforms, desophos ?
00:21:55desophosapplejugz: are you asking about cross-compilation?
00:22:06desophosbecause the compiler can compile for any platform
00:22:47applejugzYeah, that's what I meant. That's awesome though! I thougt i needed to do something fancy to get it to work on Windows or Mac
00:23:20desophosnope, `nim c myfile.nim` will produce myfile.exe on windows and a myfile binary on mac
00:23:44*pregressive quit (Ping timeout: 244 seconds)
00:23:48desophosto produce myfile.exe on mac (cross-compilation) you'd have to do something a little more specific, i haven't looked into it
00:23:58desophosapplejugz: http://nim-lang.org/docs/nimc.html might help
00:24:14desophosapplejugz: specifically http://nim-lang.org/docs/nimc.html#cross-compilation
00:26:02desophosBlueProtoman: is java not easy to read?
00:26:06desophosi thought it was
00:26:10desophosbut i haven't used it in years
00:26:26BlueProtomandesophos: No, it is. I'm saying I *do* know Java, and while I *don't* know AS3 I know that it's easy to read so I can figure it out
00:26:33desophosah okay i see
00:27:12desophoswell i might be biased but i am suspicious of anything written in AS3
00:27:21desophosalthough i don't know the language
00:27:35desophosso probably just a fear of the unknown :)
00:28:28desophosbut i imagine it would be easier to port something written in a language you know
00:28:42BlueProtomanIf it were some schmuck in a basement doing it, sure. But Richard Lord is a professional. http://www.richardlord.net/
00:29:17*BitPuffin|osx joined #nim
00:30:32BlueProtomanTrue, but there's another thing; Nim's implementation details would likely be *very* different than those of the original language. So while the interface would be the same, the implementation would be Nim-tailored
00:31:18*applejugz quit (Ping timeout: 246 seconds)
00:31:37desophoslikely indeed
00:32:26desophosbut i think re-implementing would be easier depending on the ease of understanding the original implementation
00:33:12BlueProtomanHonestly, I think a Nim port would turn out as a reimplementation anyway. I'm just describing it as a port so that my intentions are better understood.
00:37:41desophoswell, is there really such a thing as a port? unless it's java <-> C# or other very similar languages, i imagine most ports require reimplementation
00:37:44desophosmost decent ports anyway
00:38:23BlueProtomanYou're right. Again, I'm just calling it a port to make it not seem as daunting.
00:38:31desophoshaha sure
01:04:51*thotypous quit (Ping timeout: 252 seconds)
01:07:37*thotypous joined #nim
01:11:18*desophos quit (Ping timeout: 260 seconds)
01:18:18*desophos joined #nim
01:22:59*desophos quit (Ping timeout: 264 seconds)
01:29:39*ephja quit (Ping timeout: 255 seconds)
01:30:18*perturbation joined #nim
01:52:02*strcmp2 joined #nim
01:52:02*kniteli joined #nim
01:58:33*BlueProtoman quit (Quit: Leaving)
02:00:18*johnsoft quit (Quit: Leaving)
02:05:15*jaco60 quit (Ping timeout: 240 seconds)
02:05:44*perturbation quit ()
02:09:30*johnsoft joined #nim
02:37:28*distortedsignal joined #nim
02:44:56*desophos joined #nim
02:48:57*brson quit (Quit: leaving)
02:51:00*BitPuffin|osx quit (Ping timeout: 244 seconds)
03:18:46*strcmp2 quit (Read error: Connection reset by peer)
03:20:57*julian37 quit (Ping timeout: 250 seconds)
03:46:07*vendethiel joined #nim
04:17:17*julian37 joined #nim
04:17:18*xtagon joined #nim
04:22:30*julian37 quit (Ping timeout: 260 seconds)
04:33:41*darkf joined #nim
04:42:21*gunn joined #nim
04:47:30*thotypous quit (Ping timeout: 240 seconds)
04:47:39*zaquest quit (Ping timeout: 255 seconds)
04:48:38*zaquest joined #nim
04:50:37*thotypous joined #nim
04:50:42gunnHi
04:50:59gunnI'm trying to get nim to do some maths in parallel, but getting an error - cannot create a flowVar of type: BigInt
04:51:02gunnhttps://gist.github.com/gunn/9269eaf7ff001485ce5e
05:18:35*vendethiel quit (Ping timeout: 252 seconds)
05:30:49*zaquest_ joined #nim
05:31:08*zaquest quit (Ping timeout: 246 seconds)
06:00:36vegansk@gunn, the documentation says: Due to technical limitations not every type T is possible in a data flow variable: T has to be of the type ref, string, seq or of a type that doesn't contain a type that is garbage collected. This restriction is not hard to work-around in practice.
06:03:11gunnvegansk: yep, do have any ideas how to work around it?
06:05:32veganskSerialize to string in ``segment``, then deserialize to BigInt?
06:06:29veganskOr return ref BigInt?
06:14:01*Demon_Fox quit (Quit: Leaving)
06:15:54*cyraxjoe quit (Remote host closed the connection)
06:16:10*cyraxjoe joined #nim
06:18:24*julian37 joined #nim
06:19:14*cyraxjoe quit (Read error: Connection reset by peer)
06:19:37*cyraxjoe joined #nim
06:19:43gunnvegansk: ah, good ideas thanks
06:20:03gunnserializing to string and back works
06:20:18gunnI don't know enough nim yet to do it with refs
06:23:07*julian37 quit (Ping timeout: 252 seconds)
06:25:45*nande quit (Ping timeout: 255 seconds)
06:28:26*nande joined #nim
06:37:15*gour joined #nim
06:39:56*gour_ joined #nim
06:42:46*gour quit (Ping timeout: 250 seconds)
06:50:48*gour_ quit (Quit: WeeChat 1.3)
06:51:05*gour joined #nim
06:58:44*gour quit (Quit: WeeChat 1.3)
06:59:06*gour joined #nim
07:02:50*xtagon quit ()
07:02:51*gour quit (Client Quit)
07:04:57*gour joined #nim
07:06:23*gour quit (Client Quit)
07:07:19*gour joined #nim
07:11:23*gour quit (Client Quit)
07:12:54*gour joined #nim
07:23:02*desophos quit (Read error: Connection reset by peer)
07:23:09*nande quit (Remote host closed the connection)
07:23:47vegansk@gunn, like this: https://gist.github.com/vegansk/c5bfab052f3fe127ff73
07:25:09gunnvegansk: thanks!
07:25:30gunnI've been making slow trial and error progress
07:26:21gunnresult[] = sum sure beats what I'd been doing:
07:26:37gunnresult.limbs = sum.limbs; result.flags = sum.flags
07:46:35*yglukhov quit (Ping timeout: 240 seconds)
07:49:27*julian37 joined #nim
07:50:05*vqrs quit (Ping timeout: 246 seconds)
07:53:30*julian37 quit (Ping timeout: 240 seconds)
07:53:44*vqrs joined #nim
08:02:03*zaquest_ is now known as zaquest
08:05:17*distortedsignal left #nim (#nim)
08:11:16*strcmp1 joined #nim
08:22:54*gokr joined #nim
08:29:30*strcmp2 joined #nim
08:31:11*strcmp3 joined #nim
08:31:24*gokr1 joined #nim
08:32:46*strcmp1 quit (Ping timeout: 260 seconds)
08:33:45*strcmp2 quit (Ping timeout: 240 seconds)
08:35:03*gokr quit (Ping timeout: 265 seconds)
08:35:30gunnI think I've finished my parallel factorial program: https://gist.github.com/gunn/9269eaf7ff001485ce5e
08:35:42gunnThe number of threads can easily be scaled
08:37:15gunnI was hoping it would use multiple cpu cores but it seems to be just maxing out 100% of one
08:38:22Araqgunn: when this happens you're not giving each 'spawn' enough to work with
08:39:01Araqor the folding step takes all the time
08:39:10Araqbenchmark it
08:39:28gunnAraq: do you mean not enough work?
08:39:44gunnthe fold step is tiny
08:40:44Araqhrm
08:42:29*strcmp1 joined #nim
08:43:02Araqwell it should use 8 cores
08:43:09gunnAraq: the idea is each spawn processes 1/8th of the factorial each, the fold step multiplies the resulting 8 figures
08:43:18AraqI know
08:43:22gunnusing 50,000 as an input it takes 4 seconds
08:43:25strcmp1Araq: why is that? does spawn manage its own thread pool that distributes work according to load?
08:43:49Araqstrcmp1: yeah exactly but the code doesn't even rely on this fact
08:45:42*strcmp3 quit (Ping timeout: 255 seconds)
08:45:55Araqgunn: what if you use 100_000 as an input?
08:46:49*gour just watched first 30mins of Araq's video (to be continued...)
08:48:35gunnAraq: takes 0m18.108s, constant 99.9%
08:51:58Araqwell hrm. just time it against a non-parallel implementation
08:52:12*ziggypup joined #nim
08:52:13Araqand compute the speedup
08:52:14*ziggypup left #nim (#nim)
08:52:27Araq(which might even be a slowdown...)
08:53:03*Trustable joined #nim
08:53:37nivhello! i'm seeing some clang errors when compiling a specific piece of code that uses the sequtils.mapIt template. i'm new to nim so maybe im doing something wrong but this looks like a bug to me. https://gist.github.com/niv/67d0675d59ea0be77b50 can anyone confirm, and should i create a new issue?
08:55:06Araqniv: pretty sure this bug has been reported already
08:55:14Araqbut it cannot hurt to report it
08:55:23*strcmp1 quit (Ping timeout: 264 seconds)
08:55:25nivalright
08:55:40gunnAraq: updated https://gist.github.com/gunn/9269eaf7ff001485ce5e with the linear version. Almost identical times
08:56:08nivseems to be related to the object that is passed in as an arg and somehow mapIt. if i rewrite it to use map(), it compiles, even though neither of them ever references the struct
08:57:32*julian37 joined #nim
09:09:15*coffeepot joined #nim
09:10:50Araqgunn: is parallel slower or faster?
09:12:00gunnAraq: parallel is faster by about 1%
09:12:22Araqlol that's pretty bad, do you use -d:release ?
09:12:26gunnyep
09:13:07gunnI would guess it's because of the order of the numbers being multiplied rather than any parallel effect
09:13:54Araqwell instead of 1.. N/8 you can increment the step count by 8
09:14:44Araqso the first thread does 1*9*17*25 ... , the second 2*10*18*26
09:14:54Araqetc
09:15:26Araqso every thread handles equally large numbers
09:16:10gunnthat would be smarter
09:23:53*darkf quit (Ping timeout: 246 seconds)
09:24:01*s4 joined #nim
09:24:11*darkf joined #nim
09:33:07*wan quit (Quit: WeeChat 1.3)
09:40:36*yglukhov joined #nim
09:41:10gourAraq: what does it mean "Install wxnim next to a directory wxWidgets-3.0.2", iow. is it possible to install wxnim via nimble?
09:45:06gourAraq: otoh, if i try to compile example1, i get the following: https://bpaste.net/show/b351be26615d
09:47:20Araqgour: it means to have the following dir structure foobar/wxWidgets-3.0.2
09:47:24Araqfoobar/wxnim
09:47:32gourthat i have
09:47:43gourhow to invoke build?
09:48:04Araqand you need to build wxWidgets-3.0.2 via its autoconf abomination or something
09:48:17Araqso that it generates wx/setup.h
09:48:33gourahh, ok.
09:49:07*ziggypup joined #nim
09:49:10*ziggypup left #nim (#nim)
09:49:44Araqand yeah you should write a wx.nims script file to automate this process, when I wrote wxnim there was no nims
09:59:31nivI have a design question. why is nim using signed integers (int) for values that are clearly unsigned - for example, stream positions, or array offsets?
10:01:55Araqbecause there is no such thing as a "clearly unsigned integer"
10:02:18Araqyou're confusing natural numbers which cannot be zero with unsigned wrap around arithmetics
10:02:39Araqfor i in 0 .. length - 4: echo i
10:02:53Araqshould run 0 times when length == 0
10:02:59Araqnot ~4 billion times
10:03:17nivyeah, maybe so. i was thinking about stream.setPosition, for example. on 32bit, you're limited to 2GB of stream size, and it accepts negative positions too
10:04:01Araqmaybe it should indeed take int64
10:04:22nivthats all im asking, if there was a design decision why using int was the better choice that im missing
10:04:57*makoLine quit (Ping timeout: 252 seconds)
10:05:14Araqwe usually use 'int' as the default type
10:06:06nivi agree it doesn't matter in 99% of cases, and makes for simpler&prettier reading :) but w.r.t. file streams, or similar procs that use offsets, it might run into issues
10:07:03Araqpretty sure I used streams for files bigger than 4 GB
10:07:35AraqsetPosition is a weird operation for streams to begin with
10:07:40nivi havent tried it, just thinking academically here. int is sizeof(pointer) on 32bit, which would limit it to that for the given procs at least
10:08:01AraqsetPosition is limited to this, streams are not
10:08:32nivi using stream.setPos to parse a binary file format, like zip - where it reads a header and then just gives you an offset where the actual data is
10:09:34nivits not going to run into the 32bit wall anytime soon :) but i just noticed it while browsing the api docs
10:10:27nivanyways, alright, thanks :)
10:10:41Araqyeah we should change it to int64
10:12:03Araqyglukhov: your bug is dead simple to reproduce ;-)
10:12:53yglukhovAraq: and dead simple to fix i suppose? ;)
10:14:55Araqyup
10:20:13gunnAraq: I think the behaviour we're seeing is simply due to multiplying two massive BigInts taking a long time - i.e. the last 8 multiplications in the foldl actually is the bulk of execution time.
10:20:21gunnwhoops
10:20:26gunnI rewrote as you suggested, same performance.
10:21:06Araqbenchmark it to be sure
10:21:18Araqbut if so that's a really nasty problem :-/
10:21:38Araqbecause then you need to parallelize bignum multiplications
10:22:12gunnwell the figures we're talking about are very big
10:22:43gunne.g. 9e89000 * 3e110000
10:24:17gunnfigures on that scale take about 2 seconds to multiply
10:25:21*jaco60 joined #nim
10:25:43Araqwell you should see a speedup of 8 for smaller numbers then?
10:26:03Araqwith the increment by 8 version
10:26:51Araqand for tiny numbers a slowdown because of threading overhead -.-
10:27:43gunnAraq: very close for every input
10:28:15gunnI suspect this is just the nature of factorials scaling quickly
10:28:48Araqperhaps but also try multiprocessing
10:29:30gunnwhat do you mean?
10:29:30Araqmaybe there is also some serious serialization bug in the threadpool preventing it from scaling
10:29:35dom96LibMan has a good eye for videos.
10:38:11*mindriot101 joined #nim
10:53:02*Varriount-Laptop joined #nim
10:57:45*theduke quit (Ping timeout: 252 seconds)
11:09:54*theduke joined #nim
11:13:55nivis there a (high-level) evented networking library native to nim i'm overlooking? something like ruby's eventmachine
11:14:26Araqasyncdispatch ?
11:14:53nivapparently :) thanks.
11:14:58gournimqml looks interesting...
11:15:52gunnAraq: some benchmarking shows performance scales predictably, so I don't think what we
11:16:01gunnwe've seen is a problem
11:17:47gunnBigInt performance could be improved though, nim consistently takes 70% longer than ruby with a simple script: (1..n).inject(:*)
11:19:17gunnalso there seems to be a bug - have a look at the output here: https://gist.github.com/gunn/9269eaf7ff001485ce5e#file-parallel_factorial-log
11:19:34gunn824199001261152424427520000 * 6059466975437025204783660349002445834010410572229490978416137871632000000000000 = 0
11:20:17gunnunless I've missed something simple...
11:22:05*andrewvos joined #nim
11:23:11andrewvoswhy do I get this error? https://gist.github.com/AndrewVos/092acfb2e8e1826cebf4
11:23:52Araqbecause it's a tuple, not an object
11:24:32Araqactions.add((topath, frompath))
11:25:17andrewvosohhhh
11:25:26andrewvosthanks
11:25:33Araqarguably the compiler/language could allow this form of construction for tuples, but named tuples are kinda bad style
11:26:12andrewvosbut I need to name the tuple to have it in a seq right?
11:26:30andrewvostype mismatch: got (seq[SymlinkAction], (string, string))
11:26:40andrewvosI still get that error with `actions.add((toPath, fromPath))
11:26:43andrewvos`
11:27:17Araqseq[(string, string)]
11:27:25Araqis a perfectly acceptable type
11:27:41andrewvosI see, so I should change the type of my seq
11:35:14NimBotnim-lang/Nim devel edb495a Araq [+0 ±1 -0]: fixes a critical error handling bug for nimsuggest
11:35:14NimBotnim-lang/Nim devel aad4ed7 Araq [+0 ±1 -0]: allows macros to access documentation comments
11:35:14NimBotnim-lang/Nim devel 698b1fc Araq [+0 ±2 -0]: fixes #3550
11:39:11*dngad joined #nim
11:39:42dngadHello everyone :)
11:39:51*Varriount-Laptop quit (Ping timeout: 255 seconds)
11:45:51*elrood joined #nim
11:51:20federico3hi dngad
11:52:23*julian37 quit (Ping timeout: 252 seconds)
11:53:42nivwhats the best way to split up a seq of length X*N into sub-seqs (tuples) of size N with X elements each? i've found distribute(), but that doesn't seem to be the best usecase for it. is there something like the ruby-eqv of Array#each_slice?
11:57:51*Varriount-Laptop joined #nim
12:02:49nivnvm :)
12:05:10*yglukhov quit (Ping timeout: 250 seconds)
12:18:21*julian37 joined #nim
12:22:29*strcmp1 joined #nim
12:25:23*theduke quit (Ping timeout: 252 seconds)
12:39:22NimBotnim-lang/Nim devel 528cf3e Araq [+0 ±1 -0]: better error handling for bug #3550
12:39:22NimBotnim-lang/Nim devel 2f07ac7 Araq [+0 ±2 -0]: website updated
12:41:54*BitPuffin joined #nim
12:47:11*Varriount-Laptop quit (Ping timeout: 276 seconds)
13:24:41*Demon_Fox joined #nim
13:25:25*s4 quit ()
13:37:22*ephja joined #nim
13:40:31*yglukhov joined #nim
13:45:34*yglukhov quit (Ping timeout: 272 seconds)
13:48:14*Sembei quit (Ping timeout: 260 seconds)
13:51:56*gokr1 quit (Quit: Leaving.)
13:52:22*yglukhov joined #nim
14:08:36*mindriot101 quit (Remote host closed the connection)
14:11:54*Varriount-Laptop joined #nim
14:12:25Araqhrm what's our policy for bugfixes which break code?
14:12:46Araqalbeit for the bettter. what used to produce 0/1 now produces false/true
14:12:58Araqtoo dangerous for 0.12.2 right?
14:13:36yglukhovimo, lets break the code! :D
14:14:29nchambersAraq, the benefit of not having a v1.0 is this is the time to break code
14:14:54nchambersI mean, I don't think you've done it more than rust has, so you're probably entitled to it a bit
14:15:11nchambersthe other solution is to go the python route or the c++ route
14:15:18coffeepot+1 for breaking if it makes more sense/more elegant
14:15:24Araqthe benefit of having a 0.12.2 is that it is "everything works as 0.12 but sometimes better"
14:15:43Araqthere is no doubt we can have the fix in 0.13
14:17:27Araqso ... I shall releave 0.12.2 without the fix and then work continues to produce 0.13. IMHO that's how semantic versioning is supposed to work
14:26:21*Varriount-Laptop quit (Ping timeout: 255 seconds)
14:31:24*Jehan_ joined #nim
14:34:07dom96Araq: According to http://semver.org/: "Major version zero (0.y.z) is for initial development. Anything may change at any time. The public API should not be considered stable."
14:34:19dom96i.e. you shouldn't care right now
14:36:03dom96Devel already contains some changes which break backwards compat
14:39:13*julian37 quit (Ping timeout: 252 seconds)
14:39:40Araqyay -.-
14:39:46Araqok, then
14:40:13*strcmp1 quit (Quit: return 0;)
14:45:58dom96we can probably easily revert them if you really want
14:48:43Araqmeh I guess bugfixes which break weird code are fine
14:49:00AraqI doubt it will affect any code out there
14:49:09Araqexcept our own tests of course
14:50:59*gokr joined #nim
14:51:00*gokr quit (Read error: Connection reset by peer)
14:51:13*gokr joined #nim
14:52:55*yglukhov_ joined #nim
14:52:55*yglukhov quit (Read error: Connection reset by peer)
14:57:24*yglukhov_ quit (Ping timeout: 255 seconds)
15:04:11*kniteli quit (Remote host closed the connection)
15:08:28NimBotnim-lang/Nim devel 1757b84 Araq [+0 ±1 -0]: fixes #3546
15:08:28NimBotnim-lang/Nim devel ba6d0eb Araq [+0 ±4 -0]: fixes #3541
15:09:17*mindriot101 joined #nim
15:10:34*gokr quit (Quit: Leaving.)
15:11:58*askatasuna joined #nim
15:12:44*mindriot101 quit (Remote host closed the connection)
15:13:03*mindriot101 joined #nim
15:13:39*strcmp1 joined #nim
15:14:34strcmp1yeah i don't think semver really kicks in until 1.0 and after. and if you make a breaking change the next release must be v2.0, kind of annoying i guess
15:16:19*cryzed quit (Ping timeout: 240 seconds)
15:16:39*mal`` quit (Ping timeout: 240 seconds)
15:16:44*cryzed joined #nim
15:17:19*n1ftyn8_ quit (Ping timeout: 240 seconds)
15:17:39*Xe quit (Ping timeout: 240 seconds)
15:18:28*n1ftyn8_ joined #nim
15:18:36*dngad_ joined #nim
15:19:38*mal`` joined #nim
15:20:08*Xe joined #nim
15:21:53*dngad quit (Ping timeout: 276 seconds)
15:26:46*Guest48911isaway is now known as Guest48911
15:36:21*enamex joined #nim
15:54:46*yglukhov joined #nim
15:56:48*pregressive joined #nim
15:57:15*Jesin quit (Quit: Leaving)
15:59:10*Jesin joined #nim
16:03:32*yglukhov quit (Remote host closed the connection)
16:03:46*yglukhov joined #nim
16:03:53*gokr joined #nim
16:07:34*askatasuna quit (Read error: Connection reset by peer)
16:11:36*kniteli joined #nim
16:19:50*strcmp1 quit (Ping timeout: 240 seconds)
16:25:40*strcmp1 joined #nim
16:28:11*ochkoffein joined #nim
16:33:22*strcmp2 joined #nim
16:36:00*strcmp1 quit (Ping timeout: 250 seconds)
16:40:43*desophos joined #nim
16:52:09*enamex quit (Ping timeout: 255 seconds)
17:03:12*enamex joined #nim
17:06:22*Demon_Fox quit (Quit: Leaving)
17:07:09*Demon_Fox joined #nim
17:09:42*andrewvos quit (Ping timeout: 260 seconds)
17:15:47*coffeepot quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client)
17:20:38*strcmp1 joined #nim
17:20:55*strcmp1 quit (Client Quit)
17:22:26*strcmp1 joined #nim
17:24:38*strcmp2 quit (Ping timeout: 260 seconds)
17:28:40*mindriot_ joined #nim
17:31:45*mindriot101 quit (Ping timeout: 255 seconds)
17:33:02*kniteli quit (Ping timeout: 260 seconds)
17:33:10*mindriot_ quit (Ping timeout: 260 seconds)
17:36:33*gokr quit (Quit: Leaving.)
17:42:44*brson joined #nim
17:58:16*cyraxjoe quit (Ping timeout: 272 seconds)
18:03:53*yglukhov quit (Remote host closed the connection)
18:20:26*mindriot101 joined #nim
18:22:55*Gonzih joined #nim
18:25:21*mindriot101 quit (Remote host closed the connection)
18:30:07*julian37 joined #nim
18:36:05*makoLine joined #nim
18:44:27*cyraxjoe joined #nim
18:58:00*mindriot101 joined #nim
19:11:28*yglukhov joined #nim
19:13:03onionhammerok guys, someone needs to write a VS Code Nim extension
19:13:13*darkf quit (Quit: Leaving)
19:13:18*Gonzih quit (Ping timeout: 250 seconds)
19:15:35*yglukhov quit (Ping timeout: 240 seconds)
19:17:50*julian37 quit (Remote host closed the connection)
19:18:05*julian37 joined #nim
19:25:06federico3VS Code?
19:25:10onionhammervisual studio code
19:25:13onionhammerjust got open sourced
19:25:51onionhammerhttps://code.visualstudio.com/Updates
19:27:21onionhammerhttps://marketplace.visualstudio.com/#VSCode
19:33:57*exebook quit (Remote host closed the connection)
19:39:22*Matthias247 joined #nim
19:43:38*vendethiel joined #nim
19:47:51*BitPuffin quit (Ping timeout: 265 seconds)
20:07:19*vendethiel- joined #nim
20:08:46*vendethiel quit (Ping timeout: 240 seconds)
20:11:26*Sornaensis quit (Excess Flood)
20:11:51*Sornaensis joined #nim
20:12:32*str joined #nim
20:16:11gmpreussner|workonionhammer: i'm trying to understand what VS Code is
20:16:19gmpreussner|workis this some scaled down version of VS?
20:16:29gmpreussner|worki.e. just the text editor part?
20:16:56*Tahr-user quit (Quit: Leaving)
20:20:10stran atom "replacement"
20:20:22strBUT, has Go IDE support
20:20:39striirc
20:22:09gmpreussner|workso this is meant for apps that run in a web browser, yes?
20:22:38strweb/cloud apps
20:22:46strnot specifically front-end
20:26:25*mat4 joined #nim
20:26:27mat4hello
20:26:38strpersonally keeping sublime test around for a bit
20:26:45str*text even
20:26:57dngad_Personally looking forward to IntelliJ support.
20:29:58*mindriot101 quit ()
20:31:21Matthias247it's an extensible text editor, just like vi, emacs, sublime text, atom, bracket, etc.
20:32:16Matthias247but with the added debugger support for some languages it a little bit ahead of the competition in the IDE department
20:32:26*str quit (Quit: leaving)
20:32:36*dngad_ quit (Quit: Leaving)
20:32:42desophosyeah i would use it for languages it's intended to support
20:32:49Matthias247and not, it's not only suitable for web apps. In our company a lot of people use it for C
20:33:05desophosyou use VSC for C?
20:33:06desophosalready?
20:33:17desophoswasn't it just released?
20:33:21desophosor just open sourced i guess
20:33:36Matthias247it was open sourced today, it was released half a year ago
20:33:46desophosah okay
20:33:59desophosi've been thinking about moving to Atom
20:34:12desophosit has a Nim plugin but so does Sublime, which i'm currently using
20:34:23Matthias247I currently use them both, since I currently working on some typescript stuff
20:34:27desophoshm
20:34:41Matthias247Atom has currently much more plugins, but is also a lot slower
20:35:03desophosdoes it have multiple selection? cmd-D is the best
20:35:07Matthias247yes
20:35:17desophosgood
20:35:34Matthias247I think if it improves just a little bit I don't see any need for Atom anymore
20:35:55desophoswait you mean Atom is slower than VSC?
20:35:58desophosor than Sublime?
20:36:02Matthias247than VSC
20:36:05desophosah okay
20:36:15Matthias247sublime is faster then both. But barely maintained
20:36:17desophosi'd try VSC if it had a Nim plugin :p
20:36:30desophosi assume it has a python plugin
20:36:35Matthias247It would be possible since today
20:36:45desophosi pretty much only write python and nim now
20:37:16desophosand robot... sigh
20:38:20desophosthe sublime robot plugin is terrible
20:38:20Matthias247for VSC extensions you need javascript. I'm not the biggest fan of it, but with typescript it's a really good experience
20:40:12mat4Matthias247: Do you mean lexical debugging, tracer or a binary debugger?
20:40:34desophosi assumed tracing
20:41:09Matthias247mat4: not familiar with the terms, but it looks like how a debugger in an IDE looks ;)
20:41:46Matthias247but I only used it for node.js until now. There you can simply start your app with debugger attached or attach later, set breakpoints, inspect variables, ...
20:41:51*yglukhov joined #nim
20:42:15Matthias247and you can debug either js or with sourcemaps also typescript
20:42:47Matthias247debugging is also supported for C#, but I haven't used it yet
20:43:42Matthias247since today there is this: https://code.visualstudio.com/docs/extensionAPI/api-debugging
20:44:05Matthias247that means you could add debugger support for other languages with plugins
20:44:13Matthias247golang was already added
20:45:12mat4nice interface abstraction!
20:46:20*Demon_Fox quit (Remote host closed the connection)
20:48:42*mat4 probably VLC runs with Wine
20:49:08desophosVLC is cross-platform...
20:49:10desophosso is VSC
20:51:27mat4sorry, I mean Visual Studio (the IDE from Microsoft)
20:55:19mat4for Nim development I use Joe at moment (because it is a Wordstar clone, is a console application and I have added syntax highlighting)
20:56:41mat4also I like ease editing though regular expressions
20:58:35Matthias247you can create VSC syntax highlighting packages out of textmate highlighting schemes (which is what sublime uses). So that means adding NIM highlighting support to VSC should be a matter of 20 minutes
21:00:46*kulelu88 joined #nim
21:05:22*enquora joined #nim
21:09:42*enquora quit (Client Quit)
21:12:45dom96Atom is way too slow in my experience, VSC way too immature.
21:13:07dom96The googleability of VSC is particularly annoying
21:14:24mat4hmm, what's the URL for VSC by the way ?
21:18:45Matthias247https://code.visualstudio.com/
21:21:41kulelu88if you're looking for an alternate, support lime text
21:24:49mat4Matthias247: thanks
21:25:33mat4kulelu88: what's special about lime text ?
21:25:35nchambers<mat4> for Nim development I use Joe at moment (because it is a Wordstar clone, is a console application and I have added syntax highlighting) ← heh... wordstar
21:26:57kulelu88mat4: its the open-source version of sublime
21:27:13mat4thanks, I take a look
21:27:22kulelu88I don't advocate it for having advanced text-editor qualities
21:27:44kulelu88but among all the noise, I'm glad an indie project like it exists
21:41:43mat4it's written in Go as I see
21:45:36kulelu88sadly yeah
21:48:41*strcmp1 quit (Ping timeout: 246 seconds)
21:51:04*pregressive quit (Remote host closed the connection)
21:51:21*pregressive joined #nim
21:52:55*andrewvos joined #nim
22:01:22*desophos_ joined #nim
22:01:28*desophos quit (Ping timeout: 272 seconds)
22:04:56*gour quit (Quit: WeeChat 1.3)
22:07:31*tinAndi joined #nim
22:09:21*desophos_ quit (Remote host closed the connection)
22:15:43*mat4 quit (Quit: leaving)
22:22:55*heinrich5991 quit (Quit: quit.)
22:24:02*vendethiel- quit (Ping timeout: 246 seconds)
22:28:45*Guest22572 quit (Ping timeout: 255 seconds)
22:30:35*desophos joined #nim
22:31:32*elrood quit (Quit: Leaving)
22:37:27*pregressive quit (Remote host closed the connection)
22:37:44*pregressive joined #nim
22:39:27*pregressive quit (Read error: Connection reset by peer)
22:39:37*pregressive joined #nim
22:45:24*strcmp1 joined #nim
22:49:58*strcmp1 quit (Ping timeout: 250 seconds)
22:50:32*BlueProtoman joined #nim
22:52:40BlueProtomanAnyone know if it's possible to use a built-in type to represent an external type? (In this case, a basic2d.Vector2d being treated as a Box2D b2Vec2) http://stackoverflow.com/q/33789764/1089957
22:56:37*desophos quit (Remote host closed the connection)
22:59:43*desophos joined #nim
23:02:17*Trustable quit (Remote host closed the connection)
23:04:43*tinAndi quit (Quit: ChatZilla 0.9.92 [Firefox 42.0/20151029151421])
23:05:45*Matthias247 quit (Read error: Connection reset by peer)
23:09:10*desophos quit (Remote host closed the connection)
23:17:16*desophos joined #nim
23:18:19*buMPnet joined #nim
23:20:32*enamex quit (Quit: Leaving)
23:21:10*heinrich5991 joined #nim
23:21:27*heinrich5991 quit (Remote host closed the connection)
23:25:24*strcmp1 joined #nim
23:27:18*heinrich5991 joined #nim
23:32:24*Guest48911 is now known as Guest48911isaway
23:34:10*kulelu88 left #nim ("Leaving")
23:38:05desophosBlueProtoman: did you get an answer?
23:38:12desophosalso hi
23:46:03BlueProtomandesophos: Hello. Yeah, Araq posted one, but at the same time I figured out that I can't do that with Vector2d <-> b2Vec2 because they're actually of different size (floats in Nim are 64-bit, floats in C++ are usually 32)
23:46:28BlueProtomanSo I'll just have to make a private b2Vec2 wrapper and use it under the hood
23:48:21desophosporting box2d to nim?
23:48:46BlueProtomanWell, wrapping it, but yes.
23:48:52desophosawesome
23:49:10BlueProtomanDon't wanna do a full-fleged port because Erin Catto is a lot better at manual memory management optimizations than I am.
23:49:32desophosyeah no need to port a physics engine i think
23:52:30*nande joined #nim