<< 25-09-2016 >>

00:01:45*nsf joined #nim
00:09:09*chemist69 quit (Ping timeout: 272 seconds)
00:13:16*chemist69 joined #nim
00:34:06*yglukhov joined #nim
00:39:18*yglukhov quit (Ping timeout: 276 seconds)
00:40:22*sp33der89 joined #nim
00:46:08*Trustable quit (Remote host closed the connection)
00:56:24*JStoker quit (Ping timeout: 265 seconds)
01:01:48*yglukhov joined #nim
01:02:05*JStoker joined #nim
01:06:36*yglukhov quit (Ping timeout: 276 seconds)
01:42:52*chemist69 quit (Ping timeout: 265 seconds)
01:50:29*sp33der89 quit (Ping timeout: 272 seconds)
01:55:52*chemist69 joined #nim
02:23:22*nsf quit (Quit: WeeChat 1.5)
02:51:48*Senketsu_ quit (Read error: Connection reset by peer)
02:52:13*Senketsu_ joined #nim
02:57:49*ARCADIVS joined #nim
03:32:15*yglukhov joined #nim
03:37:00*yglukhov quit (Ping timeout: 260 seconds)
03:43:51*chemist69 quit (Ping timeout: 272 seconds)
03:47:32*chemist69 joined #nim
04:00:33*desophos quit (Read error: Connection reset by peer)
04:01:24*arnetheduck joined #nim
04:13:25*yglukhov joined #nim
04:18:21*yglukhov quit (Ping timeout: 276 seconds)
04:44:47*vqrs quit (Ping timeout: 250 seconds)
04:49:36*vqrs joined #nim
05:10:50*spindle quit (Read error: Connection reset by peer)
05:16:34*yglukhov joined #nim
05:21:30*yglukhov quit (Ping timeout: 264 seconds)
05:29:44*andrewo91 quit (Remote host closed the connection)
05:35:18*wgf_ joined #nim
05:51:14*chemist69 quit (Ping timeout: 265 seconds)
06:17:45*chemist69 joined #nim
06:23:51*mcc quit (Quit: Connection closed for inactivity)
06:33:30*LeNsTR is now known as lenstr
07:30:46*Trioxin quit (Remote host closed the connection)
07:33:15*Trioxin joined #nim
07:38:10*irrequietus joined #nim
07:59:01*yglukhov joined #nim
08:03:50*MyMind joined #nim
08:06:36*Sembei quit (Ping timeout: 250 seconds)
08:10:57*gokr joined #nim
08:20:35gokrHowdy all
08:21:30*gokr contemplating concurrency models...
08:57:23*yglukhov quit (Remote host closed the connection)
09:00:26dom96hello gokr
09:06:26*beatmox joined #nim
09:06:26cheatfatedom96, just to remember, i'm still waiting for your checks of tasynceverror.nim
09:07:31dom96cheatfate: oh yeah, why do you want to remove it again?
09:09:04cheatfateif you see in comments there EPIPE must be generated, but current version of asyncdispatch generates EBADF
09:10:03cheatfateon BSD/MacOS this test not working properly because BSD network allow to send() to closed socket because of timeouts...
09:10:21cheatfateso its impossible to generate EPIPE
09:10:58cheatfatecurrent version asyncdispatch not properly handles it and tries to add to kqueue empty handle so generates EBADF
09:11:27dom96alright, remove it then. We have another test that tests the PR this is related to at least partially, tasyncconnect.
09:11:41dom96This is the PR in case you're interested https://github.com/nim-lang/Nim/pull/2529
09:12:44cheatfateThis PR is also bad idea because kqueue can generate only EvError without EvRead and/or EvWrite
09:13:24cheatfateand we can't handle it
09:17:17dom96I'm not sure I understand
09:17:22dom96Why can't we handle it?
09:18:17dom96I think it makes sense to handle errors in the underlying async procs
09:24:29cheatfatedom96, now we can handle errors only if {EvRead, EvError} or {EvWrite, EvError} received, what poll() can handle if {EvError} generated only?
09:24:47*nsf joined #nim
09:26:04*gokr quit (Read error: No route to host)
09:26:10*gokr1 joined #nim
09:26:52dom96okay, I get it now.
09:27:05dom96What should happen if {EvError} is generated only?
09:27:21dom96Whatever it is, I'm sure we can add logic to do it only for that case.
09:28:09cheatfatedom96, i think it must check if there readCB or writeCB and call all of cbs found... if there no cbs at all then generate error
09:28:39dom96sure, sounds good.
09:41:36*yglukhov joined #nim
09:43:06*yglukhov quit (Remote host closed the connection)
09:50:04*gokr1 quit (Ping timeout: 265 seconds)
10:09:26*wgf_ quit (Quit: Leaving)
10:18:19*spindle joined #nim
10:21:18*nsf quit (Quit: WeeChat 1.5)
10:22:52*nsf joined #nim
10:23:27dom96hrm, so I'm benchmarking lwan (as it seems to be one of the fastest HTTP servers out there) on my MacBook, and I only get 31k req/s
10:23:43dom96Which in comparison to this guy is significantly lower https://news.ycombinator.com/item?id=12274255
10:27:48Calinougotta like how people benchmark number of requests/s on their web servers
10:27:53Calinouyet they get like 100 hits a day :^)
10:28:17dom96yep
10:28:43dom96asynchttpserver gets 19k req/s on my machine
10:28:54Calinoudid you try Caddy?
10:29:02dom96no, what is it?
10:29:04Calinouhttps://caddyserver.com/
10:29:11cheatfatedom96, how many cores your mac have?
10:29:11CalinouHTTP/2 server written in Go
10:29:15Calinoulooks like it's great for static hosting
10:29:32dom96cheatfate: I forced it to use a single thread
10:29:51dom96According to this: https://www.techempower.com/benchmarks/#section=data-r12&hw=peak&test=json
10:29:55dom96lwan is fastest for json
10:30:53cheatfatedom96, i think our json is not very fast
10:31:22dom96I'm not testing json
10:31:26dom96just plain text
10:32:05dom96jester isn't in any of the techempower benchmarks anymore :\
10:34:30dom96this seems silly
10:34:55dom96oh well
10:35:07dom96Calinou: Made any new progress on the website?
10:59:08FromGitter<gogolxdong> what is correct way to express `if 1<sigfigs or sigfigs>5:echo "sigfigs must be [1,5] was"$sigfigs`
11:03:05*aziz joined #nim
11:03:11dom96if 1<sigfigs or sigfigs>5:echo "sigfigs must be [1,5] was ", $sigfigs
11:09:24*elrood joined #nim
11:14:36hohlerdeis there a way in nim to print the abstract syntax tree in human readable form?
11:15:21*yglukhov joined #nim
11:18:07dom96hohlerde: toStrLit(node)
11:18:16dom96or treeRepr(node)
11:20:12*yglukhov quit (Ping timeout: 276 seconds)
11:23:45*krux02 joined #nim
11:26:41hohlerdethx dom96, works
11:29:43hohlerdewow, this is pretty cool, haven't worked much with macros, only templates so far
11:34:40*brechtm joined #nim
11:41:00*yglukhov joined #nim
11:41:51*yglukhov quit (Remote host closed the connection)
11:42:41dom96yep, macros are awesome :)
11:47:00*yglukhov joined #nim
11:51:52*yglukhov quit (Ping timeout: 265 seconds)
11:52:02Calinoudom96: nope :( had to work on university stuff instead
11:52:25*yglukhov joined #nim
11:57:10*miqlas-H joined #nim
11:57:14miqlas-HHi Guys!
11:57:15*yglukhov quit (Ping timeout: 276 seconds)
11:57:41miqlas-HI'm trying to get Nim first-class citizen on Haiku.
11:58:18miqlas-HI built the last 3 revision on Haiku sucessfully, but no recipe made for automatic build
12:03:24reactormonkmiqlas-H, with the linux compat layer, or native?
12:03:51miqlas-HI think we don't have linux compat layer. Do you mean POSIX?
12:04:09reactormonkyeah, sorry.
12:04:35miqlas-HWe have freebsd compat layer, and yes, i use it. But it is nothing else, just some extra header files, activated with -D_BSD_SOURCE
12:04:47miqlas-HIf i'm wrong, please, fixme. :)
12:05:36miqlas-HLast time i tried it worked really well, i has able to compile and run some sample codes.
12:05:40dom96pretty sure that Nim has some support for Haiku natively.
12:07:29miqlas-Hdom96, what "native" is? Ofc, there is some preliminary supporting code for building on Haiku, but it is just in the supporting codes (scripts and so on). For building on Haiku i use the vanilla sources
12:07:31dom96cheatfate: your tupcoming_async test is very unreliable, still getting false negatives on it.
12:08:03dom96By "native" I mean that it doesn't use any compatibility layers
12:08:59ftsfhttps://www.youtube.com/watch?v=sDop4JELAhc
12:13:05miqlas-Hnim tries to link koch with -ldl on Haiku, but we don't have -ldl. Where is it defined? I just grepped the whole source tree, no occurences.
12:20:26miqlas-Hit compiled sucessfully without -ldl, but i don't see, where it gets the -ldl
12:21:32miqlas-HAny idea?
12:22:05*nsf quit (Quit: WeeChat 1.5)
12:24:20cheatfatedom96, problems with time measuring?
12:24:30*gokr joined #nim
12:24:32dom96cheatfate: looks like it
12:24:43cheatfatei will disable timecheck then
12:26:53miqlas-Hok, it comes from nim.cfg
12:39:45*bjz quit (Max SendQ exceeded)
12:41:44*bjz joined #nim
12:42:45*gokr quit (Ping timeout: 276 seconds)
12:51:17*euantor quit (Read error: Connection reset by peer)
12:51:39*euantor joined #nim
12:52:31*Kaini quit (Quit: No Ping reply in 180 seconds.)
12:52:35miqlas-HOk, i sent my pull requests for Haiku support.
12:52:41miqlas-HI hope it is okay.
12:53:00dom96miqlas-H: Thanks! :)
12:53:39*Kaini joined #nim
12:54:27*ehmry quit (Ping timeout: 276 seconds)
12:54:27*M-Quora quit (Ping timeout: 276 seconds)
12:55:06*hohlerde quit (Ping timeout: 276 seconds)
12:56:11*yglukhov joined #nim
12:58:39*aziz quit (Quit: Ex-Chat)
13:00:59*yglukhov quit (Ping timeout: 265 seconds)
13:03:29*gokr joined #nim
13:06:16cheatfatemiqlas-H, if you dont have ldl then maybe you can't run 50% of nim's stuff
13:07:01miqlas-Hcheatfate: we have dl function, but not in -ldl.
13:07:30miqlas-HIt is in -lroot, and it is implicitly linked to everything.
13:07:47cheatfatemiqlas-H, what type of executables uses haiku?
13:09:02miqlas-HELF
13:09:25miqlas-His it an answer for your question?
13:10:14cheatfateyep
13:10:27miqlas-Hi sent a PR for csources too. It is also required to get it compiled on Haiku.
13:10:58miqlas-Hkoch buit without problems on Haiku with this patches. I think it is a good proof.
13:11:18*vqrs quit (Ping timeout: 264 seconds)
13:12:00miqlas-Hhttp://chunk.io/f/1485a274ff5940a89472ce0ddb071aca
13:12:07miqlas-HIt is a screenshot
13:13:14miqlas-HYeeee, haiku screenshot, frickin awesome, you just got the feeling, that you should download ad install it.
13:13:24miqlas-HHave no fear, it is normal.
13:13:49miqlas-HEverybody feels the same as looking at a Haiku screenshot at first time.
13:16:13*vqrs joined #nim
13:17:25*brechtm quit (Remote host closed the connection)
13:21:25cheatfatemiqlas-H, i have seen BeOS :)
13:22:48*M-Quora joined #nim
13:25:12*ehmry joined #nim
13:26:31miqlas-HGuys, is there any GUI toolkit, what we can use with Nim?
13:27:38gokrmiqlas-H: There are several. Aporia uses Gtk2, there is also a Gtk3 wrapper, and a wx wrapper that I am not sure anyone uses. And then there is also a libui wrapper which I actually use a teeny bit.
13:27:51gokrAnd there is also ... Qt stuff.
13:28:20miqlas-HQt is okay, because we don't have any GTK.
13:28:34gokrIMHO the libui wrapper is nice, since it maps to native on all supported backends.
13:29:02miqlas-HI need to go Guys, i talk to you later
13:29:04miqlas-HBye
13:29:12*miqlas-H quit (Quit: Vision[0.9.7-H-20140108]: i've been blurred!)
13:33:53*hohlerde joined #nim
13:38:34*spindle quit (Remote host closed the connection)
13:53:26*sp33der89 joined #nim
13:55:27*yglukhov joined #nim
13:56:41*vqrs quit (Ping timeout: 240 seconds)
14:03:16*vqrs joined #nim
14:24:03*nsf joined #nim
14:29:40*brechtm joined #nim
14:29:52*yglukhov quit (Remote host closed the connection)
14:48:15*yaiyan is now known as Yaiyan
15:05:57*nsf quit (Quit: WeeChat 1.5)
15:09:23*PMunch joined #nim
15:17:54*wgf_ joined #nim
15:21:21*vqrs quit (Ping timeout: 276 seconds)
15:24:23*vqrs joined #nim
15:29:21*desophos joined #nim
15:31:05*PMunch quit (Quit: leaving)
15:46:45*ARCADIVS quit (Quit: ARCADIVS)
15:57:37*yglukhov joined #nim
15:58:48*yglukhov quit (Remote host closed the connection)
16:03:09*irrequietus quit ()
16:03:22*irrequietus joined #nim
16:03:53*brechtm quit (Ping timeout: 240 seconds)
16:15:59*yglukhov joined #nim
16:21:09*yglukhov quit (Ping timeout: 276 seconds)
16:23:09*andrewo91 joined #nim
16:25:51*kulelu88 joined #nim
16:25:59*nsf joined #nim
16:33:33*zielmicha[m] joined #nim
16:39:32*yglukhov joined #nim
16:44:33*yglukhov quit (Ping timeout: 276 seconds)
16:46:10*brechtm joined #nim
16:52:02*brechtm quit (Remote host closed the connection)
17:04:42*askatasuna quit (Ping timeout: 276 seconds)
17:05:25*askatasuna joined #nim
17:06:21*wgf_ quit (Remote host closed the connection)
17:16:01*irrequietus quit ()
17:20:41*irrequietus joined #nim
17:32:15*kulelu88 quit (Ping timeout: 272 seconds)
17:42:33dom96Araq_: Is there a blacklist to prevent the tester from compiling certain modules in lib/pure/?
17:44:36*kulelu88 joined #nim
17:49:58*vqrs quit (Ping timeout: 255 seconds)
17:52:44*vqrs joined #nim
18:00:04*mcc joined #nim
18:03:36*desophos quit (Read error: Connection reset by peer)
18:04:00*desophos joined #nim
18:08:17*andrewo91 quit (Remote host closed the connection)
18:08:44*andrewo91 joined #nim
18:11:08*libman joined #nim
18:13:42*andrewo91 quit (Ping timeout: 264 seconds)
18:22:41*andrewo91 joined #nim
18:35:15*PMunch joined #nim
18:38:08Araq_dom96: no but you can move modules into lib/upcoming or something
18:41:20dom96or implement a blacklist ;)
18:42:05*yglukhov joined #nim
18:42:24*chemist69 quit (Quit: WeeChat 1.5)
18:44:15*andrewo91 quit ()
18:44:28*chemist69 joined #nim
18:47:17cheatfateis Nim has construction which follows semantic of C's 'switch()` but with access to more then one case?
18:47:47cheatfatei mean in switch() i must `break` after case
18:47:55cheatfateor all cases will be performed
18:48:09Araq_yeah, there is no such concept in Nim
18:48:34Araq_it's just a fancy if-elif-elif with a full coverage check
18:54:03cheatfateok thanks
18:54:23dom96So it looks like the C source tarballs will only include the install_nimble script instead of actual Nimble sources.
18:54:27dom96What do you guys think of this?
18:56:24Araq_I think the "official source based way to install" Nim needs to be "run these instructions and add dir to PATH"
18:56:38Araq_because then nimsuggest, nimble, nim work out of the box
18:57:09Araq_and the non-source based installation is "use your OS'es package manager"
18:59:02federico3dom96: install_nimble is runnig git clone from GH. A source tarball should contain everything needed to build and be able to build without network access.
18:59:12libmanThere's always the Super Secret Secret Goto Pragma. ;)
19:00:47dom96federico3: Agreed. But can you give a practical reason why that is important?
19:02:08Araq_ah I didn't seee the obvious solution
19:02:24Araq_ship Nimble as *Nim* source code in the tarball, not as generated C
19:02:36dom96omg lol
19:02:43Araq_:P
19:02:54dom96So you're happy with that?
19:03:03federico3dom96: without that, there would be no way to guarantee that the same sources can build the same application (on different architectures, at different times)
19:03:20dom96Glad I suggested we discuss this here instead of over PM :P
19:04:47*brechtm joined #nim
19:09:21federico3dom96: and apart from the pratical and security problems, various licenses require people that provide binary packages to make sources available. If a build script downloads even only a little script that is not part of the source packages, the distribution might end up in violation of the license (especially when the upstream repo disappears)
19:11:32dom96Thanks, those are very good reasons.
19:14:01Calinougokr: libui maps to GTK+3 on Linux, not Xlib
19:14:41Araq_dom96: yeah sure
19:18:53dom96okay, great.
19:26:36dom96Araq_: Have you seen this? https://www.reddit.com/r/nim/comments/547inw/why_wasnt_nim_optimized_for_code_reading/
19:28:49Araq_skimmed it
19:29:24*desophos quit (Read error: Connection reset by peer)
19:31:15cheatfateAraq_, could we mark realloc bug as `showstopper`?
19:31:38euantorThe whole `result` argument makes little sense to me
19:32:01euantorYou'd be assigning the array to a variable anyway, Nom just has a convention to make that more obvious
19:32:41euantorOtherwise he'd be doing `var foo = bar` and `return foot`
19:32:51euantor*foo
19:32:56Araq_cheatfate: no because we don't want to delay the release any further
19:33:07Araq_but yeah, it's a pretty weird bug
19:40:39cheatfateits a very weird bug
19:41:10cheatfateand i think fixing this bug can fix many issues about leaks
19:44:00Araq_I doubt it
19:44:26Araq_at the end of the day it's just a stupid snippet allocating ever bigger blocks of memory
19:44:43Araq_so yeah, it fails too early with out of mem
19:44:58Araq_but the allocation pattern is just absurd
19:49:10cheatfateAraq_, but for what reason this memory manager allocates 20GB of ram?
20:00:39libman>wow<
20:03:05Araq_cheatfate: address fragmentation
20:03:12Araq_still my bet.
20:03:28Araq_it works better when you disable the "return mem back to OS" feature
20:03:39Araq_but as I said, I'm still investigating
20:07:09*irrequietus_ joined #nim
20:07:21*irrequietus_ quit (Client Quit)
20:07:55*irrequietus_ joined #nim
20:08:27*irrequietus_ quit (Client Quit)
20:10:42*irrequietus quit (Ping timeout: 264 seconds)
20:12:15cheatfateAraq_, one more question about arrays, with --verbosity:2 i got every time warnings about uninitialized arrays, is it possible to initialize array properly?
20:18:23krux02I get weired error messages in my code
20:18:27krux02tensormath.nim(222, 13) Error: cannot instantiate Tensor
20:18:27krux02got: (Dimensions)
20:18:27krux02but expected: (Dimensions)
20:19:40krux02the type tensor has a static value Dimensions that must be set on declaration so the tensor size is statically known
20:23:29krux02ok thanks I did some rubber duck debugging with you ;D
20:29:44dom96krux02: that deserves a bug report, the error message needs to improve.
20:32:01*planhths joined #nim
20:34:20Araq_hi planhths welcome
20:34:35planhthshello Araq_!!
20:34:44*libman quit (Read error: Connection reset by peer)
20:35:19*pie__ joined #nim
20:36:29*libman joined #nim
20:44:58krux02dom96: yes but it is kind of hard to isolate the bug, I have a macro that creates a macro call that creates a type expression
20:46:02*starbux joined #nim
20:46:17*Trioxin quit (Ping timeout: 240 seconds)
20:46:29krux02I realized that static doesn't work stable with any dynamically sized value type, so I faked a dynamically sized array but making everything the same constant length plus a length value in an object
20:47:19krux02object( N: int, dimensions: array[16, int])
20:47:52krux02but that thing is not nice to construct as type parameter, so I have a macro that does that
20:48:00krux02but aparently there is something not nice in there
20:48:30*xet7 joined #nim
20:51:25*libman quit (Remote host closed the connection)
21:06:09*Trustable joined #nim
21:06:19*planhths quit (Quit: Konversation terminated!)
21:07:18*elrood quit (Remote host closed the connection)
21:10:19*spindle joined #nim
21:12:34*desophos joined #nim
21:15:16*yglukhov quit (Remote host closed the connection)
21:18:40krux02I have a question, I have a case where the ast passed to a typed macro still has some macro calls in the argument ast. Could this be intentional or is it a bug?
21:26:48Araq_intentional
21:30:35*yglukhov joined #nim
21:34:18krux02ok, thanks
21:35:15*yglukhov quit (Ping timeout: 265 seconds)
21:36:56*kunev quit (Ping timeout: 250 seconds)
21:39:50*kunev joined #nim
21:53:39*krux02 quit (Quit: Verlassend)
22:06:32*pie__ quit (Ping timeout: 252 seconds)
22:13:17*yglukhov joined #nim
22:18:10*yglukhov quit (Ping timeout: 260 seconds)
22:22:32*PMunch quit (Quit: leaving)
22:25:25*starbux quit (Remote host closed the connection)
22:26:07*starbux joined #nim
22:26:20*bjz quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
22:28:07*Trustable quit (Remote host closed the connection)
22:54:18*Demon_Fox joined #nim
22:55:36*yglukhov joined #nim
22:55:44*girvo joined #nim
23:02:41*yglukhov quit (Ping timeout: 240 seconds)
23:32:31*yglukhov joined #nim
23:36:59*yglukhov quit (Ping timeout: 265 seconds)
23:45:15*brechtm quit (Remote host closed the connection)
23:45:29*gokr quit (Ping timeout: 240 seconds)