<< 05-10-2018 >>
00:00:24*thomasross joined #nim
00:00:38*rnrwashere joined #nim
00:01:08*vivus quit (Remote host closed the connection)
00:02:35*rnrwashere quit (Remote host closed the connection)
00:04:15*rnrwashere joined #nim
00:06:40*rnrwashere quit (Remote host closed the connection)
00:10:33*rnrwashere joined #nim
00:16:57*rnrwashere quit (Remote host closed the connection)
00:17:11*rnrwashere joined #nim
00:23:00*rnrwashere quit (Remote host closed the connection)
00:24:02*rlr quit (Read error: Connection reset by peer)
00:26:17*rnrwashere joined #nim
00:26:24*tiorock joined #nim
00:26:25*rockcavera quit (Killed (adams.freenode.net (Nickname regained by services)))
00:26:25*tiorock is now known as rockcavera
00:42:54FromGitter<gogolxdong> In a sdl2+nuklear GUI context, which to set the font, sdl2 or nuklear?
00:45:46FromGitter<zacharycarter> https://twitter.com/Hahaitsfunny/status/1047996844872224768?s=19 ⏎ Very personal - but hopefully sharing will give someone else encouragement / hope ⏎ Please also RT if you think my message / sentiments are valid. Id appreciate the support in many ways.
00:50:50FromGitter<zacharycarter> @gogolxdong sdl2 is backend agnostic so the draw calls would come from sdl2 / opengl
00:54:07FromGitter<gogolxdong> ok, thanks.
00:58:44*owner_ joined #nim
00:59:08*owner_ is now known as Guest74242
00:59:46FromGitter<zacharycarter> Err i meant nuklear is backend agnostic - @gogolxdong - my bad
00:59:54FromGitter<zacharycarter> Not sdl2
01:00:21*Tortice quit (Ping timeout: 252 seconds)
01:00:31FromGitter<zacharycarter> Sdl2 is very oriented towards the OpenGL drawing api
01:00:48*vlad1777d quit (Ping timeout: 272 seconds)
01:03:13*Guest74242 quit (Remote host closed the connection)
01:03:40*Guest74242 joined #nim
01:22:01*Guest74242 left #nim (#nim)
01:22:22*Tortice joined #nim
01:24:51FromGitter<gogolxdong> Is this the way to set font ? ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bb6bd63271506518dbb2929]
01:25:10FromGitter<gogolxdong> got ⏎ A s s e r t i o n f a i l e d ! ⏎ ⏎ P r o g r a m : D : \ n i m n u k l e a r \ d e m o \ s d l _ o p e n g l 3 \ g u a r d i a n . e x e ⏎ F i l e : C : \ U s e r s \ l x d o n \ . n i m b l e \ p k g s \ n i m n u k l e a r - 0 . 1 . 0 \ n i m n u ... [https://gitter.im/nim-lang/Nim?at=5bb6bd766e5a401c2d1afa08]
01:40:03FromGitter<zacharycarter> @gogolxdong - I guess not - I'm guessing you need to invoke some other function which is going to produce a pointer or something to pass to the font_atlas_add - I think a lot of your frustration/ friction with nuklear is the lack of docs behind the lib - which I undrestand because it's composed by a single author / a huge library
01:40:31FromGitter<zacharycarter> I'd suggest looking at how other backends (and really you should be looking at the straight OpenGL / other drawing API backends - not ones that leverage frameworks like SDl2) to see how they're doing things
01:40:49FromGitter<zacharycarter> most likely that's the behavior you need to replicate - so it might be a C function pointer or a pointer to a C struct
01:41:10FromGitter<zacharycarter> but you're going to have to figure out what is needed - and then make sure you're passing the correct consrtuct to the API
01:41:40FromGitter<zacharycarter> it's difficult to reason about what is wrong from a Nim view - because the original API is written in C, and whatever Nim interop code exists - needs to adhere to that interface
01:42:35FromGitter<zacharycarter> I know the example I constructed isn't a full representation of the nuklear API - not to mention it's extremely outdated - but you should be able to find other usages of the library / see how it's being used / how other bindings call it
01:44:30*rnrwashere quit (Remote host closed the connection)
01:56:45*rnrwashere joined #nim
01:58:01*xet7 quit (Remote host closed the connection)
02:04:13shashlickTheLemonMan, ng0: so nimzbar now compiles fine on Windows and Linux
02:05:56*Tortice quit (Remote host closed the connection)
02:06:43shashlick@zacharycarter, @gogolxdong, @AlexMax: are we there yet as far as replacing nuklear-nim with nimnuklear in nimble?
02:07:26*xet7 joined #nim
02:10:03FromGitter<zacharycarter> I has no idea
02:10:17FromGitter<zacharycarter> I've been so out of touch with Nim and game dev since I started trying to care about work
02:10:34FromGitter<zacharycarter> but that care is quickly fading / pretty much non-existant at ths point
02:10:43FromGitter<zacharycarter> so I'm hopefully shortly going to be focusing again more on Nim game dev
02:11:12FromGitter<zacharycarter> and hopefully - if I can pony up - that means more maintenance and usage of these bindings / libs I've authored
02:11:28FromGitter<zacharycarter> and also - hopefully I can help out with their new incarnations / stabilization
02:12:04FromGitter<zacharycarter> but I've been going through a massive amount of personal issues - like I'm already a very unstable individual and I deal with a lot of mental health issues - so I'm not great at managing time and resources
02:12:33FromGitter<zacharycarter> and what little time I can find to concentrate on personal projects - often ends up taking a very sparatic / non-focused form
02:13:18FromGitter<zacharycarter> but I'm trying to work through these things and find a new job / opportunity that allows me to grow in the areas I want to but also allows me to flex my strengths - so I think once I get through this personal hell - I'll be a much more useful contributor to Nim and its ecosystem
02:13:23FromGitter<zacharycarter> especially in regards to game development
02:13:46FromGitter<zacharycarter> until then - I'm pretty much just useful for creating libraries and bindings and not being able to spend much time keeping them up to date / maintaining / improving them
02:14:03FromGitter<zacharycarter> I'm just spread very thin in terms of personal pursuits and not getting fired at work
02:18:09shashlickno worries, we'll see how it goes then
02:24:53*rnrwashere quit (Remote host closed the connection)
02:25:46FromGitter<gogolxdong> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bb6cba9435c2a518e533e94]
02:25:57FromGitter<gogolxdong> https://github.com/vurtun/nuklear/issues/505
02:25:58*rnrwashere joined #nim
02:32:29FromGitter<gogolxdong> change msyh.ttc to other font.ttf works.
02:33:09FromGitter<gogolxdong> but doesn't display the input method. blind typing.
02:37:19*xet7 quit (Remote host closed the connection)
02:45:07*tiorock joined #nim
02:45:07*tiorock quit (Changing host)
02:45:07*tiorock joined #nim
02:45:07*rockcavera is now known as Guest8873
02:45:07*Guest8873 quit (Killed (tolkien.freenode.net (Nickname regained by services)))
02:45:07*tiorock is now known as rockcavera
02:46:46dom96hello guys
02:49:05FromGitter<malkomalko> 👋
02:52:03*rockcavera quit (Ping timeout: 268 seconds)
03:02:57FromGitter<malkomalko> TIL about the `gui` lldb command. Pretty neat as I explore debugging nim applications.
03:07:00*rockcavera joined #nim
03:07:00*rockcavera quit (Changing host)
03:07:00*rockcavera joined #nim
03:07:52*rnrwashere quit (Remote host closed the connection)
03:08:59*rnrwashere joined #nim
03:11:05*rnrwashere quit (Remote host closed the connection)
03:13:06*rockcavera quit (Ping timeout: 252 seconds)
03:14:25*rockcavera joined #nim
03:14:25*rockcavera quit (Changing host)
03:14:25*rockcavera joined #nim
03:16:55*rnrwashere joined #nim
03:19:09*rockcavera quit (Ping timeout: 252 seconds)
03:20:05*rockcavera joined #nim
03:20:05*rockcavera quit (Changing host)
03:20:05*rockcavera joined #nim
03:20:55FromGitter<bung87> what’s the `x[i]` operate for float in c++ ? bit operation?
03:21:28*dddddd quit (Remote host closed the connection)
03:21:51*tiorock joined #nim
03:21:51*rockcavera is now known as Guest3840
03:21:51*tiorock quit (Changing host)
03:21:51*tiorock joined #nim
03:21:51*Guest3840 quit (Killed (barjavel.freenode.net (Nickname regained by services)))
03:21:51*tiorock is now known as rockcavera
03:23:45*rnrwashere quit (Remote host closed the connection)
03:26:22*rockcavera quit (Ping timeout: 246 seconds)
03:29:03*rnrwashere joined #nim
03:31:01*rnrwashere quit (Remote host closed the connection)
03:31:58*endragor joined #nim
03:33:25*rnrwashere joined #nim
03:33:34*yemelBz joined #nim
03:34:37*darithorn quit ()
03:34:40*yemelBz quit (K-Lined)
03:45:15*chemist69 quit (Ping timeout: 252 seconds)
03:45:45FromDiscord_<treeform> I kind of expected a way to generate query string like "?a=1&b=2" from {a: 1, b: 2} some place in nim stdlib, maybe https://nim-lang.org/docs/uri.html, but I can't find it?
03:47:08*chemist69 joined #nim
03:48:15*gangstacat quit (Quit: Ĝis!)
03:49:12*rockcavera joined #nim
03:49:12*rockcavera quit (Changing host)
03:49:12*rockcavera joined #nim
03:53:22*rnrwashere quit (Remote host closed the connection)
03:55:37*rnrwashere joined #nim
03:55:45*rnrwashere quit (Remote host closed the connection)
03:55:47dom96treeform: don't think so. I've got a PR open to go the other way: https://github.com/nim-lang/Nim/pull/7761/files
03:56:02dom96Blocked by Araq :(
03:56:36dom96but it should be easy to fix
03:56:42dom96feel free to do that if you'd like
03:59:04FromGitter<7sDream> Does anyone know which online nim ice/code runner has been upgraded to 0.19.0?
04:01:55FromGitter<malkomalko> anybody use vscode/lldb to debug? I'm trying https://github.com/vadimcn/vscode-lldb but not having much luck
04:05:34dom96Join the echo-debug-master-race :P
04:08:02*rnrwashere joined #nim
04:09:49FromGitter<7sDream> I can’t create a async proc whose name is a nnkSym from macro/template, is this expected behavior? ⏎ Working on a MWE.
04:11:48dom96seems like a bug
04:11:59dom96but yeah, these macro-macro interactions are tricky
04:12:03FromGitter<7sDream> http://ix.io/1ol4/nim
04:12:29*leorize quit (Quit: WeeChat 2.2)
04:13:50FromGitter<7sDream> Seems we can’t get symbol’s final output name in compile time
04:15:07FromGitter<7sDream> line #15 of that MWE is an typo error, please delete it if you want try...
04:17:43*leorize joined #nim
04:23:40*narimiran joined #nim
04:23:50FromGitter<7sDream> Issue #9201 for this bug(?)
04:26:59FromGitter<bung87> `c_memset` and `nimSetMem` doest exported ?
04:29:15*nsf joined #nim
04:42:05*rnrwashere quit (Remote host closed the connection)
04:43:27*gangstacat joined #nim
04:44:11*rnrwashere joined #nim
04:46:18*joshbaptiste quit (Ping timeout: 252 seconds)
04:47:38*joshbaptiste joined #nim
04:52:08*redlegion quit (*.net *.split)
04:52:08*deepend quit (*.net *.split)
04:57:08FromGitter<kaushalmodi> I am able to init a nested table, but unable to figure out how to assign to it
04:57:08FromGitter<kaushalmodi> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bb6ef0182893a2f3ba08128]
04:57:09*redlegion joined #nim
04:57:09*deepend joined #nim
05:03:52leorizeI'm surprised that even worked
05:06:42FromGitter<kaushalmodi> yeah.. as I later find.. the inside should not have initTable
05:07:02FromGitter<kaushalmodi> but still.. unable to assign values to it
05:07:04FromGitter<7sDream> I usually make value type to a table, not key type
05:07:23FromGitter<kaushalmodi> @7sDream I am trying that too.. no success
05:07:33FromGitter<7sDream> https://glot.io/snippets/f5fge4wv6i
05:07:36FromGitter<7sDream> Try this
05:08:05leorizethat shouldn't work :/
05:08:07leorizeweird
05:08:23leorizeinside the those brackets `[]` should only contains typedesc
05:10:09leorizekaushalmodi: well it's not possible to have table key as an another table if you don't have a hash function for it :P
05:10:31FromGitter<bung87> why not use tuple inside?
05:10:45*Tongir joined #nim
05:11:07FromGitter<kaushalmodi> @bung87 I need nested hash
05:11:07shashlickWhy would you use a table as a key
05:11:25shashlickI did a nested table in snip, but as a value
05:11:29FromGitter<kaushalmodi> shashlick: I don't have to.. I was just trying out weird things
05:11:37FromGitter<kaushalmodi> in the end, I need nested hash
05:11:38FromGitter<kaushalmodi> :)
05:11:51shashlickhttps://github.com/genotrance/snip/blob/master/src/snip/compile.nim
05:12:00FromGitter<bung87> what do you mean nested hash ?
05:12:01shashlickSee the variable MODES
05:12:19shashlickDeclared in globals.nim
05:12:43FromGitter<kaushalmodi> @bung87 ⏎ ⏎ https://perlmaven.com/multi-dimensional-hashes
05:12:59FromGitter<kaushalmodi> shashlick: will have a look
05:13:14FromGitter<kaushalmodi> @7sDream I am going through your example too
05:13:36FromGitter<kaushalmodi> trying to understand it, and see if a simpler sugar syntax can be created
05:14:06leorizehere's my take on it: https://ptpb.pw/cSPB
05:15:17FromGitter<bung87> I cant get the point , the data struct you declared can simply using a tuple as key ,string as value.
05:15:31FromGitter<7sDream> (https://files.gitter.im/nim-lang/Nim/7HfE/Xnip2018-10-05_13-14-59.jpg)
05:15:43FromGitter<7sDream> add a new `[]` func to simplify usage
05:17:33FromGitter<kaushalmodi> Thanks shashlick: https://ptpb.pw/hzms/nim
05:17:58FromGitter<kaushalmodi> @7sDream Nice! Let me try that
05:18:10FromGitter<gogolxdong> https://github.com/gogolxdong/guardian
05:18:28FromGitter<7sDream> https://glot.io/snippets/f5fgoyumwx
05:18:41FromGitter<bung87> I dont think it’s productive data structure.
05:18:44*leorize quit (Quit: WeeChat 2.2)
05:21:27shashlick@gogolxdong nice work! Screen shot would be cool
05:22:07FromGitter<kaushalmodi> @bung87 I want to do something like .. if nestT[1]["b"] exists, use that, else use nestT[0]["b"], else nestT[1][""], else nestT[0][""]
05:22:42FromGitter<kaushalmodi> may be I can use tuple as key ..
05:23:05FromGitter<kaushalmodi> though I am not a data structure efficiency expert.. so I'll try out your approach
05:23:05FromGitter<bung87> in this case you can simply usiing tuple key.
05:23:47FromGitter<bung87> when you assign inner table that will assign by copy
05:24:22FromGitter<gogolxdong> (https://files.gitter.im/nim-lang/Nim/sKYj/image.png)
05:26:08FromGitter<gogolxdong> (https://files.gitter.im/nim-lang/Nim/t5Va/image.png)
05:26:52shashlick👍
05:28:58shashlickI've been thinking of writing a chat native client in Nim
05:29:21shashlickFront end for slack, gitter and so on
05:30:13*rnrwashere quit (Remote host closed the connection)
05:30:29Tongirshashlick, Oooh, cool. I banged out some code for hooking up to the Slack RealTime API for v0.18 which would probably still work
05:34:14FromGitter<kaushalmodi> Thanks @bung87, this is as per your suggestion: https://ptpb.pw/0u01/nim
05:34:20FromGitter<kaushalmodi> I think this will work.
05:34:49FromGitter<kaushalmodi> Thanks all: shashlink, @7sDream and @bung87, I learned quite a bit about nested tables :)
05:35:53*styx_ joined #nim
05:35:54shashlickTongir: link?
05:36:44FromGitter<bung87> you’are welcome :)
05:37:27FromGitter<7sDream> 😉
05:40:17*styx_ quit (Remote host closed the connection)
05:40:44FromGitter<7sDream> Does anyone have time to look at #9201, Is there a possible solution?
05:41:54Tongirshashlick, https://github.com/Tangdongle/nim-slack-client
05:42:48TongirSome features are half complete. But the general authorisation and websocket connection to the slack RTM worked fined last time I checked
05:43:05Tongir~1-2 months ago or something
05:43:26shashlickNice work
05:48:21TongirIf you do end up doing it, I'd love to contribute to the slack integration :D
05:49:23*narimiran quit (Quit: Konversation terminated!)
05:50:29*leorize joined #nim
05:53:26*stefanos82 joined #nim
06:06:09FromGitter<kaushalmodi> .. coming back to that tuple-key table example: https://ptpb.pw/0u01/nim
06:06:33FromGitter<kaushalmodi> if I change the int type of orgLevel field in that tuple to Natural, it doesn't compile
06:07:01FromGitter<kaushalmodi> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bb6ff855331811c2e50ff5f]
06:08:20leorize`0.Natural`
06:08:25leorizeThat's what you're missing
06:08:30FromGitter<kaushalmodi> yeah, now I need to do `pnt[(0.Natural, "nim")] = "abc"`
06:09:06FromGitter<kaushalmodi> seems counter intuitive..
06:09:26FromGitter<kaushalmodi> if I have `var foo: Natural`, I can do `foo = 0`
06:09:44FromGitter<bung87> @7sDream hi,what the use case looks like ? seems you declared a proc that you dont have the proc name
06:10:17leorizekaushalmodi: The problem with tuples is that they don't have type information
06:10:29FromGitter<bung87> even you macro work out , I cant see how you get the proc to use
06:11:09leorizewhen you create a tuple like `(0, "string")` it won't infer to the tuple type you created, but instead an anonymous `(int, string)` tuple instead
06:11:21FromGitter<kaushalmodi> leorize: hmm, ok.. so to avoid such verbose code, I should stick to base types in tuples?
06:11:50FromGitter<kaushalmodi> and then some assert later in the code to prevent illegal field vals
06:12:11leorizeor you could write a `[]` proc that takes a `(int, string)` tuple
06:13:00FromGitter<kaushalmodi> but `[]` already accept `(int, string)`
06:13:08FromGitter<kaushalmodi> s/accept/accepts
06:13:45FromGitter<kaushalmodi> hmm, I think I understood.. let me try that
06:18:33FromGitter<7sDream> @bung87 I
06:18:44FromGitter<7sDream> I’m write a DSL for a class
06:19:36FromGitter<bung87> yeah I can see, you want a anoymous async function like in js
06:19:43FromGitter<7sDream> a class has method like `addOnMessageListener`
06:20:09FromGitter<7sDream> so I write a macro, this marco create a proc and set it as a listener
06:20:23FromGitter<7sDream> so, every time I use this macro, the name should be different
06:20:47FromGitter<bung87> but when you call `genSym` you did not get the generated symbol.
06:21:11FromGitter<bung87> then how could you use it ?
06:21:35FromGitter<7sDream> I reuse the nnkSym node when pass it the the `addOnMessageListener` function
06:22:28FromGitter<7sDream> So it will use same name in func definition and as a proc variable passed into addOnMessageListener
06:22:29FromGitter<bung87> you cant , you have no the generated symbol of proc
06:23:23FromGitter<7sDream> emm, wait for a minute, I’m writing a example(without async)
06:29:21*nimrodfrfr joined #nim
06:43:15*Tongir quit (Remote host closed the connection)
06:43:36FromGitter<7sDream> @bung87 https://glot.io/snippets/f5fj19hfbr
06:44:09FromGitter<7sDream> uncomment line 60 to see the code generated by macro
06:50:04FromGitter<bung87> yeah runs without problem
06:52:28FromGitter<gogolxdong> @bung87 and @7sDream join our Nim in china community and talk in chinese.
06:53:35FromGitter<gogolxdong> https://discord.gg/pEDQeK
06:54:24FromGitter<7sDream> I need install discord for my new laptop first...
06:55:26FromGitter<gogolxdong> It's a piece of cake for you.
06:57:57FromGitter<gogolxdong> Is it out of blocked ?
06:59:07FromGitter<7sDream> no, I think. But my shadowsocks is running now(and never stopped)
07:03:52FromGitter<bung87> `listenerFuncDef.addPragma ident("async”)` you can try do like this
07:04:26FromGitter<bung87> I cant follow the whole codes, just have some thoughts
07:05:11FromGitter<bung87> first you code gen listener is inline proc, seems closure proc more suitable.
07:05:54FromGitter<bung87> second the code gen,set listenner before the listenner proc declared.
07:06:31FromGitter<7sDream> closure is easy to add, just add it in template listenerFunc
07:07:18FromGitter<7sDream> but add {.async.} to it will case the #9201 issue
07:07:39FromGitter<7sDream> not sure about use.addPragma, I’ll try it
07:09:54FromGitter<7sDream> But I think it won’t change anything, because the error is happened when compile the generated code, whatever the {.async.} is generate by template or macro
07:09:57FromGitter<bung87> seems more like it calls before it generated.
07:10:38FromGitter<bung87> maybe needs change the macro parse order.
07:20:52*Tongir joined #nim
07:23:29*jjido joined #nim
07:31:51*jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
07:48:35*kapil____ joined #nim
07:55:14*krux02 joined #nim
07:56:16*PMunch joined #nim
07:57:19*Vladar joined #nim
07:59:29FromGitter<ephja> is it possible to statically link against a library with the required lib files but without C files declaring the symbols?
08:00:40leorizewhat do you mean?
08:01:50FromGitter<ephja> it should be possible but I dunno maybe it's difficult to implement. of course, the symbols would have to be declared in Nim then
08:02:11FromGitter<ephja> leorize: the libs that support static linking against a C lib currently require both lib files and the C sources that define the symbols
08:02:32FromGitter<ephja> which seems like duplicated effort to me
08:03:15leorizeI still don't get it. Could you show me an example?
08:03:18*gmpreussner quit (Ping timeout: 252 seconds)
08:03:44*gmpreussner joined #nim
08:07:25FromGitter<ephja> leorize: like this https://github.com/ephja/nim-glfw/blob/master/glfw/wrapper.nim#L14 but with only static libs (.a/.lib, ...)
08:08:33*MBS joined #nim
08:09:46leorizeI think just need to `{.passL: "/path/to/lib.a".}`, no C sources necessary
08:12:25*MBS quit (Remote host closed the connection)
08:33:16FromGitter<bung87> @7sDream after debuging , the `getName` get nnkSym node, you can try add ` of nnkSym: ⏎ ⏎ ```return node.strVal` in``` [https://gitter.im/nim-lang/Nim?at=5bb721cc3844923661e9587b]
08:34:07leorizebung87: https://github.com/nim-lang/Nim/pull/9204
08:34:15leorizeI've already made a PR for it ;)
08:34:19FromGitter<bung87> and will pass this raise another issue
08:35:16FromGitter<7sDream> I notice the reason，but not sure get the prefix of sym node will make other part of asyncmacro works
08:37:21FromGitter<bung87> @leorize cool
08:50:40FromGitter<tim-st> if that can still be changed without problems, I think the `nimcache` should be named `.nimcache` (prefix dot)
08:51:08FromGitter<tim-st> chossenim and nimble also have the dot
08:51:58FromGitter<ephja> leorize: good to know that it's supposed to work. I've had issues with dynamic linking on windows before, and now static linking fails. who knows why it never seems to work. well, I'm gonna try it with Linux
08:52:01leorizetim-st: nimcache has been moved to ~/.cache/nim so...
08:52:12FromGitter<tim-st> not on windows
08:53:26leorizedotfiles doesn't mean anything on windows...
08:53:52FromGitter<tim-st> it means sort order
08:54:28leorizeis it that important?
08:55:03FromGitter<tim-st> sure, do you want the three nim folder next to each other or do you want to search in the directory?
08:55:21leorizethe only purpose of dotfile is to hide stuff. The fact that choosenim and nimble use dotfile on windows can be considered a bug
08:56:55FromGitter<tim-st> VirtualBox, vscode, texlive, ssh, jetbrains, rust, eclipse do it too with prefix dot on windows
09:00:14*leorize quit (Ping timeout: 272 seconds)
09:08:27*floppydh joined #nim
09:22:27FromGitter<mratsim> @bung87 import system/ansi_c -> https://github.com/nim-lang/Nim/blob/devel/lib/system/ansi_c.nim
09:23:27FromGitter<mratsim> and you can access c_memset
09:23:59FromGitter<mratsim> or just use reset
09:24:09FromGitter<mratsim> https://nim-lang.org/docs/system.html#reset%2CT
09:24:39FromGitter<ephja> doesn't it have to be included since it isn't not exported?
09:26:07FromGitter<mratsim> it’s not exported by system.nim but the proc live in another nim files with a * marker
09:55:11*zakora joined #nim
10:03:26*seni joined #nim
10:07:40*kapil____ quit (Quit: Connection closed for inactivity)
10:13:05*Vladar quit (Remote host closed the connection)
10:21:55FromGitter<bung87> I get confused ,it is not exported
10:32:28FromGitter<bung87> does all system things implicitly exported? the vscode hints `undeclared identifier: 'c_memset’` even I `import system/ansi_c`
10:36:00*elrood joined #nim
10:37:58PMunchbung87, c_memset is not exported from the system/ansi_c module it looks like. Try to `include system/ansi_c` instead of importing it
10:39:38FromGitter<bung87> wow so any module not exported I can using include to get to use?
10:40:31PMunchNote that in Nim we have two different ways to interact with modules. Import is the normal one, it reads the file and allows you to use the exported modules without executing anything in the module. Include on the other hand simply copies the module into your code, this means you can now use all procedures (and the exported ones will be visible to modules importing your module) and any code in the main scope will get executed in your module.
10:40:44PMunchbung87, well yes in a way
10:41:10PMunchBut note that that's not what it's intended purpose is
10:41:33PMunchIf someone has designed a library with non-exported procedures it's probably for a reason
10:42:26FromGitter<bung87> I also trying to using Nim/lib/system/memory.nim also not exported..
10:42:36PMunchinclude is more intended to be able to split a logical module into multiple files, and to allow certain kinds of things to work (like having a global state variable)
10:43:35PMunchbung87, all of these modules are included into system
10:43:47PMunchhttps://github.com/nim-lang/Nim/blob/master/lib/system.nim#L2981
10:44:41PMunchHmm, but since that's imported into your own module you won't see the symbols
10:45:08PMunchNot sure why they aren't exported tbh
10:45:36*leorize joined #nim
10:45:44FromGitter<bung87> hmm so they just not exported, actually be used.
10:46:05*vlad1777d joined #nim
10:54:18*dddddd joined #nim
10:58:11FromGitter<bung87> ```const uint8_t* codes ⏎ codes[i]``` ⏎ ⏎ `[i]` is it a bit index operatoration, how can I do it in nim ? [https://gitter.im/nim-lang/Nim?at=5bb743c35331811c2e52b22b]
10:58:31FromGitter<bung87> operation
11:02:29FromGitter<bung87> found `let d = castptr UncheckedArray[byte (dest)` in /system/memory.nim
11:09:10*nimrodfrfr quit (Quit: Page closed)
11:21:44*yoplaid left #nim (#nim)
11:27:30*rnrwashere joined #nim
11:31:57*rnrwashere quit (Ping timeout: 252 seconds)
11:55:38*Vladar joined #nim
12:05:46*xace quit (Remote host closed the connection)
12:14:50*Vladar quit (Remote host closed the connection)
12:15:27*Vladar joined #nim
12:37:15*seni quit (Quit: Leaving)
12:46:24*matthewprengerKo joined #nim
12:51:28*matthewprengerKo quit (Remote host closed the connection)
12:53:06FromGitter<krux02> @bung87 no in ``codes[i]`` the ``[i]`` is a byte index operation.
12:54:14FromGitter<krux02> but generally you found a good example, you can cast any pointer to ``ptr UncheckoudArray[byte]`` to interpret the content as bytes
12:59:45*halirc joined #nim
13:01:40FromGitter<bung87> @krux02 thanks! if I need define `[]` proc what’s the self type should be?
13:02:43FromGitter<bung87> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bb760f2ae7be94016062c9d]
13:03:04FromGitter<bung87> currently I have.
13:03:17FromGitter<krux02> real?
13:03:30FromGitter<krux02> what do you mean by that?
13:03:32FromGitter<bung87> yeah `type ⏎ real* = cfloat`
13:03:33FromGitter<krux02> there is no real type in nim
13:03:53FromGitter<krux02> why don't you use float32/float64
13:04:46halircHello. What is the nim way to check if an element is contained in a sequence? I could not find something like a contains/includes proc, do I just iterate through the sequence?
13:04:53FromGitter<krux02> I think the name ``real`` is very confusing, because it is a float, and there is not a single floating point number that isn't either a rational number or neither rational nor real
13:05:00FromGitter<bung87> well , wraped cpp source but I want port it to nim
13:05:20FromGitter<krux02> ok
13:05:54FromGitter<bung87> in the wraped version it can simplely conversation to cpp
13:06:00FromGitter<krux02> but still you shouldn't do that
13:06:15FromGitter<krux02> when real is just an alias for cfloat, you should not overload the [] operator for that type
13:06:24FromGitter<krux02> it can cause a lot of trouble
13:06:26FromGitter<bung87> basicly it just follow as the cpp lib as is.
13:06:47FromGitter<krux02> are you sure?
13:06:53FromGitter<krux02> can you show me the source?
13:07:27FromGitter<krux02> It really is a bad coding style to create an operator for such an obscure operation.
13:07:36FromGitter<krux02> it's better to give it a proper name
13:07:41FromGitter<bung87> https://github.com/facebookresearch/fastText/blob/master/src/real.h
13:09:17FromGitter<krux02> something like ``proc readByte*(self: float64, index:int):byte = castptr UncheckedArray[byte (self.unsafeAddr)``
13:09:35FromGitter<krux02> and use unsafeAddr to get the addr of the argument
13:09:59FromGitter<krux02> only with ``unsafeAddr`` you get the addr of a non var argument.
13:10:09FromGitter<krux02> and the address is what you should cast
13:10:25FromGitter<krux02> never cast a floating point value to a pointer.
13:11:10FromGitter<krux02> yea, just ignore that it is called real. It's a bad name, and it's an optional alias anyway.
13:11:20FromGitter<krux02> a very bad naming convention.
13:11:28FromGitter<bung87> what if define `[]=`
13:12:34FromGitter<krux02> ``proc writeByte*(self: var float64, offset:int, value: byte): void = (castptr UncheckedArray[byte (self.addr))[offset] = value``
13:13:12FromGitter<krux02> halirc: iterating is definitively what you can do, it makes it very visual that the operation has linear runtime
13:13:25FromGitter<krux02> otherwise check if sequtils has something for you.
13:14:28FromGitter<krux02> @bung87 calling a floating point value "real" is not just misleading, it is technically wrong.
13:14:41FromGitter<krux02> I know some programming languages do that
13:15:32FromGitter<bung87> good example,interesting codes.
13:15:34FromGitter<krux02> and aparently there are people who want that (wrong) naming convention in c++ as well. I see myself as a guard to prevent those naming conventions to leak into Nim :P
13:16:15FromGitter<mratsim> So you’re areal guardian?
13:16:32FromGitter<krux02> yea
13:16:37FromGitter<krux02> I am a real guardian :P
13:17:04FromGitter<bung87> hmmm, since I just want port a lib ,I dont want to judge the author’s codes.
13:17:21FromGitter<krux02> yea, but that part is really just an alias for a type
13:17:30FromGitter<krux02> there is no sematic to that type alias
13:17:34FromGitter<krux02> you can just ignore it.
13:18:22FromGitter<krux02> it's not like that type is in any way different from a float
13:18:52FromGitter<krux02> (when I say float, I don't mean float from C, I mean floating point number in general, including float32 and float64)
13:19:08FromGitter<krux02> (and float16 and float128)
13:21:10FromGitter<bung87> maybe after I fully port the lib ,I could change it float32
13:21:26FromGitter<krux02> yea I just wanted to recommend it
13:22:58FromGitter<bung87> thanks, you make the codes no red underline in vscode:)
13:25:06*FlummiRJ joined #nim
13:25:46shashlick@bung87: are you not using the wrapper?
13:26:23FromGitter<bung87> no, I ‘m trying to port it to pure nim, almost 5 files now.
13:29:12FromGitter<bung87> the wrapper version I still can not implements the string representation of vector, have no idea about vector<T> and vector<const T> conversation
13:30:27shashlickOh that's unfortunate
13:31:02shashlickCan you share what you did so far? I'll check later today
13:31:39FromGitter<bung87> wrapper version? pure nim version?
13:31:50shashlickWrapper
13:31:56*FlummiRJ quit (Ping timeout: 272 seconds)
13:32:57FromGitter<bung87> https://github.com/bung87/cpp_nim_interop/blob/master/src/cpp_nim_interop/std/vector.nim
13:34:55shashlickOk basically interop between seq and vector?
13:35:01FromGitter<malkomalko> I'm reading Nim in Action and noticed that I can now access the value from `json` by just using `[]` and not needing to add a `str` or `getStr`. Did that change in a recent version of Nim?
13:35:01shashlickI'll take a look later today
13:35:12FromGitter<bung87> currently It has memory leak,when you call `toSeq(a)` twice you will see.
13:36:36FromGitter<zacharycarter> @malkomalko probably
13:36:37FromGitter<bung87> yeah, when it get ready can simplely replace the fastest proc arg with these friendly types.
13:36:57FromGitter<zacharycarter> I think you could figure that out for sure - if you looked at the json module's history on github
13:37:28FromGitter<zacharycarter> @malkomalko - https://github.com/nim-lang/Nim/commits/devel/lib/pure/json.nim
13:37:28FromGitter<malkomalko> Good call, I'm going to go do that
13:39:23halirckrux02, thank you. This is what I came up with, does that seem ok to you? https://pastebin.com/xVvjDqhM
13:40:12*PMunch quit (Remote host closed the connection)
13:40:23krux02that's perfect
13:40:51*TheLemonMan joined #nim
13:41:09TheLemonManhalirc, we have the `in` operator (or `contains()`) for that purpose
13:41:51krux02http://ix.io/1ooM
13:42:06krux02TheLemonMan, internally it does the same thing
13:42:35TheLemonManyeah but it's much more compact
13:42:57krux02it's beginners code
13:42:58*vlad1777d quit (Ping timeout: 245 seconds)
13:43:03krux02it can be less compact
13:44:06krux02in the beginning it's important to learn just a few concepts that can be applied very flexibly. Later then you can extend your vocabulary to non essential functions
13:44:30krux02when you have a bigger project then you start compressing your code by using more expressive terms
13:44:43krux02but for an addIfNotExists, the loop is perfectly fine
13:46:36halircOoh, in or contains was what I was looking for. The documentation did not really make it easy to find. I see now that for Sets it is mentioned as an operator in the tutorial but not for sequences specifically.
13:47:28halircI thought it was strange one had to write this ;)
13:47:34TheLemonMan`if x notin y: y.add x`
13:48:00halircnice
13:48:07TheLemonMankrux02, you're assuming someone that's just learning Nim doesnt know jack shit about programming in general
14:01:01halircI do have some programming experience but I can see how my question and the simple example can lead to thinking otherwise. I'm new to nim though and not using it often.
14:11:06*couven92 quit (Quit: Client disconnecting)
14:24:22*Tortice joined #nim
14:26:18*endragor quit (Remote host closed the connection)
14:26:46FromGitter<arnetheduck> btw @Araq, if you find yourself in a refactoring mood, it would be nice if you could come up with something more structured for the type nesting resolution - `skipTypes` and friends - it's been the source of innumerable bugs in the compiler, as it's very easy to forget to peel away some special generic or indirected type, leading to issues when the system is stressed. this is something I've struggled with as well in
14:26:46FromGitter... `nlvm` - it's hard to know exactly which types should be skipped in which situation, leading to time-consuming analysis every time - perhaps there's something type-based that could be done to limit the generality/flexibility in exchange for fewer ways to get it wrong
14:35:59*narimiran joined #nim
14:42:30TheLemonMangah, the zip library ships with a copy of libzip 0.8
14:42:43TheLemonManthat's...a 11 year old release
14:44:51*endragor joined #nim
14:46:53*endragor quit (Remote host closed the connection)
14:47:29*endragor joined #nim
14:47:59krux02arnetheduck: Araq is on vacation
14:48:28FromGitter<arnetheduck> @krux02 oh bummer. thanks!
14:48:37*krav_Yu joined #nim
14:48:44krux02but I am here to listen to your stories
14:49:29FromGitter<xmonader> In the httpclient how do i encode query string with request proc?
14:49:37*krav_Yu quit (K-Lined)
14:50:47*nsf quit (Quit: WeeChat 2.2)
14:52:04*floppydh quit (Quit: WeeChat 2.2)
14:52:30*rockcavera quit (Ping timeout: 252 seconds)
14:54:16*tribly quit (Ping timeout: 272 seconds)
14:54:27halircI'm back with another sequence question. Now I'm looking for flatten, is there something like that?
14:54:48*tribly joined #nim
14:55:29*natrys joined #nim
14:57:44krux02halirc, I would do it with a loop
14:57:57krux02but there is an add proc that appends a collection to a colleciton
14:58:02krux02so you don't need nesting.
14:58:11krux02but first check the index
14:58:53narimiranhalirc: https://narimiran.github.io/itertools/#chain.i,varargs[seq[T]] :)
14:59:04krux02halirc, there is someone who implemented `flatten` for ``Option[Option[T]]`` not really what you were looking for
14:59:12krux02that is the state of functional programming in Nim
14:59:44TheLemonMannarimiran, urlencode your urls please
14:59:44krux02there were people contributing to Nim who wanted functional programming and all contributed their way to a different subset of the collecitons
15:00:21narimiranTheLemonMan: this is copy-paste from my browser, as it is. that is what nim-doc creates
15:00:26krux02so I don't recommend to use any of the functional programming functions from the standard library, they are not thought through to the end.
15:01:01FromDiscord_<demotomohiro> Nim 0.19.0 is available on this online compiler:
15:01:02FromDiscord_<demotomohiro> https://wandbox.org/
15:01:55TheLemonMannarimiran, it's what it used to create :) backport this ( https://github.com/nim-lang/Nim/commit/de8cb0a1cf06d091dbe586fed85fb1cf904a33ac#diff-302297dd1a064075c52c27153e5a6b3a ) and you're good to go
15:02:38narimiranso `choosenim update devel` should do the trick?
15:02:43*natrys quit (Quit: natrys)
15:03:18TheLemonManI think so
15:03:31narimiranok, let's see in couple of minutes....
15:04:27halircOk,thanks. I'll go with a loop or see if I can avoid the flatten altogether.
15:05:00*endragor quit (Remote host closed the connection)
15:05:32halircHow do I make nim code run here in the chat? I saw that a while ago, something with eval.
15:08:12Tortice!eval echo ""
15:08:14NimBot<no output>
15:09:15FromGitter<brentp> is there a simple way that I'm missing to turn an `HSlice` into a `seq[int]` ?
15:09:26halirc!eval import sequtils; concat(@["TEST?"], @[@["a", "b"], @["c"]])
15:09:27NimBotCompile failed: in.nim(1, 24) Error: expression 'concat(@[@["a", "b"], @["c"]])' is of type 'seq[string]' and has to be discarded
15:10:09halircin nim 0.18 that works and I just get @["a", "b", "c"]
15:10:48TheLemonManbrentp, countup(slice.a, slice.b).toSeq ?
15:11:07Tortice!eval echo NimMinor
15:11:10NimBot18
15:12:02TheLemonMan!eval let x = 1..10; from sequtils import toSeq; echo countup(x.a, x.b).toSeq()
15:12:04NimBotCompile failed: in.nim(1, 56) Error: attempting to call undeclared routine: 'countup'
15:12:27TheLemonMan!eval let x = 1..10; from sequtils import toSeq; echo toSeq(countup(x.a, x.b))
15:12:30NimBot@[1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
15:12:41TheLemonManbleh, UFCS my ass
15:13:40TheLemonManmind you that the H in HSlice means Heterogeneous
15:14:48FromGitter<brentp> @TheLemonMan thanks!
15:14:49narimiranTheLemonMan: freshly updated devel again gives infinite loop when i try to create docs ;)
15:15:09halircah, I see the difference
15:15:22halirc!eval import sequtils; echo concat(@["why is this gone?"], @[@["a", "b"], @["c"]])
15:15:25NimBot@["a", "b", "c"]
15:15:31krux02TheLemonMan, technically toSeq is not a function, so UFCS does not apply :P
15:15:34narimiran(it says that it is nim v0.19.0, is that correct? shouldn't it be 0.19.1?)
15:15:34TheLemonMannarimiran, give me a minimal test case and I'll check it out for you
15:15:48TheLemonMankrux02, touché heh
15:15:53halircthats what I did in my code and I didn't get why that happened
15:16:09narimirankrux02: well, `toSeq` fits so naturally to be behind its argument :'(
15:16:25halircwith the echo there is no error but I think I understand now
15:16:50TheLemonMannarimiran, how can I reproduce the infinite loop?
15:17:13narimiranTheLemonMan: first, do you see 0.19.1 when you're running devel?
15:17:35TheLemonManyep
15:17:46narimiranthen something is wrong on my end
15:18:01TheLemonManoh, I meant no
15:18:54narimiranand to reproduce it, here is my file: https://raw.githubusercontent.com/narimiran/itertools/master/src/itertools.nim
15:19:59narimiranjust run `nim doc thatfile.nim` and there is an infinite loop
15:23:30TheLemonManworks fine here
15:26:05*Lavos joined #nim
15:28:31*drewmuttff joined #nim
15:28:48*drewmuttff quit (K-Lined)
15:28:58narimirangod dammit :P
15:30:27TheLemonManyou're not trying hard enough!
15:33:38narimiranthat is also true :D
15:47:33*flyx quit (Quit: ZNC - http://znc.in)
15:48:08*flyx joined #nim
15:54:55*rnrwashere joined #nim
15:55:11TheLemonManhm, if RST code-blocks were runnable would the ongoing forum-fight settle down?
15:56:23TheLemonManbecause getting `nim doc` to support the :test: annotation (like rst2html and rst2tex do) is trivial
16:02:16narimiranyeah, that might help
16:02:48narimiranbtw, what about just putting `runnableExamples` inside of a ## block, so there is no highlighting?
16:03:45narimiranwell, i guess if we would do that, we can also just replace "runnableExamples" with ".. code-block:: test" or whatever is the syntax
16:04:02TheLemonManthat won't work, runnableExamples is a normal `proc`
16:06:36narimiranbut if you can make rst code-block to serve both as documentation and test, then we don't need runnableExamples and we can use "one true way" to write examples
16:07:06*Lavos quit (Quit: Leaving)
16:07:18*Cloudflarecb joined #nim
16:08:47*Tortice quit (Ping timeout: 268 seconds)
16:10:53*Cloudflarecb quit (Killed (Sigyn (Spam is off topic on freenode.)))
16:15:13FromGitter<arnetheduck> @TheLemonMan totally - and all the pieces needed for making code blocks runnable from documentation are there - it's possible to extract them from the comment, it's possible to run a compiler instance over them etc etc - it's not a question of technical feasibility - it's a matter of deciding what the most beautiful approach is
16:16:42TheLemonManarnetheduck, no doubt that's the best one, runnableExamples sticks out like a sore thumb
16:17:59FromGitter<arnetheduck> well, arguably and clearly there is doubt - someone went ahead and implemented `runnableExamples` and there are proponents for it that would prefer to see it improved and spread out instead of axed
16:20:39TheLemonManmaybe because right now you can't run RST code blocks? if you don't have any other option `runnableExamples` is better than nothing
16:21:52FromGitter<arnetheduck> well, some are arguing that running them is not all - without fixing their editors, they want syntax highlighting and code completion when writing the examples, for example
16:25:02TheLemonManyou can also write your own examples somewhere else as .nim files and ask the RST parser to include them
16:25:32TheLemonManit's not you don't have enough options
16:25:32*endragor joined #nim
16:28:52FromGitter<arnetheduck> well, there's a separate post as well where the love for rST itself seems limited - not sure how many of those advanced features of rst you want to rely on, nice as they might seem.. also, include files are complicated - out of sight, out of mind and all that
16:29:57TheLemonManthe RST ship has sailed a long ago, didn't it?
16:30:16TheLemonManif your included files are regularly compiled they're out of sight/mind but working :)
16:30:46FromGitter<mratsim> we can build a markdown hoverboard
16:31:58*endragor quit (Remote host closed the connection)
16:33:26TheLemonManfucking choosenim & travis, I had never seen a 525 error before
16:33:53FromGitter<arnetheduck> the nice thing about markdown is that it will degrade gracefully - even when parsing rst, the results will at least be legible.. also, it's not like nim has a large corpus of well-written source-code-level documentation. the manual, being fully rST, can be autoconverted.
16:39:10*narimiran quit (Quit: Konversation terminated!)
16:40:02*rnrwashere quit (Remote host closed the connection)
16:40:51*rnrwashere joined #nim
16:41:23*endragor joined #nim
16:44:09*rnrwashere quit (Read error: Connection reset by peer)
16:44:24FromGitter<xmonader> while having flu sounded interesting to create a wit.ai client https://github.com/xmonader/witai-nim
16:45:44TheLemonManyou've also committed the binary file :)
16:47:04*endragor quit (Remote host closed the connection)
16:49:07*rnrwashere joined #nim
16:50:36*rnrwashe_ joined #nim
16:53:34*rnrwashere quit (Ping timeout: 240 seconds)
16:53:38FromGitter<xmonader> i thought i ammended that should be removed now :)
16:55:31*Tortice joined #nim
16:56:19*endragor joined #nim
17:05:41*endragor quit (Remote host closed the connection)
17:06:14*rockcavera joined #nim
17:08:20FromGitter<7sDream> (https://files.gitter.im/nim-lang/Nim/WsrX/unknown.png)
17:09:25TheLemonManI don't know what I'm staring at but looks nice
17:09:26FromGitter<7sDream> My new work thoose days, a danmaku client for douyu.com(live streaming website live twitch in China)
17:09:29*darithorn joined #nim
17:10:11TheLemonMannoice
17:10:46FromGitter<7sDream> Nim is so good at creating DSL, I really enjoy it 😄
17:13:00*Tortice quit (Remote host closed the connection)
17:13:08*Tortice joined #nim
17:13:42*Tortice quit (Remote host closed the connection)
17:14:17halircSorry I have another beginner question. I would like to use something like join on a sequence of strings. Like @["a", "b", "c"].join(" ") and as a result get "a b c".
17:15:54TheLemonMando it!
17:16:38*Trustable joined #nim
17:17:03leorizehalirc: the `join()` proc in strutils should do what you need
17:17:17leorize!eval import strutils; echo @["a", "b", "c"].join(" ")
17:17:19FromGitter<arnetheduck> on DSL's and special languages, one of the nim devs at work posted this: https://erikbern.com/2018/08/30/i-dont-want-to-learn-your-garbage-query-language.html :)
17:17:19NimBota b c
17:18:03halircHaha yeah it works. My bad. I read the bit about sequence can be passed to an openarray parameter and I found strutils join but it turns out I forgot to import strutils. It just works now :)
17:21:06*rnrwashe_ quit (Remote host closed the connection)
17:21:14*TheLemonMan quit (Quit: "It's now safe to turn off your computer.")
17:25:38*rnrwashere joined #nim
17:26:56*rnrwashere quit (Remote host closed the connection)
17:27:31*rnrwashere joined #nim
17:31:38*tay_ joined #nim
17:36:50FromGitter<7sDream> Yes, DSL is always controversial, I just enjoy the process I create it!
17:37:09*ghost64 quit (Quit: See you!)
17:37:19*rnrwashe_ joined #nim
17:37:39*tay__ joined #nim
17:38:49*tay_ quit (Ping timeout: 244 seconds)
17:39:35*tay__ quit (Client Quit)
17:40:35*halirc quit (Quit: Leaving)
17:40:37*rnrwashere quit (Ping timeout: 250 seconds)
17:43:03*ghost64 joined #nim
17:46:20*xet7 joined #nim
17:51:28*shadowbane quit (Remote host closed the connection)
17:52:32*shadowbane joined #nim
17:54:12*zakora quit (Quit: WeeChat 2.2)
17:54:28*rockcavera quit (Read error: Connection reset by peer)
17:55:06*tdc quit (Ping timeout: 252 seconds)
17:56:28FromDiscord_<demotomohiro> Maybe, making DSL similar to existing language is a good idea, but making completely new DSL is a bad idea
17:57:50*vivus joined #nim
17:59:17vivuswhat is the best way to compile my program for production use?
18:01:10*rockcavera joined #nim
18:01:12*Delta_558 joined #nim
18:02:54leorizenim c -d:release <program>.nim
18:03:30*Delta_558 quit (Remote host closed the connection)
18:03:43*rnrwashe_ quit (Remote host closed the connection)
18:03:56krux02my two cents to the runable examples, I currently like the idea to have a template (for example in system) that expands to nothing.
18:04:07krux02this template is called ``example``
18:04:38krux02then you can write your example in an ``example:`` block
18:04:59FromGitter<arnetheduck> @krux02: you mean like https://github.com/nim-lang/Nim/blob/269b957e2968345a4c9a5037ab97ab726135a921/lib/system.nim#L4206? that's the one I want to deprecate / remove :)
18:05:00krux02you get syntax highlighting, code completion, etc
18:05:28krux02what is the problem with it? I never used it.
18:05:29vivusthis is off-topic, but for anybody using linux, how did you export your nim executable to make it run globally?
18:05:45krux02vivus: put it on the path?
18:05:56vivuskrux02: yeah, but how will that look?
18:06:23leorizeexport PATH="/path/to/folder/with/compiler:$PATH"
18:06:23vivusexport PATH=/home/myuser/myapp/bin:$PATH ?
18:06:43vivusso the last folder must be where the executable is?
18:06:47krux02vivus, there is ``nimble install`` and that puts the binary in ~/.nimble/bin I think
18:06:52krux02put than on the path
18:07:14krux02or do you mean the nim compiler?
18:07:30krux02for the compiler I did it like leorize
18:07:41vivusthat's what I was trying to replicate. I already have the ./.nimble/bin as a path in .profile . I wanted to add my custom program there also
18:07:51leorizecopy your prog there :P
18:08:04leorizejust make sure it's executable, and it's in PATH
18:08:16krux02vivus, did you do ``nimble install`` in your program directory?
18:08:22vivusis it better to copy the program there or just put my own custom path? cause I am still editing the program but want it globally accessible
18:08:38leorizesymlink also works
18:08:41vivuskrux02: no, this program is not a nimble package
18:08:48krux02vivus, is the binary in ~/.nimble/bin (not ./.nimble/bin as in your comment that is important)
18:08:49leorizejust symlink it to a place in PATH
18:09:20vivuskrux02: no, the binary is in: /ubuntu/myuser/myapp/myprogram
18:09:31voice_ftpWhy there are both `gorge` and `staticExec` in standard library if they do the same thing?
18:09:46leorizevivus: ln -s /path/to/app /path/to/somewhere/in/path
18:09:46krux02leorize, the better way really is to use ``nimble install`` and put the nimble bin directory on the path, then also all packages that distribute binary files work automatically
18:10:02leorizekrux02: only when you're dealing with nimble package
18:10:06*Jesin quit (Ping timeout: 264 seconds)
18:10:18leorizevivus is dealing with their own app, which is not using nimble atm
18:10:30*tdc joined #nim
18:10:32krux02vivus, that is a very non unix style path
18:10:34krux02are you sure?
18:11:04vivusI plan on making it something like: ~/.myapp/ once I release it
18:11:24vivusso it installs on Linux like other software
18:12:01leorizevoice_ftp: ask Araq
18:12:11FromGitter<arnetheduck> @krux02 https://forum.nim-lang.org/t/4279
18:12:21krux02vivus, I am sorry but this path "/ubuntu/myuser/myapp/myprogram" looks like there is a lot of stuff done in a very messy way, because this is a path convention I never ever heared about
18:12:59vivuskrux02: this is just for local testing :) see my message above, the final path can be wherever the user wants it to be
18:13:02krux02either use nimble, and then nimble install and put the nimble bin directory on the path, with is a sandard way, or build locally, and put the local directory on the path
18:13:08krux02anything else you are on your own.
18:13:40leorizekrux02: I'm pretty sure vivus know what they are doing
18:14:00*wildlander joined #nim
18:16:27*rockcavera quit (Ping timeout: 252 seconds)
18:17:19krux02I am not so sure about that. Putting things on the path is annoying, but it is kind of a basic linux/unix thing to do
18:18:07krux02vivus: but sorry for that comment
18:18:43vivuskrux02: no need to apologize friend, helping me understand how to add a path was the end goal. educating me on the unix philosophy was a bonus :)
18:20:00krux02vivus, I don't really understand what you are planning to do, but saying "the final path can be wherever the user wants it to be" is usually much worse than just putting it at right path with no configuaration possible.
18:20:22*tzui joined #nim
18:20:50*tzui left #nim (#nim)
18:20:52*rockcavera joined #nim
18:20:52*rockcavera quit (Changing host)
18:20:52*rockcavera joined #nim
18:21:01vivusyeah its a bit difficult to explain as its possible we are not on the same page concerning where the executable should reside
18:22:23krux02I can understand if you don't like nimble, I don't like it either, I think it gets often more in the way than it actually solves a problem, but when it is about distributing a binary that is built with the Nim compiler, it is still the tool to go.
18:22:36krux02I don't think there is any viable alternative.
18:22:50krux02except maybe the package manager of the distribution.
18:23:18vivuskrux02: but nim binaries don't need nimble? a lot of them can execute as self-contained programs
18:23:53krux02vivus, well that is actually the only case nimble can handle reliably.
18:24:21krux02nimble isn't very good at tracking dependencies.
18:26:23krux02nimble can do the job of putting the binary in the right folder
18:27:04krux02and people who want to have the binary in a custom folder, they know what they are doing, believe me, they can get the binary in that folder, you don't need a config option for that.
18:27:19krux02the config is just confusing the less experienced
18:27:30krux02or might get in the way differently
18:30:03elroodthat's a fairly nice way of pronouncing the opinion hat people usually don't know what they're doing and shouldn't be given the opportunity to make choices ;)
18:31:06elroodwhen did we stop pretending everyone around us wasn't stupid?
18:35:33*rockcavera quit (Ping timeout: 268 seconds)
18:41:46*rockcavera joined #nim
18:41:46*rockcavera quit (Changing host)
18:41:46*rockcavera joined #nim
18:42:53FromGitter<alehander42> @arnetheduck i've also thought about query syntax (e.g. thought of making a linq-like dsl for nim), I kinda agree that having custom query syntaxes suck, but what's the alternative
18:43:11*rnrwashere joined #nim
18:43:23FromGitter<alehander42> sql-s also have a loot of dialects
18:44:00FromGitter<alehander42> and they are still database-specific, it doesn't make a lot of sense to use this syntax for different kind of storage
18:45:00FromGitter<alehander42> "There should be a 30 year moratorium on inventing new query languages." this sounds like communists talking about how evil the free market is :D
18:45:24FromGitter<alehander42> i've wanted to write about that metaphor for a long time, i have to sit down and do it
18:48:42FromGitter<arnetheduck> @alehander42 yeah, that post is a bit of a rant, and there are of course good DSL examples out there... I guess the problem with DSL's in general is that a lot of them get developed frivolously without any sound theoretical backing or holistic thinking behind them.. that makes them more specialized and thus less general purpose - the ROI for having to learn them is low because the knowledge isn't generally
18:48:42FromGitter... applicable and you don't learn a useful model or abstraction from studying them, just have to drudge though the limited understanding some fellow in the closet came up with for their tiny special case
18:54:22FromGitter<alehander42> honestly you can say that about the design of a lot of the programming languages/mainstream API-s/protocols
18:55:11FromGitter<alehander42> I feel that most of the stuff just evolves slowly while it gets to a somehow mature (or must-be-backward-compatible) state
18:55:19FromGitter<alehander42> somewhat*
18:57:22FromGitter<alehander42> so usually in this case you have solutionA, solutionB, after a while somebody decides to fix the problems of those with solutionC, somebody else have different needs, so slightly incompatible solutionD, which is ported to a diff technology's solutionE
18:58:25FromGitter<alehander42> and after a while solutionX and solutionY are kinda good, as the "community" managed to analyze and generalize its previous mistakes and the problem domain
19:00:52FromGitter<arnetheduck> oh for sure - and what's good for open source developers is that no one is forcing them to use these shitty dsl's and they can die a natural death and perhaps provide some fertile ground for more exploration :) it's worse if you're stuck in a coercive setup where you *have* to use them.
19:01:26FromGitter<alehander42> but the problem is the fitness function "what is better" isn't actually well defined, you might prefer good abstractions, but a lot of people might use something that *feels* easier
19:01:40FromGitter<alehander42> in a very subjective sense
19:01:58FromGitter<alehander42> so the shitty dsl-s often keep being popular
19:01:59FromGitter<arnetheduck> point is, developing a good dsl takes a fair bit of work regardless of how that work happens (top-down, community, whatever)
19:02:53FromGitter<arnetheduck> no, it's like the somewhat elusive concept of "quality" - you inherently recognize when it's there and when it's not, but it's often hard to put a finger on what makes it so
19:04:39FromGitter<alehander42> I don't quite agree, sometimes a dsl can solve 80% of usecases in some hacky but *easier* way and get a lot of mindshare, based on the people that don't need to use it in advanced way (orm and CRUD are a good example)
19:04:51*Jesin joined #nim
19:05:04FromGitter<alehander42> which might not be cool, but the point is quality doesn't always survive
19:06:36FromGitter<alehander42> editors are another example, vim's (or kakoune) modal editing might be based on a beautiful abstraction model, but still most programmers use some editor-specific custom-shortcut hacks
19:08:42FromGitter<alehander42> it's always cost/benefit, people love to praise haskell/ocaml's type systems, but then 0.5% use them
19:19:25FromGitter<alehander42> (talking about https://news.ycombinator.com/item?id=18148749 )
19:19:33FromGitter<alehander42> ops wrong chat
19:26:36*darithorn quit ()
19:35:33*Vladar quit (Remote host closed the connection)
19:37:03*xet7 quit (Ping timeout: 252 seconds)
19:50:32FromGitter<arnetheduck> ah indeed, some of these fine abstract models need to be moderated with a health dose of pragmatism to make then truly useful, in a broader sense. should that be part of the criteria for `quality`?
19:50:44*xet7 joined #nim
19:52:15*Tortice joined #nim
19:52:29*vlad1777d joined #nim
20:01:24*rnrwashere quit (Remote host closed the connection)
20:02:28krux02elrood: there is also this american way, creating fake choices just so the customer feels more like he has choice.
20:02:47*snooopyUl joined #nim
20:03:30krux02the super market has 1000 choices, all of them are some form of processed GMO corn
20:04:12*snooopyUl quit (K-Lined)
20:04:41krux02alehander42: if you want a query syntax, nim has all the tools so that you can develo a DSL with macros.
20:08:12*xace joined #nim
20:08:12krux02alehander42: btw I tested both vim and emacs. it's not worth it.
20:09:03krux02well maybe emacs a bit, because emacs lisp is kind of neat to get a few things done, but with vscode you just get a better default editor.
20:13:56*xace quit (Read error: Connection reset by peer)
20:26:05*jjido joined #nim
20:26:54FromGitter<arnetheduck> @alehander42 @krux02 speaking of dsl's (and linq), one thing we'd need from the language is a more agile syntax for lambda functions.. the current `proc` and `do` notations are just too clumsy with all the type noise
20:28:17*Tortice quit (Remote host closed the connection)
20:28:50*Tortice joined #nim
20:31:45FromDiscord_<exelotl> arnetheduck: hey I mean this is a thing, or did you already know about it? https://nim-lang.org/docs/sugar.html
20:35:54FromGitter<arnetheduck> @exelotl ooh, nice! I missed that one, all the manual talks about is the full proc syntax
20:36:57FromGitter<arnetheduck> interesting sprawl on the functions in there
20:45:15*ropeney joined #nim
20:49:25*ropeney quit (Remote host closed the connection)
20:51:33FromDiscord_<exelotl> looks like you can only have single expressions in the body though
20:52:45*tdc quit (Ping timeout: 252 seconds)
20:54:05FromDiscord_<exelotl> so doThing(() => echo "hello") is fine but doThing(() => echo "hello"; echo "world") isn't
20:58:38FromGitter<arnetheduck> yeah well that's in line with lambdas mostly - in linq, you wanna use them for small conditions like filters, sorts and the like.. probably, something like linq should be smart enough to understand them a bit, so it can include its own "query optimizer"
21:03:00*stefanos82 quit (Quit: Quitting for now...)
21:04:34*zachk joined #nim
21:38:59*jjido quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
21:48:15*nsf joined #nim
22:11:53shashlickhaah - all nimgen wrapper tests finally pass with 0.18.0, now to try 0.19.0 and see what falls off
22:24:07*zachk quit (Quit: Leaving)
22:29:37*Trustable quit (Remote host closed the connection)
22:35:28*Ven`` joined #nim
22:36:50*Ven`` quit (Client Quit)
22:39:32krux02exelotl: you should be able to do this: ``doThing(() => (echo "hello"; echo "world"))``
23:08:00FromGitter<DanielSokil> How can I filter files in `request.body` from `asynchttpserver`?
23:09:22FromGitter<DanielSokil> Currently when uploading an image with a HTML form I get: ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5bb7ef226e5a401c2d2259dd]
23:11:54*Tortice quit (Read error: Connection reset by peer)
23:11:55*owner_ joined #nim
23:12:18*owner_ is now known as Guest29236
23:22:06*elrood quit (Quit: Leaving)
23:36:39*btbytes quit (Ping timeout: 252 seconds)
23:44:19*btbytes joined #nim
23:49:56FromGitter<DanielSokil> I have found this library: `https://github.com/tulayang/httpform`, But it does not work as expected.
23:55:57FromGitter<kayabaNerve> It was last updated three years ago.
23:56:11FromGitter<bung87> guess it in another field of request
23:56:21FromGitter<kayabaNerve> I'd be surprised if it even compiles.
23:56:24FromGitter<bung87> like other frameworks does
23:57:49FromGitter<DanielSokil> I have tried it, doesn't compile.
23:58:03FromGitter<DanielSokil> > guess it in another field of request
23:58:56FromGitter<bung87> generally request.body is pain text of http body