<< 16-09-2018 >>
00:04:04*crem quit (Ping timeout: 240 seconds)
00:06:31*crem joined #nim
00:19:48*SenasOzys quit (Remote host closed the connection)
00:19:51*MeTheFlea20 joined #nim
00:20:25*SenasOzys joined #nim
00:20:36*localhost23 joined #nim
00:21:18*SenasOzys quit (Remote host closed the connection)
00:22:00*MeTheFlea20 quit (K-Lined)
00:25:58*localhost23 quit (Ping timeout: 272 seconds)
00:29:45*SenasOzys joined #nim
00:34:06*kapil___ quit (Quit: Connection closed for inactivity)
00:38:48*SenasOzys quit (Remote host closed the connection)
00:39:24*SenasOzys joined #nim
00:40:36*SenasOzys quit (Max SendQ exceeded)
00:42:01*SenasOzys joined #nim
00:43:49*SenasOzys quit (Remote host closed the connection)
00:44:37*ftsf joined #nim
00:44:39*SenasOzys joined #nim
00:45:55*SenasOzys quit (Max SendQ exceeded)
00:46:27*xet7 quit (Quit: Leaving)
00:48:23*SenasOzys joined #nim
00:49:37*SenasOzys quit (Max SendQ exceeded)
00:50:15*SenasOzys joined #nim
01:17:16*demotomohiro quit (Ping timeout: 272 seconds)
01:29:12*butchlugrod16 joined #nim
01:32:16FromDiscord<wenzani> hey guys
01:32:59FromDiscord<wenzani> I'm getting this error since I may have permanently deleted the karax folder after installing through nimble
01:33:01FromDiscord<wenzani> The .nimble-link file is pointing to a missing file:
01:33:19FromDiscord<wenzani> Can't uninstall or do anything with the karax package anymore, anyone know a fix?
01:33:57*butchlugrod16 quit (Remote host closed the connection)
02:02:52*abm quit (Ping timeout: 272 seconds)
02:14:52FromGitter<gogolxdong> a more generic solution is strip($JsonNode, chars = {'\"'}) instead of getInt, getStr, getFloat, getBool
02:21:43*wildlander quit (Ping timeout: 250 seconds)
02:32:35*tbmb joined #nim
02:36:05*tbmb quit (Remote host closed the connection)
02:49:24FromGitter<gogolxdong> mangkhut is striking us.
03:05:16*mal`` quit (Quit: Leaving)
03:10:23*mal`` joined #nim
03:11:08*krux02 quit (Remote host closed the connection)
03:12:36FromGitter<codenoid> @Bennyelg coz nim syntax is very different from python
03:35:01*ftsf quit (Ping timeout: 244 seconds)
03:37:23FromDiscord<Shield> bruteforcing http requests won't take me anywhere it seems, 100 continue and 206 partial content seems to be the answer to stream large files without fully loading them in memory, i'd like to know how can they be used properly
03:43:29*demotomohiro joined #nim
04:10:10FromGitter<qqtop> @gogolxdong , yep , the building is swinging during strong gusts..
04:10:27*kungtotte quit (Remote host closed the connection)
04:14:16shashlick@AlexMax: sorry man not able to help, am traveling for the next week for work
04:49:47*miran joined #nim
05:10:29*Sammitch24 joined #nim
05:11:08*Sammitch24 quit (Remote host closed the connection)
05:17:13*SenasOzys__ joined #nim
05:17:42*SenasOzys quit (Read error: Connection reset by peer)
05:19:48*SenasOzys__ quit (Remote host closed the connection)
05:20:34*SenasOzys joined #nim
05:30:23*SenasOzys quit (Remote host closed the connection)
05:32:43*SenasOzys joined #nim
05:33:19*ARCADIVS joined #nim
05:34:11FromDiscord<wenzani> anyone about the .nimble-link?
05:35:41*nq joined #nim
05:37:45*nq quit (Client Quit)
05:44:53*stefanos82 joined #nim
05:47:29*Muzer9 joined #nim
05:52:21*Muzer9 quit (Remote host closed the connection)
06:04:56FromGitter<andrewchambers> Hi there, is there a way to import a CPP define then use it as an array size
06:05:25FromGitter<andrewchambers> ? I'm trying to do something like array[SOME_DEFINE, byte]
06:08:15*c15ade4 joined #nim
06:11:33FromGitter<andrewchambers> Also, is there a way in a template to stringify an identifier? I figured I could make a template to at least import it as a var, make a nim constant and then assert they are the same value at runtime.
06:13:51FromGitter<bung87> when define(somedefine)
06:16:37FromGitter<bung87> https://github.com/nim-lang/Nim/blob/master/lib/pure/htmlgen.nim#L39
06:17:42FromGitter<andrewchambers> thank you, is there documentation for define ?
06:19:09FromGitter<bung87> https://nim-lang.org/docs/nimc.html#compiler-usage-compile-time-symbols
06:20:22FromGitter<andrewchambers> Oh, my question was taking a C preprocessor define
06:20:33FromGitter<andrewchambers> Im trying to make an array of
06:20:41FromGitter<andrewchambers> size matching some C preprocessor value
06:21:36FromGitter<andrewchambers> hardcoding it in nim seems risky because on a different platform the value may change.
06:24:14FromGitter<bung87> oh ,take define from c source code, hmm I have no idea sorry
06:25:59FromGitter<bung87> maybe you can write script read from c source and then generate a nim compile command.
06:33:23FromGitter<andrewchambers> hmm yeah
06:33:57FromGitter<andrewchambers> A final question, if a statement is not inside a function
06:34:13FromGitter<andrewchambers> but not in the main module, is it still run?
06:34:31leorizeit will
06:34:46leorizethe statement will be run before the main module the last time I tried
06:34:50FromGitter<andrewchambers> ah, i think I found a way to do this with templates now :) thank you.
06:36:30*leorize quit (Quit: WeeChat 2.2)
06:41:11*stefanos82 quit (Quit: Quitting for now...)
06:43:18*SenasOzys quit (Remote host closed the connection)
06:44:13*SenasOzys joined #nim
06:45:09*demotomohiro quit (Ping timeout: 244 seconds)
06:48:17miranis there something like `varargs` which will not a return `seq[T]` but it will allow different types (returning a tuple)?
06:58:21FromGitter<Bennyelg> Interesting
07:10:41FromGitter<alehander42> @andrewchambers I think you can just define something like
07:11:14FromGitter<alehander42> `const cPreValue {.importcpp: "Cident".}: typ`
07:11:19FromGitter<alehander42> and then use that
07:11:28FromGitter<alehander42> if i am understanding you correctly
07:12:11FromGitter<alehander42> where Cident is a c expression which evaluates to your value
07:12:14FromGitter<codenoid> ‏‏‎
07:12:24zachcarterI am still struggling mightly with this backtick thing and importcpp / finding a way to reliably create JS template literals
07:12:41zachcarterAraq / dom96 - this is really putting a damper in me using Nim at work
07:13:02FromGitter<alehander42> can you give a short example? I am sure there is a way to at least workaround it
07:13:36zachcarterI have one lying around I believe - one moment
07:15:46zachcarterhttps://gist.github.com/zacharycarter/a43a326108126be9fca99325bc3135ab
07:16:47zachcarterreally - I need to plug the value that resides in the variable bar - between backticks
07:16:53zachcarterin the emitted JS
07:17:28zachcarterand surround it in backticks to create a template literal
07:20:07zachcarterideally, hello world is something like ``<h1>Hello ${this.props.name}!``
07:20:15zachcarterwhich would create a JS template literal
07:20:22zachcarterso that's what I want the macro to emit
07:20:32zachcarterthat would get passed to the foo proc
07:21:14zachcarterthis would all work if backticks didn't screw up importcpp so badly I think
07:23:15FromGitter<alehander42> so this `<h1>..` is a string in nim's proc?
07:23:32FromGitter<alehander42> e.g. ⏎ ⏎ ```proc .. ⏎ return "<h1>Hello ${..}!"``` [https://gitter.im/nim-lang/Nim?at=5b9e04f454587954f9acf253]
07:25:02FromGitter<zacharycarter> well - in the macro / DSL it will be a variable name
07:25:12FromGitter<zacharycarter> so in the example it is - `bar`
07:25:25FromGitter<zacharycarter> but the value of bar is "Hello World!"
07:25:53FromGitter<zacharycarter> so what I need to have happen is when the JS is created - the emit produces: return foo("Hello World!")
07:26:03FromGitter<zacharycarter> err actually sory
07:26:09FromGitter<zacharycarter> return foo(`"Hello world!"`)
07:26:19FromGitter<zacharycarter> hello world - surrounded in backticks
07:26:35FromGitter<zacharycarter> I realize that that's not going to happen with what I have now - it's just going to escape the variable name and put the non mangled variable name there
07:27:52FromGitter<zacharycarter> I've tried doing something like this - ⏎ ⏎ ```proc eval*(o: JsObject, v: JsObject): JsObject {.importcpp: "eval(#`#`).}``` [https://gitter.im/nim-lang/Nim?at=5b9e05f8fcba1254fac25820]
07:29:00FromGitter<zacharycarter> but calling it like - ⏎ ⏎ ````eval`(toJs("<h1>Hello {this.props.name}"))``` [https://gitter.im/nim-lang/Nim?at=5b9e063ce5c2cc56ad9848ac]
07:29:05FromGitter<zacharycarter> produces something like this in the js code -
07:30:43FromGitter<zacharycarter> ```return eval(#`#`)(...)```
07:32:27*nsf joined #nim
07:34:25FromGitter<alehander42> ok, something i don't quite understand in your example is
07:34:27FromGitter<alehander42> you directly emit
07:34:30FromGitter<alehander42> e.g. if i have
07:34:43FromGitter<alehander42> ```proc willCompileButHowBactick(): cstring = ⏎ return `foo`(@[2])``` [https://gitter.im/nim-lang/Nim?at=5b9e0793e5c2cc56ad9850f8]
07:35:07FromGitter<alehander42> this will emit `@[2]` anyway
07:35:12FromGitter<alehander42> which doesn't make a lot of sense to me
07:35:37FromGitter<alehander42> ah so it's supposed to be javascript
07:35:41FromGitter<zacharycarter> yes
07:35:43FromGitter<alehander42> but your want to replace only
07:35:44FromGitter<alehander42> the names
07:35:55FromGitter<alehander42> and the strings
07:36:35FromGitter<zacharycarter> well - in the ⏎ ⏎ ```example: ⏎ proc willCompileButHowBactick(): cstring = ⏎ return `foo`(`bar`)``` ⏎ ⏎ part - `example` is representing a macro that allows the user to define a ES2015 class [https://gitter.im/nim-lang/Nim?at=5b9e0803fcba1254fac262eb]
07:36:53FromGitter<zacharycarter> in that DSL they should be able to define methods on the class using Nim syntax for the method signature
07:37:01FromGitter<zacharycarter> and then in the body of the proc - they can use JS
07:37:10FromGitter<zacharycarter> but they need to be able to reference Nim variables
07:37:36FromGitter<zacharycarter> so ideally they can do - ```return `foo`(`bar`)```
07:37:43FromGitter<zacharycarter> and foo and bar would be Nim identifiers
07:38:33FromGitter<zacharycarter> but where I'm struggling is - let's say foo is a Nim proc that produces another proc that takes in a JS template literal
07:38:49FromGitter<zacharycarter> what I've been doing is - use importcpp to bind to the JavaScript eval function
07:39:19FromGitter<zacharycarter> and I've been making the importcpp like - ```{.importcpp: "eval(#`#`)".}```
07:39:38FromGitter<zacharycarter> which works fine - until you wrap the bound proc in backticks
07:40:03FromGitter<zacharycarter> then the JSgen seems to just put ```eval(#`#`)``` - in the compiled JS code
07:47:46*ebx1000 joined #nim
07:54:53FromGitter<alehander42> i got a version of yours working
07:54:56FromGitter<alehander42> but my only problem is
07:55:07FromGitter<alehander42> i can't get emit to not produce newlines
07:55:27FromGitter<alehander42> so now for ⏎ ⏎ ```return `foo`(\`${`bar`}\`)``` [https://gitter.im/nim-lang/Nim?at=5b9e0c6fa56b215c9745e3e8]
07:55:34FromGitter<alehander42> it generates
07:55:51FromGitter<alehander42> ```foo_141493(`${bar_141510[0]}`)```
07:55:59FromGitter<alehander42> but with newlines between each token
07:56:12FromGitter<alehander42> like ⏎ ⏎ ```foo_141493 ⏎ (` ⏎ ${ ⏎ bar_141510[0]``` [https://gitter.im/nim-lang/Nim?at=5b9e0c9ce5c2cc56ad986a55]
07:56:38FromGitter<zacharycarter> hrm okay - that's better than what I've gotten so far :) thank you!
07:59:49*Vladar joined #nim
08:03:11*gmpreussner quit (Quit: kthxbye)
08:03:28*gmpreussner joined #nim
08:21:57*TheLemonMan joined #nim
08:40:19FromGitter<bung87> ```code paste, see link``` ⏎ ⏎ how to directly `newTableint,TableRef[Hash, HashSet[string ()` ? or am I define a wrong type? [https://gitter.im/nim-lang/Nim?at=5b9e16f254587954f9ad55c7]
08:50:38FromGitter<bung87> oh ,wrong question. how to `new Bin` works as above?
08:58:55*huphtur joined #nim
09:03:36*miran quit (Quit: Konversation terminated!)
09:03:48*huphtur quit (Remote host closed the connection)
09:23:39*wildlander joined #nim
09:29:25*ftsf joined #nim
09:32:36*dddddd joined #nim
09:38:11*ofelas quit (Quit: shutdown -h now)
09:40:52*krux02 joined #nim
09:45:28*jsn-_ quit (Quit: Reconnecting)
09:45:35*jsn- joined #nim
09:55:14FromGitter<alehander42> @zacharycarter this was my solution ⏎ https://gist.github.com/alehander42/990a0a88572072768a7c85ba60fd12a3
09:55:37FromGitter<alehander42> there is another , maybe better solution based on asm instead of emit
09:55:57FromGitter<alehander42> there is a way to change the marker for asm, so then it doesn't interpret "`"
09:56:32FromGitter<alehander42> so you can have something like @foo@(`${@bar@}`)
09:56:49FromGitter<alehander42> the problem is I don't know how exactly one passed this marker to `asm`
09:56:53FromGitter<alehander42> @Araq ? ^
09:57:13FromGitter<alehander42> the second thing is I am actually wondering why would one need to write js in the body
09:57:16FromGitter<alehander42> of those methods
09:58:05FromGitter<alehander42> I don't get why would you want to mix nim with js in such a way, I think you can write everything you need in pure nim translatable to the js you need
09:58:49FromGitter<alehander42> in this case it would be also easier to generate stuff like javascript format, you'll just have a macro similar to strformat , e.g. `%"a{2}"`
10:03:09c15ade4Is there a way to dump the C code generated by nim into something like a tarball with a configure script
10:04:03c15ade4well, configure script doesn't matter, just a way to dump the C source that can be built without nim
10:07:11*kapil___ joined #nim
10:10:29FromGitter<alehander42> if you compile with `--verbosity:3` (maybe a lower number is enough) you'll get the gcc commands used to compile it, the source itself is in `~/.cache/nim/project` by default
10:10:37FromGitter<alehander42> i guess there might be a bteter way
10:19:07*zachcarter quit (Ping timeout: 240 seconds)
10:28:51*leorize joined #nim
10:35:43*xet7 joined #nim
10:39:22*tefter quit (Remote host closed the connection)
10:41:11*nsf quit (Quit: WeeChat 2.2)
10:45:41FromGitter<dm1try> ^the generated code is platform-dependent btw
10:46:16FromGitter<dm1try> for cross-compilation use https://nim-lang.org/docs/nimc.html#cross-compilation
11:00:22*elrood joined #nim
11:08:06*chrysn8 joined #nim
11:09:32c15ade4cool :)
11:10:08*chrysn8 quit (K-Lined)
11:10:19c15ade4hmm another question, how do I write an array to a file - is there a way to slice it so that write will accept it?
11:12:05FromGitter<codenoid> https://forum.nim-lang.org/t/2586
11:18:20*ebx1000 quit (Read error: Connection reset by peer)
11:18:47*ebx1000 joined #nim
11:19:52c15ade4oh 0..^ worked
11:33:28c15ade4no wait, that wrote it as a string lol
11:36:02FromGitter<alehander42> do you want to write an array of bytes
11:38:57c15ade4yeah, I have array[32, byte]
11:39:04c15ade4and want to write it to stdout
11:44:26FromGitter<alehander42> i think you can use streams module
11:44:57FromGitter<alehander42> get a stream with newFileStream(path, fmWrite)
11:45:32FromGitter<alehander42> and write with f.writeData(arr, count * <size of element>)
11:45:38FromGitter<alehander42> in your case size is 1 anyway
11:46:43FromGitter<tim-st> is there also a way to write array of chars to string without iterating each char?
11:47:41FromGitter<tim-st> like `proc add*(s: var string, data: openArray[char])`?
11:51:26c15ade4it seems stream write does this
11:51:30c15ade4thank you
11:54:26FromGitter<Clyybber> @tim-st I think you can cast the array to a string
11:54:40FromGitter<Clyybber> And then set the length
11:55:22FromGitter<tim-st> yes, but I have to create a new string first and cast to `result[0]` :\
11:59:03*stefanos82 joined #nim
12:01:42FromGitter<Clyybber> Hmm, so you want to change it in place?
12:02:39FromGitter<tim-st> no I want to concat the charArray to the string without writing to much code
12:02:55FromGitter<tim-st> and still have best performance
12:03:20FromGitter<Clyybber> var somestring = caststring (addr somecharArray[0])
12:04:06FromGitter<Clyybber> ```var somestring = caststring (addr someCharArray[0])```
12:04:13FromGitter<Clyybber> this should do it i believe
12:05:31FromGitter<tim-st> so I would need to cast array[char] to string2 and then do strings1.add(string2) ?
12:05:36FromGitter<Clyybber> And I don't know if its really neccessary but afterwards do: ⏎ ```somestring.setLen(someCharArray.len)```
12:05:56FromGitter<tim-st> currently I'm using this: https://gist.github.com/tim-st/baea83ad188d587b80a05bbd93d869a3
12:08:53FromGitter<Clyybber> ```proc add*(s: var string, data: openArray[char]) = ⏎ s.add(cast[string](addr data[0]))``` [https://gitter.im/nim-lang/Nim?at=5b9e47d4f08bc22dfb65939c]
12:09:02FromGitter<Clyybber> i think this should do it
12:09:11FromGitter<Clyybber> nvrmind
12:09:53FromGitter<Clyybber> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5b9e4810a56b215c974719fc]
12:10:11FromGitter<Clyybber> this shoud do it
12:10:52FromGitter<tim-st> it raises out of memory error
12:11:13FromGitter<tim-st> xD
12:11:55FromGitter<tim-st> https://gist.github.com/tim-st/baea83ad188d587b80a05bbd93d869a3
12:12:52FromGitter<Clyybber> Oof
12:12:58FromGitter<Clyybber> Dont know why tho
12:13:25FromGitter<tim-st> hm, only on devel, on stable it raises `SIGSEGV: Illegal storage access. (Attempt to read from nil?) ⏎ `
12:13:28FromGitter<alehander42> i think you need to cast to cstring?
12:13:39FromGitter<alehander42> i mean, string is not an array of chars
12:13:45FromGitter<tim-st> yes, but cstring is null terminated, not sure if it works
12:14:07FromGitter<tim-st> and I dont think that a cast makes this quicker than iterating each char?
12:14:41FromGitter<Clyybber> Ohh
12:14:42FromGitter<tim-st> @alehander42 yes, it works, thanks
12:14:45FromGitter<Clyybber> I know why
12:15:54FromGitter<alehander42> yeah I don't know how does it deal with non 0 terminated
12:16:15FromGitter<alehander42> i'd like to think openarrays add a `\0` by default but maybe it doesn't make sense
12:16:19FromGitter<data-man> @tim-st ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5b9e4993fcba1254fac3b5f6]
12:17:06FromGitter<tim-st> thanks @data-man looks interesting
12:17:07FromGitter<dm1try> and some benchmarks pls
12:17:12FromGitter<dm1try> =)
12:17:13FromGitter<tim-st> this works: https://gist.github.com/tim-st/baea83ad188d587b80a05bbd93d869a3
12:17:27FromGitter<tim-st> but not tested for empty openArray
12:17:42FromGitter<tim-st> well obviously it doesnt work for empty openArray
12:18:05FromGitter<tim-st> (my version)
12:19:10*ARCADIVS quit (Quit: WeeChat 2.0.1)
12:20:43FromGitter<tim-st> no, it doesnt work for const openArray :(
12:20:56FromGitter<tim-st> result = `abcdefusrToCell`
12:21:29FromGitter<tim-st> I think unsafeAddr should throw compiler error on const data
12:22:59FromGitter<tim-st> with `setLen` it works, but I think then iteration is quicker
12:23:52FromGitter<Clyybber> No it is not.
12:24:03FromGitter<data-man> Look at C code :)
12:24:15FromGitter<tim-st> tbh I'm not a c code expert :(
12:24:22FromGitter<Clyybber> One is O(1) the other is O(N)
12:24:46FromGitter<tim-st> @Clyybber the second also copies `N` chars
12:25:04FromGitter<tim-st> copying `N` chars sound very much like `O(N)`
12:26:16FromGitter<Clyybber> It is, but manually iterating over it is still slower
12:26:54FromGitter<tim-st> yes, probably, maybe somone code test and add the quicker version that works with var and const data to `system.nim`
12:29:56FromGitter<data-man> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5b9e4cc38909f71f75e08bdc]
12:32:18FromGitter<tim-st> is that good? :D
12:33:29FromGitter<data-man> Yes. No loops. :)
12:34:41FromGitter<tim-st> ok, thanks! I think this version (with setLen) also has no loops: https://gist.github.com/tim-st/baea83ad188d587b80a05bbd93d869a3
12:36:38FromGitter<tim-st> but I think your version is very good, would be nice to have it in stdlib I think
12:39:38FromGitter<data-man> My version for C/C++ target only. Your version works for JS.
12:40:07FromGitter<Vindaar> thanks to LemonBoy's awesome criterion: ⏎ ⏎ ```code paste, see link``` ⏎ ⏎ for shorter strings the difference becomes smaller [https://gitter.im/nim-lang/Nim?at=5b9e4f27e5c2cc56ad99cf32]
12:40:20FromGitter<dm1try> > ok, thanks! I think this version (with setLen) also has no loops: https://gist.github.com/tim-st/baea83ad188d587b80a05bbd93d869a3 ⏎ ⏎ sad story here: ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5b9e4f34d655361f7609d463]
12:41:04FromGitter<data-man> @tim-st Minor fix ```if likely(data.len > 0):``` :)
12:41:30FromGitter<dm1try> vs your first: `for c in data: s.add(c)` ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5b9e4f7a7dca3065030c1446]
12:41:51FromGitter<tim-st> @dm1try hahaha thanks, this is really sad
12:42:39FromGitter<tim-st> @Vindaar thanks for the comparision so @data-man 's version is much quicker
12:42:55*PrimHelios quit (Quit: Leaving)
12:45:28FromGitter<dm1try> > thanks to LemonBoy's awesome criterion ⏎ > @Vindaar thanks for the comparision ⏎ ⏎ +1 :) [https://gitter.im/nim-lang/Nim?at=5b9e50681ee2ca650224c5ae]
12:53:29FromDiscord<Shield> hey tim-st, did you try sending a file in chunks yet? i'm trying to do it, so far omitting the content-length keeps the browser waiting for more data, and you can keep sending it (you just send that raw data without any headers), but i'm having trouble making it stop, so far the browser keeps waiting for more data
12:53:37FromGitter<tim-st> @data-man what is the reason you used `moveMem` instead of `copyMem`?
12:55:39FromGitter<tim-st> Shield: no, not yet, If you want to send it in chunks I think you need to signalize to send in ranges: https://developer.mozilla.org/en-US/docs/Web/HTTP/Range_requests
12:59:29*stefanos82 quit (Quit: Quitting for now...)
13:04:05FromGitter<data-man> @tim-st Just a habit :) For movemem the destination and the source can be overlapped. For our case it doesn't matter. And copymem is a bit faster.
13:05:23FromGitter<tim-st> @data-man ok, thanks, so your example would be even fast with copyMem, I think that would be the perfect solution for stdib then :D
13:05:26FromGitter<data-man> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5b9e5516b9531f2dfa759008]
13:05:59FromGitter<data-man> test3 with copymem
13:05:59*xet7 quit (Read error: Connection reset by peer)
13:06:09FromGitter<tim-st> ok, copymem is much faster (see cycles)
13:06:30*zachcarter joined #nim
13:06:44FromGitter<tim-st> no you changed the order^^
13:08:14FromGitter<tim-st> Shield: also if the content is sent in ranges the main problem is still the same (of course it's better in ranges) but I dont need ranges my content that will be sent has max size and I add gzip compression
13:13:00*leorize quit (Quit: WeeChat 2.2)
13:13:19*leorize joined #nim
13:13:57*zachcarter quit (Ping timeout: 252 seconds)
13:14:01FromDiscord<Shield> yeah ranges seems to work with a plain text, it's time to test it with an image
13:15:44FromDiscord<Shield> btw, inserting a GC_fullCollect right after being done with a request does make the gc reuse that memory instead of allocating more, so far it cuts the memory usage by half
13:17:36FromGitter<data-man> @tim-st What about extracting the server's code from nim-zim to standalone app?
13:25:28*arecaceae quit (Remote host closed the connection)
13:26:09*arecaceae joined #nim
13:30:36*xet7 joined #nim
13:31:01FromGitter<tim-st> Shield: ok, good to hear
13:31:16FromGitter<tim-st> @data-man you mean something like a gui app, basically the same as kiwix?
13:35:10*SenasOzys quit (Ping timeout: 246 seconds)
13:39:28*SenasOzys joined #nim
13:39:43FromGitter<data-man> @tim-st Yes. Or some CLI tool like zim-tools.
13:41:09FromGitter<bung87> https://docs.python.org/2/library/itertools.html#itertools.combinations I have a HashSet I want to do this
13:41:44*FichteFoll16 joined #nim
13:43:38FromGitter<tim-st> @data-man Yes, maybe I will extend it, but I'm not very skilled with gui's ^^, if you want to add some code or fork or other things you're very welcome, atleast I wanted to add a zimwriterfs implementation because I want to build a zim file and there is no single version that works on windows :\
13:45:10FromGitter<tim-st> for me the main purpose was feeding my nlp system with data and checking correctness
13:45:20*demotomohiro joined #nim
13:46:58*stefanos82 joined #nim
13:47:47*FichteFoll16 quit (Ping timeout: 240 seconds)
13:55:54FromGitter<data-man> @bung87 https://bitbucket.org/nimcontrib/nimcombinatorics Adapt it for your needs.
13:57:17FromGitter<data-man> @tim-st I don't use Windows :( ⏎ Or :)
13:57:49FromGitter<bung87> very nice! :) thanks
14:06:59*elrood quit (Quit: Leaving)
14:11:25*miran joined #nim
14:13:28ng0I have a question about the folder 'compiler'. in my install procedure, using install.sh, it gets installed automatically to $out/compiler (and afterwards I move it).. reading the document about the internals I thought I need this, but Debian ships their nim build without 'compiler'. Is this just for development, new c_code?
14:15:08dom96yeah, it's not important. You can install it via nimble too
14:15:29FromGitter<Bennyelg> How Do I iterate over untyped inside macro ? :|
14:15:54TheLemonManBennyelg, it depends on what kind of NimNode you have
14:16:07ng0dom96: so removing it from $out will have no impact on existing applications and libaries I am building?
14:16:17dom96indeed
14:16:22federico3ng0: it's not used during application bulids
14:16:22ng0just trying to understand the exact role of 'compiler'
14:16:37federico3it's used during the build of the compiler itself
14:16:56ng0ok. thanks dom96, federico3
14:16:56FromGitter<Bennyelg> @TheLemonMan StmtList in example
14:17:33TheLemonManBennyElg, then `for x in y` works just fine there
14:18:56FromGitter<Bennyelg> @TheLemonMan `false` Invalid node kind nnkProcDef for macros.`$
14:19:44FromGitter<Bennyelg> pinging @juancarlospaco are you going to fix inim soon
14:20:06FromGitter<Bennyelg> ?*
14:21:02TheLemonManwith no context I can't help, check out the node content by using treeRepr
14:26:20*mosh\14 joined #nim
14:26:31FromGitter<Vindaar> @Bennyelg you're not just trying to `echo` the elements you iterate over, right? Because then you'd get that error ⏎ ⏎ ```code paste, see link``` ⏎ ⏎ gotta e.g. replace that `echo y` by `echo y.repr`, if echoing is all you want to do [https://gitter.im/nim-lang/Nim?at=5b9e6817b4990c30eec37d01]
14:29:01*mosh\14 quit (Remote host closed the connection)
14:35:29FromGitter<tim-st> @data-man ok^^ when I have the time I will implement some features
14:36:16FromGitter<tim-st> I would really like to use linux but I cannot since the battery usage is only 1/3 on linux
14:36:33FromGitter<tim-st> (battery time)
14:37:30TheLemonManthat makes you more productive since you have to do more work in less time
14:38:40FromGitter<tim-st> no, it sets me under more preassure, more preassure -> more errors
14:39:44TheLemonMancarbon + pressure = diamonds, how comes tim-st + pressure = more errors?
14:39:58*nsf joined #nim
14:40:29ng0seems like nimble, building outside of nim, needs 'compiler/':
14:40:31ng0/tmp/guix-build-nimble-0.8.10.drv-0/nimble-0.8.10/src/nimblepkg/nimscriptsupport.nim(8, 11) Error: cannot open 'compiler/ast'
14:40:34ng0 compiler/ast, compiler/modules, compiler/passes, compiler/passaux,
14:40:57ng0but when compiler/ exists, nim itself complains about duplicate module names
14:44:56ng0since my build is not really canonical - the canonical way to build nimble is to checkout / copy its repository into the nim repository and build it from there, assuming from the standalone errors and the documentation
15:02:54*zachcarter joined #nim
15:05:38ng0okay, got it.
15:15:13FromGitter<Bennyelg> @Vindaar thanks
15:15:27FromGitter<Bennyelg> @dom96 httpbeast not compile? ⏎ lib/system.nim(414, 10) Error: type mismatch: got <RawHeaders, nil>
15:15:48dom96nimble install httpbeast@#head
15:17:19FromGitter<kungtotte> @bung87 casting to a uint8 array did the trick with my issue. Thanks for the help :) My code still doesn't work because it's reading a 0 where there shouldn't be a 0 in the file, but it's always something :P
15:19:05FromGitter<Bennyelg> I think I have a problem with devel :\
15:19:21FromGitter<Bennyelg> this is not compile anymore ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5b9e747954587954f9af4967]
15:20:21FromGitter<bung87> @kungtotte good to know.
15:21:10FromGitter<kungtotte> It had to be uint8 though, when I tried it with a char I got ASCII characters in the output.
15:21:13dom96Bennyelg: strings cannot be nil anymore
15:21:22FromGitter<Bennyelg> O so "" ?
15:21:41FromGitter<Bennyelg> thanks
15:22:08FromGitter<Bennyelg> Now it's jester ```/Users/benny/.nimble/pkgs/jester-0.4.0/jester.nim(82, 14) template/generic instantiation from here ⏎ lib/system.nim(414, 10) Error: type mismatch: got <RawHeaders, nil>```
15:22:32FromGitter<bung87> ah, so you got `124` instead of `\x124`
15:25:00dom96update jester too
15:25:18FromGitter<Bennyelg> done & worked thanks @dom96
15:26:21*abm joined #nim
15:38:36*sagax quit (Ping timeout: 252 seconds)
15:40:02*edcragg quit (Quit: ZNC - http://znc.in)
15:40:41*edcragg joined #nim
15:41:01*Cthalupa quit (Quit: ZNC 1.6.6+deb1ubuntu0.1 - http://znc.in)
15:46:53*Cthalupa joined #nim
15:52:07*ebx1000 quit (Quit: Leaving)
15:52:21*ebx1000 joined #nim
15:54:05*ebx1000 quit (Client Quit)
15:56:16*sagax joined #nim
15:57:11*ebx1000 joined #nim
16:01:15FromGitter<razuit> does anyone develop on OS X/macOS using Nim?
16:03:50FromGitter<bung87> ```code paste, see link``` ⏎ ⏎ how do I compare in simple way.. [https://gitter.im/nim-lang/Nim?at=5b9e7ee60cfe7f30f1a25f74]
16:08:54*ftsf quit (Ping timeout: 250 seconds)
16:09:17FromGitter<tim-st> the nim implementation for copying openArray to seq uses the slowest way: https://github.com/nim-lang/Nim/blob/0330fb2e1322e0b41d406cfbd42c34031b6bda83/lib/system.nim#L1516
16:09:34FromGitter<tim-st> is this always safe (because of unsafeAddr)? https://gist.github.com/tim-st/baea83ad188d587b80a05bbd93d869a3
16:10:28*brainproxy quit (Ping timeout: 245 seconds)
16:11:52TheLemonManI guess
16:12:32FromGitter<tim-st> I heard that gc can free it before or something and that's because it's unsafe?
16:14:09*qenthus1ast joined #nim
16:15:28TheLemonManeh? the gc doesn't free objects that are still alive (unless you hit a bug)
16:16:31FromGitter<tim-st> ok, thanks
16:20:34*zachcarter quit (Ping timeout: 240 seconds)
16:25:45FromGitter<data-man> @tim-st `T` type can have its own implementation of `=`, so the way is slowest but more correct, I think.
16:27:06FromGitter<tim-st> ok, so I'd need to check fist if `=` implementation is default for type `T`
16:31:39FromGitter<razuit> trying to compile the latest version of Nim from source, but I get this: ⏎ ⏎ ```but expected one of: ⏎ proc `==`(a, b: Port): bool`````` [https://gitter.im/nim-lang/Nim?at=5b9e856b54587954f9afa95c]
16:34:43TheLemonManmy fault (I guess?)... try changing the definition of TSa_family in lib/posix/posix_other.nim from `cuint16` to `cint`
16:35:01*qenthus1ast quit (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.)
16:35:18*SenasOzys quit (Remote host closed the connection)
16:36:01*enthus1ast joined #nim
16:37:59*brainproxy joined #nim
16:43:42*SenasOzys joined #nim
16:50:40*SenasOzys quit (Remote host closed the connection)
16:52:10dom96razuit: I do
16:52:47FromGitter<razuit> @dom96 what do you use for debugging Nim code? what’s your environment?
16:53:01dom96I use VS code
16:53:09dom96For debugging I just sprinkle echos around my code
16:53:32FromGitter<razuit> I see, so no debugger involved, no breakpoints
16:56:22FromGitter<razuit> TheLemonMan: that fixed it, thanks
17:00:53AlexMaxQuestion, what's going on here?
17:00:56AlexMaxhttps://paste.ee/p/XZ5pA
17:01:36AlexMaxI _think_ I saw something where I needed to tag certain fp struct members and functions as pragma cdecl
17:01:54AlexMaxBut is that correct?
17:02:08dom96razuit: yeah, I almost never use debuggers
17:02:19AlexMaxI also noticed that --cdecl didn't tag _all_ function pointer types as cdecl
17:02:27AlexMaxin c2nim
17:02:39AlexMaxand actually, one of the structs didn't even convert over
17:02:54AlexMaxI just saw the struct declaration, no definition
17:03:09enthus1asti want to construct jester routes from different places (to let modules register jester routes) any idea how this could work out? The closest thing to what i want to achieve seems to be "router" but the issue i have is that i must `extend` the module routes in a main routes block.
17:04:11AlexMaxAlso, why do I need to tag function pointers as cdecl in the first place? Doesn't Nim basically compile functions to normal C functions?
17:04:43AlexMaxBasically reusing the compiler's internal calling convention, which is usually cdecl?
17:05:06enthus1asti thing it uses nimcall by default
17:06:40AlexMaxI realize that pastebin leaves out the most important thing - the contents of the nimnuklear nim file
17:07:33AlexMaxBut yeah, I did a google search for the GCC error concerning ClE_0 and I found https://github.com/nim-lang/Nim/issues/8526
17:08:51AlexMaxStill doesn't answer why c2nim --cdel doesn't tag some fp's and functions as cdecl though.
17:09:11AlexMaxBut before I got into that nitty-gritty I thought I should check in here and see if I was even in the right ballpark
17:09:17FromDiscord<Shield> btw I managed to send a 320mb file with 150mb of ram only
17:10:05*SenasOzys joined #nim
17:12:34*demotomohiro quit (Quit: Leaving)
17:14:48FromDiscord<Shield> actually it only takes 20mb now and stable
17:17:57*elrood joined #nim
17:28:47FromGitter<Clyybber> @Araq @zah @TheLemonMan ⏎ ⏎ ```code paste, see link``` ⏎ ⏎ test1 succeeds, test2 fails with type mismatch, got int but expected uint64 [https://gitter.im/nim-lang/Nim?at=5b9e92cfbe4f300626a24e0d]
17:35:58ng0is there a native equivalent to python-qr already for nim?
17:37:53federico3ng0: nimble search qrcode
17:38:36federico3meh https://github.com/achesak/nim-qrcode uses a service :(
17:38:39ng0yes, the only result there is not really something I'd use for this project. I was just checking how much I'd need to write
17:39:07dom96enthus1ast: https://github.com/dom96/jester/blob/master/tests/alltest.nim#L30
17:39:34TheLemonManng0, you can just use python-qr via nimpy if you just want to get up to speed
17:40:55ng0my take is that I want to get rid of python in gnunet. new tests get to be no longer python, and we have this optional(?) 'gnunet-qr' part which uses python-qr and its dependency chain
17:41:01enthus1astdom96: the problem i have then is that i must add an extends line for every module
17:41:40dom96that's not much of a problem :P
17:42:00enthus1astmy goal is to load the modules at runtime :)
17:42:12FromGitter<Clyybber> TheLemonMan Do you have an idea why the above happens?
17:42:19dom96Then use a custom router
17:42:31dom96https://github.com/dom96/jester/blob/master/tests/example2.nim
17:42:55dom96the `router` macro creates a proc similar to this
17:43:06dom96you should be able to call it inside that proc depending on which you've loaded
17:43:51ng0TheLemonMan: we have around 172 lines of python remaining, I think the biggest file is this small qr script.
17:44:03*thor77 left #nim ("Leaving")
17:46:08federico3https://www.thonky.com/qr-code-tutorial/introduction that's quite a mess
17:48:47ng0https://gnunet.org/git/gnunet.git/tree/src/util/gnunet-qr.py.in right now we use zbar, which depends on python-qr iirc.
17:52:11TheLemonManClyybber, the first sizeof resolves to an integer literal while the second doesn't
17:52:43TheLemonMantherefore the compiler is unable to silently convert an int into a uint64
17:52:58FromGitter<Clyybber> Hmm, ok I shall report this then
17:53:27TheLemonMannot really, sizeof for objects is not constant-folded
17:54:00TheLemonManand since sizeof and len both return `int` the error is correct
17:54:00FromGitter<Clyybber> Means?
17:54:26FromGitter<Clyybber> But it is an inconsistency, right?
17:54:35FromGitter<Clyybber> Which could be considered a bug?
17:56:08FromGitter<Clyybber> Why is sizeof for objects not constant-folded?
17:56:43TheLemonManit's inconsistent because the compiler is smart, change the type to uint8 and the expression on the right into (sizeof(someNumbers[0]) * someNumbers.len * 100) and you'll see a similar error (coming from fitNode this time too)
17:57:44TheLemonMansee #5564
17:57:56TheLemonManwell fuck, I meant #5664
17:58:23*joselo joined #nim
17:58:35TheLemonManng0, the script is small because that qrcode lib is quite feature-packed heh
17:58:43ng0yes
17:59:24TheLemonManyou can probably replace it with a shell script if you really want to get rid of it
18:00:22FromGitter<Clyybber> TheLemonMan Does that mean that sizeof(objecttype) is not available at compile time, while sizeof(int) is?
18:00:37FromGitter<Clyybber> If so would #5664 fix that?
18:00:41*joselo quit (K-Lined)
18:01:48TheLemonManyes x2
18:01:55ng0TheLemonMan: I'd have to read more into what we do next time I have time for it.
18:02:13ng0but you are probably right.
18:02:40TheLemonManof course if you feel like writing a library for qrcode generation that'd be cool
18:03:46ng0i feel like it, but first I want to get the nim integration done, which should be done next week or so.
18:04:26ng0i have more than a couple of large projects, so it'd be shared time.. but fun project.
18:05:47*nsf quit (Quit: WeeChat 2.2)
18:09:41enthus1asthttps://github.com/ricmoo/QRCode/blob/master/src/qrcode.c
18:10:43ng0thanks
18:11:54TheLemonManrazuit, how did you manage to get that error?
18:12:54FromGitter<dm1try> ./koch tools ^
18:13:26FromGitter<dm1try> reproducilble on Mac OS with on latest devel branch
18:14:51*Jesin quit (Quit: Leaving)
18:16:15TheLemonManaha, the CI didn't catch it because the osx build is allowed to fail
18:18:38*Jesin joined #nim
18:21:04*enthus1ast quit (Ping timeout: 240 seconds)
18:29:26*aziz joined #nim
18:43:38*gartral3 joined #nim
18:46:47*aziz quit (Quit: Ex-Chat)
18:49:15*gartral3 quit (Ping timeout: 272 seconds)
18:49:53*miran quit (Ping timeout: 272 seconds)
18:51:47*miran joined #nim
18:59:02FromGitter<kayabaNerve> Just double checking, I can't read from stdin with async right?
19:01:07*sorressean11 joined #nim
19:01:33*lordheavy11 joined #nim
19:02:15*lordheavy11 quit (Remote host closed the connection)
19:04:22FromDiscord<Shield> i think you can only do it by spawning stdin.readLine() in a new thread
19:08:30*sorressean11 quit (Ping timeout: 252 seconds)
19:15:46FromGitter<kayabaNerve> But then you have to check the result in a while true loop which has the same issue
19:20:55FromDiscord<Shield> https://github.com/dom96/nim-in-action-code/blob/master/Chapter3/ChatApp/src/client.nim
19:21:43*TheLemonMan quit (Quit: "It's now safe to turn off your computer.")
19:22:03FromDiscord<Shield> it has no issue
19:24:30FromDiscord<Shield> i don't think you can do it in another way, getting the file descriptor for stdin is tricky depending on the os, not sure tho
19:25:37FromGitter<Bennyelg> what do I miss? ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5b9eae311ee2ca650226dc16]
19:29:58FromGitter<Vindaar> @Bennyelg you're only iterating the tree level below the `ProcDef`, so all you count is `Ident "testProcFirst"
19:31:18FromGitter<Vindaar> If you actually want to count `echo` you should probably only count `nnkCall` and check whether element 0 of that is the ident `echo`. But first you need to iterate over the `StmtList` (index 6 of `x` here). But you might wanna do that recursively
19:31:32FromGitter<Vindaar> check `treeRepr` of the input to get an idea
19:32:40FromGitter<Bennyelg> @Vindaar how do I iterate over StmtList ?
19:32:59FromGitter<Bennyelg> case e.kind ⏎ of nnkStmtList : then what?
19:33:52FromGitter<Vindaar> well, in principle in that case branch you just do `for a in e`
19:34:32FromGitter<Vindaar> but you might want to have some proc that does the traversal of one level for you, which you call whenever you want to go one level deeper..
19:34:58FromGitter<Vindaar> (says inexperienced me :P)
19:36:11FromGitter<Vindaar> talking about something like this: https://github.com/Vindaar/nimhdf5/blob/master/src/nimhdf5/util.nim#L33-L53
19:37:00FromGitter<bung87> what’s the purpose? these codes
19:37:14FromGitter<Bennyelg> practice ;
19:37:19FromGitter<Bennyelg> I am bad bad bad in procs
19:37:24FromGitter<Bennyelg> thanks @Vindaar
19:37:25FromGitter<bung87> ok
19:37:51FromGitter<Vindaar> no worries. It just takes a while to wrap ones head around it :)
19:38:02FromGitter<Bennyelg> I'll evnatually
19:46:21*smt_ joined #nim
19:49:24*smt quit (Ping timeout: 252 seconds)
20:05:48*LibrarianMage16 joined #nim
20:06:24*LibrarianMage16 quit (Remote host closed the connection)
20:08:26FromDiscord<Shield> how can I use opengl glGetString? it returns a ptr GLubyte which is a pointer to static string, how can I convert it to nim string?
20:09:27*dddddd quit (Remote host closed the connection)
20:09:38FromGitter<tim-st> is it zero terminated string or do you have the length of the string?
20:10:22FromDiscord<Shield> it doesn't have a length, so i guess it may be zero terminated
20:10:40FromGitter<tim-st> then cast ptr to ptr cstring
20:11:46FromGitter<tim-st> if you have length you would use copyMem: https://github.com/tim-st/nim-lmdb/blob/46abbfb8fcffee390ca0f230012fe7e1febd06ea/lmdb.nim#L46
20:13:10FromGitter<tim-st> Shield: it's null terminated: https://www.khronos.org/registry/OpenGL-Refpages/gl4/html/glGetString.xhtml
20:15:49FromDiscord<Shield> like this? var version = cast[ptr cstring](opengl.glGetString(GL_VERSION))
20:17:57FromGitter<tim-st> `var s = castcstring (otherPtr)` I would try, or with ptr before cstring
20:19:48FromDiscord<Shield> it works, thanks!
20:19:59FromGitter<tim-st> nice
20:20:17*dddddd joined #nim
20:20:44FromGitter<tim-st> I still have not found a solution how to zero copy buffers with length to NimString :(
20:23:12*miran quit (Ping timeout: 252 seconds)
20:25:43*freeflying16 joined #nim
20:29:47*freeflying16 quit (Killed (Sigyn (Spam is off topic on freenode.)))
20:36:02*SenasOzys quit (Remote host closed the connection)
20:37:35*SenasOzys joined #nim
20:42:08*zachcarter joined #nim
21:03:07FromGitter<Bennyelg> Where can I find any example using Channles on nim
21:25:44*Guest90010 joined #nim
21:28:32*Guest90010 quit (Remote host closed the connection)
21:34:07FromGitter<kayabaNerve> @Bennyelg Be the change you want to see
21:38:41FromGitter<Vindaar> https://github.com/nim-lang/Nim/pull/8611
21:38:44FromGitter<Vindaar> ^
21:52:09*MyMind joined #nim
21:52:10*Sembei quit (Ping timeout: 246 seconds)
22:03:42*Vladar quit (Remote host closed the connection)
22:13:52*ldlework is now known as ldlework[MIA]
22:16:17*kapil___ quit (Quit: Connection closed for inactivity)
22:35:22*sagax quit (Quit: Konversation terminated!)
22:37:36*qwertfisch is now known as hcsiftrewq
22:51:07*gangstacat quit (Ping timeout: 240 seconds)
22:55:21*gangstacat joined #nim
22:56:26FromGitter<drewp> https://gist.github.com/drewp/d299548fd28c9db22dc1c8f0b1cf8f8a main.nim(44, 14) Error: type expected ⏎ ??
22:57:18FromGitter<drewp> newKnob returns Knob, which is a ref object type, and that's all good, right?
22:58:07*wildlander quit (Quit: Konversation terminated!)
23:05:00*gangstacat quit (Ping timeout: 252 seconds)
23:06:40FromGitter<dm1try> newKnob(37,38) ?
23:06:57FromGitter<dm1try> or newKnob(pin1 = 37, pin2 = 38)
23:07:13FromGitter<dm1try> https://nim-lang.org/docs/tut1.html#procedures-named-arguments
23:13:05FromGitter<drewp> oh wow, i was only looking at the left side. thanks
23:13:07*Jesin quit (Read error: Connection reset by peer)
23:15:15*Jesin joined #nim
23:18:20FromGitter<dm1try> actually I just remembered similar error; was doing the same thing - own constructor =)
23:19:04FromGitter<dm1try> and stumbled across the same problem with colon/equal sign
23:28:37*elrood quit (Quit: Leaving)
23:29:09*PI4GV15 joined #nim
23:33:58*PI4GV15 quit (Remote host closed the connection)
23:43:28FromGitter<drewp> https://gist.github.com/drewp/d299548fd28c9db22dc1c8f0b1cf8f8a (updated) main.nim(29, 5) Error: illegal capture 'result'
23:44:30FromGitter<drewp> i see some forum talk about that error, but not a workaround i understand, like https://forum.nim-lang.org/t/2653#16411
23:49:54FromDiscord<awr> i've been thinking about maybe making an RFC for a `{.scopepush.}` pragma as an alternative to `{.push/pop.}` that implicitly pops when the scope is left. i remember if you forget to use `{.pop.}` for certain pragmas the compiler can get very fussy and give you weird non obvious error statemtns (or at least, it used to the last time i did that)
23:54:36*BlaXpirit joined #nim
23:54:52*oprypin quit (Ping timeout: 244 seconds)
23:54:58*BlaXpirit is now known as oprypin