<< 20-08-2020 >>
00:12:35*rockcavera joined #nim
00:14:40*krux02_ quit (Remote host closed the connection)
01:11:50*casaca quit (Remote host closed the connection)
01:16:24*endragor joined #nim
01:16:49*vicfred quit (Quit: Leaving)
01:33:55*casaca joined #nim
01:40:59*endragor quit (Remote host closed the connection)
01:41:07*endragor joined #nim
01:48:34*apahl quit (Ping timeout: 244 seconds)
01:50:49*apahl joined #nim
02:03:00nisstyreI have a C function that takes a pointer and a number of elements, which I want to wrap, is it safe/possible to cast an array to a pointer of that type?
02:03:21nisstyreI'm assuming yes as long as the number of elements is correct
02:06:11FromDiscord<Varriount> @nisstyre Yes, you take the pointer of the first element of the array
02:07:19nisstyreI think I actually want UncheckedArray since I don't know the size at compile time
02:07:29nisstyreVarriount: ok sounds good
02:13:17*vicfred joined #nim
02:28:46FromDiscord<Varriount> @nisstyre: You can also do the same thing with sequences (although, you obviously have to be more careful there, due to the dynamic length).
02:30:49*muffindrake quit (Ping timeout: 272 seconds)
02:31:33disruptekpeople usually tell me to be careful with my dynamic girth.
02:32:38*muffindrake joined #nim
02:35:32FromDiscord<Varriount> disruptek: You just can't help yourself, can you. 😛
02:48:14*thomasross quit (Ping timeout: 260 seconds)
02:52:53disruptekwhat's the fun in that?
03:03:56FromDiscord<flywind> Hi, disruptek. I have implemented hashed and hierarchical timing wheels and `delayQueue`, but still needs more tests.
03:03:58FromDiscord<flywind> https://github.com/xflywind/ioselectors/blob/master/src/ioselectors/timerwheel.nim#L26
03:04:09FromDiscord<flywind> https://github.com/xflywind/ioselectors/blob/master/tests/timer/t_timerwheel.nim
03:07:10FromDiscord<flywind> Insert 100,000 random timers is about 8 ms(with `--gc:arc -d:danger`.
03:08:53*vicfred quit (Quit: Leaving)
03:13:36*audiofile quit (Quit: Default Quit Message)
03:22:45*waleee-cl quit (Quit: Connection closed for inactivity)
03:27:12*rockcavera quit (Remote host closed the connection)
03:47:26*hnOsmium0001 quit (Quit: Connection closed for inactivity)
04:06:02*supakeen quit (Quit: WeeChat 2.8)
04:06:42*supakeen joined #nim
04:35:34FromDiscord<iWonderAboutTuatara> Nim is super nice
04:35:41FromDiscord<iWonderAboutTuatara> I have to say
04:35:48FromDiscord<iWonderAboutTuatara> It's rust but it's way less painful
04:35:57FromDiscord<iWonderAboutTuatara> Python-like syntax is the best syntax
05:05:12*JustASlacker joined #nim
05:13:34*JustASlacker quit (Ping timeout: 246 seconds)
05:50:27FromDiscord<flywind> Does `nimrtl` or `nimhcr` have some restrictions in libraries? I want to use logging/strformat/unittest, but they all crashed.
05:51:50Zevvhcr is tricky
05:52:15Zevvi'm not sure if it is 100% sane at this time
05:56:47*solitudesf joined #nim
05:57:20*cornfeedhobo quit (Remote host closed the connection)
06:00:29*hnOsmium0001 joined #nim
06:15:00FromDiscord<iWonderAboutTuatara> Hey, is @ing users through irc broken?
06:15:04FromDiscord<iWonderAboutTuatara> I remember it being possible
06:15:09FromDiscord<iWonderAboutTuatara> But I may be going insane
06:15:28FromDiscord<Rika> try it
06:26:21*letto quit (Quit: Konversation terminated!)
06:31:12*letto joined #nim
06:36:45*JustASlacker joined #nim
06:38:51*cornfeedhobo joined #nim
07:35:27FromDiscord<iWonderAboutTuatara> @Zevv[IRC]#0000
07:35:34FromDiscord<iWonderAboutTuatara> Just gives the @
07:40:31*l1x quit (Ping timeout: 244 seconds)
07:40:48*Vladar joined #nim
07:41:03*krux02 joined #nim
07:41:35*zielmicha__ quit (Read error: Connection reset by peer)
07:42:37*alehander92 quit (Ping timeout: 260 seconds)
07:42:46*serthujop joined #nim
07:45:33FromDiscord<Rika> what
07:45:56FromDiscord<Rika> i dont understand what you're doing, pinging discord -> irc or irc -> discord
07:46:11FromDiscord<Rika> one is simply typing the name, the other is `@username`
07:47:30Zevvzup
07:48:14Zevvdont know, i just see my nick hopping by and that makes my irc go yellow
07:48:19Zevvso that wrks
07:48:27*rayman22201 quit (Ping timeout: 260 seconds)
07:52:24*endragor quit (Remote host closed the connection)
07:52:52*endragor joined #nim
07:54:30*l1x joined #nim
07:54:34*rayman22201 joined #nim
07:54:46*zielmicha__ joined #nim
07:58:05*endragor quit (Ping timeout: 240 seconds)
07:58:28*alehander92 joined #nim
08:02:05FromDiscord<djazz> I fixed my code to work with gc:arc, yay! GC_rf/GC_unref while working with the C library helped a lot to clean up my code too!
08:02:09FromDiscord<djazz> GC_ref*
08:03:28*endragor joined #nim
08:03:39FromDiscord<djazz> Still a WIP, I'm wrapping the latest version of webview (there are other nim wrappers, like webview and webgpu, but they use an old version of the C library) https://github.com/daniel-j/nim-webview
08:04:00FromDiscord<djazz> `nimble task example && ./example`
08:04:59FromDiscord<djazz> With ARC I also had to remove async stuff, which is fine since I use threadpool instead (gui needs to run on main thread)
08:05:19FromDiscord<djazz> So callbacks from the gui gets spawn'd
08:07:14FromDiscord<Elegant Beef> Nice
08:09:33FromDiscord<djazz> You can see usage of ref/unref here https://github.com/daniel-j/nim-webview/blob/master/src/webview.nim#L158-L168
08:10:48FromDiscord<djazz> Once you figure out the quirks, coding with Nim is a joy 😄
08:11:03FromDiscord<djazz> I'm happy I'm not doing this in C++/C
08:11:53FromDiscord<Elegant Beef> what quirks?
08:13:04*NimBot joined #nim
08:13:33FromDiscord<djazz> Mostly related to gc. I come from dynamically typed languages (JS, Python, PHP)
08:14:29FromDiscord<Elegant Beef> Ah
08:14:39FromDiscord<Elegant Beef> I came from C# so it was only weird to not have classes
08:15:32FromDiscord<Elegant Beef> But wasnt really weird since the UFCS is a joy
08:16:18FromDiscord<Elegant Beef> your webview test folder is just a config.nims 😄
08:16:51FromDiscord<djazz> I'll add tests later, when I've made the API 😄
08:17:06FromDiscord<djazz> WIP, example.nim is my playground
08:17:27*hnOsmium0001 quit (Quit: Connection closed for inactivity)
08:18:14FromDiscord<djazz> The compile pragma is great btw. I used nimterop to generate it first, and it used header, and I kept getting symbol duplication.
08:19:10FromDiscord<djazz> But now I made it compile the C library as an object file and link it. Lots of things in the header that I don't need directly included
08:20:14FromDiscord<djazz> I only wish the compile pragma had an option to not cache or detect if its dependencies changed
08:20:46FromDiscord<djazz> webview.cc is a oneliner with `#include "webview.h"`, so changing the header requires the nim -f flag
08:22:02FromDiscord<Elegant Beef> *This is why you just write everything in pure nim* 😛
08:22:17FromDiscord<djazz> I might split up it later so the C api is in one file and the nim module api in one
08:23:07FromDiscord<djazz> Sure, you do the GTK, Cocoa and EdgeHTML bindings in Nim for me 😄
08:25:16FromDiscord<Elegant Beef> On second thought... webviews suck
08:28:39FromDiscord<djazz> Yeah... but pretty neat way to do cross platform apps (that doesn't use electron)
08:29:30FromDiscord<djazz> Maybe I can add Qt's WebEngine as alternative to GTK too
08:30:07*l1x quit (Changing host)
08:30:07*l1x joined #nim
08:30:10*rayman22201 quit (Changing host)
08:30:10*rayman22201 joined #nim
08:30:10*zielmicha__ quit (Changing host)
08:30:10*zielmicha__ joined #nim
08:30:10*alehander92 quit (Changing host)
08:30:10*alehander92 joined #nim
08:30:10FromDiscord<djazz> (that would be a big dependency on Windows and macOS, but small on Linux systems running a Qt desktop like KDE)
08:31:31FromDiscord<djazz> I wonder what name I'll submit this under to nimble, since webview is taken...
08:32:22*casaca quit (Ping timeout: 246 seconds)
08:40:11*maier joined #nim
08:43:12FromDiscord<Elegant Beef> But.. but.. fidget! 😄
08:44:10*serthujop quit (Quit: Lost terminal)
08:47:12FromDiscord<djazz> fidget uses opengl right?
08:47:20*cornfeedhobo quit (Quit: ZNC - https://znc.in)
08:47:50Araqiirc yes
08:49:13Zevvi think fidget is kind of agnostig, opengl is one of the backends
08:52:02FromDiscord<Elegant Beef> It runs on all platforms in theory
08:54:21FromDiscord<Elegant Beef> At the moment it only uses opengl/webgl apparently
08:55:19FromDiscord<Elegant Beef> Rather canvas
08:58:52FromDiscord<djazz> opengl and cairo for vector and font it seems
08:59:40*awe00 joined #nim
09:00:59FromDiscord<djazz> thing is, I need to render html, in a part of the app, with clickable links
09:03:21FromDiscord<lqdev> fidget doesn't use cairo for font rendering, it uses typography
09:03:27FromDiscord<Varriount> @djazz Dynamic Typers Unite!
09:03:35*cornfeedhobo joined #nim
09:03:35FromDiscord<lqdev> !repo treeform/typography
09:03:36disbothttps://github.com/treeform/typography -- 9typography: 11Fonts, Typesetting and Rasterization. 15 116⭐ 7🍴
09:03:37FromDiscord<djazz> 😄
09:03:43FromDiscord<Varriount> (I came from Python)
09:04:17FromDiscord<djazz> something like http://www.litehtml.com/ would be nice to have in fidget
09:04:49FromDiscord<lqdev> but it's against the point of fidget
09:05:31FromDiscord<djazz> does fidget have the api so I can draw my own things?
09:06:06FromDiscord<djazz> or the the backend abstracted away
09:06:35FromDiscord<Varriount> Or more accurately, what does one do if they need functionality fidget can't provide?
09:07:05FromDiscord<Varriount> (edit) 'more accurately,' => 'in other words,'
09:08:43FromDiscord<Rika> i hated dynamic typing
09:09:13FromDiscord<Rika> when i discovered 3.6 had type annotations i was looking for a way to enforce them without massive performance loss *cough mypy cough*
09:10:15supakeenThey're only for documentation/static analysis it's been quite widely debated if runtime optimizations were to be implemented based on them.
09:10:26supakeenThere are some projects that use them to compile to other languages (like C).
09:10:34FromDiscord<Rika> i know theyre only for doc
09:10:40FromDiscord<Elegant Beef> @djazz fidget is literally make your own GUI components
09:11:06supakeenWe do use them at work since they *do* catch certain classes of bugs which is very nice and the documentation is pretty great.
09:11:23supakeenBut CPython is likely to never do anything useful with them at runtime.
09:11:53FromDiscord<Varriount> As an implementation, CPython is fairly conservative
09:12:23supakeenYea, it has only one optimizer (if I recall correctly) and it's a peephole one.
09:12:31FromDiscord<Varriount> Yes.
09:12:31supakeenOn purpose.
09:13:16FromDiscord<Varriount> And it was only the most recent version that transitioned to using a more flexible parser generator
09:13:33supakeenAh yea, which is why it'll get pattern matching soon (tm).
09:13:55FromDiscord<Varriount> I'm not fully for that, honestly
09:14:00supakeenFor Guido's EuroPython Q&A I had submitted a question about if they'll use the new PEG parser to get rid of some of the 'warts'.
09:14:03supakeenBut they won't.
09:14:19supakeenFor example `1.foo` vs `(1).foo`.
09:14:23FromDiscord<Varriount> I feel like, as a language, Python is fairly feature complete
09:14:30supakeenSince I believe the old parser couldn't go back on error and the new one can.
09:14:41FromDiscord<djazz> Circuit Python is pretty cool. Running a Python REPL over serial to a microcontroller hehe 😄
09:14:46supakeenI am definitely not all for all the bits of patma but some of it is nice.
09:15:13FromDiscord<Varriount> supakeen: I guess this is something like pattern matching: https://github.com/kolypto/py-good
09:15:47supakeenMrm that looks more like validation of data?
09:16:08supakeenPattern matching as in the functional term where you can do something like Rust has.
09:16:21supakeenhttps://github.com/gvanrossum/patma that one.
09:16:30supakeenIt's a 'different thing' I think.
09:21:13FromDiscord<Varriount> Eh, I see the uses
09:21:26*alehander92 is now known as alexander92
09:21:41*alexander92 is now known as alehander92
09:21:52FromDiscord<Varriount> I'm still not sure it's needed enough to merit adding to the language
09:22:17FromDiscord<Varriount> But maybe I'm just getting old and conservative
09:24:14Zevvno man we've got enough of those
09:25:59alehander92we have a pre-rfc
09:26:06alehander92for pattern matching supakeen
09:26:13alehander92btw varriount i never got your objection
09:26:32FromDiscord<Varriount> Objection for...?
09:26:34alehander92validation is not really the same, like, i can imagine it having some related points
09:27:07alehander92https://github.com/nim-lang/RFCs/issues/245#issuecomment-663443583
09:27:08disbot➥ Early version of RFC: Pattern matching ? ; snippet at 12https://play.nim-lang.org/#ix=2sbl
09:27:38alehander92i mean, it's a great observation, just trying to say one needs it many places where one does just assignment or ifs
09:27:46alehander92so it might not be obvious by writing code
09:28:24alehander92because oftn matching can be a more succinct syntax for a sequnce of if
09:28:29alehander92or if + assignments
09:28:58supakeenalehander92: Yes I like it :)
09:29:20FromDiscord<Varriount> I don't understand what that has to do with my comment
09:30:26alehander92well, you say you haven't felt the need for it
09:30:26FromDiscord<Varriount> I guess the `binding` functionality is useful
09:30:55*NimBot joined #nim
09:31:20alehander92and it has different possible idioms which are in a way invisible if one haven't really seen them in code before
09:31:54alehander92which is why it might not be a good fit for nim(people just not recognizing it), but i still think it would be valuable
09:33:00alehander92e.g. one can use it for unification, combine it well with various ADT-s : nim nodes are a good example exactly because all kinds of variant objects should be well suited for it
09:47:29FromDiscord<Varriount> alehander92: I wasn't stridently arguing against pattern matching, although I do think an initial implementation using a macro or compiler plugin would be better
09:47:39alehander92we have those
09:47:45alehander92we have several libs
09:48:09alehander92actually my idea would be to just combine them together into a more refined api
09:48:15alehander92again using a macro-only lib if possible
09:48:56alehander92i don't want to change the core language (yeah, it would probably make some of those features better, but nim is about macros right
09:48:57alehander92)
09:52:09FromDiscord<Varriount> My biggest feature wish is to get macros that can alter (at least) the AST of the scope they are called in, that is after their invocation.
09:52:54FromDiscord<Varriount> Essentially, block macros, but without having to indent
09:55:26FromDiscord<Varriount> sent a code paste, see https://play.nim-lang.org/#ix=2uHR
10:07:07*abm joined #nim
10:11:30*nc-x joined #nim
10:13:25*vikfret quit (Ping timeout: 264 seconds)
10:14:53*nc-x quit (Remote host closed the connection)
10:15:50alehander92hm
10:16:10alehander92but why not indenting
10:16:17alehander92i honestly prefer having another keyword
10:16:34alehander92i think it's a bit confusing when behavior changes based on a directive up there
10:16:51alehander92(on the other hand, that's what async pragma etc do so, i am not very fair in that)
10:17:36alehander92othewise, people often want to change the callsite somehow
10:18:08alehander92but i kinda agree with araq that this seems like a big change: it makes it harder to know what does a macro change
10:30:02*bung joined #nim
10:31:32*casaca joined #nim
10:40:20*NimBot joined #nim
10:46:33*arecacea1 quit (Remote host closed the connection)
10:47:14*arecacea1 joined #nim
10:52:15*kinkinkijkin joined #nim
10:54:32kinkinkijkinhello, I'm using the nimgl vulkan bindings with the intent of properly learning vulkan, using the nim reference, nimgl reference, and vulkan tutorial and reference
10:55:35kinkinkijkinbut I've been stopped at the fact that I forgot what nim code corresponds to a class with public and private members from C++ lol
10:55:58kinkinkijkindone way too much functional programming of late
10:58:47kinkinkijkinnote entirely sure why they're using a class for it, since I don't know the api well
10:58:54kinkinkijkinnot*
11:01:35kinkinkijkinnevermind, actually read through the tutorial further to find why, it's not necessary
11:02:27*vikfret joined #nim
11:03:34*waleee-cl joined #nim
11:35:26*Zectbumo quit (Remote host closed the connection)
11:59:41*lritter joined #nim
12:02:31*bung quit (Quit: Lost terminal)
12:06:02*supakeen quit (Quit: WeeChat 2.8)
12:06:33*supakeen joined #nim
12:11:53*maier quit (Quit: leaving)
12:19:12*PMunch joined #nim
12:19:48*rockcavera joined #nim
12:33:54*Kaivo quit (Ping timeout: 265 seconds)
12:35:59*Kaivo joined #nim
12:37:32*vicfred joined #nim
12:38:35federico3https://analysis-tools.dev/tag/c anything useful for Nim?
12:41:01*dv^_^ quit (Ping timeout: 264 seconds)
12:51:45FromDiscord<djazz> Noticed switch("gc", "arc") in config.nims causes a ton of errors, but works fine with --gc:arc hah
12:52:34FromDiscord<Recruit_main707> thats very weird... how are your files named?
12:53:00FromDiscord<Recruit_main707> federico3 cant you just use a c/c++ one?
12:53:12FromDiscord<Recruit_main707> its basically what you are gonna find
12:53:26FromDiscord<djazz> ..assuming that's how you set switches when running nimble run/build
12:53:58FromDiscord<Rika> what errors are given
12:54:36Araqfederico3, DrNim is influenced by Frama C
12:54:55FromDiscord<vieru> sent a long message, see http://ix.io/2uIu
12:55:23federico3Recruit_main707 huh?
12:55:57FromDiscord<Recruit_main707> djazz: do you have your main.nim file, and a main.nims file in the same dir?
12:56:16*Lord_Nightmare quit (Ping timeout: 246 seconds)
12:56:26FromDiscord<djazz> no, just config.nims at root of my project
12:56:28FromDiscord<Recruit_main707> federico3: nvm... i didnt understand :p
12:56:40FromDiscord<djazz> lots of errors like `stdlib_io.nim.cpp:270:6: note: crosses initialization of ‘int T7_’`
12:57:03FromDiscord<djazz> `stdlib_tables.nim.cpp:1041:3: error: jump to label ‘LA1_’`
12:57:08FromDiscord<Recruit_main707> hmm, im not sure how its done with nimble, what happens if you do it that way?
12:58:06FromDiscord<djazz> I haven't used nims files before now
13:05:58FromDiscord<kodkuce> hmm 1 issue with dumping all in one file now i am wasting time scroling to find what i want
13:06:14FromDiscord<kodkuce> and i have only 512 line of code
13:06:48FromDiscord<kodkuce> tough i would be anble to use search to jump but i frogot how i named half stuff xD
13:07:46FromDiscord<kodkuce> and i just now figured out i can ctrl clink on error link to jump to it 🙂
13:08:45*tane joined #nim
13:18:22*vicfred quit (Remote host closed the connection)
13:18:46*vicfred joined #nim
13:19:10*vikfret_ joined #nim
13:21:23*vikfret quit (Ping timeout: 240 seconds)
13:23:00leorize@djazz that's ood, `--` is a macro that translates to `switch()`
13:23:05leorizehave you filed an issue?
13:26:47*vikfret_ quit (Quit: Leaving)
13:34:58*Vladar quit (Remote host closed the connection)
13:36:24FromGitter<jorjun_twitter> proc mytest*(): cint {.importc, dynlib:"test.so" .}
13:36:51FromGitter<jorjun_twitter> Compiles OK, but getting: test.so: cannot open shared object file: No such file or directory
13:37:19shashlickLD_LIBRARY_PATH=.
13:37:55FromGitter<jorjun_twitter> tried that.. it’s in there..
13:38:56FromGitter<jorjun_twitter> I mean I have a bin folder next door to my nim file and the test.so is in there and full path to that bin folder is in LD_LIBRARY_PATH. Still no joy
13:40:00*Vladar joined #nim
13:40:33FromGitter<jorjun_twitter> ah it says, “...bin/test.so: only ET_DYN and ET_EXEC can be loaded"
13:44:25*Lord_Nightmare joined #nim
13:45:55FromGitter<jorjun_twitter> looks like I won’t be able to leverage c libraries without getting further into being a low level c nerd..
13:46:23FromGitter<jorjun_twitter> will do more of that..
13:46:24shashlickwhat exactly is your issue
13:48:51FromDiscord<kodkuce> can i have try catch in try catche?
13:48:55FromDiscord<kodkuce> nested
13:49:54FromDiscord<kodkuce> will just nim playground it
13:51:29PMunchOf course, why wouldn't that work?
13:52:56FromGitter<jorjun_twitter> @shashlick Looking to understand the nim runtime error: “...bin/test.so: only ET_DYN and ET_EXEC can be loaded” trying to wrap a shared library that depends on lots of other shared libraries… now thinking I need to make a libxxx.a of them all and try to link to just that
13:54:12FromGitter<UrKr> Hi, ⏎ Why does this throw this error: ⏎ ⏎ ```code paste, see link``` ⏎ ... [https://gitter.im/nim-lang/Nim?at=5f3e8083efafab7e3a8831f6]
13:54:42FromGitter<UrKr> It used to work so something must have changed
13:55:21FromDiscord<Rika> `expr` is now `typed` i think
13:57:09FromGitter<UrKr> Thanks, that worked if I remove the pragma. I get a warning though ⏎ Warning: inherit from a more precise exception type like ValueError, IOError or OSError. If these don't suit, inherit from CatchableError or Defect. [InheritFromException]
13:57:20FromDiscord<haxscramper> If I create PR in the nimble package repository it still has to be manually reviewed for someone, right? If it passed CI tests
13:58:59FromDiscord<djazz> Wow, adding a proc from GTK was easy! Huh. `proc gtk_window_set_decorated(window: WebviewWindow, setting: bool) {.header: "<gtk/gtk.h>".}`
13:59:22FromDiscord<djazz> then just calling it like a nim proc. No emit pragma hacking
13:59:53FromDiscord<Rika> @UrKr that sounds like an issue from a different portion of your code now, an object that is `of Exception`, it says what to do in the error
14:00:00shashlick@djazz what issue did you have with nimterop earlier
14:01:06FromDiscord<djazz> duplicate imported symbols, might be the lib. So I use compile pragma now, instead of header/dynlib
14:01:37FromDiscord<djazz> not symbols, sorry. name collisions
14:01:44FromDiscord<djazz> in the generated C++ code
14:03:08FromDiscord<kodkuce> https://play.nim-lang.org/#ix=2uIU i want to close connection, whitout doom
14:07:59shashlick@djazz any output you can share?
14:08:04shashlickOr snippet
14:08:23shashlickDid you gave multiple cimport calls
14:09:02FromDiscord<djazz> I'll refactor it so I can put the C-nim procs in its own file, then I'll try
14:09:17FromDiscord<djazz> I think I used toast on cli
14:10:27FromDiscord<adokitkat> sent a code paste, see https://play.nim-lang.org/#ix=2uIW
14:10:55FromDiscord<adokitkat> (edit) 'https://play.nim-lang.org/#ix=2uIW' => 'https://play.nim-lang.org/#ix=2uIX'
14:11:41*waleee-cl quit (Quit: Connection closed for inactivity)
14:11:59Araqdjazz, try .importc: "...", header: "..." or .importc, nodecl
14:13:04FromGitter<UrKr> Is this no longer how you define a new xml tag or am I calling it wrong ⏎ ⏎ ```code paste, see link``` ⏎ ⏎ I got the pattern from here ... [https://gitter.im/nim-lang/Nim?at=5f3e84efee58011680bb48f1]
14:13:30FromGitter<UrKr> I get this ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5f3e850a3dac5343401ca286]
14:14:05Araqremove the line 'let e = callsite()'
14:14:40FromGitter<UrKr> Thanks, that did the trick
14:14:56FromDiscord<djazz> ah, I was messing with nodecl when trying to get my callbacks to work, that might be it.
14:14:59Araqand please update your Nim compiler so that you can look at a modern htmlgen.nim file
14:15:37FromGitter<UrKr> Ok!
14:16:33FromDiscord<djazz> right now they are just importc heh `proc webview_create(debug: cint = 0; window: pointer): webview_t {.importc.}`
14:18:25FromDiscord<djazz> I'm learning! This is fun!
14:19:16*Vladar quit (Ping timeout: 265 seconds)
14:19:34FromGitter<jorjun_twitter> cracked it… extra gcc in makefile to make a single giant shared library, then worked fine.
14:19:48FromDiscord<Recruit_main707> cool
14:20:35*Vladar joined #nim
14:33:08FromDiscord<djazz> oh, I got it working with header and cdecl/nodecl https://github.com/daniel-j/nim-webview/blob/master/src/webview.nim#L7-L8
14:33:39FromDiscord<djazz> it works with both cdecl and nodecl. What do they do exactly?
14:34:39disruptekno one knows.
14:36:07FromDiscord<djazz> I'll put nodecl there, just in case
14:38:03Araqdjazz: use .header rather than .nodecl
14:39:18FromDiscord<djazz> https://media.discordapp.net/attachments/371759389889003532/746015504782589952/DBcUtoo.png
14:41:46Araq.header implies .nodecl
14:43:08FromDiscord<djazz> ah
14:43:10FromDiscord<djazz> makes sense
14:43:23FromDiscord<djazz> screenshot of the example app https://media.discordapp.net/attachments/371759389889003532/746016534010265640/Screenshot_from_2020-08-20_16-42-58.png
14:44:46FromDiscord<djazz> best way to invoke the js compiler is outside of the main cpp nim compile? or can js be compiled as part of the cpp compile?
14:47:25disruptekwhat are you talking about?
14:53:06disruptekmy word-salad parser is offline, i guess.
14:55:08FromDiscord<mratsim> @djazz look at https://github.com/Vindaar/NeuralNetworkLiveDemo
15:00:21FromDiscord<djazz> There is no build script. And the html loads the js file from a local nimcache folder
15:00:39shashlick@djazz - https://github.com/nimterop/nimterop#header-vs-dynlib also describes some details around this
15:02:17*audiofile joined #nim
15:09:18*hnOsmium0001 joined #nim
15:12:14*dv^_^ joined #nim
15:16:09FromDiscord<djazz> Thanks
15:19:27*audiophile joined #nim
15:20:30*audiofile quit (Ping timeout: 260 seconds)
15:20:35*audiophile is now known as audiofile
15:36:56disruptekleorize: the problem with nim.nvim is that when nimsuggest doesn't work, which is frequently, i really feel the pain. 🤣
15:37:59disruptekwe need a wiki entry that reminds me how to do logging.
15:42:00shashlickno takers for nimble localdeps testing?
15:42:32disruptekmaybe status are the only customers?
15:44:33shashlickmany have grumbled about shared deps
15:53:30krux02disruptek, I think that is a good point. Logging is something that comes up again and again. But pay respect that there is no universal logging solution. For my personal projects printing with echo does work fine already.
15:54:12krux02some logging you want to switch on/off at compile time, some you want to control at runtime. And some you want to remove again from source if debugging is done.
15:55:27FromDiscord<Kiloneie> There is no good debugger for Nim is there? (i mean i know the tracebacks are really good) Does anyone use a debugger?
15:55:47FromDiscord<lqdev> i sometimes use gdb when dealing with C interop
15:55:57FromDiscord<lqdev> but otherwise echo-debugging is good enough
15:57:10FromDiscord<Kiloneie> I read something similar for profiling.
15:58:08FromDiscord<Kiloneie> Nim doesn't seem to need a debugger to be honest, at least when using pure Nim.
15:58:48leorizedisruptek: yea, I'm planning to implement some easy debugging option for nim.nvim
15:59:00leorizewill do after this month, I'm wrapping up some stuff atm
15:59:13*endragor quit (Remote host closed the connection)
15:59:19leorize@Kiloneie: nim-gdb works well enough
15:59:34FromDiscord<Kiloneie> Okay
15:59:41*endragor joined #nim
15:59:52FromDiscord<lqdev> leorize: error reporting when
15:59:58leorizeit doesn't understand seqs though, so you have to do the arcane magic of `print seq->data[0]`
16:00:03shashlickThe profile module works pretty well
16:00:10shashlickAnd gdb is great for pure Nim as well
16:00:24leorize@lqdev soon™
16:00:31FromDiscord<lqdev> 'kay
16:00:36leorizeI'll have to add something to make nimsuggest prints to stdout/stderr
16:00:43disruptekkrux02: yeah, i wrote up my thoughts on logging in the rfc on the subject; i was talking about nvim logging...
16:00:47leorizelog file is fine and all but integration with that is a pain
16:00:47FromGitter<awr1> i use gdb occasionally but echo is good enough. obviously this is you know, "not good enough" for a lot of prospective people looking to play with nim
16:01:01FromGitter<awr1> `system.locals` is also neat for debugging if somewhat obscure
16:01:28FromGitter<awr1> for graphics debugging i use renderdoc but this is not really relevant to most people.
16:01:50*dddddd_ joined #nim
16:01:54krux02awr1: well I use renderdoc as well
16:02:10krux02so it is relevant.
16:03:44*dddddd quit (Ping timeout: 256 seconds)
16:03:47FromGitter<awr1> yeah renderdoc is quite good, i use it quite frequently, great for Vulkan, GL, D3D dev. i have only sort of ever played with NSight and i've never used PIX or the microsoft gpu debugging utils so i can't really compare it to them. renderdoc got some interesting SPIR-V shader debugging options quite recently that i've been itching to play with
16:03:48krux02Sorry, I use apitrace not renderdoc. Every time I used renderdoc it segfaulted instantly, I could never use it.
16:04:22krux02was a few years ago, could have been fixed by now.
16:04:29FromGitter<awr1> i have never seen apitrace before
16:04:40*FromDiscord quit (Remote host closed the connection)
16:04:44krux02it was my only option.
16:04:45*endragor quit (Ping timeout: 240 seconds)
16:04:54*FromDiscord joined #nim
16:04:57krux02It is not a real debugger
16:05:01*awe001 joined #nim
16:05:11krux02it is a recording device
16:05:33krux02you record OpenGL commands and then you replay them and it shows you buffer content and stuff like that.
16:06:06krux02since it records all OpenGL calls it really shows you how something is drawn in a game.
16:06:13*awe00 quit (Ping timeout: 264 seconds)
16:06:18krux02I'ven run it on several games by now.
16:06:41FromGitter<awr1> yeah it's not a "debugger" in the traditional sense like gdb, although maybe that has technically changed with the recent SPIR-V addons
16:07:10FromGitter<awr1> but it's still a "debugger" insofar that it is useful for diagnosing rendering issues and profiling
16:08:32FromGitter<awr1> is @mratsim here?
16:13:50FromDiscord<Varriount> Depends on who's askin'
16:14:03FromDiscord<Varriount> You wit da cops?
16:15:31PMunchHaha :P
16:16:16FromGitter<awr1> lol
16:16:59PMunchHe's not on IRC at least
16:17:06*PMunch shuffles mratsim out the back door
16:17:28PMunchThat's extra fun as you can't see those messages :P
16:17:39*PMunch says that this is our secret IRC language!
16:17:54FromDiscord<Hearthstone> What?-
16:17:59FromDiscord<lqdev> he's away on discord
16:18:04FromDiscord<Hearthstone> https://media.discordapp.net/attachments/371759389889003532/746040364221333574/Screenshot_20200820_171755_com.cutthecord.base.jpg
16:18:16FromGitter<awr1> it's The French Connection, but with nim instead of contraband
16:18:34disruptekman, i really need plumbing.
16:18:39FromDiscord<lqdev> PMunch: was that a /me?
16:18:56PMunchUhm, no...
16:19:10*PMunch yes it totally was
16:19:25FromDiscord<lqdev> you can't lie to me. i have the power of the log! https://media.discordapp.net/attachments/371759389889003532/746040704404553839/unknown.png
16:19:44PMunchOh damn it, I've been made!
16:20:02*PMunch runs after mratsim to go hide
16:20:20PMunchBut wait, the bot actually relays those over to Discord now?
16:21:08FromDiscord<lqdev> no
16:21:14FromDiscord<lqdev> i just guessed
16:21:25FromDiscord<lqdev> discord doesn't have /me
16:21:33FromDiscord<lqdev> _it literally looks like this lol_
16:21:49FromDiscord<lqdev> just wraps your text in __
16:21:50FromGitter<awr1> IRC is insecure
16:21:54PMunchBut Hearthstone shared a screenshot where they did show up?
16:21:58FromGitter<awr1> switch to superior military grade ROT13 encryption immediately
16:22:01disruptekawr1 is insecure.
16:22:13FromDiscord<lqdev> PMunch: they're relayed by ircord
16:22:15FromGitter<awr1> true
16:22:41PMunchYeah that's what I meant, so you can see what I type there (although it misses my nick, which kinda is the whole point..)
16:23:32*def-- joined #nim
16:23:39FromGitter<awr1> ROT13 is technically "military-grade" when you think about it because it is a Caesar Cipher
16:23:59*def- quit (Ping timeout: 256 seconds)
16:23:59*def-- is now known as def-
16:24:32*awe001 quit (Ping timeout: 256 seconds)
16:24:49disruptekshashlick: has anyone at all tested it other than yourself?
16:25:43*leorize quit (Ping timeout: 240 seconds)
16:26:21*leorize joined #nim
16:35:18*awe001 joined #nim
16:37:46shashlickLocal deps? Just me - is still a pr
16:38:08disruptekwhat should i do to test it?
16:38:15shashlickHave some minor changes still to push and test cases to make
16:40:28disruptekokay, lemme know. i will test it... somehow.
16:41:18*awe002 joined #nim
16:43:48*awe001 quit (Ping timeout: 256 seconds)
16:51:25*awe002 quit (Ping timeout: 240 seconds)
16:52:29shashlickdisruptek - okay, `nimble install https://github.com/genotrance/nimble@#localdeps` and then clone any project
16:52:34shashlickmkdir nimbledeps
16:52:51shashlickafter that all nimble operations should use that nimble dir
16:52:55shashlickglobal shouldn't be used
16:52:56*endragor joined #nim
16:53:32shashlickmaybe nimble install --nimbleDir:xyz - that way you don't install nimble globally
16:53:59disrupteki'm using my git branch.
16:54:55disruptekwow, it builds so fast.
16:55:11shashlickno more compiler right
16:55:35*awe002 joined #nim
16:55:39disruptekyou took the compiler out?
16:56:01shashlicklong ago
16:56:04*rockcavera quit (Remote host closed the connection)
16:56:05shashlicknimble just calls `nim e`
16:56:10disruptekhuh.
16:56:37disrupteki'm actually embedding the compiler into nimph to do nimscript.
16:57:11disruptekeverything is so much easier and faster without nimble requirements.
16:58:11shashlickthat's what nimble used to do
16:58:18shashlicki got rid of that and simply call nim e
16:58:24disruptekyeah.
16:58:28shashlickthis was April last year
16:58:49disrupteki got confused between you pulling the compiler out of nimble and pushing it into nimterop. 😁
16:58:57shashlickah k
16:59:41disruptekit's weird, it used to be that nimph was around 600k lines. now it's like 500-550k. so somehow our deps are getting tighter.
16:59:42shashlickideally you could play with pnodes both in macros and runtime
17:00:16disrupteki'm trying to build paranim but it's gonna take ages because i have so little bandwidth.
17:00:17shashlickwithout having to import the compiler
17:00:24disruptekah, it crashed.
17:00:31shashlickthe compiler should be stdlib
17:00:46disruptekideally, we don't have macros.
17:01:06disruptekwell, it seems that it did not use my localdeps directory.
17:01:23shashlickits called nimbledeps
17:01:27disruptekoh shit.
17:01:44shashlickcome on, confusion is useful
17:02:00disrupteki guess you know what the idiots are going to try.
17:02:36shashlickits hard enough keeping up with normal folks, idiots are also known as manual fuzzers
17:03:18disruptekwhat use could it be to be in localdeps mode and /NOT/ download packages.json from teh interwebs?
17:04:16shashlicki don't see why that question is useful in any scenario
17:04:21disruptekTip: 365 messages have been suppressed, use --verbose to show them.
17:05:52shashlickwhat say - does it work
17:06:01*endragor quit (Ping timeout: 265 seconds)
17:06:33shashlickyou can try your chmod 000 ~/.nimble trick
17:06:39disruptekturns out i'm dumb. paranim is just a lib.
17:06:51disrupteki'm trying paravim now.
17:07:08disruptekso far, so good. it takes awhile because my network is almost as slow as my brain.
17:07:44*natrys joined #nim
17:08:19disruptekseems like it worked.
17:09:56disruptekhmm, it didn't build libvim.
17:11:32disrupteki gave it a url and it tried to install it. seems like this is working great, shashlick.
17:12:17*Trustable joined #nim
17:12:51shashlicki'm wondering about `nimble develop pkgname`
17:13:01shashlickshould we add a flag to that to set it up in local deps mode
17:13:34disruptekdevelop mode is weird because of the chain of .nimble-links.
17:13:51shashlickif you are in a project with localdeps, it will clone to pkgname directory and link into localdeps so that work swell
17:14:17disruptekright.
17:14:17shashlickbut if you are going to nimble develop the top level project from scratch, it will end up installing deps in ~/.nimble
17:14:47disruptekeven if there's a nimbledeps directory?
17:15:29disruptekoh, you mean after the initial clone it will continue to install the deps...
17:16:05shashlickya - if you are in a project, it works fine
17:16:11shashlickbut if it is the project, it doesn't
17:16:24shashlickso one option is to add a flag
17:16:42shashlickor instead detect if the directory is empty and clone into it instead of a subdir
17:16:47disruptekaraq told me he loves environmental variables.
17:16:53disruptekhe prefers them to flags, always.
17:17:30disrupteki know, i was surprised, too.
17:30:27shashlickam propagating to child processes with NIMBLE_DIR
17:31:16disruptekthat should make him very happy.
17:48:57*audiophile joined #nim
17:49:05FromDiscord<mratsim> @awr1 yes?
17:49:10shashlick@dom96 - you around
17:51:35*audiofile quit (Read error: Connection reset by peer)
17:51:41*audiophile is now known as audiofile
17:58:22FromDiscord<--HA--> When I declare a variable and then only use it in the scope of a when statement I get a warning about declared but not used. Is there a good way to get rid of that warning?
18:02:28FromDiscord<mratsim> let foo {.used.} = blah()
18:18:41FromDiscord<dom96> shashlick: now I am
18:21:54FromGitter<awr1> hey @mratsim what are your thoughts on this? https://github.com/mratsim/weave/issues/163
18:21:55disbot➥ [Request] Optional per-thread startup procedure
18:22:19FromDiscord<--HA--> @mratsim thanks
18:22:43*rockcavera joined #nim
18:23:16FromDiscord<mratsim> @awr1 yes that's a good idea, and similarly, exit should allow for a cleanup proc
18:23:37FromGitter<awr1> allright! i think i might do a PR then tomorrow
18:30:38shashlick@dom96 - so should we add a custom nimbleDir/bin (command line, env var or nimbledeps) to the path for child processes
18:31:47shashlicki am forwarding nimbledir to child processes - https://github.com/nim-lang/nimble/pull/834/files#diff-8f5588fb3c673d679a2ed04e79cdc74eR253
18:31:48disbot➥ Local deps mode
18:32:02shashlickso if you call nimble in a task, it will still use the same nimbleDir
18:32:10shashlickhttps://github.com/nimble-test/issue280and524/blob/master/issue280and524.nimble is a good test case
18:32:35shashlickbut now that test breaks since `generator` isn't in the path anymore
18:32:55shashlickso the question is whether to add them to path at least within nimble and its child tasks
18:47:32*Jesin joined #nim
18:50:21*waleee-cl joined #nim
19:01:02FromDiscord<dom96> shashlick: sure, I don't see why not.
19:01:47*Jesin quit (Quit: Leaving)
19:06:10*dddddd_ is now known as dddddd
19:09:09shashlickok cool
19:21:51FromDiscord<fracasse> Heyo!
19:22:10FromDiscord<dom96> hiya!
19:22:13FromDiscord<lqdev> hi
19:22:17FromDiscord<fracasse> I have a question: how do I initialize a seq of strings so that the compiler doesn't deem it a `seq[TaintedString]`?
19:22:41FromDiscord<lqdev> !eval TaintedString is string
19:22:43NimBotCompile failed: ???(0, 0) Error: expression 'true' is of type 'bool' and has to be discarded
19:22:53FromDiscord<fracasse> hmmmm
19:23:17FromDiscord<fracasse> It may be the same, but I have a type that requires a seq[string]
19:23:29FromDiscord<fracasse> Should I change the field's type to `seq[TaintedString]` then?
19:23:36FromDiscord<lqdev> hold on
19:23:45FromDiscord<lqdev> !eval echo seq[string] is seq[TaintedString]
19:23:48NimBottrue
19:23:54FromDiscord<lqdev> nah should be fine
19:24:12FromDiscord<lqdev> by default `type TaintedString = string`
19:24:21FromDiscord<fracasse> cause this is what i get: https://media.discordapp.net/attachments/371759389889003532/746087239217447032/Zrzut_ekranu_2020-08-20_o_21.24.13.png
19:24:31FromDiscord<fracasse> oh no wait
19:24:34FromDiscord<fracasse> no non o
19:24:36FromDiscord<lqdev> you're passing a seq[string] to where a single string is expected
19:24:39FromDiscord<fracasse> yes yes
19:25:05FromDiscord<fracasse> xD you're right
19:25:17FromDiscord<fracasse> so this was just my mistake
19:25:36FromDiscord<fracasse> On that note, what's the semantic difference between strings and tainted strings? Just so I know.
19:26:19FromDiscord<lqdev> there's a compiler switch --taintMode:(on|off) which, if enabled, turns `type TaintedString = string` to `type TaintedString = distinct string`
19:26:21Zevvtainted strings come from the big bad world
19:26:36FromDiscord<lqdev> yeah zevv is right
19:26:44Zevvas always
19:26:56FromDiscord<lqdev> TaintedStrings are user input from outside of your program's source
19:27:01FromDiscord<Clyybber> you know what also comes from the big bad world
19:27:08FromDiscord<fracasse> Aah, I see
19:27:08Zevvmy mooom
19:27:11FromDiscord<Clyybber> the bug that keeps my typed fix from bootstrapping
19:27:26FromDiscord<Clyybber> or rather the fact that I can't spot the mistake
19:27:54FromDiscord<fracasse> Thanks everyone!
19:28:20FromDiscord<fracasse> By the way this is @aachh here.
19:28:35Zevvclyybber: so what happens exactly then?
19:29:43FromDiscord<fracasse> When?
19:32:47FromDiscord<Clyybber> Zevv: Basically instead of backing up the current scope I now open and close a scope
19:33:27FromDiscord<Clyybber> so that this works: https://github.com/nim-lang/Nim/pull/15158/files#diff-8ebf85b0c488068cf672b3cc15f5d0d0R345-R351
19:33:28disbot➥ Fix #5691
19:34:05FromDiscord<Clyybber> but the issue is that I'm getting an undeclared identifier error in guards.nim
19:34:28FromDiscord<Clyybber> maybe the right time to use disrupteks minimizer
19:41:37Zevvhmm you asked ar4q i assume
19:42:12FromDiscord<Clyybber> not yet
19:42:16disrupteki doubt it'll work but i'd be interested if you try it.
19:42:33FromDiscord<Clyybber> disruptek: Do you think it can work on the whole compiler?
19:42:53disruptekit only cares about the input file, so, sure, maybe.
19:43:05FromDiscord<Clyybber> I mean on multiple input files
19:43:05Zevvyou know dust works by means of mechanical turk right? you run it, it sends a copy to disruptek, he quickly minimizes and sends it back to you
19:43:14disrupteklol
19:43:39Zevvyou cant imaging how much work I have keeping npeg running these days
19:43:48FromDiscord<Clyybber> disruptek: Can it work on multiple files?
19:44:00disrupteki need it to work on multiple inputs, but i haven't figured out how to hook includes properly. i want imports to flow through.
19:44:09disruptekno, it only works on the main input file.
19:44:10FromDiscord<Clyybber> hmm ok
19:44:17disruptekbut, it's like `nim check`.
19:44:29disruptekcopy guards.nim to goats.nim and then dust goats.nim
19:45:38disruptekalso i would probably only enable to removeIndex yield in dust/mutate.nim
19:45:49*awe002 quit (Ping timeout: 264 seconds)
19:46:00disruptekthe others aren't super useful and they will probably make it too slow.
19:47:03FromDiscord<Clyybber> how long does it take usually?
19:47:23disruptekit's exhaustive, so it takes a very long time.
19:47:37disruptekbut, it will keep showing you its best effort.
19:47:49disrupteki mean, a couple minutes for a small file.
19:47:54FromDiscord<Clyybber> ok
19:47:59*awe002 joined #nim
19:48:06FromDiscord<Clyybber> I will to finish my manual appraoach
19:48:06disrupteki does like 10 compiles/sec.
19:48:12FromDiscord<Clyybber> and see if it leads me anywhere
19:48:14FromDiscord<Clyybber> and the ndust
19:48:27disruptekyeah, anything you can remove ahead of time will help.
19:49:55FromDiscord<tomck> Hello! Is there support for weak refs? I'm currently using the traditional GC but might switch to ARC at some point
20:04:12*abm quit (Read error: Connection reset by peer)
20:04:38*abm joined #nim
20:09:40disrupteksorta; you can read about cursors.
20:10:26disruptek^ tomck
20:10:59FromDiscord<tomck> hmmm so i might as well just use ptr?
20:12:52FromDiscord<tomck> https://nim-lang.github.io/Nim/destructors.html#cursor-inference-slash-copy-elision↵Hmmm, does this mean that `tab` is also an inferred cursor?
20:14:10disruptekno, ptr doesn't get any management under arc.
20:15:00disruptektab is a value type, although due to its size it'll be passed by reference.
20:15:56disruptekwhat do you think of nim after having used it for a few weeks?
20:16:56ZevvI'm getting the hang of it
20:17:01disrupteknice.
20:21:05ZevvI do not understand the first sentence in https://nim-lang.github.io/Nim/destructors.html#swap: "The need to check for self-assignments and also the need to ..."
20:21:08Zevvcan someone elaborate on this?
20:21:57disruptekimagine that you put x into y; that involves destroying y.
20:22:05disruptekto make room for x, right?
20:22:08Zevvright
20:22:35disruptekbut we don't want to run destructors for a swap.
20:22:53disruptekso we just cheat. because cheating with copyMem matches the semantic we actually want.
20:23:01Zevvno. but if you do 'tmp = a; a = b; b = tmp', there will be no destructors?
20:23:18Zevvoh there will
20:23:54ZevvI'm still thinking too reffy
20:23:59FromDiscord<Clyybber> well ideally
20:24:01FromDiscord<Clyybber> there would be none
20:24:19FromDiscord<Clyybber> but that requires my unfinished firstWrite lastRead analysis
20:24:25FromDiscord<Clyybber> that does the whole thing in one go
20:24:40FromDiscord<Clyybber> and thus can detect lastRead/firstWrite pairs
20:24:58disruptekgah i don't even know if i buy that that's safe.
20:25:04FromDiscord<Clyybber> it is
20:25:11FromDiscord<Clyybber> oh no
20:25:13FromDiscord<Clyybber> wait
20:26:28Zevvso basically you want to properly detect this and tread these assignments as sinks? does that make sense?
20:26:32FromDiscord<Clyybber> you will need one copy there
20:26:40FromDiscord<Clyybber> hmm
20:26:50FromDiscord<Clyybber> unless we come up with some real smart shit
20:27:23disruptekwe could technically optimize them out in cases where there's no backend interaction.
20:27:44disruptekdoesn't seem worth the complexity.
20:27:58FromDiscord<Clyybber> sink(tmp, a), sink(a, b), sink(b, tmp)
20:28:05Zevvright?
20:28:37FromDiscord<Clyybber> its not safe for every type though
20:28:47Zevvfor example?
20:29:34FromDiscord<Clyybber> a buffer whose sink operation moves the memory from src to dest and gives src back to the os
20:29:50Zevvwhen there's a real thing in there like a file descriptor, at a normal `=` you'd need a dup(), but at sink() you can just move the fd.
20:30:20Zevvwell no, for a sink you keep the same memory buffer?
20:30:29Zevvso any refs or ptrs to there are still valid
20:30:33FromDiscord<Clyybber> its only safe for those that leave the src safe to reuse
20:30:41FromDiscord<Clyybber> Zevv: Well yeah, but imagine
20:31:03FromDiscord<Clyybber> maybe I'm overthinking it and it is really safe
20:31:09FromDiscord<Clyybber> oh yeah
20:31:10FromDiscord<Clyybber> it is
20:31:23FromDiscord<Clyybber> since because of lastread all of those sinks are just memcpys
20:31:44FromDiscord<Clyybber> eh firstWrite
20:31:55Zevvso
20:33:25*natrys quit (Quit: natrys)
20:34:15disruptekclyybber is trying to confuse you.
20:34:25disruptekit's simpler than it sounds.
20:34:36Zevvoh but I have been in a perpetual state of confusion for a long, long time already
20:34:38FromDiscord<Clyybber> yeah, I'm not in the right phase right now
20:34:45FromDiscord<Clyybber> semphase it is
20:44:35ZevvI understand this can print 'foo' zero times, or one time. http://ix.io/2uKu
20:44:43Zevvbut two times seems magical to me
20:45:06disruptekjust lucky, i guess.
20:45:26FromDiscord<Clyybber> lol?
20:48:56Zevvnot sure. is it funny?
20:49:07FromDiscord<Clyybber> yeah
20:49:10FromDiscord<tomck> sent a long message, see http://ix.io/2uKx
20:49:17disruptekhar har
20:49:17FromDiscord<Clyybber> its as funny as the bug I now tracked down
20:49:22FromDiscord<Clyybber> it fails with and
20:49:25Zevvoh you did
20:49:28FromDiscord<Clyybber> but not with anand = a and b
20:49:40disruptektomck: right, cursor is weak in that it doesn't get managed.
20:49:54disruptekbut it allows you to use refs and it may get some compile-time checks.
20:50:00disruptekptr is completely unsafe.
20:50:20Zevvclyybber: lol?
20:50:40disruptekcursors have no refcount and cannot contribute to the refcounts of refs they ref.
20:51:01FromDiscord<Clyybber> Zevv: lol indeed, but tracking that whole thing down, I kinda want to delay to tomorrow hehe
20:51:15FromDiscord<Clyybber> at least now I got a 10 line repro
20:51:25Zevvdustified?
20:51:28Zevvor manual
20:51:31FromDiscord<Clyybber> manuallified
20:51:51*Trustable quit (Remote host closed the connection)
20:51:52disruptektomck: you mean about exceptions that you wish you didn't have to catch a KeyError?
20:52:01Yardanicothere's getOrDefault for that :P
20:52:07Zevvman I'm sitting outside with my laptop in some kind of nature reserve, the amount and size of critters on my screen is amazing
20:52:20FromDiscord<Clyybber> heh
20:52:28disrupteki just had to have a fight with a dragonfly.
20:52:36ZevvI smack some of then, and this hedgehog under my chair then goes crunchcrunch
20:52:43Zevvi feel so one with nature
20:52:54FromDiscord<Clyybber> disruptek: Lol, one just flew into my room half an hour ago
20:52:57Zevvdisruptek: who won
20:53:10disruptekbentley is so useless.
20:53:20FromDiscord<Clyybber> my cat too
20:53:22FromDiscord<Clyybber> just gtfo
20:53:36FromDiscord<Clyybber> but its cool, dragonflies are cool
20:53:36disrupteka dragonfly in the rv is like having a bird in a normal-sized room.
20:53:44Zevvhaha
20:53:46FromDiscord<Clyybber> haha
20:53:53ZevvI got an owl in the house once
20:53:59Zevvit was a small brand, but still
20:54:05Zevvit makes your living room look pretty small
20:54:07disruptekshit, that's a big bird.
20:54:18Zevvcame in through the chimney, with all the soot
20:54:19disruptekand speaking of, they leave quite a mess.
20:54:20disruptekwell, probably not owls.
20:54:33disruptekwell, maybe sooty owls.
20:54:37Zevvthey do :)
20:55:54FromDiscord<Clyybber> at least you don't have weasels screeching in front of your window at 3 am
20:56:05FromDiscord<Clyybber> they make a horrifying sound
20:56:11FromDiscord<Clyybber> when they fight
20:56:22disrupteki have a pack of coywolves in the woods.
20:56:30FromDiscord<Clyybber> oh, thats better
20:56:31FromDiscord<Clyybber> lol
20:56:42disruptekcan't let bentley out at night, not that he'd go after them. the pussy.
20:57:01FromDiscord<Clyybber> smol boye
20:57:15disruptekfoxes sound pretty crazy. what do weasels sound like?
20:57:32FromDiscord<Clyybber> like those typical woman horror movie screams
20:57:40disruptekawesome.
20:57:46FromDiscord<Clyybber> just a bit less human
20:57:55Zevvnever heard them. did hear the cursing of my dad though when they chewed through the wiring of his car
20:58:17FromDiscord<Clyybber> oh yeah
20:58:18Zevvhm or were that ferrets, can't remember
20:58:53FromDiscord<Varriount> 🦊
20:58:57FromDiscord<Clyybber> marder
20:59:01Zevvyeah those
20:59:59Zevv"According to Gesamtverband der Deutschen Versicherungswirtschaft, a German auto insurance company, martens account for tens of millions of dollars cumulatively in damage to car cables each year."
21:00:04FromDiscord<Clyybber> weasels are a kind of marder
21:00:38FromDiscord<Clyybber> or whatever you call it in english
21:00:46FromDiscord<Clyybber> the dutch understand
21:00:51FromDiscord<Varriount> Why do they chew car cables? Are they environmental activists?
21:01:16FromDiscord<tomck> sent a long message, see http://ix.io/2uKC
21:01:19Zevv"Their territorial nature is the most likely explanation, where martens feel the need to exhibit dominance by marking objects, including cars, within their territorial area.
21:01:34Zevvi.e. "because they can"
21:01:37Zevvfuckers by nature
21:01:43FromDiscord<Clyybber> they also like the smell
21:01:53FromDiscord<Clyybber> and the warmth of the cars inside
21:02:07ZevvI do like the warmth of the cars inside
21:02:20Zevvnot in summer tho. no airco :(
21:02:20FromDiscord<Clyybber> oh, actually, that seems to be debunked
21:02:33disruptektomck: yeah, it's not the same as a weak ref. it's the closest thing arc has at the moment.
21:02:36FromDiscord<Clyybber> so your explanation is the right one
21:02:57disruptek!repo badresults
21:02:59disbothttps://github.com/disruptek/badresults -- 9badresults: 11like results but worse 15 0⭐ 0🍴
21:03:02disruptek!repo nimresults
21:03:03disbotno results 😢
21:03:09disruptek!repo results
21:03:10disbothttps://github.com/xyb/googlesearch.nim -- 9googlesearch.nim: 11Nim library for scraping google search results 15 4⭐ 0🍴 7& 8 more...
21:03:10FromDiscord<Clyybber> nice sales pitch
21:03:28disruptekthere's another results lib that you may prefer.
21:03:29FromDiscord<Varriount> Zevv: Leave the windows slightly down in the summer. It keeps the car cool
21:03:34FromDiscord<Clyybber> you searching for arnes one?
21:03:44disruptekyeah.
21:04:04FromDiscord<tomck> Less about having results, i can implement those myself, but if it's not propogated through the stdlib/community libs they're basically useless (or at least very very awkward)
21:04:22disruptektomck: if key in collection: ... is the same as if collection.contains key: is the same as if collection.hasKey key:
21:04:34ZevvVarriount: oh sure, having no arco made me learn a lot of tricks. I also have these rad 70's beads seats that totally match the beige/brown interior of my car
21:05:14FromDiscord<tomck> Yeah but that'll do two lookups - unless the optimiser is aware of Table? In which case, it'll do two lookups with any *other* data structures that aren't a Table
21:05:38FromDiscord<tomck> ```↵if key in myTable:↵ use(myTable[key]) ## Second lookup↵```
21:06:06disruptekthen you /do/ want getOrDefault or whatever.
21:06:37disruptekthere is a convention of containsOrIncl on sets which does what you want, most like.
21:07:17disruptekanyway, it's a fair criticism and the stdlib is always the weak link.
21:07:30Zevvis there not some kind of block that only runs with the key of found?
21:07:41FromDiscord<tomck> Is the stdlib up for change? or is nim too stable now?
21:07:55disruptekthe stdlib is always up for change.
21:08:01FromDiscord<tomck> The *big* issue is lack of good iterators imo, rust seems to handle them better
21:08:03Zevvtomck: new additions are frowned upon, changes and fixes are cool
21:08:30Zevvpersonally I find the Tables api already beyond hilarious
21:08:47disruptekit's pretty annoying that add was deprecated.
21:08:59FromDiscord<tomck> add is deprecated?
21:09:08disruptekin favor of []=, yes.
21:09:13Zevvwe have one hundred and thirteen different exporetd procs in the Table module
21:09:20disruptekwhatfer catering to people that cannot read docs.
21:09:28Zevvtomck: there was too much confusion over the ability of tables to store more then one value for a key
21:09:42FromDiscord<Elegant Beef> Yea i didint realize add was deprecated and found it weird that it was
21:10:02FromDiscord<Elegant Beef> but this is weird behaviour for add anywho↵> Puts a new (key, value) pair into t even if t[key] already exists.
21:10:08disruptekshoulda just made a new dupe-free table that was more optimal.
21:10:11FromDiscord<tomck> not deprecated in the online docs, any other big deprecated things lol?
21:10:16*PMunch quit (Quit: leaving)
21:10:22disruptekit's deprecated in devel.
21:10:36disruptekprobably won't be removed before 1.6 at the earliest.
21:10:56FromDiscord<tomck> oh i see, fair
21:11:18Zevvtomck: I'm not familiar with rust iterators, what's Nim missing you think?
21:12:02FromDiscord<tomck> rust doesn't have inline iter, it can just optimised closure iters very well ( i think?)
21:12:05disrupteki think he means first-class iterator operations like python.
21:12:07*Vladar quit (Quit: Leaving)
21:12:10FromDiscord<tomck> yes ^
21:12:19FromDiscord<tomck> wait, closure iters aren't allocated right?
21:12:44disruptekwhat do you mean?
21:12:58FromDiscord<tomck> they can just be normal values too? (so in theory they *could* be optimised as well as rust's iters?)
21:13:16FromDiscord<tomck> wait, are closure iters even first class?
21:13:37disrupteki think so.
21:13:44Zevv1.5
21:13:45FromDiscord<tomck> i've never used closure iters, i don't know if there's some weird allocations / function pointers happening
21:14:15Zevvwell they kind of are what the name implies. they are heavier then the skinny iterators
21:14:37FromDiscord<tomck> interesting, can you `.map` a closure iterator? or is it a completely dynamic thing?
21:15:00disruptekit's weird in that you get a ptr to the env, but that's hardly surprising.
21:15:05Zevvyou need to toSeq them first, so they're not kind of 1.5 class indeed
21:15:13disruptekwut
21:15:25Zevvno?
21:15:28disruptekno.
21:15:32Zevvno!
21:15:34disruptekno!
21:15:42FromDiscord<Clyybber> sey
21:15:44disruptekyou can map closure iterators.
21:15:48FromDiscord<Recruit_main707> yesn't
21:15:57disrupteki mean, you can pass them around and apply funcs and whatnot.
21:16:17disruptekbut they aren't as full-featured as, say, python's generators.
21:16:38disruptekand they don't work at compile-time, which is kinda a big deal for nimions.
21:17:14ZevvI feel some cps coming up
21:17:22disrupteklol yeah
21:17:24disruptek!repo cps
21:17:24FromDiscord<tomck> hmmm so i'm going from this atm, not sure if it's outdated https://nim-by-example.github.io/for_iterators/↵Just looks like there's a generic `iterator()` type, which seems to imply there's some kind of dynamic function pointer stuff going on?
21:17:24disbothttps://github.com/disruptek/cps -- 9cps: 11Continuation-Passing Style for Nim 🔗 15 29⭐ 1🍴
21:17:32disruptekthis is our solution to the problem.
21:17:59Zevvif only clyybber would not sleep so much. like. every day.
21:18:01Zevvwaste of time
21:18:13FromDiscord<tomck> wait they don't work at compile time?
21:18:30FromDiscord<Clyybber> truly
21:18:33disruptekclosure iterators do not, no.
21:18:52FromDiscord<Clyybber> once you find out you don't need sleep
21:19:03FromDiscord<tomck> why? aren't they just a struct?
21:19:09FromDiscord<Clyybber> your mind goes wiiide
21:19:17Zevvdude. I never knew that
21:19:20ZevvTIL
21:19:26Zevvbet I'm not a nimion then
21:19:27FromDiscord<Recruit_main707> till it explodes
21:19:34disrupteki don't know enough about the front-end to say why this limitation exists.
21:19:46disruptekweirdly, i'm learning the compiler back-to-front.
21:20:02FromDiscord<Clyybber> https://www.youtube.com/watch?v=jANThtZJGrs
21:20:20FromDiscord<Clyybber> disruptek: The limitation exists because nobody implemented it
21:20:22FromDiscord<Clyybber> :P
21:20:29disruptek~features
21:20:29disbotfeatures: 11begin in the unfortunate state of /unimplemented/ -- disruptek
21:20:54FromDiscord<Recruit_main707> this is extremely fitting lol, how many times have you used it?
21:21:32Zevv"sorry state"
21:21:33disruptekit's one of my favorite quotes by araq.
21:21:45Zevvmine is "I don't know you"
21:21:52disruptek~araq
21:21:53disbotAraq: 11👑Andreas "What Else?" Rumpf👑 -- disruptek
21:21:56FromDiscord<Recruit_main707> -- disruptek 🤔
21:22:02Zevvdon't do that, you'll wake him
21:22:10FromDiscord<tomck> so if i use closure iters, is it going to dynamic dispatch something, or call a function pointer?
21:22:23disruptekcall a function pointer.
21:22:41*Jesin joined #nim
21:22:45FromDiscord<tomck> hmmmmm i see
21:23:00disruptekbut cps gives you the best of both worlds.
21:23:16FromDiscord<tomck> i have no idea what a continuation is lol, i'm not that deep yet
21:23:31disrupteki thought you came from a functional lang.
21:23:42Zevvit's just a function pointer :)
21:23:53FromDiscord<Clyybber> \> mine is "I don't know you" hahahaha
21:23:59FromDiscord<tomck> Errr not so much
21:24:07FromDiscord<Clyybber> hey Zevv whats up
21:24:19FromDiscord<tomck> i do a bit of haskell, but don't know the lingo
21:24:21Zevvhehe
21:24:31disrupteka continuation is a fn* as well as the data it is passed.
21:24:43disruptekinside that data is the fn*. 😉
21:25:51FromDiscord<tomck> ohhhh are these just futures / promises?
21:25:58FromDiscord<tomck> maybe at a lower level?
21:26:05disrupteklower level, yes.
21:26:29FromDiscord<tomck> is this all static?
21:26:53disrupteki think the answer is yes but i'm not sure what you're asking.
21:27:32FromDiscord<tomck> Wait i haven't connected how these relate to first-class iterators
21:27:34disruptekit's static in that it's typed, but dispatch itself is dynamic. however, the model can be statically analyzed to assert that it's /safe/ for some measure of /safety/.
21:27:44FromDiscord<tomck> i see
21:28:24FromDiscord<tomck> does nim have something similar to typeclasses? So that I can implement a proc on all types which satisfy some typeclass?
21:28:29disrupteksure.
21:28:44FromDiscord<tomck> how do i go about doing that
21:28:46disruptekwe also have concepts, which are aka interfaces in other languages.
21:28:51FromDiscord<Clyybber> proc(a: int or float)
21:29:05disruptek~concepts
21:29:05FromDiscord<tomck> errr wait let me explain more
21:29:05disbotconcepts: 11user-defined type classes; documented in the experimental section of the manual: https://nim-lang.org/docs/manual_experimental.html#concepts -- disruptek
21:29:39FromDiscord<tomck> ahhhhhh yes i think this is what i want - are concepts static dispatch?
21:29:46disruptekyes.
21:30:05FromDiscord<tomck> amazing! when're these getting stabilised?
21:30:15FromDiscord<Clyybber> they basically are already
21:30:20FromDiscord<tomck> Holy shit these are way more powerful than typeclasses i think
21:30:20disruptekthey haven't changed in a couple years.
21:30:41disrupteki'm starting to use them because i'm tired of waiting for them to "stabilize". they will be stable enough.
21:30:45FromDiscord<Clyybber> tomck: They are quite literally infinitely powerful
21:30:59*audiofile quit (Ping timeout: 240 seconds)
21:31:11FromDiscord<Clyybber> because you can put all kinds of code into their body
21:31:51disruptekconcepts * generics * statics do produce a pretty small head of the needle, however.
21:32:38FromDiscord<Varriount> It's only a shame that vtrefs never materialized
21:33:20FromDiscord<tomck> well this is mind blowing stuff
21:33:31FromDiscord<tomck> i shudder to consider the compile time lol
21:33:45FromDiscord<Elegant Beef> But sadly they arent runtime 😄
21:33:48Zevvhttps://www.lightningmaps.org/?lang=nl oooh coming straight at me, amazingly spooky having this much lightning on the horizon, but not hearing a thing yet
21:34:13FromDiscord<Clyybber> nice!
21:34:36*eery joined #nim
21:35:04disruptekthe compile-time isn't too bad.
21:35:28disrupteknim compile-time is comparable to python's runtime.
21:35:50FromDiscord<Clyybber> heh, 2 lines now
21:37:49FromDiscord<tomck> This is great! I had no idea concepts were a thing
21:37:54FromDiscord<tomck> time to ruin my codebase
21:38:02FromDiscord<Clyybber> heh, to elevate you mean
21:38:09FromDiscord<Clyybber> they really are a great *concept*
21:38:41disruptekthey're a little rough around the edges, but they work.
21:39:42FromDiscord<Clyybber> theres one thing you should know, concept refinement is in the manual, but not yet implemented
21:39:51FromDiscord<tomck> What's refinement?
21:40:10FromDiscord<tomck> oh like, 'this concept but with X too'?
21:40:36FromDiscord<Clyybber> yeah, basically inheritance/composition
21:41:07FromDiscord<tomck> sent a code paste, see https://play.nim-lang.org/#ix=2uKH
21:41:16FromDiscord<Clyybber> yep
21:42:02FromDiscord<Clyybber> you don't need the "hasFieldCalled"
21:42:04FromDiscord<tomck> do any other langs have concepts? (other than c++20)
21:42:17FromDiscord<Clyybber> probably, there are many langs in the sea
21:42:21FromDiscord<tomck> oh yeah i see, makes sense
21:42:47FromDiscord<tomck> any other 'legit' langs, that have a reasonable community
21:42:59FromDiscord<Clyybber> not that I'm aware of
21:43:21FromDiscord<tomck> cool! i'm surprised the uber-functional people haven't been flooding in
21:43:33FromDiscord<Clyybber> heh, they have been
21:43:42FromDiscord<Clyybber> even going as far as to request a subset of nim
21:43:50FromDiscord<Clyybber> that disallows procs and var
21:44:09FromDiscord<Clyybber> but we told them they can write a macro that disallows that stuff
21:44:16FromDiscord<tomck> lol
21:44:33disrupteknewbies gonna newb.
21:45:20FromDiscord<tomck> this is cool, i never hear about nim anymore, does the GC push people away? this seems like a way better version of most of the current systems langs
21:45:23*solitudesf quit (Ping timeout: 265 seconds)
21:45:37FromDiscord<Clyybber> nim doesn't have coroporate backing
21:46:00disruptekit's ridiculous how many people are turned off by the fact that nim supports GC. and, it's like, you don't even have to use it.
21:46:09disruptekand there are several to choose from.
21:46:11disruptekand none of them suck.
21:46:18FromDiscord<Clyybber> its also ridiculous how many people think they need a gc
21:46:55disrupteki see the code these people write.
21:46:59disruptekc'mon, man.
21:47:01FromDiscord<tomck> i think it's the same problem that D had, where (afaik) the whole stdlib relied on the GC
21:47:12FromDiscord<tomck> that's what put me off all these years
21:47:34disrupteki'm surprised how competitive D and zig are.
21:47:37FromDiscord<tomck> ngl i still don't use the arc annotations, i don't understand how they work
21:47:46FromDiscord<Clyybber> you don't have to
21:47:58FromDiscord<Clyybber> sink can be inferred
21:48:01FromDiscord<tomck> Zig feels like it's growing huge, i thought D was dead now (?)
21:48:29FromDiscord<tomck> yeah, but i don't like not knowing lol
21:48:38FromDiscord<Clyybber> heh, read destructors.md
21:48:40FromDiscord<tomck> It's mainly `lent`
21:48:47FromDiscord<Clyybber> lent is like a view
21:48:54FromDiscord<Clyybber> but its not finished yet
21:49:00FromDiscord<tomck> i have done, what about this part: `lent T is like var T a hidden pointer. It is proven by the compiler that the pointer does not outlive its origin`
21:49:08FromDiscord<tomck> without lifetime annots i'm not even sure that's possible
21:49:25*lritter quit (Ping timeout: 264 seconds)
21:49:25FromDiscord<Clyybber> we will inferr that stuff
21:49:35FromDiscord<Clyybber> you don't need to explicitly annotate lifetimes
21:50:07FromDiscord<Clyybber> also because a lent is not guaranteed to stay a lent
21:50:10FromDiscord<Clyybber> its an optimization
21:50:16FromDiscord<tomck> sent a code paste, see https://play.nim-lang.org/#ix=2uKJ
21:50:26FromDiscord<tomck> Assuming `proc []` returns a `lent int` here
21:50:34FromDiscord<Clyybber> if you would do something that would outlive the lent we may copy
21:50:41FromDiscord<tomck> ohhhhh
21:50:50FromDiscord<Clyybber> but thats not set in stone
21:50:57FromDiscord<Clyybber> we are still designing it
21:51:03FromDiscord<tomck> ah i see, ok
21:51:11FromDiscord<Clyybber> but you can use it already
21:51:23FromDiscord<tomck> If there was a `lent {.nocopy.}` that'd work for me
21:51:27FromDiscord<Clyybber> it will just get less restrictive in the future
21:52:11FromDiscord<Clyybber> tomck: You can make your `=` {.error.} and that will prohibit all copies
21:52:37FromDiscord<tomck> eh?
21:52:46FromDiscord<Clyybber> but yeah, something like a lentNoCopy thing might be useful
21:53:19FromDiscord<tomck> ohhhhh, so the lent version of `[]` is implemented like↵```↵proc `[]`[T](x: lent seq[T]): lent T = ...↵```
21:53:22FromDiscord<tomck> the seq itself is lent
21:53:41FromDiscord<tomck> so if you just make the non-lent version error, it'll show you at compiletime?
21:53:54FromDiscord<Clyybber> it should yeah
21:54:01FromDiscord<Clyybber> it behaves like var currently
21:54:07FromDiscord<Clyybber> except that its not mutable
21:54:08FromDiscord<Clyybber> basically
21:54:27FromDiscord<Clyybber> currently cursors are this COW lent thing
21:54:35FromDiscord<Clyybber> IMO lent/view/cursor its all the same thing
21:54:47FromDiscord<Clyybber> I plan on unifying it
21:55:43FromDiscord<tomck> so if i have a `seq[T]`, can i get a `lent T`?
21:55:52FromDiscord<Clyybber> yeah
21:55:54FromDiscord<tomck> or do i need to explicitly 'borrow' the `set[T]`?
21:56:00FromDiscord<tomck> (edit) '`set[T]`?' => '`seq[T]`?'
21:56:01FromDiscord<tomck> how?
21:56:03FromDiscord<Clyybber> lent is a borrow
21:56:41FromDiscord<tomck> oh so nim will just try and call the `lent` version of the proc if possible, then if it decides it can't (via magic) it'll call the non-lent version?
21:57:05FromDiscord<Clyybber> I don't think that it does currently
21:58:17FromDiscord<Clyybber> yeah, currently it won't
21:58:30FromDiscord<tomck> currently it won't do what, call the lent version, or magically call the non-lent version?
21:58:39FromDiscord<Clyybber> magically the non-lent
21:59:04FromDiscord<tomck> ah ok, so it will segfault currently, but semantically it won't - just pending compiler changes?
21:59:11FromDiscord<Clyybber> no it won't segfault
21:59:15FromDiscord<Clyybber> it will error at compiletime
21:59:26FromDiscord<tomck> ohhh if it can't do the copy? i see
22:00:05FromDiscord<tomck> that's actually great, you should pop in a compiler flag to turn off auto-copy project wide, i wouldn't be touching that lol
22:01:15FromDiscord<tomck> and you could explicitly copy with .deepCopy to get around the compiler error i assume?
22:05:02disruptek#13804
22:05:04disbothttps://github.com/nim-lang/Nim/issues/13804 -- 3_ (underscore) is gensym'd inside templates, therefore gives unused warning ; snippet at 12https://play.nim-lang.org/#ix=2uKO
22:05:06disruptekthis again
22:05:20FromDiscord<Clyybber> oh, yeah that ones been on my easy-fix lsit
22:05:31FromDiscord<Clyybber> but I'm so stuck with #5961
22:05:31disbothttps://github.com/nim-lang/Nim/pull/5961 -- 5'newEnum' proc
22:05:35FromDiscord<Clyybber> not stuck
22:05:38FromDiscord<Clyybber> and not that one
22:05:39FromDiscord<Clyybber> lol
22:05:45disruptekyeah, i know.
22:05:51FromDiscord<Clyybber> well, tomorrow will enlighten me hopefully
22:06:06FromDiscord<Clyybber> sleep/awake well people
22:06:09FromDiscord<Clyybber> cee ya
22:06:12disruptekgn
22:15:00*tane quit (Quit: Leaving)
22:15:27*arecacea1 quit (Remote host closed the connection)
22:15:51*arecacea1 joined #nim
22:16:58*thomasross joined #nim
22:17:29FromDiscord<brainproxy> is there a nim equiv of golang's `image`, `image/draw` packages?
22:18:13disruptekhard to say without being familiar with those packages.
22:21:38FromDiscord<brainproxy> this seems promising: https://github.com/Ethosa/nimage
22:22:03disruptek!repo imageman
22:22:04disbothttps://github.com/SolitudeSF/imageman -- 9imageman: 11Image manipulation library 15 41⭐ 3🍴
22:26:33FromDiscord<stisa> https://github.com/treeform/flippy too maybe?
22:34:34shashlicklocal deps only needs readme updates, then it's a go
22:36:49*abm quit (Quit: Leaving)
22:36:51*zyklon joined #nim
22:51:49FromDiscord<hobbledehoy> would anyone know why dochack.js is not getting outputted with `nim doc`?
23:13:14FromDiscord<hobbledehoy> found the issue, 1.2.6 doesn't output, switching to devel fixed it
23:13:36disruptekshashlick: nice job, dude.
23:18:19*selk joined #nim
23:19:30*arecacea1 quit (Remote host closed the connection)
23:19:49*arecacea1 joined #nim
23:22:47shashlickhttps://github.com/nim-lang/nimble/blob/0dc46e92d33fa6ed6e8e52de4dc75cf99947fb95/readme.markdown#nimbles-folder-structure-and-packages
23:25:53*endragor joined #nim
23:26:32*rockcavera quit (Remote host closed the connection)
23:30:06*Jesin quit (Quit: Leaving)
23:34:23*opal quit (Ping timeout: 240 seconds)
23:39:34disruptekwhy can't you just put --nimblePath="nimbledeps" into your nim.cfg so the compiler picks them all up?
23:39:49disrupteknimble should be allowed to clobber that file, right?
23:40:11disruptekshashlick ^
23:40:46*endragor quit (Ping timeout: 258 seconds)
23:42:53*rockcavera joined #nim
23:43:15*krux02_ joined #nim
23:45:23FromDiscord<dom96> pretty sure that's the plan for the future
23:46:08disruptekthe text makes it sound like nimble doesn't support that, but afaik it does.
23:46:17*krux02 quit (Ping timeout: 260 seconds)
23:51:36disruptekyou know the worst mistake i made with nimph?
23:51:51disrupteknot using AbsoluteDir and AbsoluteFile from the start.
23:53:37*awe002 quit (Ping timeout: 264 seconds)
23:55:36*hnOsmium0001 quit (Ping timeout: 246 seconds)
23:56:35*endragor joined #nim
23:59:58*hnOsmium0001 joined #nim