<< 13-05-2020 >>

---

00:11:01	*	konvertex quit (Ping timeout: 264 seconds)
00:24:31	disruptek	~nimcache
00:24:31	disbot	nimcache: 11the best way to figure out where it is located is to specify that location with --nimcache -- disruptek
00:24:40	Prestige	Made a little devlog about the WM I'm making in nim: https://www.youtube.com/watch?v=cD7xBqEdqgo
00:24:46	Prestige	pretty topical
00:30:29	disruptek	neat.
00:30:39	disruptek	no wayland support, right?
00:30:40	*	zacharycarter quit (Ping timeout: 272 seconds)
00:30:44	*	gokr quit (Quit: Leaving.)
00:30:48	Prestige	Right, just xlib for now
00:30:59	Prestige	when wayland gets wider adoption I'll port to wayland
00:35:32	disruptek	the thing that would attract nimions is some kinda /really/ easy client api.
00:37:04	Prestige	client api?
00:37:47	disruptek	some way for me to control it from code.
00:37:52	*	silvernode joined #nim
00:38:14	disruptek	here's my current hack:
00:38:15	Prestige	I was thinking of setting up IPC but then you could use any language
00:38:17	disruptek	!repo swayipc
00:38:18	disbot	https://github.com/disruptek/swayipc -- 9swayipc: 11swayipc (i3ipc) for Nim 15 4⭐ 0🍴
00:42:02	Prestige	Might be something nice to have for v1.0
00:46:14	*	lritter quit (Ping timeout: 240 seconds)
00:47:14	*	lritter joined #nim
00:53:23	FromDiscord	<exelotl> Man, I'm really surprised there's not something that's just like strformat but with loops and conditionals
00:54:39	disruptek	sounds easy.
00:54:52	*	lritter quit (Quit: Leaving)
00:54:59	FromDiscord	<exelotl> Yeah
00:55:56	*	ryan_ is now known as honger
00:57:06	*	zacharycarter joined #nim
00:58:46	*	silvernode quit (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.)
00:59:57	FromDiscord	<exelotl> 'nim-templates' seems close but didn't work for me. Generated code contained errors, maybe I was misusing it somehow
01:01:40	*	zacharycarter quit (Ping timeout: 246 seconds)
01:02:02	FromDiscord	<exelotl> nim-jade also seems to give good output but it's 6 years old so I don't have high hopes for it compiling at all xD
01:02:16	FromDiscord	<exelotl> Plus I don't want to use jade
01:04:30	FromDiscord	<exelotl> The other things I found are trying to be too clever. Either some DSL or requiring you to create context objects and stuff, doing extra work at runtime
01:15:58	*	poohman joined #nim
01:38:22	*	chemist69 quit (Ping timeout: 260 seconds)
01:40:06	*	chemist69 joined #nim
01:58:53	*	bok joined #nim
02:13:41	*	muffindrake quit (Ping timeout: 246 seconds)
02:16:11	*	muffindrake joined #nim
02:20:52	*	dadada quit (Ping timeout: 272 seconds)
02:27:00	*	dadada joined #nim
02:27:23	*	dadada is now known as Guest24748
02:51:00	*	Guest24748 quit (Ping timeout: 256 seconds)
02:57:00	*	dadada joined #nim
02:57:23	*	dadada is now known as Guest5361
03:03:30	*	zacharycarter joined #nim
03:07:54	*	zacharycarter quit (Ping timeout: 240 seconds)
03:21:15	*	Guest5361 quit (Ping timeout: 258 seconds)
03:27:06	*	dadada__ joined #nim
03:33:50	FromDiscord	<Varriount> timotheecour: I really like your recent PRs relating to tests.
03:34:37	FromDiscord	<Never Listen To Beef> So this is going to be a silly question for a ugly implementation, but can case not be used in a similar matter to if for ternaries?
03:34:50	FromDiscord	<Never Listen To Beef> Where non discarded values are the output logic
03:34:56	FromDiscord	<Varriount> exelotl: What about source filters?
03:34:58	FromDiscord	<Never Listen To Beef> are outputted*
03:35:40	FromDiscord	<Varriount> @Elegant Beef what do you mean "not be used"?
03:35:59	FromDiscord	<Never Listen To Beef> i mean can case be used similar to how we do ternary logic in nim
03:36:00	FromDiscord	<Varriount> You can not use case for a lot of things
03:36:25	FromDiscord	<Never Listen To Beef> I'm figuring not, since i couldnt get it to work
03:36:34	FromDiscord	<Never Listen To Beef> And im good at bodging stuff
03:37:30	FromDiscord	<Varriount> You might have to use a template to get the compiler to accept it.
03:37:48	FromDiscord	<Never Listen To Beef> Yea it's just to remove a bit of reptetitive logic
03:37:52	FromDiscord	<Never Listen To Beef> So it's not really worth it
03:48:58	*	ftsf_ joined #nim
03:49:26	*	ftsf quit (Ping timeout: 256 seconds)
03:51:06	*	dadada__ quit (Ping timeout: 260 seconds)
03:55:43	disruptek	king: please drop a link to your project.
03:56:59	*	dadada joined #nim
03:57:23	*	dadada is now known as Guest47494
03:57:29	FromDiscord	<Yardanico> https://github.com/cnlohr/rawdrawandroid
03:58:56	FromDiscord	<Yardanico> Although this was already kinda possible with jnim and "android" package
04:00:15	leorize	@Elegant Beef: uhh... case expressions exists?
04:00:20	leorize	and yes they are in the manual also
04:01:46	*	nsf joined #nim
04:02:00	disruptek	~news
04:02:00	disbot	news: 11the #nim-news channel has a Nim news feed of updates to pull requests, issues, and packages. -- disruptek
04:02:50	*	thomasross quit (Ping timeout: 272 seconds)
04:05:48	*	rockcavera quit (Remote host closed the connection)
04:06:02	*	supakeen quit (Quit: WeeChat 1.9.1)
04:06:40	*	supakeen joined #nim
04:21:06	*	Guest47494 quit (Ping timeout: 256 seconds)
04:21:38	leorize	disruptek: merged indent-rewrite to master
04:21:50	leorize	please switch to it :)
04:22:50	*	dddddd quit (Ping timeout: 260 seconds)
04:27:02	*	dadada__ joined #nim
04:31:32	FromDiscord	<Technisha Circuit> I've tried reading up on how to use C++ libs in Nim but I can't figure it out, how does it work exactly?
04:31:47	FromDiscord	<Yardanico> You use C++ backend
04:31:50	FromDiscord	<Yardanico> That's how it works
04:31:57	*	ftsf joined #nim
04:32:08	FromDiscord	<Yardanico> "nim cpp"
04:32:51	FromDiscord	<Technisha Circuit> Oh :P
04:33:28	FromDiscord	<Yardanico> Although there's another way - to write C bindings for a C++ library yourself or maybe someone else has done that already
04:33:33	FromDiscord	<Technisha Circuit> By using the C++ backend, does that break any existing Nim modules?
04:33:37	FromDiscord	<Technisha Circuit> Oh?
04:33:42	*	ftsf_ quit (Ping timeout: 258 seconds)
04:33:48	FromDiscord	<Yardanico> Well C++ backend is pretty stable
04:33:49	FromDiscord	<Technisha Circuit> I was looking at import cpp
04:33:55	FromDiscord	<Technisha Circuit> So it should be fine?
04:34:04	FromDiscord	<Yardanico> Yeah
04:34:53	FromDiscord	<Yardanico> And about another way - for example there's SFML in C++ and C bindings to it "CSFML", there's imgui and C bindings "cimgui"
04:35:06	FromDiscord	<Technisha Circuit> Oh nice
04:35:10	FromDiscord	<Yardanico> https://github.com/Extrawurst/cimgui
04:35:13	FromDiscord	<Technisha Circuit> Thanks
04:36:13	FromDiscord	<Technisha Circuit> How would i do it with Importcpp? I want to use Sleepy discord just so i can test Cpp libraries in Nim
04:36:51	*	Lord_Nightmare quit (Quit: ZNC - http://znc.in)
04:37:43	FromDiscord	<Yardanico> https://nim-lang.org/docs/manual.html#implementation-specific-pragmas-importcpp-pragma
04:37:47	FromGitter	<kaushalmodi> Here's one simple example of binding Nim with C++ lib: https://github.com/kaushalmodi/std_vector/blob/master/src/std_vector.nim
04:38:10	*	Lord_Nightmare joined #nim
04:39:49	FromGitter	<kaushalmodi> here's the crux: ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5ebb7a152baa19185466d194]
04:50:50	*	dadada__ quit (Ping timeout: 260 seconds)
04:57:08	*	dadada joined #nim
04:57:31	*	dadada is now known as Guest83621
05:00:53	FromDiscord	<Benumbed> false != False :/
05:01:36	FromDiscord	<Benumbed> I had a real moment trying to figure out why `False` was not defined
05:03:38	Yardanico	nim has partial case insensitivity
05:03:49	Yardanico	"partial" <- everything except first letter (because of types)
05:03:59	Yardanico	if you really want to have False you can just do ""const False = false"
05:04:04	FromDiscord	<Benumbed> Oh I know
05:04:06	FromDiscord	<Benumbed> I was just griping
05:04:47	FromDiscord	<Benumbed> partial case sensitivity doesn't bite me much, but when it does it gives me a bit of pause trying to figure out WTF happened 🙂
05:13:16	leorize	I'm adding mixed highlighting to nim.nvim (ie. highlight C code in emit as C)
05:14:00	leorize	question is: should I highlight only the C code and not the various Nim string escape characters?
05:14:57	*	waleee-cl quit (Quit: Connection closed for inactivity)
05:21:14	*	Guest83621 quit (Ping timeout: 256 seconds)
05:24:13	FromDiscord	<Technisha Circuit> Thanks!
05:26:41	*	narimiran joined #nim
05:26:57	*	dadada__ joined #nim
05:29:57	FromDiscord	<Technisha Circuit> How hard would it be to make the Nim JS backend use JavaScript's IO?
05:30:13	FromDiscord	<Technisha Circuit> So console.log instead of raw echo
05:30:19	Yardanico	???
05:30:26	Yardanico	-d:nodejs for some simple IO
05:30:31	Yardanico	no stuff like stdio/stdout though
05:30:36	FromDiscord	<Technisha Circuit> Oh okay, thanks :p
05:30:39	FromDiscord	<Technisha Circuit> Oh okay
05:30:39	Yardanico	because nim's JS backend is specifically targeted at frontend
05:31:12	FromDiscord	<Technisha Circuit> Oh okay
05:31:16	FromDiscord	<Technisha Circuit> For the web then
05:44:46	*	leorize[m] uploaded an image: 1589348633_grim.png (23KB) < https://kde.modular.im/_matrix/media/r0/download/asra.gr/LLZXumrEYWvRYVvfVPAvClSY >
05:45:00	leorize	narimiran, disruptek: new nim.nvim feature ^ anyone interested?
05:45:43	leorize	this basically let you choose what language to highlight in `emit`
05:45:52	leorize	by default I set it to C
05:46:05	leorize	but you can customize it to JS or whatever as wanted
05:46:51	leorize	also feedback wanted :P
05:47:01	leorize	when should I apply this kind of highlighting?
05:47:08	leorize	only for trible quoted literals?
05:47:10	leorize	triple*
05:51:04	Yardanico	does "mod" in Nim behave the same as "%" in Python? I'm trying to port a simple math script for (fast) decomposition of a number into two prime cofactors (for telegram mtproto), but my Nim version doesn't really work :D
05:51:08	*	dadada__ quit (Ping timeout: 272 seconds)
05:51:57	Yardanico	python version - https://play.nim-lang.org/#ix=2lR6
05:52:00	Yardanico	my attempt at nim version - https://play.nim-lang.org/#ix=2lR7
05:54:19	Yardanico	basically I need to find the biggest prime cofactor of a number and then to find the second one I can just divide the number by the first cofactor
05:57:00	*	dadada joined #nim
05:57:23	*	dadada is now known as Guest89109
05:58:06	Yardanico	I mean I can do the naive algo but it'll be quite slow :P
05:58:55	narimiran	leorize: oh, nice! i personally don't use emit, but nice feature
05:59:01	narimiran	Yardanico: it is not the same as python's %
05:59:05	Yardanico	uh
05:59:11	Yardanico	that's bad :P
05:59:25	narimiran	wait a bit, there's a python-like version, let me find it
05:59:27	Yardanico	maybe I should look into nimpylib
05:59:37	narimiran	(the difference is for negative numbers)
05:59:39	Yardanico	https://github.com/Yardanico/nimpylib/blob/master/src/pylib/ops.nim#L37
05:59:43	Yardanico	this one maybe
06:00:05	narimiran	https://nim-lang.github.io/Nim/math.html#floorMod,T,T
06:00:29	Yardanico	oh thanks a lot
06:01:45	Yardanico	hmm still doesn't correctly work, maybe I ported something the wrong way
06:01:53	narimiran	it might be that :)
06:05:08	narimiran	you sure it's not `random` producing different values for each version? :D
06:05:19	Yardanico	nah now I tested with same predefined numbers
06:05:25	Yardanico	and it's still broken for nim
06:05:29	Yardanico	I'll try porting again
06:05:48	Yardanico	I'll log all numbers to compare where the difference starts to appear
06:05:54	narimiran	try to put some more parentheses
06:06:02	narimiran	maybe the precedence is different
06:06:14	narimiran	e.g. here: `(y * y) mod N + c`
06:06:33	Yardanico	well I replaced it by floorMod anyway
06:06:36	Yardanico	which has parens
06:07:21	Yardanico	the biggest thing I'm afraid of is that Python version goes over uint64 somewhere :P
06:08:27	leorize	don't worry, python don't have arbirtary length precision int iirc
06:08:33	narimiran	it does
06:08:47	Yardanico	leorize: sadly it actually does :)
06:08:52	narimiran	so: worry :D
06:08:57	Yardanico	2**1000 works fine in python
06:09:01	Yardanico	(2^1000)
06:09:45	narimiran	`y * y` would be my first suspect
06:09:48	leorize	you gotta baing out stint then :P
06:10:18	Yardanico	narimiran: well that y is a "random" number, I can make it smaller
06:10:56	narimiran	i know, i'm just saying it's that place that i would be worried about overflowing
06:10:58	leorize	well 2 ^ 64 * 2 ^ 64 == 2 ^ 128
06:11:03	leorize	so stint[128] :P
06:12:05	Yardanico	this algo I'm porting is apparently Brent's modification of Pollard's rho algorithm
06:12:49	Yardanico	https://en.wikipedia.org/wiki/Pollard%27s_rho_algorithm
06:13:10	Yardanico	I'll try porting from wiki :D
06:16:41	Yardanico	oh wait it actually works, although it's slower than this python script, but I'll improve it later probably
06:18:26	Yardanico	or I'll try to implement brent's algo from the paper https://maths-people.anu.edu.au/~brent/pd/rpb051i.pdf
06:20:52	*	Guest89109 quit (Ping timeout: 246 seconds)
06:21:00	Yardanico	oh, seems like 93% of time is spent in gcd
06:21:03	Yardanico	makes sense
06:27:04	*	dadada__ joined #nim
06:32:24	*	PMunch joined #nim
06:39:36	*	gokr joined #nim
06:45:15	*	liblq-dev joined #nim
06:46:19	*	solitudesf joined #nim
06:46:44	FromGitter	<gogolxdong> @kaushalmodi what does std::vector<'*0>() mean?
06:46:55	Yardanico	it's explained in nim manual
06:47:03	Yardanico	https://nim-lang.org/docs/manual.html#implementation-specific-pragmas-importcpp-pragma
06:47:13	Yardanico	"An apostrophe ' followed by an integer i in the range 0..9 is replaced by the i'th parameter type."
06:50:46	*	dadada__ quit (Ping timeout: 260 seconds)
06:50:50	FromGitter	<gogolxdong> ah ,manual has been updated, ours is outdated.
06:51:19	Yardanico	ours?
06:51:47	Yardanico	and idk how old your manual is
06:51:48	FromGitter	<Willyboar> chinese edition
06:51:54	Yardanico	that thing was in nim manual in nim 0.15
06:51:58	Yardanico	https://nim-lang.org/0.15.0/manual.html#importcpp-pragma-importcpp-for-procs
06:52:07	Yardanico	heck even in 0.13 https://nim-lang.org/0.13.0/manual.html#importcpp-pragma-importcpp-for-procs
06:52:22	Yardanico	so it was probably just not translated :P
06:53:48	FromGitter	<gogolxdong> missed that part probably then. :)
06:56:59	*	dadada joined #nim
06:57:23	*	dadada is now known as Guest46880
06:58:26	FromGitter	<gogolxdong> nah, it's there, I've fogotten them, dear.
06:59:16	FromGitter	<gogolxdong> but it's true they are rarely used.
07:02:42	FromGitter	<bung87> translate js class to `ref object of T` is right choice?
07:05:54	Yardanico	if you don't need inheritance you can just have "ref object"
07:07:23	FromGitter	<bung87> ok, I think I also need inheritance
07:10:52	*	fredrikhr joined #nim
07:13:23	PMunch	JS has inheritance?
07:13:32	PMunch	Last I checked it doesn't
07:13:40	Yardanico	yes, kinda
07:13:52	PMunch	Hmm
07:13:59	Yardanico	https://developer.mozilla.org/en-US/docs/Learn/JavaScript/Objects/Inheritance
07:14:06	PMunch	I do remember that there were some ways to hack around it..
07:14:11	Yardanico	uhh I guess I give up on this prime factorization for now
07:14:21	Yardanico	i can't make my nim version work as fast as python and it doesn't even work right now :D
07:14:26	Yardanico	gonna post a question on the forumn
07:15:13	PMunch	Hmm, do we have RW locks in Nim?
07:17:50	PMunch	Hmm, so there are some in the "posix" module
07:17:55	PMunch	I guess I can use those
07:21:06	*	Guest46880 quit (Ping timeout: 260 seconds)
07:26:15	PMunch	Are there no cross-platform RW locks?
07:27:00	*	dadada__ joined #nim
07:33:09	FromDiscord	<mratsim> it's not a synchronization primitives
07:33:26	FromDiscord	<mratsim> fancier locks can go in a library
07:39:09	Yardanico	is there any nim library to use streams with sockets?
07:39:20	Yardanico	so I don't have to write in a stringstream, then read it and send the data as string? :P
07:50:04	FromGitter	<bung87> maybe you could implements a `pipe` api
07:51:07	*	dadada__ quit (Ping timeout: 260 seconds)
07:56:36	FromDiscord	<mratsim> @Yardanico, pretty sure we have that in either Chronos or FastStreams
07:57:00	*	dadada joined #nim
07:57:23	*	dadada is now known as Guest97932
07:59:10	leorize[m]	I did the impossible and implemented C highlighting for emit in nim.nvim: https://github.com/alaviss/nim.nvim/commit/dc868703122cfe39d52a0d192ab24b7cbcbbd878 :)
07:59:40	PMunch	Neat!
08:09:02	*	zacharycarter joined #nim
08:13:25	leorize	PMunch: you might want my indentation script now, I built the thing to perfection :P
08:14:30	*	zacharycarter quit (Ping timeout: 258 seconds)
08:14:49	leorize	narimiran, liblq-dev, Zevv, disruptek: please switch back to master, I've merged indent-rewrite there and added some small goodies
08:14:59	narimiran	alrighty
08:15:15	leorize	please do report any issue with indentation and/or highlighting
08:15:26	leorize	now I'm gonna go to bed
08:16:12	PMunch	leorize, huh?
08:16:16	PMunch	Indentation script?
08:16:25	leorize	the auto indentation script for vim
08:16:29	PMunch	Ooh
08:17:03	PMunch	And where can I find that?
08:17:16	leorize[m]	https://github.com/alaviss/nim.nvim/blob/master/indent/nim.vim
08:17:24	PMunch	Thanks :)
08:17:37	*	konvertex joined #nim
08:18:00	leorize	depending on your plugin, you might need to tweak some regex I use to filter unneeded syntactic elements
08:18:09	leorize	but I think my regex is universal enough
08:25:50	*	krux02 joined #nim
08:56:00	FromGitter	<bung87> https://github.com/bung87/ts2nim please give me advise , check the table not sure am I correctly translate it
08:56:47	liblq-dev	TS `number` is more close to a Nim `float`
08:58:04	liblq-dev	TS let -> Nim var, TS const -> Nim let
08:58:13	liblq-dev	TS var doesn't have a Nim counterpart
08:58:37	FromGitter	<bung87> `TS number is more close to a Nim float` yeah,someone also mentioned this.
08:59:56	FromGitter	<bung87> `TS let -> Nim var, TS const -> Nim let`, no ts can declare any thing to const, then modify members, but nim doest
09:00:51	Yardanico	you can in nim if an object is a ref type
09:01:01	Yardanico	because then "let" means that the reference is immutable, not the underlying data
09:01:15	Yardanico	and afaik all "objects" in TS are ref
09:01:15	FromDiscord	<Recruit_main707> https://github.com/zielmicha/go2nim
09:01:15	FromDiscord	<Recruit_main707> has someone tried this=
09:01:45	Yardanico	considering number of stars and last commit I doubt it :P
09:01:47	FromGitter	<bung87> Yardanico ok, that's really helped me
09:02:25	FromGitter	<bung87> go2nim that's I thought before, and the name `go2nim` is cool
09:04:29	FromGitter	<bung87> Yardanico `and afaik all "objects" in TS are ref` , there's also literal
09:04:47	Yardanico	so can you modify "literals" declared with let?
09:04:52	Yardanico	like "let a = 5; a = 3"
09:06:00	*	dddddd joined #nim
09:06:08	FromGitter	<bung87> yeah , in js it is
09:06:16	Yardanico	but we're talking about TS
09:09:34	FromGitter	<bung87> you can check it here https://www.typescriptlang.org/v2/en/play
09:10:07	Yardanico	I don't want to write in any type of JS anyway :)
09:10:26	FromGitter	<bung87> It's superset of js , not entirely another language.
09:10:41	Yardanico	but do it as liblq-dev said
09:10:44	Yardanico	ts let -> nim var
09:10:46	Yardanico	ts const -> nim let
09:13:16	FromGitter	<bung87> `let a:seq[int] = @[] ⏎ a.add 1`
09:13:33	FromGitter	<bung87> ```const a= [] ⏎ a.push(1)``` [https://gitter.im/nim-lang/Nim?at=5ebbba3d20eaac185307bf54]
09:13:59	Yardanico	well of course, the mappings can't always be the same
09:14:07	Yardanico	but the one liblq said will probably be the closest one
09:14:12	Yardanico	or just always use var
09:14:39	FromGitter	<bung87> always use `var` I think better, no need to infer type
09:15:16	FromGitter	<bung87> like js before es5
09:15:25	liblq-dev	but all variable declaration statements infer the type
09:15:33	liblq-dev	there's no problem in doing let a = @[1, 2, 3]
09:15:54	liblq-dev	it doesn't work for an empty seq because there are no elements in the seq to infer the type from
09:16:17	liblq-dev	also making a let a: seq[int] = @[] is kind of pointless since you can't add to the seq
09:16:49	*	silvernode joined #nim
09:17:11	FromGitter	<bung87> I just says in ts `const a = [];a.push(1)` is valid
09:18:46	FromGitter	<bung87> during translating I don't infer the type, just using estree ast
09:18:58	Yardanico	we understand that
09:19:08	Yardanico	but in nim all of var/let/const can infer the type
09:20:05	FromGitter	<bung87> ok , I get it, using var seems fine for now
09:20:54	*	liblq-dev quit (Ping timeout: 240 seconds)
09:21:24	*	WilhelmVonWeiner quit (Ping timeout: 272 seconds)
09:22:28	FromGitter	<bung87> the only obvious problem now, is tuple init parsed as `ArrayExpression` node.
09:24:06	FromGitter	<bung87> I may store variable declaration during translating , and infer closest variable type
09:26:28	alexander92	hey
09:26:36	alexander92	not using types might be hard
09:27:17	alexander92	we did translate some python and ruby to nim and iirc we did try to infer at least some of the local types as well
09:27:46	alexander92	our experiments never got to any mature state, so not really a great example
09:27:59	alexander92	but i love to talk about lang 2 lang translation
09:28:06	FromGitter	<bung87> :)
09:28:20	alexander92	one important thing is
09:28:25	alexander92	do you want idiomatic nim output
09:28:28	alexander92	or just valid output
09:28:51	FromGitter	<bung87> current approach seems fine, I will using type infer when actually needs.
09:29:38	FromGitter	<bung87> yeah, I thought validate whole project after translating, may just `nim c`?
09:29:46	alexander92	no no
09:29:48	alexander92	my point is
09:29:53	Yardanico	no, he asks you if you want to get good nim code
09:29:56	Yardanico	or just working nim code
09:29:58	alexander92	yeah
09:31:06	FromGitter	<bung87> ah, current target is valid code looking , not for running
09:31:08	alexander92	otherwise it's good to try to do it in the simplest possible way, so it might be wise to not infer for now
09:31:27	alexander92	ok, so the other thing is e.g.
09:32:01	alexander92	how do you translate stdlib
09:32:04	alexander92	invocations
09:32:22	alexander92	because at one point you need to map "idioms" from one language to another
09:32:42	alexander92	or at least map some api-s from X to Y
09:33:16	FromGitter	<bung87> it just for simple task, eg. input string output string
09:33:32	alexander92	(on the other hand this is probably skippable for javascript in the browser! as nim targets the same api there anyway)
09:33:45	alexander92	hm, ok
09:33:54	alexander92	well for algo-s you can really get a long way
09:34:00	FromGitter	<bung87> so my goal is very small
09:34:18	alexander92	with mostly translating syntax
09:34:33	alexander92	ok, would be glad to follow along
09:34:34	FromGitter	<bung87> but still usefull for translating tiny lib
09:34:53	alexander92	yeah but
09:35:14	alexander92	you'd almost always need some manual work on top of the result
09:35:23	alexander92	but i think this is absolutely fine
09:35:37	alexander92	and even more realistic than many other approaches
09:35:39	FromGitter	<bung87> yeah, that's right and expected.
09:35:43	alexander92	ok
09:37:17	*	Trustable joined #nim
09:38:37	*	Guest97932 quit (Ping timeout: 246 seconds)
09:42:01	*	dadada joined #nim
09:42:25	*	dadada is now known as Guest18922
09:50:41	Yardanico	is it wrong to parse json at compiletime to generate Nim definitions for types? :P
09:57:01	PMunch	Why would it be?
09:57:07	Yardanico	just asking :)
09:57:23	PMunch	I wrote jsonschema which parses something like typescript definitions into Nim type-checking
09:57:30	PMunch	Not quite into Nim types, but eh
09:58:29	alexander92	PMunch actually
09:58:50	alexander92	if one can separate the typescript defs => nim types part
09:58:58	alexander92	this can be reused maybe by bung87
09:59:16	alexander92	and by a newer dts2nim tool
09:59:18	PMunch	Well it doesn't actually convert them to types
09:59:27	alexander92	ah ok
09:59:37	alexander92	ah i see
09:59:40	alexander92	it makes sense
09:59:47	PMunch	It just implements a "verify" procedure that verifies if some JSON matches the typescript definition
10:00:09	PMunch	And a distinct JsonNode type that has accessors for all the defined fields.
10:00:10	alexander92	btw bung87
10:01:17	Yardanico	github seems laggy today again
10:01:48	alexander92	one other thing: its possible to just dump json from ts and load it from nim : and just write your compiler in nim
10:02:01	alexander92	because then you can just convert the Es Tree jsonnodes to
10:02:03	alexander92	PNodes
10:02:10	alexander92	and use renderer.nim from the compiler
10:02:13	alexander92	to generate the nim code
10:02:34	alexander92	but this might be not needed of course, just another option
10:02:36	FromGitter	<bung87> wow, that's goood idea
10:02:53	alexander92	yeah, it shouldn't be hard
10:03:14	alexander92	just a word of caution renderer.nim does generate invalid code for some cases
10:03:22	alexander92	but it should be rare
10:03:26	livcd	Is the fastest json pkg the packedjson?
10:04:03	alexander92	and as the tool would require a bit of manual review of build errors thats ok
10:04:18	alexander92	but i think treeform helped improving renderer.nim even more
10:04:27	alexander92	so it should be ok for most normal code probably
10:04:35	alexander92	hm, if its used for nimpretty even for most
10:04:57	alexander92	PMunch ohh good
10:05:16	FromGitter	<bung87> hmm, but first write estree parser in Nim ?
10:05:18	alexander92	but still i imagine one can reuse the parsing part
10:05:22	alexander92	bung87 no no
10:05:39	alexander92	you do parse in typescript, but then you just dump the resulting data structure as json in a file
10:05:59	*	Guest18922 quit (Ping timeout: 260 seconds)
10:06:06	alexander92	and then you start the nim compiler and load the json from it
10:06:58	alexander92	basically `compiler => run ts_parser (generate ast.json) => run nim_compiler (load ast.json generate nim code)`
10:07:24	FromGitter	<bung87> ok , json parse part could be a dependency , and then I translate the transpiler itsself ,depends on it , seems work
10:07:54	alexander92	well, the estree thing might be harder to translate , no idea what kind of code is it
10:08:03	alexander92	but even this still looks ok to me
10:08:17	alexander92	after all if people translate typescript its ok to require them to run some typescript
10:09:16	alexander92	but of course, think about it, you might decide that just producing directly strings with code from typescript is easier
10:09:19	alexander92	its your choice
10:10:38	FromGitter	<bung87> I considering it , maybe just make it as a nimble package
10:11:36	FromGitter	<bung87> when this done,I might write something could import nodejs module from nim
10:12:07	*	dadada joined #nim
10:12:30	*	dadada is now known as Guest60258
10:12:47	FromGitter	<bung87> first narrow it to node module write in ts.
10:13:19	alexander92	hm, i have to tell you that i'd just import directly nodejs module using the javascript backend
10:13:31	alexander92	if i use the c backend, i'd wrap a c lib or something else
10:14:03	alexander92	i'd also thought about more nodejs integration and i received backlash in the past and it's correct: it doesnt make much sense for nim
10:14:10	FromGitter	<bung87> how do you do that
10:14:19	alexander92	using a clib ?
10:14:33	FromGitter	<bung87> `import directly nodejs module `
10:14:48	alexander92	i just import it using `require` :)
10:14:55	alexander92	and type it using `importcpp`
10:15:03	FromGitter	<bung87> no manually work?
10:15:05	alexander92	or directly just use `.toJs` and cast
10:15:11	alexander92	err, no
10:15:16	alexander92	but this is for the javascript backend
10:15:41	alexander92	but thats what i am trying to explain, right: you already can import any typescript library in nim if you use javascript backend
10:16:20	FromGitter	<bung87> I saw the nim module for js backend, think better using name `importjs` `exportjs`
10:16:46	PMunch	Hmm, I wonder what would happen if I took all the procedures in the "minilib" and implemented them through vm.registerCallback..
10:17:04	alexander92	bung87 ! you're right
10:17:12	alexander92	i think they did add those aliases
10:17:16	alexander92	but not super sure
10:17:31	PMunch	So that every call in the minimal nimscript would actually be calls to the VM
10:17:39	PMunch	Err, calls to compiled code
10:17:41	alexander92	but again: you can already do that, so thats way i'd use that instead of your tool usually
10:18:01	FromGitter	<bung87> not just aliases, they may transform something during compile time.
10:18:20	FromGitter	<bung87> mapping object to js object
10:18:41	FromGitter	<bung87> not the plain object in nim..
10:19:29	alexander92	sorry, i was wrong
10:19:31	alexander92	yes!
10:19:51	alexander92	but again, that's why i wouldn't transpile a typescript lib, right
10:20:08	alexander92	because its easier to just translate typescript definitions to those importjs ones!
10:20:15	*	dddddd_ joined #nim
10:20:17	alexander92	without translating the whole code
10:20:21	Yardanico	kostya/benchmarks , also tested with arc
10:20:22	Yardanico	https://gist.github.com/Yardanico/8a78eaf4e74ede614c26baec7c6f0850
10:20:50	alexander92	(except if you want to generate C code or to really port a typescript lib to nim : then i'd use your tool!)
10:21:42	FromGitter	<bung87> https://github.com/nim-lang/Nim/commit/9acbf99efb42d7d3f98f3d1d6677923fc4dd3a28#diff-42ba1d994f4fa8c6ad17a7efae7936ccR129
10:22:01	*	dddddd quit (Ping timeout: 246 seconds)
10:22:08	FromGitter	<bung87> there's importjs 16 hours ago.
10:22:21	*	dddddd__ joined #nim
10:22:29	alexander92	bung87 yes, ok, this is not my point, i know about it
10:23:39	FromGitter	<bung87> ok I get your point
10:23:40	federico3	startProcess outputStream().readAll(), on a process that exited, sometimes hangs forever on the read syscall
10:23:45	federico3	any help?
10:25:12	alexander92	bung87 ok, otherwise i am excited about whatever you decide on :D
10:25:25	*	dddddd_ quit (Ping timeout: 264 seconds)
10:25:44	alexander92	PMunch sounds intriguing, would it really optimize
10:26:57	FromGitter	<bung87> I think both way can moving forward...
10:26:59	livcd	federico3: uhmm did not happen to me but that was way before 1.2...What process did you start?
10:27:22	federico3	nim c or nim doc
10:27:34	*	dddddd__ quit (Ping timeout: 240 seconds)
10:28:31	FromGitter	<bung87> I ' ll translate whole ts project first see what's next to do
10:28:54	alexander92	bung87 what do you mean
10:29:40	*	dddddd__ joined #nim
10:29:52	FromGitter	<bung87> I dont make choise, I may want both :)
10:29:55	*	dddddd__ is now known as dddddd
10:30:58	*	fredrikhr quit (Ping timeout: 258 seconds)
10:31:58	FromGitter	<bung87> translating, binding, for js ,for c, import, export , bundle, gen modules.
10:33:17	*	silvernode quit (Ping timeout: 260 seconds)
10:34:47	*	monokrom joined #nim
10:39:27	PMunch	alexander92, I don't know..
10:40:09	FromGitter	<bung87> oh, I got a idea, for std libs , I could just translating deno's
10:41:41	FromGitter	<bung87> https://github.com/denoland/deno/tree/master/std
10:42:39	PMunch	alexander92, problem is that "grep -Hnr proc minilib | wc -l" is still over 2000
10:42:48	PMunch	So it's not something I really want to do by hand :P
10:43:56	alexander92	bung87 ok! for both
10:44:19	alexander92	otherwise it depends, i'd guess most people would use the node's lib if you transpile node libs
10:46:29	FromGitter	<bung87> they may or may not, if they want compile to c
10:47:14	*	chemist69 quit (Ping timeout: 244 seconds)
10:48:09	FromGitter	<bung87> or they nim similar nodejs module api that's no nimble package has
10:48:13	*	chemist69 joined #nim
10:48:14	Yardanico	oh nice found another arc bug
10:48:39	FromDiscord	<Yardanico> @clyybber did you encounter Error: unhandled exception: injectdestructors.nim(405, 13) `not containsGarbageCollectedRef(n.typ)` [AssertionDefect] ?
10:49:12	Yardanico	I hit it when trying to compile havlak benchmark from https://github.com/kostya/benchmarks with orc (same with arc ofc)
10:50:22	alexander92	bung87 ok, but they either use directly the nodejs module
10:50:30	alexander92	without transpilation
10:51:00	alexander92	ok, i get what you mean, it might be easier to translate if there is a mapping of the node stdlib in native nim
10:51:08	alexander92	but this is too much work imho for very very little usecases
10:51:28	alexander92	and i'd just look for a c/rust lib to use usually if thats the case
10:51:50	alexander92	i'd focus on the browser maybe
10:52:52	*	silvernode joined #nim
10:53:06	PMunch	Hmm, is there a way to profile nimscript?
10:53:11	FromGitter	<bung87> ok
10:53:35	PMunch	It would be interesting to try a couple of the most used procedures first
10:54:18	skrylar[m]	hoi
10:57:57	narimiran	Yardanico: have you seen https://en.wikipedia.org/wiki/Modular_arithmetic#Example_implementations ?
10:58:42	Yardanico	narimiran: thanks for that link!
10:59:02	narimiran	sorry for not bringing it up earlier :)
11:00:11	narimiran	i'll post it on the forum too, if there will be somebody else in the future with a similar problem....
11:00:25	Yardanico	not that forum's search is good enough to find stuff you need :P
11:00:44	narimiran	:X
11:01:50	*	silvernode quit (Ping timeout: 260 seconds)
11:04:00	*	Guest60258 quit (Ping timeout: 256 seconds)
11:05:10	*	dadada joined #nim
11:05:28	dadada	remember a discussion on here about a feature that makes compile time variables available at runtime too
11:05:34	dadada	this is one of the first times that I actually need something like this :-)
11:05:36	dadada	var foo {.compileTime.}: bool .... -> so how can I now have this at runtime with the same symbol?
11:05:41	dadada	... sorry if reposting, there was a net issue
11:06:23	Yardanico	uh
11:06:26	alexander92	you can always have var fooRT = foo # right
11:06:33	Yardanico	!eval var foo {.compileTime.} = true; echo foo
11:06:35	NimBot	true
11:06:44	Yardanico	¿que
11:07:08	alexander92	!eval var foo {.compileTime.} = true; foo = false; echo foo
11:07:10	NimBot	true
11:07:15	alexander92	hmm
11:07:19	Yardanico	huh
11:07:34	Yardanico	!eval var foo {.compileTime.}: NimNode; echo foo
11:07:36	NimBot	Compile failed: /usercode/in.nim(1, 40) Error: type mismatch: got <NimNode>
11:07:45	FromDiscord	<Recruit_main707> isnt var {.compileTime.} the same as const??
11:07:49	Yardanico	no
11:07:56	alexander92	!eval var foo {.compileTime.} = true; static: foo = false; echo foo
11:07:57	Yardanico	with var compiletime you can change var at compile-time
11:07:58	NimBot	<no output>
11:08:11	Yardanico	alexander92: it did echo in the compiler log
11:08:18	Yardanico	static: (foo = false, echo foo)
11:08:20	alexander92	yeah but i dont know how to access
11:08:25	alexander92	wow
11:08:34	Yardanico	wow?
11:08:39	alexander92	hm, but how to write
11:08:44	alexander92	two lines to !eval ?
11:08:48	Yardanico	you can't
11:08:54	alexander92	:D
11:08:56	Yardanico	but let me show my ultimate ability
11:09:09	dadada	is there a when compileTime: code1 else: code2 ... I know static:, but there's no else part for it
11:09:15	Yardanico	when nimvm
11:09:21	Yardanico	when defined(nimvm) or something like that
11:09:28	dadada	ya, thanks
11:09:42	alexander92	when nimvm ? or defined?
11:09:53	Yardanico	https://nim-lang.org/docs/manual.html#statements-and-expressions-when-nimvm-statement
11:09:58	Yardanico	when nimvm
11:10:15	alexander92	<3
11:10:22	alexander92	guys
11:11:15	alexander92	debugging is fun
11:11:46	Yardanico	alexander92: btw did you see https://www.youtube.com/watch?v=QXneLboQMB0 ?
11:12:11	Yardanico	https://github.com/hediet/vscode-debug-visualizer/pull/51
11:12:13	disbot	➥ Add Nim demo
11:12:42	Yardanico	that extension also supports graphs
11:15:17	*	kitech1 quit (Ping timeout: 260 seconds)
11:16:30	*	kitech1 joined #nim
11:21:32	alexander92	Yardanico <3
11:21:36	alexander92	that's superb
11:21:47	FromDiscord	<Rika> !eval var foo {.compileTime.} = true; (static: foo = false); echo foo
11:21:49	NimBot	Compile failed: /usercode/in.nim(1, 47) Error: expected: ')', but got: '='
11:21:52	FromDiscord	<Rika> 😮
11:21:54	Yardanico	nah I already tried that Rika
11:21:57	Yardanico	alexander92: and the idea is quite "simple stupid"
11:22:00	FromDiscord	<Rika> i wouldnt know
11:22:00	alexander92	we dont do stuff like this
11:22:04	alexander92	but we
11:22:07	Yardanico	the extension just reads json from the debuggger :D
11:22:09	alexander92	i mean we have a similar feature
11:22:24	FromDiscord	<Recruit_main707> Yardanico: does it work with the c backend?
11:22:27	alexander92	but the way it functions is different
11:22:30	Yardanico	wdym? @Recruit
11:22:33	Yardanico	that's what I used it with
11:22:41	Yardanico	with lldb as a debugger
11:22:51	Yardanico	because it's simple to define NCSTRING as string there
11:22:59	alexander92	i'd show you when we can get to testers
11:23:08	FromDiscord	<Recruit_main707> it says c++ on throw
11:23:15	Yardanico	??
11:23:19	alexander92	but its a great extension
11:23:19	FromDiscord	<clyybber> @Yardanico nope
11:23:27	Yardanico	ah I see what you mean
11:23:34	Yardanico	@Recruit nah it's just from lldb
11:23:38	Yardanico	default breakpoints I guess
11:24:18	Yardanico	see https://github.com/hediet/vscode-debug-visualizer/blob/master/demos/nim/.vscode/launch.json how I tell LLDB to understand that NCSTRING is c-string
11:24:33	Yardanico	I guess with enough lldb-fu we can make something a bit similar to nim-gdb
11:24:46	Yardanico	http://lldb.llvm.org/use/variable.html
11:24:51	alexander92	oh nice
11:24:58	alexander92	otherwise it should be possible
11:25:07	alexander92	if they have support for something like pretty-printers
11:25:31	Yardanico	seems so
11:25:41	alexander92	i can take a look
11:25:44	alexander92	one day
11:25:54	Yardanico	it's python as well
11:25:58	*	Yardanico left #nim ("https://quassel-irc.org - Chat comfortably. Anywhere.")
11:26:01	*	Yardanico joined #nim
11:26:03	Yardanico	https://github.com/fantaosha/LLDB-Eigen-Pretty-Printer/blob/master/LLDB_Eigen_Pretty_Printer.py
11:26:05	alexander92	ugh i really want to show some of our ideas
11:26:08	Yardanico	found some example
11:26:13	alexander92	otherwise another open source debug idea i have is a
11:26:21	alexander92	simple nim vm debugger
11:26:29	alexander92	but i am not sure how useful that is
11:26:32	Yardanico	oh you're talking about nimconf?
11:26:41	alexander92	not really, just like for a project
11:26:42	alexander92	in general
11:26:43	*	zama quit (Ping timeout: 260 seconds)
11:27:02	*	ftsf quit (Ping timeout: 260 seconds)
11:27:12	alexander92	ah yeah this does look similar
11:27:12	Yardanico	https://lldb.llvm.org/use/python.html
11:27:19	alexander92	it should be maybe possible to reuse some of the logic
11:27:35	Yardanico	and lldb can be used by binaries compiled with gcc too of course
11:27:42	Yardanico	I wonder if the integration is better if I compile with clang though :P
11:27:53	alexander92	but it isnt too important
11:27:55	alexander92	yes btw
11:28:01	alexander92	i think they do have some column info in dwarf
11:28:14	alexander92	but not sure if this is useful for the nim -> c thing
11:28:26	alexander92	on the other hand, gcc also can generate column info
11:28:38	alexander92	but there might be differences between both
11:28:50	*	zama joined #nim
11:42:02	PMunch	Hmm, slight problem with implementing the entire stdlib in native code with callbacks for nimscript.. Data is sent as PNodes, which will probably be a PITA to decode into the required Nim code..
11:43:34	PMunch	I wish there was a way though to take these files and strip all documentation comments and evaluate the `when not defined(nimscript)` cases to remove the procedures you wouldn't be able to use anyways..
11:46:55	Araq	huh? do you think it's reasonable to do that? why do you need the "entire" stdlib within nimscript
11:49:07	*	poohman quit (Ping timeout: 260 seconds)
11:49:18	*	zacharycarter joined #nim
12:01:22	FromDiscord	<Prodigle> What's the keyword to do the opposite of addr? E.g from a pointer to the data
12:01:31	Yardanico	operator []
12:01:35	Yardanico	yourptr[]
12:01:38	FromDiscord	<Prodigle> ty
12:01:41	dadada	expandMacros works for macros, is there something that works in the same way for pragmas?
12:01:53	dadada	to know what user defined macros expand to when used as pragmas?
12:01:55	Yardanico	@Prodigle see https://nim-lang.org/docs/manual.html#types-reference-and-pointer-types
12:02:04	Yardanico	it's explained in more detail there :P
12:03:48	*	hax-scramper joined #nim
12:03:51	PMunch	Araq, I'm trying to boil it down as much as possible. I already have a "minilib" which is only the parts that you need in order to compile a very basic example. But a lot of the stuff in these files are obviously never used (eg. the "from unicode import toLower, toUpper" leads to a lot of stuff being "required" while never actually used)
12:04:14	PMunch	Or were you talking about implementing it as callbacks?
12:04:43	PMunch	That was just a thought I have. Instead of shipping the stdlib sources, why not just compile the procedures into the binary and expose them as callbacks.
12:04:55	*	joshbaptiste quit (Ping timeout: 246 seconds)
12:04:56	PMunch	But that would be a massive PITA..
12:05:44	PMunch	(By the way, I plan to have a switch/detector to see if you have Nim installed, and if so try to use the stdlib it ships. So you won't be "stuck" with only the minimal stuff I ship)
12:05:51	dadada	nothing to show expanded pragmas?!
12:06:02	*	supakeen quit (Quit: WeeChat 1.9.1)
12:06:20	Yardanico	dadada: never heard of this, expandMacros doesn't work?
12:06:21	PMunch	Wouldn't that just be the same as expandMacros?
12:06:28	Yardanico	then you'll have to modify the pragma definition I guess :P
12:06:34	Yardanico	so it prints the resulting code
12:06:44	*	supakeen joined #nim
12:07:00	dadada	Yardanico: okay, maybe it does work, ah, dammit, it's just not how I expect things to work... that's all, yes after all it did work, sorry for using your time
12:07:00	PMunch	One thing that irks me about expandMacros, it only works if the code is valid.. So it's useless for debugging
12:07:12	lqdev[m]	echo result.repr should work
12:07:15	PMunch	Unlike "echo result.repr"
12:07:32	hax-scramper	How I can convert macro argument to `typedesc`? I want to use `typetraits.arity` in macro.
12:07:33	Yardanico	it depends on what kind of "debugging" we're talking about though :P
12:08:20	dadada	okay, I made this very small macro for something like var myVar {.doStuff.}: bool = true ... and inside the macro there's no way to know whether the expression is a var or a let you just get
12:08:50	dadada	Arglist: Ident "myVar" Ident "bool" Ident "true"
12:09:08	dadada	so if I want to produce different code depending on var/let or something else, how is it done?
12:09:57	Yardanico	don't do it as a pragma and wrap the whole code block in the macro instead?
12:10:05	Yardanico	or maybe both at the same time
12:10:14	Yardanico	check if a variable has your own pragma in a macro
12:10:44	dadada	yeah, but it'd be much cooler as a pragma, and it'd fit in with the other pragmas, doesn't want to be a misfit, you know how those are treated in school
12:10:56	Yardanico	how would it "fit"? :D
12:11:20	lqdev[m]	hex-scramper: generate a call to arity() in your macro?
12:11:34	dadada	it's analogous to .compileTime. ... want to write a pragma that does some magic so a var works at compile and runtime ... just trying some stuff
12:11:57	lqdev[m]	or maybe generate a call to another macro that gets a typedesc, and pass typeof(yourNimNode) to that other macro from your main macro
12:14:00	hax-scramper	lqdev[m]: No, just call arity inside of a macro body. Something like this: `macro printArity(t: typedesc) = echo arity(t)`.
12:14:07	dadada	well, the brutish way I know could work is to use lineInfo and then parse the code by hand
12:17:04	hax-scramper	I can of course just parse AST and get necessary parts but I wanted to know if there any way I can just convert between `NimNode` and `typedesc`. Latter one is defined as compiler magic and I can't just write my own converter because of that.
12:25:51	*	liblq-dev joined #nim
12:27:45	Araq	PMunch, minilib is system.nim plus the system/ directory
12:27:54	Araq	what else would be required?
12:28:23	PMunch	This stuff: http://ix.io/2lSo
12:28:42	PMunch	In order to run this code: http://ix.io/2lSp
12:28:49	FromGitter	<bung87> fatal: unable to access 'https://github.com/bung87/ts2nim.git/': The requested URL returned error: 429
12:28:59	Yardanico	?
12:29:06	Yardanico	what are you trying to do?
12:29:10	PMunch	addFloats is declared in "exposed" and is implemented in the code
12:29:23	Yardanico	ah nvm @bung87 it's probably due to github behaving badly today
12:29:25	FromGitter	<bung87> travis results.
12:29:32	Yardanico	restart the build then
12:29:56	FromGitter	<bung87> never thought 429 would happen to github..
12:30:12	Araq	huh, what is "import exposed"?
12:30:16	FromGitter	<bung87> ok,then
12:31:36	*	kradnoel joined #nim
12:31:37	PMunch	Araq, http://ix.io/2lSr as I said, it's only a declaration that is implemented in the program
12:33:10	PMunch	It's basically just this: https://github.com/nim-lang/Nim/tree/devel/tests/compilerapi
12:33:43	Araq	why does it need more than system.nim
12:33:52	Araq	that's the interesting question for me
12:34:20	PMunch	With only "system.nim" and the "system" folder I get this error: system.nim(1132, 19) Error: cannot open file: std/private/since
12:35:45	Araq	gah, that's a regression
12:35:55	Araq	this must be system/since.nim
12:36:11	PMunch	Adding that file I get this: system/assertions.nim(4, 19) Error: cannot open file: std/private/miscdollars
12:36:26	Araq	-.-
12:36:35	Araq	same story
12:36:58	PMunch	Adding that file I get this: system/nimscript.nim(275, 6) Error: cannot open file: os
12:37:35	PMunch	And that brings in pretty much the rest
12:38:00	Araq	ah and that's my fault. that's what I get for fixing bugs. however
12:38:08	PMunch	Through strutils and pathnorm
12:38:09	Araq	don't define NimScript in tcompilerapi
12:38:54	PMunch	I don't, it's set in createInterpreter in nimeval.nim
12:39:09	PMunch	It sets nimscript and nimconfig
12:39:18	Araq	yeah, bad idea
12:40:53	PMunch	Undefining it gives me this error: system.nim(529, 11) Error: cannot open file: system/ht
12:41:01	Yardanico	ht?
12:41:02	PMunch	Erm, that should be "hti"
12:41:03	Araq	ht?
12:43:41	PMunch	Not having nimscript define leads it to require stuff like ansi_c, memory, atomics, cgprocs etc.
12:43:52	PMunch	And that hti thing
12:44:22	*	NimBot joined #nim
12:44:36	Araq	so add a new define nimTiny and ensure nimscript isn't included when nimTiny is defined
12:48:13	FromGitter	<wrq> I think the nim forum should be moved to a Discourse instance. It'd be a lot easier to have a full-featured platform for organizing medium to large-sized efforts.
12:48:20	Yardanico	uhhh
12:48:21	Yardanico	:DDDD
12:48:39	Yardanico	one of the reasons it was created is to not use discourse
12:48:43	Yardanico	https://github.com/nim-lang/nimforum
12:49:08	FromGitter	<wrq> Right, but the quality of it is severely lacking. I am signed in right now and there's no sign-out function?
12:49:35	FromGitter	<wrq> I don't see any controls except New Thread, and upon trying that, I am told my email is not confirmed, though I have no email about Nim forum.
12:49:58	Araq	well most of us can login/logout so there is that ;-)
12:50:13	zacharycarter	so you've never used the forum but you want to switch to discourse to organize large efforts?
12:50:27	zacharycarter	also - I can logout
12:50:28	Araq	in order to confirm your email, confirm your email, discord won't help with that problem
12:50:46	Yardanico	@wrq to sign out you have to click on your profile pic in right upper corner and press logout
12:50:50	Yardanico	seems simple enough
12:51:00	FromGitter	<wrq> I don't see a profile in in the corner
12:51:06	FromGitter	<wrq> let me share a screenshot
12:51:08	Yardanico	then are you sure you're logged in?
12:52:51	alexander92	@wrq there are still some bugs in the forum, but it seems surprisingly ok to me for a language forum, even if not as powerful as discourse
12:53:07	Yardanico	yeah, and also @wrq - current version deployed doesn't have all latest features nimforum has
12:53:19	Yardanico	check https://forum.my-toolbox.xyz/ to see how latest nimforum looks and works :P
12:53:27	Yardanico	it's my test instance
12:53:34	alexander92	and dogfooding/example software in a language are useful enough for nimforum to be a better option in this case
12:53:42	Yardanico	yeah that too
12:54:49	FromDiscord	<mratsim> Plus it's the best test for async/httpbeast
12:54:59	Yardanico	well it's single threaded with httpbeast anyway
12:55:03	FromGitter	<wrq> https://imgur.com/a/RMqmURr
12:55:04	Yardanico	nimforum uses some global variables as of now
12:55:15	FromDiscord	<mratsim> httpbeast is multithreaded though no? cc @dom96
12:55:16	Yardanico	@wrq it's there on the right
12:55:17	Yardanico	I see it
12:55:18	Yardanico	@mratsim it is
12:55:20	Yardanico	with --threads:on
12:55:31	Yardanico	but nimforum doesn't compile because it uses some global vars as I said :) I tried
12:55:51	FromGitter	<wrq> ah, so it is
12:55:54	PMunch	Araq, if I leave the nimscript symbol defined, but don't define the nimconfig symbol I end up with this list of required files: http://ix.io/2lSA which I'm guessing is closer to what you had in mind
12:55:55	FromDiscord	<mratsim> global mutable vars?
12:55:59	Yardanico	I think so, yes
12:56:05	FromDiscord	<mratsim> ah, meh
12:56:09	Yardanico	lemme check the error again to find where that happens
12:56:10	FromGitter	<wrq> nevertheless, I think it's a fine effort, but a discourse instance would be much more pleasant to use
12:57:06	FromDiscord	<mratsim> Can we add Nim highlighting to a Discourse?
12:57:32	alexander92	@wrq it's always a tradeoff: after all on the other hand, it's better to have a healthy ecosystem of forums
12:57:35	FromDiscord	<mratsim> my biggest grip with nimforum is choosing RST instead of markdown for post syntax
12:57:48	PMunch	nimhcr is included because of minilib/system.nim:3029: include nimhcr which is behind a "when not defined(createNimHcr)" check
12:57:58	alexander92	the same way there is healthy ecosystem for chats/slack-like apps
12:58:08	alexander92	mratsim yeah
12:58:14	FromDiscord	<mratsim> otherwise it's in an OK enough state that dogfooding is beneficial
12:58:19	alexander92	there could be at least an option
12:58:26	alexander92	but that's a flamewar topic :( :D
12:58:29	Araq	*if* we switch (and it's a big if), we should move to github's forum stuff
12:58:31	FromGitter	<wrq> I'm not saying remove nimforum
12:58:32	Yardanico	well our rst parser is also a markdown parser anyway :P
12:58:37	Yardanico	Araq: well it's not ready yet anyway
12:58:51	alexander92	huh is this new https://github.community/
12:58:56	FromGitter	<wrq> Nimforum should remain as an example project, beneficial to many users if they pursue webdev
12:59:03	alexander92	but i still think nimforum is good enough for now
12:59:07	FromGitter	<kaushalmodi> > <mratsim> my biggest gripe with nimforum is choosing RST instead of markdown for post syntax ⏎ ⏎ Same here
12:59:26	zacharycarter	+1 on that
12:59:29	FromGitter	<wrq> and, I think a full-featured platform as an alternative would be preferred by most users
12:59:30	zacharycarter	other than that I think the forum is great
12:59:34	PMunch	I guess if you add "and not defined(nimscript)" to that check you would have only system and system.nim (plus the two files from std that you said were regressions).
12:59:39	PMunch	Araq ^
12:59:42	Yardanico	zacharycarter: well rst parser is also a markdown parser :P
12:59:56	Araq	also, there could be more than one forum, maybe Python has more than one forum too
13:00:08	FromGitter	<kaushalmodi> Yardanico: It is.. but not a proper markdown parser, right?
13:00:10	Araq	you can summon an unofficial one
13:00:19	dom96	what. You're actually entertaining the idea of getting rid of the Nim forum?!
13:00:30	*	Hideki joined #nim
13:00:31	Yardanico	dom96: he's not
13:00:45	Araq	it is a "proper" markdown parser, except for the fact that markdown is a family of implementations with an unclear spec
13:00:53	FromGitter	<wrq> python has hundreds of forums, depending on your perspective
13:00:59	FromDiscord	<mratsim> just use Github markdown and done
13:01:12	*	Hideki is now known as Guest38545
13:01:14	dom96	mratsim: write a parser in Nim for it and I'll do it
13:01:33	Araq	there is no need, you can simple patch the existing parser
13:01:38	FromDiscord	<mratsim> I'm busy trying to write a multithreading runtime 😉
13:01:54	FromGitter	<kaushalmodi> Yardanico: By "proper markdown parsing", I meant by something like this Nim markdown library that has Commonmark compliant tests etc: https://github.com/soasme/nim-markdown
13:01:57	Yardanico	@mratsim - about global vars, there's not many, but the biggest is probably the DB
13:02:00	dom96	hm, this would be a fun project to be fair.
13:02:03	FromDiscord	<mratsim> anctually making it play well with servers and async
13:02:04	Araq	but I'm tired of this myth fighting
13:02:05	*	dom96 adds it to his project ideas
13:02:12	FromGitter	<kaushalmodi> dom96: https://github.com/soasme/nim-markdown
13:02:22	FromGitter	<wrq> discourse uses highlightjs, which has Nim support
13:02:43	Yardanico	i doubt it works properly for *all* cases anyway :P
13:02:44	dom96	kaushalmodi: ooh
13:03:02	Araq	Yardanico, so what, you can use ```nim
13:03:13	Araq	and # heading and the link syntax
13:03:20	Yardanico	Araq: I mean discourse
13:03:24	Yardanico	highlight.js
13:03:54	Araq	plus all the stuff that RST has in common with markdown. which is a lot...
13:04:12	*	rockcavera joined #nim
13:04:36	FromGitter	<bung87> `prism` is better
13:05:02	PMunch	prism?
13:05:28	Yardanico	ok time to minimize another compiler crash with arc :P
13:05:55	FromGitter	<kaushalmodi> Yardanico: This ⏎ ⏎ ```code paste, see link``` ⏎ ⏎ gives ... [https://gitter.im/nim-lang/Nim?at=5ebbf0b32cf0da0ad9fd80c2]
13:06:05	FromGitter	<wrq> so, dogfooding for the sake of dogfooding seems to be the conclusion of the discourse/nimforum discussion
13:06:15	Yardanico	@wrq it helps the ecosystem
13:06:22	*	WilhelmVonWeiner joined #nim
13:06:23	FromGitter	<wrq> well
13:06:35	FromGitter	<wrq> it sends the users the message "MY IDEA is better"
13:06:38	Yardanico	you can then ask "why create the language if *insert language name* already exists" ? :)
13:06:45	*	WilhelmVonWeiner is now known as Guest38566
13:07:24	Araq	wrq: the message is also "at least our own stuff has different security flaws"
13:08:21	FromGitter	<bung87> @PMunch https://prismjs.com/
13:08:25	FromGitter	<wrq> I mean, it would be if it was functioning, I haven't actually gotten the email yet, so I really can't even interact with nimforum presently
13:08:30	alexander92	wrq sorry, but the community already uses the forum without too much complaint
13:08:40	alexander92	it is true that this email issue should be fixed
13:08:50	FromGitter	<wrq> nah, dontt worry about it
13:09:02	FromGitter	<wrq> araq is fine with it that way
13:09:34	dom96	"this email issue" is likely gmail being overly strict with our emails
13:09:39	alexander92	well, if enough people weren't, we/they would just create a discource instance
13:09:49	alexander92	democracy
13:09:54	dom96	discourse won't fix that
13:10:07	Araq	wrq: you can also ask questions on the github issue tracker
13:10:12	alexander92	hm, but why doesn't this happen with most sites
13:10:15	FromGitter	<wrq> I've received a password reset email, but no activation email
13:10:35	dom96	just tell us your nickname and we can activate it for you
13:10:59	FromGitter	<wrq> "haydenjones"
13:11:09	FromGitter	<wrq> is the username
13:11:20	*	cgfuh joined #nim
13:12:19	Araq	it's not just about the tech we would all like to have instead, the forum is a 100MB database of useful information and code snippets.
13:12:44	Araq	and migrating the info to discord is more work than improving our forum
13:12:55	alexander92	btw there is also a subreddit, but its not so active iirc
13:13:00	Yardanico	well yeah
13:13:06	Araq	there is also Stack Overflow
13:13:12	Yardanico	alexander92: sometimes it gets questions though
13:13:13	Yardanico	like
13:13:13	Yardanico	https://www.reddit.com/r/nim/comments/giaeev/what_are_the_biggest_weaknesses_of_nim_in_your/
13:13:26	dom96	Araq, discord != discourse lol
13:13:35	Yardanico	dom96: well you know what he meant anyway :)
13:13:42	FromGitter	<wrq> why would it need to be migrated?
13:13:46	FromGitter	<wrq> these could be parallel
13:13:48	Yardanico	uh
13:13:53	PMunch	Yardanico, seems to be the same old stuff as replies, so that's good
13:13:58	PMunch	At least no big surprises
13:14:04	alexander92	hm, a web framework indeed
13:14:31	PMunch	Another one of my back-burner projects..
13:14:38	dadada	why does proc(name: "something") sometimes work and at other times you neeed proc(name = "something")
13:14:43	dadada	bug/feature/why?
13:14:49	dadada	I like : a little more
13:14:53	Yardanico	dadada: : never works in this case afaik
13:14:57	alexander92	PMunch same :D
13:15:00	Yardanico	proc(name: ) is object construction syntax
13:15:05	PMunch	If it works that's a bug :P
13:15:08	Yardanico	Object(field1: "a", field2: "b")
13:15:09	Araq	: is object construction, = is named args
13:15:15	PMunch	alexander92, maybe we should compare notes sometime?
13:15:20	FromGitter	<bung87> `SomeNumber* = SomeInteger|SomeFloat` mapping to js number seems more closer?
13:15:23	alexander92	yeah, i was about to say that
13:15:34	dadada	ah, I got confused because I used : in an initObject() proc, that makes it look like object construction IMO
13:15:46	alexander92	for some needed projects it might be more useful to combine forces instead of more PoC-s
13:15:48	FromGitter	<wrq> I see that my Nimforum "rank" has been changed, thank you, whomever did that
13:15:56	Yardanico	@bung87 well you have to understand that it'll only exist at compile-time
13:15:59	PMunch	alexander92, exactly
13:16:06	Yardanico	so if you have code which can use either floats or numbers it won't go well
13:16:40	FromGitter	<bung87> @Yardanico em, ok, let me think more
13:16:42	PMunch	Hmm, curiously enough without the "nimconfig" switch I don't require "system/nimscript" :P
13:18:52	dom96	wrq: you're welcome
13:20:35	FromDiscord	<Prodigle> Is there a quick way to debug output the contents of a multi-level deep sequence? Something akin to var_dump in Php?
13:21:59	FromDiscord	<mratsim> echo?
13:22:20	Yardanico	Araq: is that code supposed to compile with arc? I mean it does with the default gc https://play.nim-lang.org/#ix=2lSK
13:22:27	Yardanico	and "Error: unhandled exception: injectdestructors.nim(405, 13) `not containsGarbageCollectedRef(n.typ)` [AssertionDefect]" with arc
13:22:40	Yardanico	found in https://github.com/kostya/benchmarks/blob/master/havlak/havlak.nim
13:26:03	PMunch	Hmm, Araq you might know this. When I use minilib I can't use registerAdditionalOps. This is because it registers the callback as "stdlib.hashes.hashVmImpl". If I change it to "*.hashes.hashVmImpl" it works, so it seems like it somehow detects that minilib is not in fact stdlib. How is this done, and can I override it somehow?
13:29:39	Yardanico	@kaushalmodi https://forum.my-toolbox.xyz/t/6#10
13:29:59	disruptek	i'll be late today.
13:30:04	Yardanico	disruptek: how late? :(
13:31:11	FromDiscord	<Prodigle> Anyone any ideas on this? Fails at line 80, setting a var to the value in an array that is definitely there because it's statically initialized https://play.nim-lang.org/#ix=2lSP
13:31:13	FromDiscord	<Prodigle> Line 80
13:31:38	Yardanico	@Prodigle yes
13:31:42	Yardanico	you have an off-by-one mistake
13:31:51	Yardanico	you iterate from 0 to 6 inclusive
13:31:58	*	Guest38545 quit (Remote host closed the connection)
13:32:04	narimiran	Yardanico: there are 7 values there ;)
13:32:10	Yardanico	oh
13:32:12	Yardanico	my bad
13:32:12	FromDiscord	<Prodigle> toCalculate is 7 big
13:32:13	FromDiscord	<Prodigle> yeah
13:32:31	Yardanico	does stuff change if you change let to const? :)
13:32:39	disruptek	probably stream in 2-3hrs.
13:32:44	FromDiscord	<Prodigle> Same issue
13:33:05	FromDiscord	<Prodigle> a .len() output confirms 7
13:33:32	narimiran	remove `;` on line 78, just to be sure it's not that :)
13:34:35	*	PMunch contemplates adding --threads:on support for playground
13:34:44	FromDiscord	<mratsim> the joy of Nim templates into the stacktraces
13:34:51	FromDiscord	<mratsim> eating*
13:34:53	FromDiscord	<Prodigle> Removed ; and same error
13:35:00	PMunch	What's with all the whitespace by the way Prodigle?
13:35:03	Yardanico	@PMunch while you're at it also add --gc:arc and --gc:orc please
13:35:19	PMunch	To the playground?
13:35:20	FromGitter	<bung87> what equals js operator `>>>` in nim?
13:35:23	Yardanico	yes, why not? :P
13:35:29	narimiran	PMunch: and while at it, please fix adding indentation where not needed ;)
13:35:30	PMunch	I mean I guess..
13:35:46	Yardanico	@bung87 seems to be unsigned right shift
13:35:49	FromDiscord	<Prodigle> @Pmunch just copy pasted from my IDE, I split things up into sections when debugging
13:35:56	PMunch	narimiran, PRs welcome :)
13:36:02	narimiran	:)
13:36:06	Yardanico	https://stackoverflow.com/questions/7718711/javascript-triple-greater-than/7718947
13:36:11	PMunch	That goes for you too I guess Yardanico :P
13:36:15	Yardanico	haha
13:36:26	PMunch	@Prodigle, ah I see
13:36:28	Yardanico	I guess we can juts add more dropdowns :P
13:36:40	FromDiscord	<mratsim> I can reproduce, very weird bug
13:36:46	PMunch	They don't render very well on mobile unfortunately..
13:36:58	FromGitter	<bung87> oh it's called ` zero-fill right shift.`
13:36:59	PMunch	The whole layout should be rewritten with grids or something
13:37:04	PMunch	To make it a bit more robust
13:37:19	FromDiscord	<Prodigle> Possibly to do with the time macro?
13:37:19	Yardanico	wait lol I didn't know nim playground also uses karax nowadays
13:37:24	FromDiscord	<Prodigle> Though I can't see how
13:37:24	*	fanta1 joined #nim
13:37:32	PMunch	bung87, that means cast to unsigned and shr in Nim
13:37:42	Yardanico	PMunch: seems like JS people abuse it though
13:37:46	narimiran	wait wait
13:37:52	Yardanico	"Ignoring its intended meaning, this is most likely where you'll see it used: >>> 0 is unique in that it is the only operator that will convert any type to a positive integer: "
13:37:53	narimiran	`results`
13:37:54	PMunch	Yardanico, it's been that way since I took over :P
13:38:00	FromDiscord	<mratsim> you didn't initialize `results` @Prodigle
13:38:01	PMunch	Or almost that long
13:38:16	narimiran	what is `len(results)` before line 80?
13:38:28	FromDiscord	<Prodigle> Ah
13:38:31	FromDiscord	<mratsim> This is unnecessary: results[i] = newSeq[int64]()
13:38:38	Yardanico	you can just do resuilts.add, yeah
13:38:42	FromGitter	<bung87> @PMunch ok , thanks!
13:38:45	narimiran	and the answer is: zero :)
13:39:27	Yardanico	"/* Generated by the Nim Compiler v1.0.0 */" at least not as old as the nim forum :P
13:39:45	*	xet7 quit (Remote host closed the connection)
13:39:50	narimiran	yep, change line 82 to `results.add newSeq[int64]()` and it works
13:39:58	PMunch	That's just an indicator of how long it's been since I've done anything to the front-end :P
13:39:59	FromDiscord	<mratsim> you don't need to
13:40:05	FromDiscord	<mratsim> just remove the line altogether
13:40:31	FromDiscord	<mratsim> the forst for loop sends the work, the second collect the results
13:40:34	FromDiscord	<mratsim> first*
13:40:42	FromDiscord	<Prodigle> Yup! that fixed it, thanks a lot
13:40:48	FromDiscord	<Prodigle> Silly mistakes cost time 😅
13:41:15	FromDiscord	<mratsim> well, programming is 80% of the time fixing typos 😛
13:41:30	FromDiscord	<mratsim> though in the original code I sent you it wasn't there
13:42:14	dadada	ladies and gentleman let me introduce to you a new pragma that I wrote today that allows to create variables that work on compile and runtime
13:42:17	dadada	https://play.nim-lang.org/#ix=2lST
13:42:26	dadada	obviously it's not perfect yet, as it is brandnew
13:42:29	FromDiscord	<Prodigle> 😅 Would you believe me if I said I've worked in C++ for years. Learning the bare minimum to write Nim code has really made me dumb
13:42:45	dadada	and I want to improve it :D but it's pretty cool that it already basically works
13:42:49	FromDiscord	<mratsim> @dadada, {.UberTime.} ??
13:42:51	dadada	leorize: I have done it!
13:42:58	dadada	mratsim: yep
13:43:06	dadada	uber = above
13:43:37	FromGitter	<kaushalmodi> Yardanico: The Markdown parsing on the new nimforum looks much better
13:43:43	FromDiscord	<mratsim> what problem does it solve?
13:43:56	FromGitter	<kaushalmodi> I signed up for an account on the test nimforum site and am waiting for email
13:44:05	Yardanico	well you should receive it really fast
13:44:14	Yardanico	what's your username so I can confirm you manually? :P
13:44:14	*	someunknownuser joined #nim
13:44:19	*	xet7 joined #nim
13:44:23	Yardanico	the problem is that I use gmail's SMTP (yes)
13:44:40	FromGitter	<kaushalmodi> I signed up with my same username: kaushalmodi
13:44:41	Yardanico	nvm confirmed you
13:44:45	dadada	mratsim: I have written a module that needs to work at compile and runtime and with something like this it would, currently I need a manual solution (using when nimvm)
13:44:45	FromGitter	<kaushalmodi> thanks
13:45:14	Yardanico	it's running on my vps and doesn't really use a lot of resources
13:45:30	Yardanico	~34mb ram as of now
13:45:30	FromGitter	<kaushalmodi> I replied: https://forum.my-toolbox.xyz/t/6#12
13:45:31	disbot	no footnotes for `34mb`. 🙁
13:45:44	someunknownuser	is hasData from the osproc module supposed to block the current thread until data is available?
13:45:44	Yardanico	and it's not my work btw :P
13:45:44	FromGitter	<kaushalmodi> do we get email notifications for replies on the new nimforum?
13:45:45	FromDiscord	<mratsim> I'm still not sure what's the difficulty, I have plenty of code that needs to support both
13:45:55	Yardanico	@kaushalmodi I honetsly don't know, never tried, but probably not
13:45:59	FromGitter	<kaushalmodi> Yardanico: your work for setting up that test site
13:46:05	Yardanico	well it was not that hard
13:46:37	FromDiscord	<mratsim> ah it's for pointer types?
13:46:44	Araq	mratsim: http://okmij.org/ftp/papers/Macros-talk.pdf
13:46:53	Araq	"The reason macros are so difficult to comprehend is
13:46:53	Araq	that they in general, do not compose."
13:46:56	dadada	mratsim: are you speaking to me?
13:47:06	FromDiscord	<mratsim> yes
13:47:37	FromDiscord	<mratsim> you want to store a proc in a variable and you need a workaround for the vm?
13:48:01	Araq	"We will see however that macros written in a
13:48:02	Araq	continuation-passing style (CPS) always compose."
13:48:06	Araq	lol, CPS again
13:48:10	Yardanico	haha
13:48:21	dadada	mratsim: again yes in general modules will work at both times, but I had a case where I needed a global variable to store state, and this won't simply work, it either worked in compile or runtime depending if I defined it as {.compileTime.} or not, and then there's the when nimvm: solution, but this would mean you'd have to repeat this trick for every module that needs this, and with .uberTime. it just gets
13:48:27	dadada	very very simple
13:48:44	Yardanico	@kaushalmodi maybe I need to play with nimforum a bit so this forum is not just some throwaway test instance :P
13:48:51	Yardanico	like trying to use https://github.com/soasme/nim-markdown
13:48:57	shashlick	@Yardanico did your libvlc wrapper work? I tried it yesterday and vlc_object.h was required
13:49:07	FromGitter	<kaushalmodi> Yardanico: That would be awesome
13:49:16	Yardanico	well the thing is that it worked when I was testing it with older nimterop @shashlick, I'll try again now
13:49:55	dadada	mratsim: no I need simple variables, I just store them in procs, to workaround Nim's limitations
13:49:59	shashlick	Ok cool
13:50:23	dadada	but I assume uberTime could work with any type of variable proc or not, but I haven't thoroughly tested it yet, it's just the prototype that finally runs
13:50:37	FromDiscord	<mratsim> @Araq, I know this guy, he is the one behind tagless final, and also behind a DSL for linear algebra in OCAML
13:50:37	FromDiscord	<mratsim> - http://okmij.org/ftp/meta-programming/HPC.html
13:50:37	FromDiscord	<mratsim> - https://github.com/mratsim/Arraymancer/issues/347#issuecomment-461009747
13:50:37	FromDiscord	<mratsim> - http://okmij.org/ftp/meta-programming/lift/convolution.ml
13:50:37	FromDiscord	<mratsim> - http://okmij.org/ftp/tagless-final/
13:50:40	disbot	➥ discussion: porting Halide to nim ; snippet at 12https://play.nim-lang.org/#ix=2lSX
13:50:42	dadada	I've to do some test cases and I know that assignments probably don't work yet, but that should be solvable
13:50:56	FromDiscord	<mratsim> @dadada, just store the symbol in tables
13:51:25	dadada	mratsim: what if I don't want the overhead of a table? and why would a table magically work at compile and runtime?
13:51:28	Yardanico	shashlick: yeah it works for me
13:51:33	Yardanico	how can I try with ast2?
13:51:46	Yardanico	the same -f:ast2 ?
13:51:58	Yardanico	ah maybe I have the old toast binary, let's see
13:52:00	shashlick	Do you have latest nimterop?
13:52:05	FromDiscord	<mratsim> But since 2002, DSL composition has significantly evolved: https://github.com/mratsim/Arraymancer/issues/347#issuecomment-461009747
13:52:05	FromDiscord	<mratsim> especially since the combining Deep and Shallow Embedding DSL paper: http://www.cse.chalmers.se/~josefs/publications/TFP12.pdf
13:52:07	disbot	➥ discussion: porting Halide to nim ; snippet at 12https://play.nim-lang.org/#ix=2lSX
13:52:08	*	Senketsu joined #nim
13:52:23	FromGitter	<bung87> @PMunch should I using `uint32()` or `uint64()` or just aways map to uint64?
13:52:25	shashlick	Also, someone tried yesterday on windows
13:52:35	Yardanico	shashlick: yeah I know I saw it
13:52:41	dadada	it looks all you ever get here is negativity when you don't go the same path as everybody else
13:52:55	FromDiscord	<mratsim> @dadada, tables work at runtime and compile-time.
13:53:10	PMunch	bung87, depends on what the underlying type is..
13:53:18	shashlick	You might want to import vlc_common.h
13:53:23	dadada	mratsim: of course they do, but do they work at both at the same time?!
13:53:47	FromGitter	<bung87> em , it's mapping from js to nim, so I dont know the type
13:53:47	PMunch	dadada, why do you say that?
13:53:50	FromDiscord	<mratsim> I'm not sure why you are talking about negativity
13:54:00	Yardanico	shashlick: on latest nimterop it seems to fail, yeah
13:54:08	Yardanico	after adding "elif sym.name[0] == '_': sym.name = sym.name[1..^1]" it complains it can't find media_track_t
13:54:11	dadada	the interesting part of .uberTime. is at the bottom in the expandMacros section ... this is what becomes easier
13:54:42	dadada	the rest is just implementation details
13:54:49	Yardanico	so how do I use ast2 and stuff?
13:55:07	FromDiscord	<mratsim> I don't understand what's different at the bottom from {.compileTime.}
13:55:30	FromDiscord	<mratsim> that's why i asked what problem does it solve
13:55:56	Yardanico	ah nvm sorry
13:56:40	Yardanico	shashlick: it seems to actually have worked with ast2 lol
13:57:17	Yardanico	nice work :)
13:58:04	Araq	mratsim: well time for me to get up to research
13:58:08	dadada	mratsim: actually you're right, for this example, this is because they're both evaluluated at compile time I assume, the compiler figures out that test2 never changes and precompiles it all?! but I had a module where you don't know the value of test2 beforehand, so real runtime, and also needed it for compiletime, too, and there it would fail at compile time either for the one or the other case
13:58:33	dadada	mratsim: and this SHOULD solve it, although I'm not sure I guess because I haven't really tried it on the module that inspired this work
13:59:20	FromDiscord	<mratsim> ah, seems similar to importc constants
13:59:29	PMunch	Araq, did you reply to my stdlib vs. * question by the way?
14:00:16	dadada	PMunch: because I know that I don't write code simply because "let's see where this gets me", I always have some personal use case that currently doesn't work without workarounds and I ask in this channel to find the best known solutions, and when I don't find something that I like, I try to package something for easy future use
14:00:18	FromDiscord	<mratsim> Research is all there, and my ML/Linear algebra DSL is inspired by it
14:00:18	FromDiscord	<mratsim>
14:00:19	FromDiscord	<mratsim> Developping extensible shallow embeddings (functions in host lang) on top of deep embeddings (fixed AST/IR):
14:00:19	FromDiscord	<mratsim>
14:00:19	FromDiscord	<mratsim> Folding Domain Specific Language: Deep and Shallow Embedding
14:00:19	FromDiscord	<mratsim> Combining Deep and SHallow Embedding for EDSL and paper.
14:00:21	FromDiscord	<mratsim> Yin-Yang: Concealing the Deep Embedding of DSLs and project report.
14:00:23	FromDiscord	<mratsim> Uniting Language Embedding for Fast and Friendly DSLs (thesis)
14:00:25	FromDiscord	<mratsim> Rewriting a Shallow DSL using a GHC extension
14:00:27	FromDiscord	<mratsim> An Image Processing Language: External and Shallow/Deep Embeddings
14:00:31	FromDiscord	<mratsim> bottom of this issue: https://github.com/mratsim/Arraymancer/issues/347#issuecomment-461009747
14:00:34	PMunch	@mratsim, you of all people!
14:00:35	disbot	➥ discussion: porting Halide to nim ; snippet at 12https://play.nim-lang.org/#ix=2lSX
14:00:41	dadada	PMunch: and then it's always met with: but why would you need this! ...
14:00:54	Yardanico	shashlick: https://github.com/Yardanico/nim-libvlc/commit/660a1cdc648c5b0901203571e934f8e05ce9bc16
14:00:56	Yardanico	works for me now
14:01:20	PMunch	dadada, we're just curious :P
14:01:22	dadada	I can understand it though... I guess you don't technically need this... I've written more than 20 modules now and the need has only come up once or twice so far
14:01:37	FromGitter	<bung87> `abs(-5 shr 1 shr 1)` == `5 >>> 1`
14:01:47	Yardanico	uhh no?
14:01:49	FromDiscord	<mratsim> I can't help you if I don't know what you want to do
14:01:52	Yardanico	isn't there a simpler way
14:01:57	Araq	mratsim: so ... write an RFC? how to fix Nim's template/macro system
14:02:19	PMunch	There are a lot of people who comes in here and asks for stuff that might be technically possibly, but after asking them why and showing them the "royale" way of doing it then they go "oh that's much easier!"
14:02:21	FromDiscord	<mratsim> I'm OK with the macro system? What's wrong with it? Besides maybe needing "breeze"
14:02:25	dadada	mratsim: you mean me? :D I didn't ask for help?!
14:02:33	shashlick	@Yardanico try removing some of your skips and overrides
14:03:07	FromDiscord	<mratsim> yeah but We can't share your joy if we don't understand what you accomplished as well
14:03:22	PMunch	dadada, what is it that you're trying to do now by the way? Haven't been following the chat closely
14:03:29	FromDiscord	<mratsim> it's not stackoverflow, we won't say, please use OOP, SOLID, Factory instead 😉
14:03:41	dadada	PMunch: well I asked leorize for the royale way earlier, and I assume he is a top nim guy, and his solution was to use when nimvm, I needed when nimvm for it and it indeed worked, but I don't like to have to use when nimvm, when I can do it with a simple pragma, and now there's one
14:04:24	Yardanico	shashlick: can't remove video_setup_device_info_t and 3 objects which it's using
14:04:27	Araq	we should though, 'SOLID', it's superb, just look at the term, how can you be against "solid"?
14:04:31	Yardanico	https://git.videolan.org/?p=vlc.git;a=blob;f=include/vlc/libvlc_media_player.h;h=a6f5fbf8e58113f3f9b0e1a5ca96ae21588b63e1;hb=HEAD#l522
14:04:32	dadada	PMunch: difficult question, I'm developing multiple modules, this is just a small piece of the puzzle, it's for another module that already worked, but I needed it to work at compile and runtime
14:04:41	dadada	and that's where .uberTime. came in
14:04:43	PMunch	Ah, so you wanted a pragma that applies "when defined(nimvm)"? That seems quite useful actually
14:05:09	dadada	admittedly you don't need it very often
14:05:13	PMunch	uberTime?
14:05:19	dadada	uber = above
14:05:34	dadada	it's above run and compile, so it's uber them :D
14:06:21	PMunch	Haha, I know what über means, still not sure I understand what it does..
14:06:30	dadada	PMunch: https://play.nim-lang.org/#ix=2lST
14:06:49	dadada	I've no real great examples yet, besides my one module this was made for
14:07:01	disruptek	you guys should use the mumble whenever it's useful.
14:07:04	disruptek	it's free.
14:07:07	PMunch	Oh, like not compileTime or runTime but uberTime
14:07:08	PMunch	:P
14:07:12	*	Hideki joined #nim
14:07:13	PMunch	I thought it was some kind of timing thing :P
14:07:29	dadada	the expandMacros section can also be compiled with .compileTime. as mrratsim pointed out, but it works for broader cases
14:07:32	*	someunknownuser quit (Quit: someunknownuser)
14:07:35	*	Hideki is now known as Guest60418
14:07:39	FromDiscord	<clyybber> disruptek: its also empty :d
14:07:46	dadada	no, it's not about timing it's about access and modification of variables at all times
14:07:52	disruptek	well, ima gonna stream early, i guess.
14:08:25	PMunch	dadada, so what happens if you change a uberTime variable on compileTime and then use it at runtime?
14:08:29	PMunch	Do you get the updated value?
14:08:53	dadada	PMunch: no it's not meant for that, it's just for simplification of module code, the values are reset for runTime
14:09:02	PMunch	Ah, right
14:09:22	dadada	when I release this I guess I've to write an FAQ :D
14:09:48	dadada	basically think of compile time and run time as targets
14:09:49	PMunch	So it's basically just a way to re-use the name/type of a variable on both run- and compiletime?
14:10:00	dadada	and uberTime allows for a global var to work the same way in both targets
14:10:06	dadada	exactly
14:10:25	FromDiscord	<Technisha Circuit> How would i call Nim from Python?
14:10:48	dadada	nimpy I think, I haven't used it though
14:11:17	PMunch	@Technisha Circuit: https://robert-mcdermott.gitlab.io/posts/speeding-up-python-with-nim/
14:11:30	*	solitudesf- joined #nim
14:11:33	PMunch	~pythonnim
14:11:34	disbot	no footnotes for `pythonnim`. 🙁
14:11:36	FromDiscord	<Technisha Circuit> That's for calling Python from Nim
14:11:44	PMunch	~pythonnim https://robert-mcdermott.gitlab.io/posts/speeding-up-python-with-nim/
14:11:44	disbot	no footnotes for `pythonnim`. 🙁
14:11:48	Yardanico	@Technisha no
14:11:56	Yardanico	that's for calling Nim from Python
14:11:56	PMunch	disruptek, how do I register a footnote again?
14:12:02	Yardanico	"footnote is" afaik
14:12:06	FromDiscord	<Technisha Circuit> What?
14:12:08	Yardanico	with ~ probably
14:12:11	Yardanico	@Technisha check the article again
14:12:14	PMunch	~pythonnim is https://robert-mcdermott.gitlab.io/posts/speeding-up-python-with-nim/
14:12:14	disbot	pythonnim: 11https://robert-mcdermott.gitlab.io/posts/speeding-up-python-with-nim/
14:12:15	FromDiscord	<Technisha Circuit> Okay
14:12:16	PMunch	Ah
14:12:20	PMunch	~pythonnim
14:12:21	disbot	pythonnim: 11https://robert-mcdermott.gitlab.io/posts/speeding-up-python-with-nim/ -- PMunch
14:12:25	FromDiscord	<Technisha Circuit> That's for writing modules in Nim
14:12:26	dadada	PMunch: it's possible I overcomplicated it and there's an easier solution, I wanted this to work for all types of types, that's why I used ptrs, because NimScript can't deal with refs to refs for example
14:12:30	Yardanico	@Technisha yes
14:12:34	Yardanico	but for using them in Python
14:12:43	Yardanico	it's https://github.com/yglukhov/nimpy
14:12:44	FromDiscord	<Technisha Circuit> I want to call Python from Nim without making a moduke
14:12:47	FromDiscord	<Technisha Circuit> Module*
14:12:55	Yardanico	pls decide
14:12:59	Yardanico	python from nim or nim from python?
14:13:06	PMunch	Haha, that was the opposite of what you asked for :P
14:13:08	Yardanico	https://github.com/yglukhov/nimpy works for both anyway
14:13:08	FromDiscord	<Technisha Circuit> Wait
14:13:14	Yardanico	https://github.com/yglukhov/nimpy#calling-python-from-nim
14:13:14	FromDiscord	<Technisha Circuit> I'm stupid
14:13:26	FromDiscord	<Technisha Circuit> Thanks :P
14:13:55	PMunch	~pastespam is Please don't paste code and/or multiline content on Discord. It comes through on IRC and is very annoying
14:13:55	disbot	pastespam: 11Please don't paste code and/or multiline content on Discord. It comes through on IRC and is very annoying
14:14:06	Yardanico	PMunch: that should change soon :P
14:14:08	PMunch	Nice, that'll save me some typing :P
14:14:19	PMunch	Wasn't it active here?
14:14:24	*	solitudesf quit (Ping timeout: 256 seconds)
14:16:14	Yardanico	PMunch: only for a bit
14:16:20	Yardanico	but when I found out discord stopped sending member list to me
14:16:26	Yardanico	it's all fine now about the member list though
14:16:28	PMunch	Right..
14:16:40	PMunch	So it can be made active here again?
14:16:47	Yardanico	well ~maybe`
14:16:48	Yardanico	~
14:16:48	disruptek	~paste
14:16:49	disbot	no footnotes for ``. 🙁
14:16:49	disbot	paste: 11a frowned-upon behavior in chat; please use a service such as https://play.nim-lang.org/ or http://ix.io/ or https://gist.github.com/ and supply us a URL instead. -- disruptek
14:17:11	PMunch	Ah we already had one for that
14:17:47	PMunch	If I add a footnote to something that's already registered, will it make another one or overwrite?
14:17:52	disruptek	maybe i should expose delete for footnotes today.
14:18:05	disruptek	it's additive; it only overwrites your own contributions.
14:18:24	disruptek	it's good to have both; it's like autocorrect.
14:18:27	*	hpyc9 quit (Ping timeout: 260 seconds)
14:18:54	PMunch	Ah right
14:19:00	PMunch	Just make sure delete does the same :P
14:19:26	Yardanico	PMunch: I can try launching it for a bit in this chat, not from nim vps though because that'll require me to cross-compile and stuff
14:19:36	PMunch	~playground is https://play.nim-lang.org the official Nim playground, can run Nim in the browser and share snippets
14:19:37	disbot	playground: 11an online in-the-browser IDE for simple Nim experiments at http://argentina-ni.ml/
14:19:37	disbot	playground: 11https://play.nim-lang.org the official Nim playground, can run Nim in the browser and share snippets
14:19:42	*	hpyc9 joined #nim
14:19:44	disruptek	heh
14:19:48	PMunch	Hihi
14:19:51	Yardanico	hoho
14:20:15	FromGitter	<Clyybber> @cooldome Hey, do you maybe have a smaller snippet that causes the compiler to hang with my ping-pong PR?
14:20:28	disruptek	maybe when the bot sees a naked github link it can expand it like it does for repo calls.
14:20:29	dadada	PMunch: okay, looks like testcases for this will be difficult to make, unittests aren't prepared for compile time testing? check() for example doesn't seem to work there
14:20:37	disruptek	!repo naked
14:20:41	disbot	https://github.com/jostmey/NakedTensor -- 9NakedTensor: 11Bare bone examples of machine learning in TensorFlow 15 2471⭐ 158🍴 7& 29 more...
14:21:08	PMunch	Where does it pull those from?
14:21:17	Yardanico	github search
14:21:24	Yardanico	so I can try launching my bridge here guys
14:21:25	PMunch	Just the top one?
14:21:34	Yardanico	if it crashes a lot pls ping me and I'll start the old one :P
14:21:46	disruptek	just the top one?
14:21:56	Yardanico	~changing bridges~
14:21:57	disbot	no footnotes for `changing`. 🙁
14:21:59	*	FromDiscord quit (Remote host closed the connection)
14:22:04	PMunch	Haha :P
14:22:10	*	Senketsu quit (Ping timeout: 244 seconds)
14:22:28	PMunch	Yeah disruptek I was wondering if !repo word just searches for word on github and shows the top one
14:22:30	*	nl joined #nim
14:22:30	*	nl quit (Remote host closed the connection)
14:22:35	Yardanico	wait wrong nickname
14:23:18	*	FromDiscord joined #nim
14:23:18	*	krisppurg joined #nim
14:23:22	Yardanico	done
14:23:27	PMunch	Hello discord!
14:23:31	PMunch	(again)
14:23:47	Yardanico	yeah you can ping now, although it's still not the best it works quite well
14:23:53	FromDiscord	<Technisha Circuit> If I'm making a python module in Nim, is it better to use the print function from `pyBuiltinsModule` or is it better to use echo from Nim
14:23:55	FromDiscord	<Technisha Circuit> Nice
14:24:06	Yardanico	@technisha try pasting code pls
14:24:07	FromDiscord	<Technisha Circuit> Hi PMunch
14:24:12	FromDiscord	<Technisha Circuit> Okay
14:24:36	FromDiscord	<Technisha Circuit> ```python↵print("""This↵Is↵Redundant""")```
14:24:44	Yardanico	that's too small
14:24:53	FromDiscord	<Technisha Circuit> Okay :P
14:25:05	FromDiscord	<Yardanico> sent a code paste, see http://ix.io/2lTb
14:25:22	FromDiscord	<Technisha Circuit> sent a code paste, see http://ix.io/2lTc
14:25:25	FromDiscord	<Technisha Circuit> That fine?
14:25:30	Yardanico	yes
14:25:43	zacharycarter	just use echo?
14:25:45	dadada	so what do I do when I need at compile time testing?
14:25:55	dadada	putting unittest in a static block doesn't seem to work
14:26:12	dadada	https://forum.nim-lang.org/t/5261 <- they also didn't seem to get far with it
14:26:19	FromDiscord	<Technisha Circuit> Okay
14:27:57	*	PMunch quit (Quit: leaving)
14:28:08	Yardanico	~stream
14:28:08	disbot	stream: 11https://twitch.tv/disruptek (live video/audio) and mumble://uberalles.mumbl.io/ (live voice chat)
14:28:10	Yardanico	hop on guys
14:29:45	FromDiscord	<Yardanico> pic test https://media.discordapp.net/attachments/371759389889003532/710136704899481650/2020-05-13-11-28-25_grim.png
14:31:09	*	nsf quit (Quit: WeeChat 2.8)
14:32:19	FromGitter	<bung87> @dadada just using assert
14:33:40	*	Guest60418 quit (Ping timeout: 256 seconds)
14:37:14	*	waleee-cl joined #nim
14:38:11	*	silvernode joined #nim
14:47:50	FromGitter	<bung87> am I offline? no new message
14:48:57	Yardanico	no you're online, don't expect people to always answer :)
14:51:12	FromGitter	<bung87> ok, I thought this channel getting hot ,suddenly no one talks..
14:51:39	dadada	bung87: no you're there! thanks for the tip, assert works, but for the most part
14:51:57	*	kradnoel quit (Remote host closed the connection)
14:52:39	FromGitter	<bung87> ok
14:52:50	*	kradnoel joined #nim
14:55:28	*	liblq-dev quit (Quit: WeeChat 2.8)
15:01:37	FromDiscord	<Anuke> How does one convert a `ptr uint8` (returned glGetString) to a `cstring`?
15:01:40	*	kradnoel quit (Ping timeout: 265 seconds)
15:02:10	Yardanico	@Anuke cast[cstring](ptr uint8)
15:02:11	Yardanico	?
15:02:53	FromDiscord	<Anuke> thanks
15:03:37	FromDiscord	<Anuke> Assuming I convert it to a string with `$` and use it somewhere else, what will happen to that `cstring` / memory?
15:03:50	FromDiscord	<Anuke> do I need to manage it at all?
15:05:22	dadada	hmm, one of my procs doesn't seem to register as an overload for a nim generated []= assignment operator
15:05:25	FromDiscord	<mratsim> the cstring is collected at end of function
15:05:46	dadada	I exported it, so why
15:05:52	FromDiscord	<mratsim> it's just a stak variable
15:05:57	*	gokr quit (Quit: Leaving.)
15:06:22	FromDiscord	<mratsim> ooh IRC replies are not bundled together anymore into a blur
15:06:39	Yardanico	yeah feels weird right? :P
15:06:46	FromDiscord	<FromIRC> Uptime - 43 minutes, 34 seconds, 298 milliseconds, 720 microseconds, and 164 nanoseconds
15:06:46	FromDiscord	<Yardanico> !status
15:06:49	FromDiscord	<mratsim> @dadada I use static: doAssert
15:07:03	FromDiscord	<mratsim> @Yardanico https://status.im/
15:07:08	FromDiscord	<Yardanico> hahaha
15:07:13	*	fredrikhr joined #nim
15:07:22	dadada	mratsim: thanks man! that works bung87 also told me to
15:07:37	Yardanico	dadada: you can use @ to mention him btw :P like @mratsim
15:07:49	Yardanico	*showing off ircord*
15:07:58	dadada	not really need in irssi
15:08:05	dadada	needed
15:08:11	Yardanico	dadada: no I mean so that he sees it on Discord
15:08:20	FromDiscord	<mratsim> did you solve your prime factorization problem btw?
15:08:22	dadada	ah, that's cool then
15:08:53	Yardanico	@mratsim well not yet, I did other stuff :P
15:08:58	FromDiscord	<codic> ```nim↵echo("yay i can paste code now. lorem ispum dolor")↵```
15:09:12	Yardanico	ah right I'm stupid
15:09:17	Yardanico	need to send playground link instead of ix.io
15:09:27	Yardanico	because it can use ix.io as well
15:09:44	FromDiscord	<codic> it sends pastes to an external service?
15:09:49	FromDiscord	<codic> nice
15:09:51	Yardanico	ix.io if that works
15:09:55	Yardanico	if it fails - gives link to a discord message
15:10:02	FromDiscord	<codic> ah cool
15:10:07	dadada	Yardanico: can I force a builtin operator to either not be generated, be removed, or be overloaded with my own, I've tried with `[]=`[T](t: var Object[T], newVal: T) = ... but the thing is never found ... also tried `[]=`[T](t: var Object[T], newVal: Object[T]) = ... but the internally generated one is still used
15:10:33	Yardanico	sorry, I don't really know about that :( maybe you can try without generics?
15:10:33	dadada	oh and I exported of course with `[]=`*[T](...
15:10:37	krisppurg	owo
15:10:44	FromDiscord	<codic> sent a code paste, see https://discordapp.com/channels/371759389889003530/371759389889003532/710147015635501106
15:11:19	dadada	also tried without generics, not working
15:13:22	*	silvernode quit (Ping timeout: 260 seconds)
15:16:22	FromDiscord	<Recruit_main707> https://play.nim-lang.org/#ix=2lTx is this possible?
15:16:41	Yardanico	yes why not?
15:16:46	Yardanico	oh wait
15:16:49	Yardanico	what is "index"
15:17:14	FromDiscord	<Recruit_main707> whatever its specified later
15:17:34	FromDiscord	<Recruit_main707> (the index of the root Bot class)
15:17:41	Yardanico	well what exactly you think is not possible?
15:18:04	Yardanico	you can store non-ref types in container types
15:18:10	Yardanico	even if it's like a recursive thing
15:18:23	FromDiscord	<Recruit_main707> `undeclared identifier: 'index'`
15:18:31	FromGitter	<kaushalmodi> `me*: index` <-- doesn't the right side need to be a type?
15:18:33	Yardanico	yes, that's not possible at all
15:18:39	Yardanico	I asked you about "index"
15:18:42	FromGitter	<kaushalmodi> here you are using a variable element from parent obj
15:18:46	FromDiscord	<Recruit_main707> thx
15:18:56	FromDiscord	<Recruit_main707> good call 😅
15:20:56	*	poohman joined #nim
15:32:58	dadada	I've just one final issue and that's really destroying all my hopes, nim always executes macros first, so macros don't see the world as the result of my pragma (which ironically is also a macro)
15:33:24	dadada	so, if I just could force nim to evaluate my pragma before the macros, everything should work alright
15:34:15	dadada	hmm, o ooh, it actually get executed before, but why doesn't it work... my hypothesis was wrong
15:39:29	dadada	ah, just another tiny mistake looking far worse than it was
15:47:22	dadada	works! hahaaa
15:48:08	krux02	dadada: evaluation order can be controled via typed and untyped macros.
15:48:31	FromGitter	<kaushalmodi> dadada: That much be some emotional rollercoaster
15:48:35	FromGitter	<kaushalmodi> s/much/must
15:49:51	dadada	kaushalmodi: basically I doubted myself, I didn't know if this was doable, and then very shortly before the finish line I thought all was for naught, but with my last changes, this works on my modules I initially wanted it for, beautifully
15:50:35	dadada	so yes, total rollercoaster
15:52:02	krux02	yea there are no macros like Nim macros.
15:53:35	FromDiscord	<mratsim> I yes the "I didn't know it was doable, but then somehow it worked" happens to me everytime
15:54:48	*	rockcavera quit (Remote host closed the connection)
15:55:17	dadada	do we really need old nim manuals like 0.17.2 online? I keep googling and clicking on wrong links and then I search the old manual ... without noticing
15:55:26	dadada	only to later regret it because so much has changed
15:55:34	dadada	their entries appear very high on google
15:55:49	dadada	if someone wants those old manuals, they should just download them from an archive
15:55:55	Yardanico	dadada: well it still might be useful to someone
15:56:05	Yardanico	and anyway the results to actual newest versions are the first ones
15:56:16	dadada	Yardanico: never doubted that, but you should put the needs of the majority first... and this gets confusing
15:56:38	FromGitter	<sheerluck> you guys have manuals? o_0
15:56:45	dadada	Yardanico: not in all cases, sometimes the old ones come first, depends on the search terms
15:56:56	Yardanico	@sheerluck nim manual is really useful
15:57:06	Yardanico	I give it to people asking questions here from time to time :P
15:57:11	FromGitter	<kaushalmodi> dadada: FWIW I just bookmark https://nim-lang.github.io/Nim/theindex.html
15:57:14	Yardanico	dadada: well that didn't happen for me
15:57:17	narimiran	dadada: just bookmark this: https://nim-lang.github.io/Nim/manual.html and enjoy the latest and greatest
15:57:31	*	silvernode joined #nim
15:57:32	narimiran	haha, almost the same message as kaushalmodi :)
15:57:35	Yardanico	oh we're at 1.3.5 now, didn't notice
15:57:42	narimiran	neither did i
15:57:46	Yardanico	I was still thinking it was 1.3.3
15:57:52	*	ptdel joined #nim
15:58:07	*	abm joined #nim
15:58:25	narimiran	and i immediately guessed correctly who bumped the version number :)
15:58:36	*	ptdel quit (Remote host closed the connection)
15:58:38	Yardanico	:DDD
15:59:12	FromGitter	<kaushalmodi> Yardanico: 1.3.5 happens few mins back in Timothee's PR
15:59:31	FromGitter	<kaushalmodi> s/happens/happened
15:59:41	FromGitter	<kaushalmodi> *too many typos today*
15:59:49	Yardanico	well I already have it locally
15:59:57	Yardanico	living on the edge
16:00:00	FromDiscord	<mratsim> I often get 0.13.0 manual from some reason
16:00:44	FromDiscord	<mratsim> but we can add "canonical" url to the manual or maybe even a latest version field if GOogle supports that?
16:01:09	FromDiscord	<mratsim> because the content is duplicated and Google will lower the ranking
16:01:19	*	abm quit (Client Quit)
16:01:27	FromGitter	<kaushalmodi> (https://files.gitter.im/nim-lang/Nim/Kc15/image.png)
16:01:29	FromGitter	<kaushalmodi> about the nim manual, I have set this bookmark in Firefox
16:01:56	FromGitter	<kaushalmodi> with that, I just type Ctrl+L `nimd manual` and I jump to the manual
16:02:05	FromGitter	<kaushalmodi> similarly `nimd sequtils`, and so on
16:02:50	*	ptdel joined #nim
16:03:06	FromGitter	<kaushalmodi> > but we can add "canonical" url to the manual or maybe even a latest version field if GOogle supports that? ⏎ ⏎ +1 ⏎ ⏎ ^ @narimiran [https://gitter.im/nim-lang/Nim?at=5ebc1a39852efc0adad7336a]
16:03:21	narimiran	PR away
16:07:47	FromGitter	<kaushalmodi> I haven't used Jekyll; you basically need to add this one metadata tag for each page: https://moz.com/learn/seo/canonicalization
16:08:09	FromGitter	<kaushalmodi> where each older version page's rel canonical points to an equivalent page on nim devel docs
16:12:16	*	Hideki joined #nim
16:12:39	*	Hideki is now known as Guest73990
16:12:59	shashlick	does set -e not work as expected on osx?
16:16:52	*	Guest73990 quit (Ping timeout: 272 seconds)
16:17:00	FromGitter	<kaushalmodi> may be you need to install GNU bash?
16:17:52	FromGitter	<kaushalmodi> from a random search: https://itnext.io/upgrading-bash-on-macos-7138bd1066ba?gi=78a91d651343
16:18:32	*	rockcavera joined #nim
16:19:17	shashlick	okay thanks
16:31:49	krux02	does macos lack behind in bash as well?
16:33:32	dadada	getAst apparently doesn't get you the raw original content of a template, you get procs and variable names with gensymfoo attached instead
16:33:47	dadada	is there something that does the same as getAst but without meddling with ident names?
16:33:54	dadada	or can I teach getAst to behave
16:34:32	dadada	krux02: yes, they don't want gplv3 on mac, and therefore they stick to the old gplv2 bash
16:35:01	Yardanico	dadada: make a template dirty?
16:35:21	krux02	ok, then there is at least a reason behind it that is not to make gnu software look poor compared to shiny apple solutions.
16:35:52	dadada	Yardanico: I'll embed the thing in a quote do anyway, where I assume the gensyms will then get added twice
16:36:56	dadada	krux02: well, that's one of the reasons, they do not really like the GNU, and especially with GPLv3's opinions of not forcing users behind closed software on Tivos an the like, Apple doesn't want a part of it
16:38:08	krux02	Well, apple does sort of the opposite of microsoft.
16:38:19	dadada	meaning?
16:39:01	krux02	I would expect bash on windows to get all sorts of extensions that rely on windows builtin feautures that can't be reproduced on other platforms.
16:39:16	krux02	embrace extend extinguish
16:39:54	dadada	hmm, I think WSL is pretty cool, I've never heard it adds anything of the embrace extend extinguish fashion, deb packages just seem to work on Windows... who would have ever thought that?
16:40:10	dadada	still not making me consider Windows as a OS though :-)
16:40:39	dadada	s/as a/as my/
16:41:05	krux02	yea, I kind of like minimalism. And I would really like to know if there are any operating systems out there that try to be as minimal as possible?
16:41:15	krux02	probably minix
16:41:21	Yardanico	well there's KISS but it's a linux distro at the end of the day
16:41:36	Yardanico	https://k1ss.org/
16:41:36	FromGitter	<sheerluck> krux02 gentoo
16:41:52	FromDiscord	<Clyybber> krux: OpenBSD, minix, plan9
16:41:52	krux02	it is still a Linux distro
16:41:56	FromDiscord	<Clyybber> redux
16:42:02	dadada	krux02: the Linux foundation is supporting a new microkernel, I think it's pretty minimalistic and well tested, there was an article about it
16:42:23	krux02	The problem I see with linux is, they keep adding stuff to the kernel. It always grows.
16:42:31	Yardanico	new drivers
16:42:34	FromDiscord	<Clyybber> because the world grows
16:42:36	FromDiscord	<Clyybber> and the needs
16:42:40	FromDiscord	<Clyybber> of the people
16:42:46	krux02	I don't have a problem with new drivers. That is a good thing.
16:43:09	krux02	But probably it some point in the future it might become unmaintainable because there is just too much code written by people long dead.
16:43:29	Yardanico	then we'll have a new kernel/os :P
16:43:39	dadada	actually I see it as the great thing, the network effect is basically what makes Linux successful, and if you look at for example what's going on with Linux filesystems, you'd rather have one too many than one too few, personally all I need is ext4, but bcachefs will be very interesting for many once it's there
16:43:52	Yardanico	bcachefs is meh
16:43:56	FromDiscord	<Clyybber> nah
16:43:56	Yardanico	btrfs has been around for years
16:43:59	FromDiscord	<Clyybber> its btrfs but better
16:44:01	dadada	Yardanico: why?
16:44:07	krux02	There is this rust operating system with a true microkernel. I might take a look into that at some point in the future.
16:44:14	dadada	Yardanico: did you read about all the design shortcoming of btrfs?
16:44:44	FromDiscord	<Clyybber> they have been fixed
16:44:48	FromDiscord	<Clyybber> btrfs is fine nowadays
16:44:52	FromDiscord	<Clyybber> but bcache is faster
16:44:56	FromDiscord	<Clyybber> *fs
16:45:15	FromDiscord	<Clyybber> krux02: That *is* redux
16:45:18	dadada	Clyybber: it's not btrfs, it has made completely different design choices... but I don't want to go into this, because I wanted to get work done
16:45:23	FromDiscord	<Clyybber> *redox
16:45:51	FromDiscord	<Clyybber> dadada: Of course its not btrfs, but it serves the same purpose
16:46:01	FromGitter	<sheerluck> in gentoo you can remove everything from kernel. It will boot, print "Hello krux02" and then poweroff. No microkernel will ever compete with that
16:46:04	*	poohman quit (Ping timeout: 265 seconds)
16:46:18	krux02	I always mix up redox os with react os
16:46:50	krux02	sheerluck: that is interesting.
16:47:03	krux02	and Gentoo has been around for quite some time and it doesn't seem to die.
16:47:43	FromDiscord	<Clyybber> void linux is pretty nice
16:48:05	FromDiscord	<Rika> hello
16:48:30	krux02	I am actually pretty happ with Arch Linux as a linux distribution.
16:48:54	*	couven92 joined #nim
16:49:19	krux02	I just don't like the concept of an always growing kernel.
16:51:36	*	fredrik92 joined #nim
16:51:44	*	fredrikhr quit (Disconnected by services)
16:51:48	*	fredrik92 is now known as fredrikhr
16:52:08	*	xet7 quit (Remote host closed the connection)
16:53:01	*	poohman joined #nim
16:54:14	*	xet7 joined #nim
16:55:25	*	couven92 quit (Ping timeout: 264 seconds)
16:58:44	*	xet7 quit (Remote host closed the connection)
16:59:19	FromGitter	<sheerluck> this is how fresh gentoo looks like -- https://i.imgur.com/vKdJaOf.jpg nothing is running, nothing is installed. Even mlocate, even dhcpcd
16:59:36	Yardanico	well gentoo doesn't define what you install
16:59:41	Yardanico	so you can't really say gentoo is "minimal"
16:59:51	Yardanico	if you install systemd + plasma with all apps it won't be exactly "minimal"
16:59:57	livcd	sheerluck: yeah but why would you use it?
17:01:06	*	xet7 joined #nim
17:02:26	dadada	I know I can't be cool as a Fedora user, but I've never found a compelling reason to switch, the only thing is maybe that Arch has newer packages faster, but it's possible to install Arch package on Fedora with some tricks... I think people want to have a distro they believe in, it's not only about the technical aspects, they want to be different from everybody else
17:05:27	zacharycarter	is everyone 12?
17:06:03	FromGitter	<sheerluck> livcd That is a philosophical question. You must decide for yourself why you need Gentoo. Most people do not need Gentoo. They usually say "I don't like compiling". And they are right. Nobody likes compiling. Well, except me.
17:07:04	dadada	zacharycarter: it can be seen as a good thing to keep the inner child alive, so yes, many are and they're not
17:07:12	livcd	sheerluck: most of my devices are not suitable for compiling anything bigger than the nim compiler
17:07:28	livcd	Araq: do you think Nim would be more popular if you were an american?
17:08:02	Yardanico	lol
17:08:11	zacharycarter	is this a serious question?
17:08:17	FromDiscord	<Rika> ok what
17:08:20	FromDiscord	<Rika> lmao
17:08:23	FromDiscord	<Clyybber> ?
17:08:32	livcd	it is
17:09:02	FromDiscord	<exelotl> let's use a bald eagle as the nim mascot :p
17:09:09	zacharycarter	bald eagle with guns
17:09:11	FromGitter	<kaushalmodi> ah crap, Araq isn't American? /s
17:09:17	Yardanico	smh
17:09:20	Yardanico	im going to zig
17:09:21	zacharycarter	yeah I'm leaving guys sorry
17:09:25	alexander92	hehehehe
17:09:26	zacharycarter	thought Araq was US born and bred
17:09:28	Yardanico	zig creator is a true american at least
17:09:34	zacharycarter	peace out
17:09:37	alexander92	new yorker!
17:09:47	FromDiscord	<exelotl> @zacharycarter Heck yeah
17:10:06	alexander92	andrewk are you a true american
17:10:12	Yardanico	he's not here
17:10:15	livcd	I am serious. I am not trolling.
17:10:24	*	couven92 joined #nim
17:10:27	zacharycarter	it's a ridiculous question though isn't it?
17:10:29	*	fredrikhr quit (Disconnected by services)
17:10:30	FromDiscord	<Generic> I would question whether Nim would have ever come to be in America
17:10:32	*	lritter joined #nim
17:10:32	*	couven92 is now known as fredrikhr
17:10:33	alexander92	well
17:10:37	FromDiscord	<Generic> considering the Pascal influence
17:10:38	Yardanico	I remember when andrewk joined to congratulate nim on 1.0 :P
17:10:40	alexander92	many designers are not american
17:10:42	Yardanico	and guess what 4raq said xDD
17:10:43	alexander92	so i dont think so
17:10:54	alexander92	japanese, danish, dutch, german
17:11:01	livcd	zacharycarter: Why would it be ridiculous?
17:11:03	Yardanico	R U S S I A N
17:11:08	Yardanico	xd
17:11:16	alexander92	cyrillic language man
17:11:19	Yardanico	да
17:11:20	zacharycarter	because I don't think anyone considers the nationality of a technology creator when they pick a technology
17:11:32	zacharycarter	and most people who stumble across Nim don't even know of 4raq
17:11:32	livcd	It's a serious issue that a lot of decisions in the Ruby world happen in Japan. And a lot of people simply feel "excluded".
17:11:32	alexander92	`ако ен > 2: излез(
17:11:37	Yardanico	what
17:11:39	dadada	livcd: C++ was created by a dane, so from a very very small country, it became very successful, maybe Araq in the American Araq Parallel Universe (someone please write that book) would've made better connections to promote his stuff, but apart from that I don't see why a western European country can't be good starting ground, case in point the Linux kernel was started in Finland, which is also much smaller
17:11:45	dadada	(in population) than Germany
17:11:52	zacharycarter	let's stop pinging him guys
17:12:01	Yardanico	just write it as 4raq
17:12:02	Yardanico	ez
17:12:07	alexander92	livcd eh in a way, but its about env not so much as original country
17:12:33	FromDiscord	<exelotl> I still wish we used that minim creature as the nim mascot :(
17:12:38	alexander92	now with the whole internet work, maybe it matters less
17:12:42	livcd	Just imagine someone needs support during the US working hours.
17:12:44	livcd	Most of us are asleep.
17:12:50	livcd	There's only disruptek LOL
17:12:54	Yardanico	XD
17:13:06	zacharycarter	Nim doesn't have support
17:13:12	zacharycarter	it's not like anyone gets paid to help anyone in here
17:13:14	FromDiscord	<exelotl> such a good design but rejected for being too cute and too similar in proportions to the gopher
17:13:17	Yardanico	zacharycarter: well it kinda does
17:13:27	alexander92	well, its about communities not so much crreator
17:13:35	Yardanico	zacharycarter: "Commercial support includes priority bug fixes and feature requests. All interested parties should email [email protected]."
17:13:39	livcd	zacharycarter: you know what I meant.
17:13:40	zacharycarter	and there are more Americans than disruptek
17:13:50	alexander92	zachary is almost an american
17:13:52	zacharycarter	I'm sure if you pay someone for support during US horus you can get it
17:13:56	alexander92	he want to finland
17:14:06	alexander92	so he got a little europe probably
17:14:10	zacharycarter	but I'm back in the states baby!
17:14:30	zacharycarter	but yeah Europe rubbed off on me, I'm at least like 1 percent European now
17:14:36	alexander92	i need to put some springsteen on
17:14:48	livcd	just imagine newcommers stopping by and the channel is silent
17:14:58	Yardanico	~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
17:15:01	alexander92	livcd we can just tell disruptek to spam more
17:15:24	alexander92	i dont mean that he spams, just he can use stuff like disbot
17:15:29	zacharycarter	I just think the question is a bit ridiculous
17:15:31	solitudesf-	livcd, and how does that relate to 4raq's origin?
17:15:36	*	solitudesf- is now known as solitudesf
17:15:43	alexander92	if only he was bulgarian
17:16:22	livcd	solitudesf: he would have connections to FB,Google,Amazon and not to SAP :D
17:16:39	alexander92	google have many offices in europe
17:16:42	solitudesf	.
17:16:52	alexander92	we got dom96 in facebook
17:17:12	alexander92	we just need to put yardanico in yandex
17:17:18	alexander92	and make it yardandex
17:17:22	livcd	OO
17:17:29	livcd	or jetbrains :)
17:17:43	alexander92	which also have office in munich btw
17:17:52	FromGitter	<sheerluck> Most people are Chinese
17:18:08	FromGitter	<bung87> why says that?
17:18:18	livcd	alexander92: you do realize that these big companies have offices almost in every tier1 city in the world right?
17:18:18	alexander92	it's a good thing for you man
17:18:28	livcd	yet you dont hear from Google Tokyo, Google Seoul that much
17:18:35	FromDiscord	<Rika> > livcd: It's a serious issue that a lot of decisions in the Ruby world happen in Japan. And a lot of people simply feel "excluded".
17:18:44	FromDiscord	<Rika> this is an issue of japan, not ruby, afaik
17:19:00	FromGitter	<sheerluck> I love Japan
17:19:02	FromDiscord	<Rika> anything that's from japan tends to be fairly exclusive of intl issues
17:19:11	alexander92	but you
17:19:14	FromDiscord	<Rika> same thing in the other way
17:19:17	alexander92	do you imagine how much support japan gives them
17:19:23	livcd	btw we can switch to #nim-offtopic if you mind :P
17:19:23	alexander92	they had like industrial marketing and stuff
17:19:25	dadada	someone from Japan writes moe, the vim like editor written in Nim, highly recommended that you check it out
17:19:27	alexander92	on rubykaigi
17:19:36	dadada	it's neat for editing small files without the overhead of other editors
17:19:46	alexander92	japanese people are proud of their tech, so i guess they help their own
17:19:54	FromDiscord	<Rika> but vim doesnt have much overhead either no?
17:20:20	dadada	Rika: depends on how many plugins you load, and I bet the installed size of vim is much bigger than moe
17:20:35	FromDiscord	<Rika> @Yardanico the "@nim-lang" message you sent has a ping, you should check it and fix
17:20:40	dadada	I see moe as a replacement for joe/nano, not vim
17:21:19	livcd	But I am also curious if zig builds momentum just because Andrew is an american :P
17:21:39	alexander92	we can go and ask them :D
17:21:49	FromDiscord	<Rika> > too cute↵@exelotl wtf how does a design get rejected for being too cute xd
17:22:22	livcd	alexander92: you will see that in some stats i guess...
17:22:39	FromDiscord	<Rika> speaking of vim i'm more inclined to move to it because i realized i could ssh from my laptop to my desktop and use vim on there instead of on my laptop
17:22:55	FromDiscord	<Rika> not sure if that works too well on vscode...
17:23:02	livcd	Rika: switch to emacs
17:23:08	livcd	You can use "tramp" for that
17:23:17	FromDiscord	<Clyybber> switch to kakoune
17:23:32	FromDiscord	<Rika> @_@
17:24:14	alexander92	livcd i asked them
17:24:30	livcd	someone needs to ask the tabnine guy to give it for free for Nim
17:25:25	alexander92	is it good for nim?
17:25:43	FromDiscord	<exelotl> @Rika this one https://github.com/nim-lang/RFCs/issues/104
17:25:44	disbot	➥ Nim's mascot proposal
17:25:57	livcd	alexander92: even andreas said it's not that bad
17:26:04	FromDiscord	<exelotl> Its so good!! people have bad taste smh
17:26:11	alexander92	we do need a different mascot yeah!
17:26:32	alexander92	i want to get rid of the crown too ..
17:26:45	dadada	what the heck
17:26:48	dadada	the crown rules
17:26:56	dadada	alexander92: you can't be serious
17:27:00	FromDiscord	<Rika> oh
17:27:05	FromDiscord	<Rika> i see why they rejected it
17:27:38	*	poohman quit (Ping timeout: 265 seconds)
17:28:59	zacharycarter	mascot discussions have been going on for years
17:29:11	FromDiscord	<Technisha Circuit> What's a simple GUI library for Nim?
17:29:19	zacharycarter	what kind of GUI library?
17:29:29	zacharycarter	immediate mode? retained?
17:29:42	zacharycarter	what platform?
17:29:57	dadada	Technisha Circuit: I'm using gintro/gtk it works well, though there're rough edges, gtk is a solid toolkit
17:29:57	FromGitter	<bung87> !repo webgui
17:29:59	disbot	https://github.com/juancarlospaco/webgui -- 9webgui: 11Web Technologies based Crossplatform GUI Framework with Dark theme 15 25⭐ 3🍴 7& 1 more...
17:30:17	FromDiscord	<Technisha Circuit> Thanks
17:30:58	alexander92	dadada i am serious! no crown!
17:31:19	dadada	Technisha Circuit: I use glade to design the UI -> load it with gintro and add my signals/connectors in code, then manipulate stuff in code, it has a learning curve though, glade also has some nasty bugs, but you can adjust to them, liststores/treestores are particularly difficult to work with, but it can be done
17:31:24	FromDiscord	<Technisha Circuit> Does webgui need to use the js backend?
17:31:34	FromDiscord	<Technisha Circuit> Oh?
17:32:10	*	couven92 joined #nim
17:32:13	FromGitter	<bung87> no
17:32:21	dadada	alexander92: you are my enemy then!
17:32:37	FromDiscord	<Clyybber> nimisis
17:32:53	FromDiscord	<Technisha Circuit> How does webgui work then? Does it display html and css?
17:33:20	FromGitter	<bung87> use os webview api
17:33:27	FromDiscord	<Technisha Circuit> Oh okay
17:33:35	alexander92	livcd btw
17:33:40	alexander92	timothee is american as well
17:33:47	alexander92	which changes the balance a bit :D
17:33:49	FromDiscord	<Technisha Circuit> So how would i use it for making a complex gui?
17:33:54	alexander92	or at least he lives in the usa
17:33:58	alexander92	his name sounds french
17:34:18	zacharycarter	didn't you just say you wanted a simple gui library?
17:34:20	zacharycarter	lol
17:34:56	FromGitter	<bung87> @Technisha Circuit write your html , give the file path as newWebview first param
17:35:42	FromGitter	<bung87> using bindProcs interation between js and nim
17:35:57	alexander92	he is french! sorry.
17:38:52	dadada	why doesn't anybody take gtk seriously, I got a lot of work done with gintro/gtk in just a few days... it's like everything has to be done with web technologies today, even on the desktop... why? Yes I get it everybody already knows CSS/JS/HTML, but that doesn't mean gtk or Qt are obsolete, you can use CSS with GTK (and I think with Qt, too), and there are webviews/webengines for both if you really need it,
17:38:58	dadada	also constructing GUIs gets very easy with glade/qt-designer, they're tried and tested with huge communities behind them, ..., there's a bit of a learning curve but it's not really hard
17:39:37	dadada	+ at least on Linux everybody has those two toolkits
17:40:34	dadada	every GTK/QT question you can imagine is already answered on the net/can be googled
17:40:38	FromDiscord	<Rika> i want to be able to make a qt app in nim xd but i dont know how xddd
17:40:43	FromGitter	<bung87> maybe the company just want hire one guy write on project runs on everywhere.
17:41:27	FromDiscord	<Technisha Circuit> Thanks!
17:41:32	dadada	bung87: gtk/qt both run on Windows/Linux/Mac/BSDs, Qt also runs on Android(iPhone, too I think?), Qt even runs on the web
17:42:20	FromGitter	<bung87> yeah, they just dont want hire a guy write c++
17:42:23	dom96	dadada, I built an IDE in GTK, it was a pain in the ass
17:42:41	dadada	dom96: in what language?
17:42:44	dom96	In Nim
17:42:47	dadada	when?
17:42:51	FromDiscord	<Rika> who said html would make it easier tho 😛
17:42:55	*	krisppurg quit (Quit: Connection closed for inactivity)
17:43:00	dom96	Web technologies may be slow, but they allow basically limitless flexibility
17:43:08	FromDiscord	<Rika> dom are you talking about aphoria? did you make aphoria? i dont remember
17:43:22	dom96	dadada, in the gtk2 days, now you'll tell me that Gtk3 is better
17:43:40	zacharycarter	Why not just use electron?
17:43:44	dom96	show me an app that uses gtk and runs well on MacOS
17:43:45	zacharycarter	if you're going to use web technologies
17:43:51	dadada	dom96: I'm developing something with the latest gtk3 gintro bindings now and apart from very stupid liststore/treestore stuff, it is pretty smooth, yes
17:43:51	dom96	Rika: yeah, Aporia :)
17:44:00	FromDiscord	<Rika> i dont know the name sorry
17:44:09	*	couven92 quit (Quit: Client disconnecting)
17:44:11	FromGitter	<Willyboar> ~stream
17:44:11	disbot	stream: 11https://twitch.tv/disruptek (live video/audio) and mumble://uberalles.mumbl.io/ (live voice chat)
17:44:25	dadada	dom96: I've written apps in both gtk/qt, I like Qt better in general, but it's not like gtk is impossible bad, I think it's quite decent
17:44:27	*	couven92 joined #nim
17:44:39	dom96	dadada, it definitely depends on what you're making
17:44:49	dom96	gtksourceview has always been a pain
17:44:57	dom96	so for IDEs I wouldn't recommend it
17:45:01	dom96	and VS Code's success speaks for itself
17:45:11	*	fredrikhr quit (Disconnected by services)
17:45:27	*	couven92 is now known as fredrikhr
17:45:41	*	couven92 joined #nim
17:45:50	alexander92	ys
17:46:01	alexander92	electron is just good enough for many cases
17:46:13	FromDiscord	<Rika> how do you use electron with nim? would that require using the js backend?
17:46:14	zacharycarter	there's also this - https://deplinenoise.files.wordpress.com/2017/03/webtoolspostmortem.pdf
17:46:28	dadada	dom96: oh gtksourceview is still weird, but you get a lot for free there too, and don't forget that VS code was developed with MS paid people (nothing wrong there), they could do a good job with a classical toolkit as well, I don't use MS products, but I've heard the old visual studio is also pretty good
17:46:56	FromGitter	<sheerluck> Nim's mascot proposal -- https://i.imgur.com/XImwlWy.png
17:47:03	disruptek	~stream
17:47:03	disbot	stream: 11https://twitch.tv/disruptek (live video/audio) and mumble://uberalles.mumbl.io/ (live voice chat) -- disruptek
17:47:07	disruptek	compiler stuff.
17:47:08	FromDiscord	<Recruit_main707> no
17:47:14	FromDiscord	<Rika> sheerluck: ok weeb xd
17:47:22	FromDiscord	<Rika> ~~i am joking~~
17:47:31	FromDiscord	<Recruit_main707> ba-nim-e
17:47:36	zacharycarter	no thanks on that mascot
17:47:39	FromDiscord	<Rika> also that's touhou? i think?
17:48:07	dadada	dom96: do not underestimate gtk, gtk4 is on the horizon, I think a lot of the work is financed by RedHat/IBM
17:48:13	alexander92	ok, so
17:48:17	dadada	since they're a gnome shop
17:48:40	alexander92	PMunch i'll send you tomorrow some notes for what i thought about a bit more general web framework
17:48:47	alexander92	hm, time to openmp
17:49:34	FromGitter	<Willyboar> alehander92 you decide to do it?
17:50:04	alexander92	oh, not really sure, just wanted to brainstorm a bit and see what happens
17:50:35	alexander92	i dont have much time, but i can help a bit if several people agree on some design
17:51:50	Prestige	I keep typing "echo" at work instead of "console.log" now .-. I'm stuck in Nim-mode
17:52:12	FromGitter	<Willyboar> i can help a lot if the framework targets from beginner to pro
17:52:35	FromDiscord	<Rika> prestige: but nim has a js backend
17:52:51	Prestige	tis true but I have to use JS for work D:
17:53:25	FromGitter	<Willyboar> also i have a lot of ideas too :P
17:53:40	FromDiscord	<exelotl> bruh
17:53:51	FromDiscord	<exelotl> The mascot was the crown all along
17:53:55	zacharycarter	make a web framework like microsoft's blazor
17:54:37	FromDiscord	<Rika> prestige: easy, just write nim and use the generated output xddddd
17:54:39	alexander92	willyboar absolutely
17:54:43	alexander92	i'll link you as well
17:55:01	Prestige	Lol @Rika code review will be fun after doing that :P
17:55:17	alexander92	zachary hmm, i dont really know well .net world
17:55:19	alexander92	the *
17:55:26	FromGitter	<Willyboar> me neither
17:56:03	FromGitter	<Willyboar> i have used rails, sinatra, flask and a little django
17:56:11	FromDiscord	<Rika> i'm not serious dont worry
17:56:28	zacharycarter	well just check it out and thinka bout how it would work in Nim
17:56:40	FromGitter	<Willyboar> rails and flask are the coolest
17:56:45	zacharycarter	also when you mentioned web framework I thought you meant frontend
17:56:52	zacharycarter	https://dotnet.microsoft.com/apps/aspnet/web-apps/blazor
17:57:01	*	waleee-cl quit (Quit: Connection closed for inactivity)
17:58:51	kungtotte	dadada: if RH are forking up for gtk4 it's probably going to depend on systemd :P
17:59:30	dadada	kungtotte: they were a Gtk shop for forever, it's not new, they also developed for gtk3 and probably gtk2
17:59:35	FromGitter	<Willyboar> zachary i mean mostly server side
18:01:19	FromGitter	<Willyboar> It will be interest to see something like razor to nim. But this is not what i really like.
18:01:41	dom96	dadada, the thing with VS Code is that it's fundamentally extensible in a really easy way, this is the killer feature of Electron IMO
18:02:01	zacharycarter	maybe I'll work on something like blazor for Nim
18:02:05	dom96	the fact that MS has created Visual Studio, written in a native toolkit, and that it's not used as widely as VS Code proves that it's not just about money
18:02:07	zacharycarter	I've said that before though :P
18:02:47	dadada	dom96: hmm, vscode is free though, wasn't visual studio restrictively licensed, and not available on Linux? only reason I use vscode is that I now can, before I couldn't
18:04:06	FromGitter	<Willyboar> zachary i would love to see it. I am watching anything about nim-web area
18:04:37	dadada	dom96: agree on importance of extensions, but again, look at the gnome desktop, it has a javascript engine at its heart and can be extended to your hearts content, no reason that you can't do the same with a any desktop app/toolkit, I think the real reason behind electron's success is that people already have all the knowledge of JS/HTML/Web/CSS, when with other toolkits you have to relearn
18:05:30	*	tane joined #nim
18:07:35	dom96	yeah, for someone that's already got very little time though, electron sounds very good
18:07:53	dom96	there is tons of resources out there for how to work around weird html/css issues
18:07:56	dom96	not so with GTK
18:08:06	FromGitter	<Willyboar> Razor looks a lot like a template engine
18:08:34	FromDiscord	<dom96> Like, I spent lots of my time debugging issues like this for example: https://picheta.me/articles/2013/08/gtk-plus--a-method-to-guarantee-scrolling.html
18:10:14	dadada	dom96: thanks for writing that down, now people can find it :-)
18:11:42	dom96	might be about time I revive this blog
18:12:35	dadada	btw. is there any inbuilt method for loading code from a string directly like you would use include for code from a file?
18:12:41	*	PMunch joined #nim
18:12:52	dadada	I know I can use a macro ...
18:13:02	*	silvernode quit (Ping timeout: 256 seconds)
18:13:38	leorize	parseStmt?
18:13:51	dadada	I know, I would use that for the macro
18:14:04	dadada	so the answer is no
18:14:54	leorize	if all you need is parseStmt, use parseStmt :P
18:15:06	leorize	!eval import macros; parseStmt("echo \"Hello\"")
18:15:08	NimBot	Compile failed: /usercode/in.nim(1, 25) Error: expression 'parseStmt("echo \"Hello\"")' is of type 'NimNode' and has to be discarded
18:15:20	leorize	welp, guess you can't cheat like this :P
18:15:26	leorize	!eval import macros; eval("echo \"Hello\"")
18:15:28	NimBot	Compile failed: /usercode/in.nim(1, 16) Error: undeclared identifier: 'eval'
18:15:58	leorize	!eval import macros; emit("echo \"Hello\"")
18:16:00	NimBot	Hello
18:16:13	leorize	but emit is deprecated anyway :P
18:16:57	FromDiscord	<Recruit_main707> !eval import macros; discard parseStmt("echo "Hello"")
18:16:59	NimBot	Compile failed: /usercode/in.nim(1, 41) Error: undeclared identifier: 'Hello'
18:17:24	leorize	I think discord cancelled your backslashes :P
18:17:38	FromDiscord	<Recruit_main707> !eval import macros; discard parseStmt("echo \"Hello\"")
18:17:40	NimBot	Compile failed: /usercode/in.nim(1, 24) Error: request to generate code for .compileTime proc: parseStmt
18:17:57	FromDiscord	<Recruit_main707> it cancelled yours actually
18:18:33	leorize	this will be fixed once Yardanico got the new bridge to work correctly I hope :)
18:18:35	dadada	leorize: is there any way to get dirt y results from quote do or getAst? I want to get Nim code as NimNodes, but with no gensym symbol magic
18:18:59	Araq	dadada, template {.dirty.} plus getAst on it
18:19:05	leorize	dadada: well getAst will return whatever the template returns
18:19:16	dadada	finally, Araq saves the day
18:19:59	leorize	disruptek: did you catch the latest nim.nvim update I posted earlier?
18:20:11	disruptek	i forgot to update.
18:20:20	FromGitter	<Willyboar> do it live
18:21:33	FromGitter	<Willyboar> he did it :P
18:21:40	leorize	yay
18:22:08	PMunch	leorize, I think this is the new bridge..
18:22:26	leorize	nah, Yardanico rolled it back because his bridge crashed too much
18:22:45	PMunch	Today?
18:22:52	FromDiscord	<Clyybber> sick
18:23:02	leorize	he rolled it back shortly after introducing it :P
18:23:13	leorize	dunno if he sorted the issue out
18:23:14	FromDiscord	<Clyybber> but its on
18:23:31	leorize	ah, it does appear so
18:24:07	leorize	I think there might be leaks in nimsuggest regarding template/macro instantiations
18:25:13	leorize	https://github.com/alaviss/nim.nvim/issues/18#issuecomment-572722537 <- some log from one of nim.nvim users
18:25:15	disbot	➥ Highlighting sometimes disappears
18:25:29	leorize	> Error: template instantiation too nested
18:25:57	leorize	I definitely need to spend sometime sorting this out...
18:28:22	*	poohman joined #nim
18:29:33	PMunch	leorize, he switched back to his bot today: https://irclogs.nim-lang.org/13-05-2020.html#14:21:56 I went through the logs and doesn't look like he reverted it
18:29:52	leorize	I stand corrected then :)
18:31:13	*	superbia joined #nim
18:33:57	*	vqrs quit (Ping timeout: 272 seconds)
18:35:18	FromDiscord	<mratsim> sent a long message, see http://ix.io/2lUD
18:35:18	*	vqrs joined #nim
18:36:00	FromDiscord	<Clyybber> submit/complete
18:37:16	*	poohman quit (Ping timeout: 246 seconds)
18:38:29	FromDiscord	<Rika> same as cly
18:38:30	*	fanta1 quit (Quit: fanta1)
18:38:50	*	silvernode joined #nim
18:39:21	zacharycarter	shashlick looks like some enums weren't handled correctly for cimgui by nimterop
18:40:17	FromDiscord	<mratsim> mmmm, I'll think about it, thanks
18:40:21	zacharycarter	`ImGuiMouseCursor_` being one
18:41:13	FromDiscord	<mratsim> I was leaning toward settle, because the thread that "submits" doesn't actually do any "completion" it just waits in a loop until the job is ready
18:41:38	FromGitter	<Willyboar> submit/complete too
18:42:02	zacharycarter	I like submit/waitFor
18:43:21	FromDiscord	<Rika> "ready" can be used as a verb
18:43:56	zacharycarter	you could have a non-blocking `testIfDone` and then a blocking `waitForDone` or something
18:44:06	shashlick	@zacharycarter can you please open an issue? I'm really behind on stuff
18:44:12	zacharycarter	sure
18:44:15	FromDiscord	<FromIRC> Uptime - 4 hours, 21 minutes, 1 second, 227 milliseconds, 166 microseconds, and 95 nanoseconds
18:44:15	FromDiscord	<Yardanico> !status
18:44:16	livcd	alexander92: i thouht timothee is french
18:44:34	leorize	PMunch: looking forward to your semantic highlighting to nimlsp :)
18:44:43	FromDiscord	<mratsim> I have isReady for that already @Zachary Carter
18:44:59	leorize	then after that you just need to get nimlsp into nim itself and I can happily drop a ton of my code :P
18:45:37	FromDiscord	<mratsim> 4s: spawn sync, submit, settle
18:45:38	livcd	dom96: port sciter
18:45:48	FromDiscord	<mratsim> sciter is not opensource
18:47:59	zacharycarter	why port sciter? someone wants to write an antivirus app's gui wtih Nim?
18:50:33	PMunch	leorize, don't hold your breath for that. It's probably going to take a while :P
18:50:59	leorize	I'm already looking into implementing all the goodies directly myself :P
18:51:00	FromDiscord	<mratsim> HTML/CSS/JS compat is interesting for visualization for example
18:51:13	FromDiscord	<mratsim> but you can use webview and build on UI on top
18:51:37	livcd	mratsim: why does it matter if it's open source or not?
18:51:54	livcd	zacharycarter: you can build whatever you want with that not only antivirus UI
18:52:01	zacharycarter	it was a joke
18:52:21	livcd	sorry I am drunk already :X
18:52:50	zacharycarter	just don't use web tech for native gui's and if you really want to just use electron
18:52:59	zacharycarter	that would be my advice
18:53:10	*	gsingh93 quit (Ping timeout: 260 seconds)
18:53:14	FromGitter	<Willyboar> I think there is a nim - electron lib
18:53:22	livcd	zacharycarter: sciter is not webtech
18:53:53	zacharycarter	I'm pretty sure you use HTML / JS / CSS with sciter
18:53:54	PMunch	@Willyboar, I did some stuff with Nim <-> Electron a while back
18:54:02	PMunch	But that was just a PoC, not a library
18:54:16	livcd	zacharycarter: and does that mean that indicates it's a webtech?
18:54:18	FromGitter	<Willyboar> yeap is yours i was checked a while ago
18:54:41	zacharycarter	well it's probably using some web technology to allow guis to be built with web technologies no?
18:54:43	livcd	zacharycarter: Is XAML webtech? Is QML webtech?
18:54:51	PMunch	Why not use wxnim though?
18:54:57	FromGitter	<Willyboar> Also this https://github.com/bluenote10/nim-electron-karax
18:55:09	livcd	zacharycarter: not at all
18:55:12	zacharycarter	what???
18:55:19	PMunch	If it smells like webtech and it moves like webtech it's probably webtech
18:55:27	zacharycarter	I'm pretty sure you're mistaken livcd
18:55:32	*	gsingh93 joined #nim
18:55:37	*	fredrikhr quit (Quit: Client disconnecting)
18:56:06	livcd	zacharycarter: I am pretty sure I am not.
18:56:12	zacharycarter	okay show me proof then
18:56:15	FromGitter	<Willyboar> @PMunch i don;t want to use it i just check it. I am interesting more on web
18:56:31	livcd	zacharycarter: easy https://sciter.com/developers/engine-architecture/
18:56:49	zacharycarter	I see HTML DOM
18:56:52	zacharycarter	web tech
18:57:22	livcd	none of which is inherently webtech -,-
18:57:43	zacharycarter	okay I'm not sure if you're so drunk you don't understand that HTML is a web technology
18:57:55	zacharycarter	it's a markup langauge that's only used in browser engines
18:57:59	livcd	zacharycarter: is XAML a web technology?
18:58:02	livcd	answer me
18:58:24	zacharycarter	A) I don't use XAML and B) From my understanding XAML isn't used in the browser
18:58:32	zacharycarter	or on the web
18:58:50	zacharycarter	this is like saying client side JS or CSS aren't web technologies
18:58:54	livcd	zacharycarter: the guy just chose HTML/CSS to define the UI the same way QML/XAML work
18:59:05	livcd	there's no v8 or chrome or electron or whatever
18:59:29	zacharycarter	I didn't say there was v8 or chrome or electron did I?
18:59:36	zacharycarter	I said it was built using web technolgoies
18:59:55	zacharycarter	which it is if it's using CSS for styling and HTML for presentation
19:00:22	zacharycarter	and I imagine it has the same issues that every other GUI solution buit with web technologies does
19:00:26	livcd	zacharycarter: I dont get you. If he was using something that was similar to QML would you still say the same?
19:00:35	zacharycarter	No because QML isn't used in a gd browser!
19:00:41	zacharycarter	jesus
19:01:17	livcd	zacharycarter: sciter is just as native as QT is
19:01:25	zacharycarter	okay this argument is pointless
19:01:40	zacharycarter	you're conflating my arguments
19:01:50	zacharycarter	and putting words in my mouth
19:02:05	zacharycarter	go port and use sciter - I don't care
19:02:11	livcd	zacharycarter: i cant port it XD
19:02:17	FromDiscord	<Rika> what is happening o_o
19:02:17	livcd	I can use it sure..
19:02:21	zacharycarter	then go use it
19:02:26	federico3	PMunch: If it smells like webtech better open the window
19:02:35	FromDiscord	<Rika> what are you two even arguing about
19:02:41	zacharycarter	nothing it's over
19:05:38	livcd	zacharycarter: You can read what the author has to write about it. Search for c-smile: https://www.reddit.com/r/programming/comments/fu4r3a/sad_state_of_cross_platform_gui_frameworks/fmdl8tv/?context=3
19:06:02	livcd	Sciter is not a browser engine. It is rather UI engine that uses HTML/CSS in places where, for example WPF, uses XAML.
19:06:22	FromDiscord	<Rika> bruh
19:06:26	FromDiscord	<Rika> so is it over or not
19:07:44	PMunch	HTML and CSS is still web-tech though. It is tech designed for the web
19:07:47	shashlick	@zacharycarter - replied to the issue - just ridiculous stuff people do in C
19:07:50	PMunch	That by itself doesn't mean sciter is bad though
19:07:59	PMunch	It actually looks pretty compelling
19:09:06	PMunch	Although it might suffer a bit in the accessibility realm due to not being OS native
19:10:26	livcd	Sciter uses GPU accelerated backend for rendering by default (e.g. on Windows it is Direct2D/DirectX ). On comparable UIs, Sciter will render stuff significantly faster than Qt (mixed CPU/OpenGL backend) or wxWidgets (GDI rendering primarily).
19:10:58	livcd	zacharycarter: I am sorry if i irritated you. I did not mean to :X
19:16:34	*	waleee-cl joined #nim
19:16:52	*	silvernode quit (Ping timeout: 265 seconds)
19:16:56	*	silvernode_ joined #nim
19:18:23	zacharycarter	livcd: it's fine. I just don't enjoy an argument that there is no logical end to
19:18:55	zacharycarter	thanks for the reply shashlick
19:19:26	shashlick	no worries - not sure if we can do anything for this in nimterop - let me know if you have any ideas
19:19:52	livcd	zacharycarter: I was just trying to relay what the author is saying about it but I did not really do a good job
19:20:05	zacharycarter	does the cPlugin step run before cImport shashlick?
19:20:20	zacharycarter	if the underscore was stripped first, wouldn't that solve the issue?
19:20:28	shashlick	cPlugin creates a plugin.nim file which is sent to toast in the cImport call
19:20:33	zacharycarter	ah
19:20:57	shashlick	the underscore'd symbol occurs later in the code
19:21:32	*	silvernode_ quit (Ping timeout: 256 seconds)
19:21:37	*	silvernode joined #nim
19:22:10	zacharycarter	well it's better than nothing - until I have time to automate the generation of hand written bindings I can live with this
19:22:34	shashlick	well the other option is to not simply strip underscores
19:22:34	zacharycarter	I need to brush up on perl at some point anwyay
19:22:59	zacharycarter	yeah - I could use regex instead I guess to do it
19:23:03	shashlick	they are two different symbols, just that removing underscore loses information so they look the same
19:23:09	zacharycarter	yeah
19:23:39	shashlick	anyway, you have the flexibility to choose and nimterop allows you to do it any particular way
19:23:54	zacharycarter	yeah that is very true
19:23:59	zacharycarter	better than not being able to do anything about it :)
19:31:16	*	Senketsu joined #nim
19:32:24	*	natrys joined #nim
19:34:33	*	silvernode quit (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.)
19:35:40	*	Hideki joined #nim
19:35:53	*	Senketsu quit (Ping timeout: 265 seconds)
19:36:04	*	Hideki is now known as Guest9018
19:39:55	*	Guest9018 quit (Ping timeout: 246 seconds)
19:54:53	*	alexander92 is now known as alehander92
19:59:07	alehander92	so
20:10:37	FromGitter	<Knaque> Out of curiosity, what's the difference between `re` and `nre`? (Besides the broken `sequtils.toSeq()` with `nre`)
20:11:06	*	Jesin quit (Quit: Leaving)
20:11:30	*	NimBot joined #nim
20:15:17	*	Jesin joined #nim
20:17:27	FromGitter	<kaushalmodi> They are similar with different API
20:17:54	FromGitter	<kaushalmodi> Now nim has both because we couldn't decide between them before Nim 1.0.0 got released
20:18:17	Araq	and the best one is an external module called nim-regex
20:18:22	Araq	I think
20:18:25	FromGitter	<kaushalmodi> There's also `regex` available via nimble (`nimble install regex`) which is a strong contender
20:18:35	FromGitter	<kaushalmodi> re/nre depend on pcre
20:18:56	FromGitter	<kaushalmodi> nim-regex aka regex is all made in and made for nim (no pcre dependency)
20:28:00	*	silvernode joined #nim
20:29:20	silvernode	Afternoon
20:29:28	superbia	afternoon sir
20:30:00	silvernode	I'm just on for a quick question, and then I'll be back later tonight for a bit before work
20:30:29	silvernode	Just wondering what the official method is for getting offline docs and references to functions is.
20:31:42	Araq	the installer ships the docs but you can also do 'koch docs'
20:31:52	FromDiscord	<Varriount> Hm, a new bot
20:33:54	silvernode	Araq: I used choosenim, I'll take a look in the dot folder I suppose. Thanks
20:35:29	*	dadada quit (Ping timeout: 258 seconds)
20:36:05	silvernode	koch docs
20:36:47	*	natrys quit (Quit: natrys)
20:36:53	silvernode	well I must have done that wrong, ha
20:39:19	*	narimiran quit (Ping timeout: 258 seconds)
20:40:11	FromDiscord	<Technisha Circuit> What's `cImport`?
20:40:32	FromGitter	<kaushalmodi> Technisha Circuit: It's a macro from a nimble installed package `nimterop`
20:40:59	FromGitter	<kaushalmodi> It's used to auto generate Nim wrapper for types and proc signatures from C header files
20:42:25	*	dadada joined #nim
20:42:30	*	gokr joined #nim
20:42:48	*	dadada is now known as Guest51870
20:43:19	FromDiscord	<Technisha Circuit> Oh okay
20:46:18	leorize	I need a regex expert
20:47:05	leorize	so I'd like to match the comma in a Nim's parameter list: (a: string, b: int)
20:47:46	leorize	the problem is: (a: Table[string, int], b: int)
20:47:48	FromDiscord	<mratsim> .\*\\,.\*
20:48:02	FromDiscord	<mratsim> ah
20:48:13	leorize	I'd like to match just the , between the params :/
20:48:15	FromDiscord	<mratsim> you had one problem, with regex, you have .... 😉
20:48:27	FromGitter	<bung87> no you dont need regex
20:48:28	Araq	leorize: regexes cannot parse nested structures
20:48:28	leorize	splitting a Nim param list is hard :P
20:48:44	Araq	it's not, only with regexes because regexes suck
20:49:05	alehander92	leorize can't you use the nim parser
20:49:10	leorize	so I'm off writing a small parser for this?
20:49:14	leorize	I'm doing this for nim.nvim
20:49:23	leorize	split the param list so I can provide parameter completion
20:49:27	alehander92	Araq
20:49:51	alehander92	i am using the nim parser
20:49:59	alehander92	in a toy compiler for a subset of nim
20:50:06	alehander92	(trying to make it parallel )
20:50:09	alehander92	so i wondered
20:50:13	alehander92	can i make it pull much less somehow
20:50:16	leorize	ie. got proc (a: Table[string, int], b: int) -> call(Table[string, int], int) for the user to fill in
20:50:36	alehander92	it really slows down the compilation which is a bit strange to me
20:50:42	alehander92	because its just a parser
20:51:14	alehander92	but to use it i import modules which pull stuff like sempass and other stuff
20:51:22	alehander92	do i use it in a wrong way
20:52:13	Araq	yes.
20:53:00	FromDiscord	<mratsim> https://media.discordapp.net/attachments/371759389889003532/710233151871385650/unknown.png
20:53:22	alehander92	i do the config ref, ident cache, condsyms , extccomp init stuff
20:53:27	alehander92	is any of this not needed
20:54:05	FromDiscord	<mratsim> alehander92 started following alexander92 😛
20:54:06	Araq	import compiler/parser
20:54:15	Araq	is what you need
20:54:22	alehander92	mratsim yeah i found out i took his username
20:54:24	alehander92	in irc
20:54:26	alehander92	:(
20:54:55	FromDiscord	<Rika> lmaoooo
20:54:55	alehander92	then even PNode is undefined
20:54:57	dom96	TO͇̹̺ͅƝ̴ȳ̳ TH̘Ë͖́̉ ͠P̯͍̭O̚​N̐Y̡ H̸̡̪̯ͨ͊̽̅̾̎Ȩ̬̩̾͛ͪ̈́̀́͘ ̶̧̨̱̹̭̯ͧ̾ͬC̷̙̲̝͖ͭ̏ͥͮ͟Oͮ͏̮̪̝͍M̲̖͊̒ͪͩͬ̚̚͜Ȇ̴̟̟͙̞ͩ͌͝S̨̥̫͎̭ͯ̿̔̀ͅ
20:54:59	Araq	leorize: writing a parser is what you should do
20:55:27	Araq	alehander92: so add 'import ast' etc until it compiles
20:55:33	FromGitter	<bung87> maybe just about hundred line
20:55:33	alehander92	but i did that before
20:55:42	alehander92	and then it loads all those modules
20:55:51	alehander92	thats what i am trying to fix
20:56:20	*	letto joined #nim
20:57:21	leorize	Araq: or I can be lazy and hack nimsuggest to do this for me :)
20:57:31	Araq	sure
20:57:49	Araq	writing a parser makes you a better person though
20:57:51	leorize	something like: after a `sug`, produce `decode` lines
20:58:00	leorize	well the compiler got the AST
20:58:16	dom96	or just get the compiler to output the AST as JSON?
20:58:16	superbia	I wrote a parser, but I don't feel any better..
20:58:32	leorize	nimsuggest should reduce work for its user, not creating more :P
20:58:42	Araq	dom96: but then you need to parse the json :P
20:59:19	*	pbb quit (Remote host closed the connection)
20:59:20	*	letto_ quit (Ping timeout: 272 seconds)
20:59:27	dom96	you have to parse something anyway
20:59:32	dom96	json seems most trivial
20:59:34	FromGitter	<bung87> @superbia makes your code looking smart
20:59:55	Araq	well you cannot parse json with regexes either
21:00:16	superbia	regex is old technology, should avoid it
21:00:32	leorize	I wish I could
21:00:42	FromGitter	<bung87> so the better choice is ??
21:01:16	dom96	I'm sure vim can parse json
21:01:20	FromGitter	<bung87> or I need a editor not support regex finding
21:02:25	*	pbb joined #nim
21:03:18	FromDiscord	<Technisha Circuit> Is there a js to nim transpiler?
21:04:02	leorize	dom96: I can parse whatever, but it has to be fast or the user will notice :P
21:04:49	dom96	seriously doubt json parsing will be a problem
21:05:00	FromGitter	<bung87> I have ts to nim
21:05:13	dom96	Technisha Circuit: that's impossible
21:06:02	*	Guest51870 quit (Ping timeout: 256 seconds)
21:06:18	FromGitter	<bung87> https://github.com/bung87/ts2nim
21:06:25	FromDiscord	<Technisha Circuit> Okay
21:06:26	FromDiscord	<Technisha Circuit> O
21:07:26	leorize	Araq: is there any reason why we don't have reversible name mangling?
21:08:04	*	silvernode quit (Ping timeout: 256 seconds)
21:08:28	Araq	yeah, I don't like mangled names that take up kilobytes
21:10:54	leorize	yea but it's hard to support external tooling without this :/
21:10:54	*	solitudesf quit (Remote host closed the connection)
21:11:11	leorize	.ndi have the info, but it doesn't have the full type def
21:11:47	leorize	so we can't easily distinguish foo(int) and foo(string)
21:12:03	leorize	makes it hard to write debugger helpers
21:12:14	*	dadada__ joined #nim
21:12:24	Araq	do you think `string` is a unique type? well it's not. do you think `system.string` is? nope. so how about `utils.system.string`? nope, people want to be able to reuse the `utils` module name
21:12:40	Araq	er, I mean package name
21:12:57	Araq	so in the end only <full path here>.nim/string is a unique name
21:13:23	Araq	and now you "only" have to do that for all types in the signature
21:15:15	leorize	so in the end we will still need the compiler :P
21:17:52	*	abm joined #nim
21:20:11	Araq	well be my guest and implement something like https://en.wikiversity.org/wiki/Visual_C%2B%2B_name_mangling
21:21:08	Araq	and prove the mangling is sound
21:25:54	*	letto_ joined #nim
21:27:49	*	letto quit (Ping timeout: 264 seconds)
21:29:16	*	superbia quit (Quit: WeeChat 2.4)
21:30:02	*	couven92 is now known as fredrikhr
21:35:21	*	thomasross joined #nim
21:36:02	*	dadada__ quit (Ping timeout: 260 seconds)
21:39:48	*	everfree joined #nim
21:40:20	*	couven92 joined #nim
21:40:45	*	dadada joined #nim
21:41:08	*	dadada is now known as Guest71164
21:41:34	*	everfree is now known as arty_everfree
21:42:31	*	arty_everfree is now known as arty
21:48:52	*	pbb quit (Ping timeout: 256 seconds)
21:54:33	*	hax-scramper quit (Remote host closed the connection)
21:55:15	Araq	https://internals.rust-lang.org/t/pre-rfc-a-new-symbol-mangling-scheme/8501/87 of course, others have the same problems
21:57:52	*	tane quit (Quit: Leaving)
21:58:02	*	pbb joined #nim
21:59:45	*	abm quit (Ping timeout: 265 seconds)
22:02:14	*	arty left #nim (#nim)
22:02:38	*	Hideki joined #nim
22:03:03	*	Hideki is now known as Guest865
22:04:34	*	abm joined #nim
22:09:21	alehander92	scriptconfig
22:09:24	alehander92	loads sem
22:09:28	alehander92	which loads vm
22:09:33	alehander92	and so much other stuff
22:09:36	*	pbb quit (Ping timeout: 265 seconds)
22:10:12	alehander92	and nimconf loads it, and i
22:10:19	alehander92	need it `loadConfigs`
22:10:51	alehander92	hm, but if i just miss loadConfigs
22:10:54	alehander92	it seems to wok
22:10:56	alehander92	work*
22:11:02	alehander92	so that was my error, sorreh!
22:14:22	*	pbb joined #nim
22:14:27	PMunch	alehander92, what are you working on?
22:14:33	*	abm quit (Read error: Connection reset by peer)
22:14:42	alehander92	importing the parser
22:14:49	alehander92	and running a parse on a file
22:14:52	PMunch	Ooh
22:14:56	alehander92	i was doing something i didnt need
22:15:05	alehander92	so i was pulling too much of the compiler :D
22:15:05	PMunch	I was thinking about that earlier today
22:16:04	PMunch	To see if I was able to "minify" the stdlib parts nimscript needs. To ie. remove all comments, doc comments, and stuff that's behind when not defined(nimscript) switches
22:16:22	PMunch	I'm still stuck on this annoying issue that I can't get nimscript to see my own library as stdlib though..
22:16:59	PMunch	If I register stdlib.hashes.hashVmImpl it doesn't work, I have to register it as *.hashes.hashVmImpl
22:17:20	Prestige	Sup PMunch
22:17:39	PMunch	Which means that calling registerAdditionalOps doesn't work if I'm not using the actual stdlib..
22:17:46	PMunch	Prestige, actually on my way to bed :P
22:17:54	PMunch	It's 12AM here..
22:18:24	PMunch	Well, I say on my way to bed.. I should be on my way to bed..
22:20:19	Prestige	Oh, well gn :P
22:20:27	PMunch	What are you up to?
22:20:41	PMunch	I'm still playing with NimScript as a config language
22:22:01	Prestige	PMunch: working on nimdow
22:22:17	PMunch	That's the name of your WM right?
22:22:19	Prestige	I've made good progress, only a few things left to implement.. status bar and multihead, mainly
22:22:22	Prestige	Yeah
22:22:27	Prestige	https://github.com/avahe-kellenberger/nimdow#roadmap
22:22:43	PMunch	Do you plan on supporting different bars and such?
22:23:26	Prestige	Yep planning on polybar first
22:23:54	PMunch	Oh I thought your were planning on writing your own
22:24:03	PMunch	What's your story for multihead like?
22:24:35	Prestige	I'm planning on using xrandr to split up the displays, haven't looked into it much
22:26:04	PMunch	Well yeah, xrandr is what I use to set my screen layout
22:26:16	PMunch	I was more thinking how your WM would handle it
22:26:32	PMunch	Like i3 just treats it as more places to put workspaces
22:26:45	Prestige	Just creating "Screen" objects that are mapped to a certain region, each getting their own tags
22:28:12	PMunch	Yeah I guess that's similar
22:35:26	*	couven92 quit (Quit: Client disconnecting)
22:39:20	*	oats joined #nim
22:39:27	FromDiscord	<Elegant Beef> PMunch if you do get a functional embbeded version of nimscript that can do actual scripting ensure to send it to me 😄
22:40:01	PMunch	Oh, that I have already
22:40:03	PMunch	Kind of
22:40:22	FromDiscord	<Elegant Beef> My WM i have each screen with it's own layout and it just layouts in a first in first drawn layout
22:40:28	FromDiscord	<Elegant Beef> Ah
22:40:28	PMunch	And expect an article on it in the next couple of days
22:40:35	FromDiscord	<Elegant Beef> Ok cool
22:43:36	Prestige	Sweet
22:49:52	PMunch	Hmm, getting one step closer..
22:50:37	FromGitter	<bung87> when assign literal int to float ,why does it not convert it internally?
22:51:00	PMunch	I figured out that matches(s: PSym; x: string) in vmgen is apparently what is failing to equal stdlib.hashes.hashVmImpl when I'm using my own stdlib. And that's because s.owner when it comes to the "stdlib" part is set to "unknown"
22:51:24	*	monokrom quit (Quit: Leaving)
22:51:25	PMunch	So I need to figure out how I can get it to set owner to "stdlib" when I'm using my own library instead
22:51:45	PMunch	(By the way, conf.libpath is set to my path, so it has to be something earlier..)
22:52:06	PMunch	s/earlier/different
22:55:48	FromDiscord	<Rika> bung87: it should afaik?
22:55:58	FromDiscord	<Rika> !eval let x: float = 23; echo x
22:56:00	NimBot	23.0
22:56:06	FromDiscord	<Rika> yeah it does
22:56:16	FromDiscord	<Rika> @bung87
22:59:29	PMunch	Halleluja!
22:59:33	PMunch	I figured it out
22:59:40	PMunch	And now I feel like a complete idiot :P
23:00:33	*	Guest865 quit (Remote host closed the connection)
23:00:48	PMunch	So guess why my minilib wasn't recognised as the actual stdlib?
23:01:24	PMunch	The bug that I've now tracked through the compiler with echo statements and figured had to do with how modules are resolved throughout the compiler
23:01:51	PMunch	The bug that I've spent hours trying to wrap my head around earlier and tried to pry an answer from 4raq for
23:02:14	*	Hideki joined #nim
23:02:15	PMunch	It was simply missing the "stdlib.nimble" file..
23:02:39	*	Hideki is now known as Guest83618
23:02:42	PMunch	stdlib is a nimble module like everything else
23:03:08	PMunch	My minilib was seen as "unknown" because it didn't have a stdlib.nimble file in it...
23:03:16	*	PMunch flips table
23:03:44	*	PMunch puts table back carefully. Damn these standing desks are heavy..
23:03:55	*	gokr quit (Quit: Leaving.)
23:06:58	*	Guest83618 quit (Ping timeout: 260 seconds)
23:11:49	*	PMunch quit (Quit: leaving)
23:12:03	*	krux02_ joined #nim
23:14:27	*	krux02 quit (Ping timeout: 240 seconds)
23:15:19	FromDiscord	<Rika> (╯°□°）╯︵ ┻━┻
23:15:25	FromDiscord	<Rika> ┬─┬ ノ( ゜-゜ノ)
23:15:54	zacharycarter	ugh - I thought that setting the content-disposition header in a HTTP response would trigger the browser to download the response as a file based on the content-type header
23:15:58	zacharycarter	but apparently not?
23:18:11	FromGitter	<awr1> Is @dom96 here? I keep running into a strange bug with Nim 1.2.0 using webdriver, keeps telling me 'takeScreenshot' is undeclared despite it being kinda obviously there and public
23:18:13	FromGitter	<awr1> https://github.com/dom96/webdriver/blob/master/src/webdriver.nim#L250
23:18:49	FromDiscord	<Rika> maybe we can help if you elaborate on the error
23:18:56	FromGitter	<awr1> my code is
23:18:57	FromGitter	<awr1> https://play.nim-lang.org/#ix=2lWb
23:19:31	FromDiscord	<Rika> what about the error message?
23:19:55	FromGitter	<awr1> it just tells me takeScreenshot is undeclared
23:20:33	FromGitter	<awr1> also yes, technically i think that is the wrong way to save a screenshot from webdriver, i think it's like base64 encoded or something
23:20:43	FromGitter	<awr1> but that's besides the point
23:20:55	FromDiscord	<Rika> can you do check the version? `nimble list -i | grep webdriver` or something
23:21:05	FromDiscord	<Rika> might be somehow outdated
23:21:12	FromGitter	<awr1> 1) 2.0
23:21:15	FromGitter	<awr1> i just downloadeded it
23:21:21	FromGitter	<awr1> i can try head i guess
23:21:23	FromDiscord	<Rika> hmm
23:22:26	FromGitter	<awr1> yeah i needed to get head
23:22:36	FromGitter	<awr1> `nimble install webdriver@#head`
23:22:38	FromGitter	<awr1> fixed it
23:22:55	FromDiscord	<Rika> why though? the nimble file is still 2.0
23:23:39	FromGitter	<awr1> i don't know, it's strange
23:24:55	FromDiscord	<Rika> https://github.com/dom96/webdriver/blob/c2fee574c650abb706af40dee317b85ad3820cbe/src/webdriver.nim the version back in september also had this proc...
23:25:29	FromDiscord	<Rika> the version without it was back in 2018
23:25:34	FromDiscord	<Rika> thats super weird
23:38:40	*	mal`` quit (Quit: Leaving)
23:59:01	*	ftsf joined #nim