<< 07-09-2020 >>

---

00:04:34	*	SenasOzys quit (Ping timeout: 260 seconds)
00:11:48	*	Tlangir quit (Quit: Leaving)
00:29:35	*	vicfred quit (Ping timeout: 265 seconds)
00:39:28	*	Jjp137 quit (Ping timeout: 244 seconds)
00:40:57	*	Jjp137 joined #nim
00:41:35	*	bouzu__ quit (Quit: Leaving)
00:51:23	*	abm quit (Quit: Leaving)
00:52:19	disruptek	Error: internal error: expr: param not init input_19385020
00:55:59	*	vicfred joined #nim
01:13:51	*	arecacea1 quit (Remote host closed the connection)
01:14:48	*	arecacea1 joined #nim
01:29:11	*	apahl quit (Ping timeout: 272 seconds)
01:29:47	FromDiscord	<Avatarfighter> what is an actual use case of the "ropes" module in the stdlib? I'm having trouble understanding why someone would want to store large strings in memory.
01:30:31	*	apahl joined #nim
01:31:09	FromDiscord	<Elegant Beef> No clue what "Very long" means but if you're using a large jsonfile/html file and appending to it it'd be a pretty nice thing
01:31:36	FromDiscord	<Avatarfighter> Well I'
01:31:40	FromDiscord	<Avatarfighter> whoops
01:32:13	FromDiscord	<Avatarfighter> I'm not too sure as to why one would want to have "very large strings" in memory
01:32:31	FromDiscord	<Avatarfighter> I'm just curious to see if someone has used the module before to be honest I guess
01:34:47	wiml	Avatarfighter: It's handy if you want to modify large strings, compose large strings out of pieces, or extract many substrings. The classic use case is a text editor, but any time you have a large string, the time spent copying substrings around can sometimes be significant.
01:35:09	FromGitter	<DylanModesitt> Text editors, DNA sequences come to mind as uses
01:36:04	disruptek	what are we talking about? gap buffers?
01:36:12	FromDiscord	<Avatarfighter> wiml, DylanModesitt: I definitely didn't think about those use cases, those are really good use cases
01:36:22	FromDiscord	<Avatarfighter> tell disruptek if he unmutes discord he can know 😛
01:36:47	FromDiscord	<Rosen> don't tell him anything, if he prefers to only see half of a chatroom power to him lmao
01:36:55	wiml	ha
01:36:55	FromDiscord	<Avatarfighter> haahh
01:37:13	FromDiscord	<Avatarfighter> we just have to peak his interest and he'll unmute
01:38:11	FromDiscord	<Avatarfighter> Thank you again wiml and DylanModesitt 🙂
01:42:53	FromDiscord	<Rika> It’s pique, not peak
01:43:10	*	Tanger joined #nim
01:43:21	disruptek	!last clyybber
01:43:21	disbot	clyybber spoke in 12#nim 3 hours ago 12https://irclogs.nim-lang.org/06-09-2020.html#22:00:26
02:06:03	*	gmaggior quit (Quit: Leaving)
02:49:01	FromDiscord	<PizzaFox> sent a code paste, see https://play.nim-lang.org/#ix=2wnA
02:49:08	FromDiscord	<PizzaFox> i assume its not classes
02:52:44	FromDiscord	<Avatarfighter> Have you tried using the random library @PizzaFox ?
02:52:54	FromDiscord	<PizzaFox> i am not asking how to write this code
02:53:07	*	muffindrake quit (Ping timeout: 260 seconds)
02:53:08	FromDiscord	<PizzaFox> i am asking how the finished product would be used as if a real nim developer had made it
02:53:27	FromDiscord	<Avatarfighter> you wont find a "Die" object but you will have a `rand()` proc which takes an int as an input
02:53:29	FromDiscord	<Avatarfighter> ohh
02:53:32	wiml	PizzaFox: In nim you could use an object type for that ( see for example https://nim-lang.org/docs/tut1.html#advanced-types-objects )
02:53:45	FromDiscord	<PizzaFox> will research
02:53:56	FromDiscord	<Avatarfighter> I thought you were asking if there is a nim equivalent
02:54:05	FromDiscord	<PizzaFox> that class above just stores a few properties as state used in the `roll` method
02:54:23	FromDiscord	<PizzaFox> an object seems like what i want
02:54:36	FromDiscord	<PizzaFox> ill make a `roll` proc that accepts a `Die` as an argument
02:54:40	FromDiscord	<PizzaFox> (edit) 'ill' => 'i'll'
02:55:15	*	muffindrake joined #nim
02:57:47	FromDiscord	<Avatarfighter> Yeah, I think for your `roll` proc you might like the `rand` proc inside of the random module since you can give it a number and it'll give you a number in the range of 0-number
02:58:32	FromDiscord	<PizzaFox> yeah thats what i was thinking
02:59:18	wiml	like so: https://play.nim-lang.org/#ix=2wnC (probably has bugs but gives you the idea of nim syntax for this)
02:59:50	FromDiscord	<Avatarfighter> ahaha wiml and I did the same thing
02:59:52	FromDiscord	<Avatarfighter> https://play.nim-lang.org/#ix=2wnB
03:00:08	wiml	hahaha
03:00:12	wiml	yours is better :)
03:00:24	FromDiscord	<Avatarfighter> Don't forget to `randomize()` when using the random module 🙂
03:03:13	disruptek	all the ladies say so.
03:05:01	FromDiscord	<Avatarfighter> i dont know if i enjoy the random comments disruptek says bc he has discord muted lmao
03:05:05	*	vicfred quit (Quit: Leaving)
03:20:52	FromDiscord	<PizzaFox> i am now realizing i need to run some code when someone instantiates a `Die`
03:21:12	FromDiscord	<PizzaFox> should i make a `proc` that instantiates a die with the proper values?
03:22:06	disruptek	man, term rewriting macros are nuts.
03:24:36	FromDiscord	<PizzaFox> nim compiles to js right
03:24:41	FromDiscord	<PizzaFox> just make it do the reverse
03:27:31	disruptek	just think about it and get back to me.
03:29:00	FromDiscord	<PizzaFox> here is my broken nim code: https://play.nim-lang.org/#ix=2wnF↵here is the JS code i'm trying to port: https://hasteb.in/zehobuyo.js
03:29:08	bozaloshtsh	disruptek: have any examples of useful term-rewriting macros?
03:29:19	disruptek	dude.
03:29:34	bozaloshtsh	I remember thinking it was the coolest language feature but can't think of a use for it off the top of my head
03:29:36	FromDiscord	<PizzaFox> (edit) 'https://hasteb.in/zehobuyo.js' => 'https://hasteb.in/seyakoho.js'
03:29:56	disruptek	i got carded tonight and the barmaid is looking at my id and squinting and i'm like, "what?"
03:30:37	disruptek	and she's pretty cute. she says, "you have, like, 5 little chili peppers next to you photo. what the fuck is that? i've never seen that before. is that a colorado thing?"
03:31:00	disruptek	"nah," says i, "that's just because i'm extra hot."
03:31:27	disruptek	i could tell her knees were weak as she turned around to clutch herself and repeat my address under her breath.
03:32:59	disruptek	ima rewrite some of jason to use a term rewriting macro. i think it's probably the cleanest way to do static json encoding.
03:33:02	disruptek	!repo jason
03:33:03	disbot	https://github.com/disruptek/jason -- 9jason: 11JSON done right 🤦 15 19⭐ 1🍴
03:33:45	disruptek	i just impl arrays but it's almost like there's no point until i can have any jason encoding performed at compile-time magically.
03:34:30	FromDiscord	<Elegant Beef> I cant believe you named a nim module after me
03:34:32	FromDiscord	<Elegant Beef> Im so honoured
03:38:09	disruptek	the hardest part is the naming. i want you to provide an encoder and i'll magically pick it up and use it at compile-time if i can.
03:38:25	disruptek	it's surprisingly tricky to make this all happen magically.
03:41:07	FromGitter	<gogolxdong> ```/mnt/c/nim-libp2p/libp2p/protocols/pubsub/pubsub.nim(199, 58) Error: 't' is of type <lent TopicPair> which cannot be captured as it would violate memory safety, declared here: /mnt/c/nim-libp2p/libp2p/protocols/pubsub/pubsub.nim(197, 7); using '-d:nimWorkaround14447' helps in some cases```
04:06:02	*	supakeen quit (Quit: WeeChat 2.9)
04:06:38	*	supakeen joined #nim
04:47:07	*	a_chou joined #nim
04:50:17	*	waleee-cl quit (Quit: Connection closed for inactivity)
04:57:37	*	narimiran joined #nim
05:09:24	*	solitudesf joined #nim
05:09:31	*	a_chou quit (Quit: a_chou)
05:32:34	*	vicfred joined #nim
05:58:32	*	lpy joined #nim
06:22:56	*	PMunch joined #nim
06:37:50	FromDiscord	<lantos> mostly compile-time JSON encoding↵"I don't care." -- Araq, 2020
06:37:51	FromDiscord	<lantos> lol
06:42:44	FromDiscord	<whisperdev> lol
06:54:02	FromDiscord	<whisperdev> I am still waiting for treeform :?
06:56:41	Araq	gah
06:56:56	Araq	they just cancelled my half-marathon event :-(
07:05:33	*	Vladar joined #nim
07:07:40	FromDiscord	<Elegant Beef> @whisperdev what do you need treeform for?
07:10:16	FromDiscord	<PizzaFox> sent a code paste, see https://play.nim-lang.org/#ix=2wof
07:16:44	*	ldlework joined #nim
07:16:51	ldlework	Anyone want to help me update this old nim? https://gist.github.com/dustinlacewell/54ded16bfaabbb9a7de9
07:16:57	FromDiscord	<Elegant Beef> I dont think iterating over would allocate to the heap, but i've known to be wrong about stuff like that
07:17:01	FromDiscord	<Elegant Beef> Hey ldle
07:17:53	ldlework	hello
07:18:08	ldlework	Something wrong around line 47 with the tables
07:18:25	ldlework	got <Table[system.float, proc (diff: float){.closure.}], float, proc (diff: float): Color{.noSideEffect, gcsafe, locks: 0.}> but expected one of:
07:19:54	FromDiscord	<Elegant Beef> the procs return colour
07:20:03	FromDiscord	<Elegant Beef> The table def is a void proc
07:23:11	FromDiscord	<Elegant Beef> Hopefully that explanation was informational enough, but you can see the issue here aswell↵https://play.nim-lang.org/#ix=2woi
07:29:27	ldlework	Anyone around here use NixOS?
07:29:54	ldlework	ehmry: poke
07:31:35	wiml	Idlework: I haven't gotten really deep into it but I use it for a couple servers and occasionally nixpkgs elsewhere
07:32:35	*	NimBot joined #nim
07:32:35	ldlework	wiml: the program I compiled errors when run: libcsfml-graphics.so: cannot open shared object file: No such file or directory
07:32:49	ldlework	I'm not sure how I am supposed to change where the nim program looks for libcsfml, IE the nix store
07:36:24	ldlework	wiml, any idea?
07:37:07	ehmry	ldlework: yes, thats this rpath thing...
07:37:44	ehmry	https://github.com/NixOS/nixpkgs/pull/95692
07:37:44	disbot	➥ nim: fix dynamic library loading, wrap for cross-compilation
07:37:58	wiml	Hrm, is it linked in, or is it dynamically-loaded? Does it show up if you `ldd` ?
07:38:13	ehmry	its probably dynamically loaded
07:39:00	ehmry	I'd like to always wrap the nim compiler on nixos so that things like cross-compilation are transparent, but the problem is that nimble stops working, and I haven't looked into why
07:40:12	ldlework	ehmry: do you have a shell.nix or something I can use..?
07:41:49	ldlework	ehmry: I'm just trying to compile this simple file, https://gist.github.com/56bd99a207545701179fc48100f0268d
07:41:57	ldlework	it uses csfml
07:42:08	ldlework	the library that fails to dynamically load
07:42:45	ehmry	ldlework: i'm using a nix flake these days https://git.sr.ht/~ehmry/dhall-nim/tree/8d636193933d832c277c6fe0b84a139c2438aa91/flake.nix
07:44:11	ehmry	if you build nim from the nim branch of github.com/ehmry/nixpkgs it might work, but you still need a build environment with csfml
07:44:48	ehmry	I don't want to include csfml with the compiler because it blows up the closure
07:45:27	ldlework	ehmry: I have a nix-shell with csfml in it, why doesn't the compiler find it when I build
07:45:35	ldlework	I don't know anything about flakes :/
07:47:32	ehmry	because the binary will try to dynamically load csfml at runtime, the library isn't search for at all during compile time
07:48:01	ehmry	and because there is no global library location, its not going to be found at runtime
07:48:27	*	lpy left #nim (#nim)
07:48:29	ehmry	so I've been patching the compiler to look at the nix build environment during compilation and load libraries by absolute path
07:48:51	FromGitter	<alehander92> hm
07:49:03	FromGitter	<alehander92> how do other compilers do that?
07:49:08	ehmry	and those ablsolute paths can be detected by nix as runtime dependencies
07:49:09	wiml	Is there a reason not to dynamically link the library instead of dlopening it?
07:49:33	ehmry	yea, most compilers have the linker figuring this stuff out at link time
07:49:36	ldlework	wiml, what version of the library in the store should be the one that gets dynamically linked?
07:49:50	ldlework	remember you can have N versions of the same package installed
07:50:02	wiml	The one that's visible at build time ... just like any other library
07:50:18	ldlework	There is no sense of globally available libraries
07:50:28	ldlework	in Nix
07:51:11	wiml	right, that's the advantage of dynamically linking it --- it gets a particular version instead of having to search at runtime
07:51:32	wiml	just like any other library
07:51:41	ehmry	there is talk of removing rpath as much as possible in nix to speed things up
07:51:51	ldlework	ehmry: why doesn't this work? https://gist.github.com/9f6bfded5ec1e0ce435b068095bcb811
07:51:55	ehmry	I think dynamic loading by absolute path is fine
07:52:40	ldlework	?
07:53:03	ehmry	ldlework: same error?
07:53:06	ldlework	Yeah.
07:53:49	ldlework	There should just be a compiler option to set the path for a linked library
07:53:54	FromDiscord	<whisperdev> @Elegant Beef for asyncssh
07:54:16	FromDiscord	<PizzaFox> is nimcrypto really maintained anymore
07:54:24	FromDiscord	<PizzaFox> it has like 0 docs
07:54:58	FromDiscord	<PizzaFox> and these very wacky examples using raw pointers for zero padding since the library doesnt give any padding implementations for convenience
07:55:04	ehmry	ldlework: I'll try to build it then
07:55:08	ldlework	<3
07:55:18	ldlework	actually I think there was something wrong with that shell.nix
07:55:21	ldlework	it's only now cloning your nixpkgs
07:55:29	ldlework	nah same error
07:56:03	ehmry	PizzaFox: yea, I don't use nimcrympto, the constant casting between types is stupid
07:56:59	ehmry	I mean haveing to cast every argument is dumb
07:58:24	ldlework	ehmry: I don't really see any changes here that would do that stuff you said, https://github.com/ehmry/nixpkgs/blob/master/pkgs/development/compilers/nim/default.nix
07:58:56	ldlework	no patches, etc
07:59:23	ldlework	is it on some other branch?
07:59:41	ldlework	I see that it is
08:00:13	ehmry	yes, nim branch
08:03:17	FromGitter	<alehander92> morning
08:04:38	*	krux02 joined #nim
08:05:09	Araq	PizzaFox: as far as I know it's maintained and used in production but of course, internally developed software doesn't prioritize documentation
08:06:03	ldlework	ehmry:
08:06:05	ldlework	/home/ldlework/src/nim-test/noise.nim(20) noise
08:06:07	ldlework	/home/ldlework/.nimble/pkgs/csfml-2.5.0/csfml/graphics.nim(238) newRenderWindow
08:06:09	ldlework	/home/ldlework/.nimble/pkgs/csfml-2.5.0/csfml/private/graphics_gen.nim(2197) newRenderWindow_U32
08:06:11	ldlework	SIGSEGV: Illegal storage access. (Attempt to read from nil?)
08:06:18	ldlework	Is this a result of the compiler hacks or a bug in my program?
08:06:37	ldlework	This is the line: window = newRenderWindow(videoMode(cint size, cint size), "maxnoise")
08:08:11	ehmry	ldlework: works for me
08:08:24	*	awe001 joined #nim
08:08:37	ldlework	wat
08:08:54	ldlework	did you try the shell.nix?
08:09:41	ehmry	yea, with "https://github.com/ehmry/nixpkgs/archive/8ba4bdabbac12b39ccd1bf35f814d5943c35bbac.tar.gz", but I'm on aarch64
08:10:27	ldlework	ok I used a different commit somehow
08:10:29	ldlework	recompiling nim..
08:10:52	FromDiscord	<Elegant Beef> What's your hardware ehmry?
08:11:19	ehmry	mnt reform beta https://mntre.com/
08:11:20	ldlework	same failure
08:11:51	ldlework	:(
08:12:21	ehmry	I have to go, but I'll work on this later today
08:12:37	ldlework	ehmry: real quick, nothing wrong here? https://gist.github.com/e06866c7149ba77236ffd45be2af15a0
08:12:56	ldlework	dang heh
08:13:15	ehmry	no, that matches mine
08:15:28	ldlework	ehmry: any hints of what to look into while you're gone? or is it going to be a result of your patches.
08:16:08	FromGitter	<alehander92> Araq what is that functional lang for gpu-s you mentioned
08:17:09	Araq	https://futhark-lang.org/
08:18:53	ehmry	ldlework: we should be using the same compiler and csfml, so maybe the libraries from nimble are different? FWIW the nim-gdb that comes with the shell works
08:19:22	alehander92	ah i remember seeing stuff about it by his author maybe
08:19:23	alehander92	thanks
08:22:12	ldlework	i don't know anything about gdb
08:22:29	ldlework	ehmry: do I need to add csfml to the build inputs of the nim compiler derivation?
08:25:37	ldlework	I'm trying to do:
08:25:39	ldlework	nim = ehmry.nim.overrideAttrs (old: {
08:25:41	ldlework	buildInputs = old.buildInputs ++ [ ehmry.csfrml ];
08:25:43	ldlework	});
08:25:46	ldlework	but it says old doesn't have a buildInputs attr, wtf
08:26:10	Zevv	Araq yeah that's pretty cool stuff
08:28:16	alehander92	ah this is the athas guy
08:29:36	alehander92	he hangd a lot in the #haskell channel when i was trying to learn iirc
08:33:48	*	awe002 joined #nim
08:35:59	ldlework	ehmry: hmm the way you compose nim-unwrapped and nim, makes it really hard to override nim-unwrapped's buildInputs
08:36:09	ldlework	So I imagine that is not actually what I'm supposed to do.
08:37:19	*	awe001 quit (Ping timeout: 260 seconds)
08:45:47	*	hnOsmium0001 quit (Quit: Connection closed for inactivity)
08:46:00	*	Guest_6 joined #nim
08:46:06	Guest_6	Hello everyone
08:46:19	Guest_6	I'm tuatarian from the discord server, just seeing what IRC is
08:46:36	FromDiscord	<iWonderAboutTuatara> I can confirm that is me
08:46:45	Guest_6	can we tag across clients?
08:47:03	Guest_6	iWonderAboutTuatara:
08:47:20	FromDiscord	<iWonderAboutTuatara> @Guest_6[IRC]
08:47:40	FromDiscord	<iWonderAboutTuatara> @371759389889003532-752449773629472878 test
08:47:46	Guest_6	no, that didn't work
08:47:53	*	Guest_6 is now known as Tuatarian
08:48:05	FromDiscord	<iWonderAboutTuatara> so i changed my name on irc sid
08:48:06	FromDiscord	<iWonderAboutTuatara> e
08:48:27	Tuatarian	I don't really understand how IRC works
08:48:40	Tuatarian	but it thinks I'm a different person now, on discord end anyway
08:48:44	Tuatarian	with a new PFP and all
08:51:38	alehander92	different account yeah
08:51:39	alehander92	probably
08:54:45	Tuatarian	Same account
08:54:53	Tuatarian	I just used the namechange function
08:55:35	*	Tuatarian quit (Remote host closed the connection)
08:55:44	FromDiscord	<iWonderAboutTuatara> I prefer discord
08:55:45	*	maier joined #nim
08:55:52	FromDiscord	<iWonderAboutTuatara> It's a fair bit more normal to me
08:56:11	FromDiscord	<iWonderAboutTuatara> Maybe because I've been using it a fair bit longer
08:59:26	ldlework	ehmry: I got it working. I needed to use nvidia-offload
08:59:59	*	awe003 joined #nim
09:01:15	*	arecacea1 quit (Remote host closed the connection)
09:01:38	*	arecacea1 joined #nim
09:02:53	*	awe002 quit (Ping timeout: 265 seconds)
09:08:19	*	ehmry quit (Remote host closed the connection)
09:09:34	*	abm joined #nim
09:13:12	FromDiscord	<iWonderAboutTuatara> Question: Can i use nim with Flutter?
09:13:26	FromDiscord	<Clyybber> disruptek: ping
09:13:33	FromDiscord	<iWonderAboutTuatara> that doesn't ping inot irc
09:13:39	FromDiscord	<iWonderAboutTuatara> iirc
09:14:09	*	Tua joined #nim
09:14:14	Tua	Hello
09:14:19	FromDiscord	<iWonderAboutTuatara> Tua: ping
09:14:37	FromDiscord	<iWonderAboutTuatara> Tua
09:14:44	FromDiscord	<iWonderAboutTuatara> oh, it does ping
09:14:55	FromDiscord	<iWonderAboutTuatara> Tua
09:15:27	FromDiscord	<iWonderAboutTuatara> anyway, if anyone has tried to use Nim with Flutter, please let me know
09:15:37	FromDiscord	<iWonderAboutTuatara> Or if it's not possible, please let me know as well
09:15:47	*	Tua quit (Remote host closed the connection)
09:16:05	*	snowolf quit (Ping timeout: 240 seconds)
09:22:34	Araq	iWonderAboutTuatara: not sure, if Flutter's core is written in C(++), then yes
09:22:51	FromDiscord	<iWonderAboutTuatara> It may be written in Dart
09:22:55	FromDiscord	<iWonderAboutTuatara> I'm not 100% sure
09:25:16	FromDiscord	<iWonderAboutTuatara> It's all Dart
09:25:28	FromDiscord	<iWonderAboutTuatara> https://media.discordapp.net/attachments/371759389889003532/752459506767757312/unknown.png
09:25:51	FromDiscord	<iWonderAboutTuatara> It's possible that the 0.2 + 0.0 % core is written in C++ and Obj C, but that seems unlikely
09:27:14	Araq	you can compile Nim to JS and try some Dart/JS interfacing
09:27:46	*	Tanger quit (Remote host closed the connection)
09:28:24	Araq	but I dunno, I would use a UI library not written to pitch yet-another terribly dynamic language
09:28:58	FromDiscord	<iWonderAboutTuatara> I chose flutter because it makes my life a lot easier
09:29:08	FromDiscord	<iWonderAboutTuatara> Don't need to rewrite a GUI twice, in Kotlin then Swift
09:29:34	Araq	a couple of months ago everybody chose React because it make life easier
09:29:56	FromDiscord	<iWonderAboutTuatara> I do js for my job, but I've heard too many pain stories from React Native
09:30:08	Araq	sorry, I cannot follow every fad
09:30:14	FromDiscord	<iWonderAboutTuatara> lmao fair enough
09:30:29	FromDiscord	<iWonderAboutTuatara> I do webdev stuff for my job, so I generally follow web stuff
09:30:35	FromDiscord	<iWonderAboutTuatara> and not a whole lot else
09:31:01	FromDiscord	<iWonderAboutTuatara> I would have gone electron, but it just swallows battery life
09:31:13	FromDiscord	<iWonderAboutTuatara> Which is not good for an offline mobile game
09:31:41	FromDiscord	<iWonderAboutTuatara> I haven't seen a lot of people complaining about Dart though, but admittedly there is almost nobody using it
09:31:54	FromDiscord	<iWonderAboutTuatara> too dynamic from what ive seen
09:32:20	Araq	I remember the times where they argued for unsound type co-/contravariance
09:32:58	Araq	not sure if Dart still has that, for my the better JS is called TypeScript
09:32:58	FromGitter	<gogolxdong> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5f55fe4ad4f0f55ebb093080]
09:33:19	FromGitter	<gogolxdong> How to cast seq[T] to ptr SeqHeader with --gc:arc?
09:34:40	FromDiscord	<Clyybber> SeqHeader changed in gc:arc.
09:34:52	FromDiscord	<Clyybber> A seq is now a len|pointer pair
09:35:18	FromDiscord	<Recruit_main707> no more casting between strings and seqs then?
09:35:29	Araq	and soon a len/cap/pointer triple so that primitive vectorization works better...
09:35:31	FromDiscord	<Clyybber> no, strings behave the same
09:35:47	FromDiscord	<Recruit_main707> ok
09:39:14	FromDiscord	<iWonderAboutTuatara> Araq: I'm a massive fan of TS
09:39:26	FromDiscord	<iWonderAboutTuatara> It generates pretty readable js as well, which is a huge plus
09:40:36	FromDiscord	<iWonderAboutTuatara> But I honestly don't kind dynamic langauges too much
09:41:07	FromDiscord	<iWonderAboutTuatara> I have a preference for static but I 2ill 8ee a dynamic language if using an alternative would make my life significantly more difficult
09:43:29	Araq	how exactly does shipping a JIT on iPhone not make it "difficult", they are against their policies
09:43:44	Araq	they used to enforce this too
09:44:08	Araq	ah never mind, they have an AOT too
09:50:16	FromDiscord	<iWonderAboutTuatara> It handles it for you iirc
09:50:31	FromDiscord	<iWonderAboutTuatara> They allow you to ship a JIT, you just have to use theirs
09:50:37	Araq	ah they fixed it, "Dart 2.0 implemented a new sound type system. "
09:50:41	FromDiscord	<iWonderAboutTuatara> Which is a bit BS imo, but whatever
09:50:47	FromDiscord	<iWonderAboutTuatara> Oh good
09:51:18	FromDiscord	<iWonderAboutTuatara> Wait I'm so confused
09:51:25	FromDiscord	<iWonderAboutTuatara> Is it statically typed or not?
09:53:20	FromDiscord	<iWonderAboutTuatara> It seems to be static by default, you can declare s dynamic variable with `dynamic s = "Hello world!"`
09:53:29	Araq	narimiran, LaTeX detected here: https://dart.dev/guides/language/specifications/DartLangSpec-v2.2.pdf
09:53:36	FromDiscord	<iWonderAboutTuatara> But keyword `var` is static
09:53:53	Araq	seems mostly static now, yeah.
09:54:05	FromDiscord	<iWonderAboutTuatara> That's nice
09:54:08	*	Araq enjoys detecting LaTeX documents.
09:54:22	FromDiscord	<iWonderAboutTuatara> That language spec looks like a proof
09:54:40	Araq	(they are always the ones that overwrite their margins. No other tool does this bullshit.)
09:54:53	FromDiscord	<iWonderAboutTuatara> Yeah I don't like LaTeX at all
09:55:10	FromDiscord	<iWonderAboutTuatara> But I'm not a guy who writes proofs so idk
09:55:18	FromDiscord	<iWonderAboutTuatara> Doesn't look nice though, not at all
09:57:16	narimiran	programmers. can write a new programming language and deal with very complex stuff. cannot write a normal LaTeX document.
09:57:35	FromDiscord	<iWonderAboutTuatara> https://media.discordapp.net/attachments/371759389889003532/752467594954539078/Screenshot_20200907-025716.jpg
09:57:41	FromDiscord	<iWonderAboutTuatara> Either have the space or don't, pick one
09:57:51	FromDiscord	<iWonderAboutTuatara> Lmao
09:57:52	narimiran	or at least: skim through the produced pdf and see if something needs to be changed
09:58:39	FromDiscord	<iWonderAboutTuatara> Do reactions show up in irc?
09:58:52	FromDiscord	<iWonderAboutTuatara> :nim1:
09:59:22	Araq	narimiran, you can "skim" through 200 pages a couple of times but eventually it gets tiresome
09:59:39	FromDiscord	<iWonderAboutTuatara> Also, it's a very minor thing
10:00:00	FromDiscord	<Elegant Beef> If you're curious what IRC sees just look↵https://irclogs.nim-lang.org/
10:00:46	FromDiscord	<iWonderAboutTuatara> I think that kind of defers the question though
10:00:49	*	wiml quit (Quit: Leaving)
10:01:02	FromDiscord	<iWonderAboutTuatara> Does that site have support for showing the emojis/reactions?
10:01:16	FromDiscord	<Elegant Beef> No, why would it, the reactions/emojis are discord side
10:01:26	FromDiscord	<Elegant Beef> Other than the base discord ones
10:01:33	FromDiscord	<iWonderAboutTuatara> 👍
10:02:01	FromDiscord	<iWonderAboutTuatara> That showerd up correctly in irc logs
10:02:06	FromDiscord	<Elegant Beef> Lol, irc doesnt have deletions so deleting it was pointless 😄
10:02:06	FromDiscord	<iWonderAboutTuatara> 👍
10:02:11	FromDiscord	<iWonderAboutTuatara> Oh lmao
10:02:47	FromDiscord	<Elegant Beef> It's relatively archaic but it's what irc users like 😄
10:04:08	FromDiscord	<iWonderAboutTuatara> Yeah, seems overall worse to me
10:04:20	FromDiscord	<iWonderAboutTuatara> But worse enough to switch? Probably not
10:04:38	*	letto quit (Quit: Konversation terminated!)
10:06:34	*	letto joined #nim
10:06:38	*	darwinism-paced joined #nim
10:16:45	FromDiscord	<Clyybber> Araq: Why don't we simply change the spec to allow implicit and explicit returns together?
10:17:08	Araq	because it sucks, one should make up his mind
10:20:43	*	Vladar quit (Quit: Leaving)
10:22:32	*	bouzu_ joined #nim
10:24:17	*	Vladar joined #nim
10:27:55	FromDiscord	<Clyybber> Araq: https://github.com/nim-lang/Nim/pull/15123 branching is now done in semiStmtList
10:27:56	disbot	➥ Better semiStmtList parsing
10:29:57	*	Tanger joined #nim
10:33:53	*	pietroppeter joined #nim
10:34:17	alehander92	dart
10:34:20	alehander92	uses static typing iirc
10:35:00	alehander92	i really like the typescript experiments tho
10:35:20	alehander92	https://github.com/microsoft/TypeScript/pull/40336
10:35:22	disbot	➥ Template string types and mapped type 'as' clauses ; snippet at 12https://play.nim-lang.org/#ix=2woW
10:52:01	Araq	Clyybber: great now update the grammar too ;-)
10:57:06	FromDiscord	<Clyybber> Araq: Afaict its already in there, in parsePar's grammar "| simpleExpr"
10:58:12	FromDiscord	<Clyybber> ah, no
11:16:56	Tanger	Hey folks, I'm playing around with nimsuggest and I've followed the suggestions here https://nim-lang.org/docs/nimsuggest.html and ran nimsuggest --stdin --debug <myfile>.nim. With the query, should my first argument (file.nim) be "." or something to query on the file I supplied as an argument?
11:17:56	Tanger	Nevermind
11:18:11	Tanger	I was using `sug` instead of `def`, which is what I wanted, haha
11:19:22	FromGitter	<gogolxdong> how to check a seq is literal or not then.
11:21:58	FromDiscord	<Rika> a var seq can never be a literal afaik
11:23:30	FromDiscord	<Clyybber> Araq: Done https://github.com/nim-lang/Nim/pull/15123/commits/f2fac28a9b903c37d567d4cd035c752639ce9687
11:23:31	disbot	➥ Better semiStmtList parsing
11:23:51	FromDiscord	<Clyybber> gogolxdong: Why would you need to check that?
11:23:59	FromGitter	<gogolxdong> ``` (cast[ptr SeqHeader](s).reserved and (1 shl (sizeof(int) * 8 - 2))) != 0```
11:24:42	FromDiscord	<Rika> thats just unsafe as heck
11:26:38	FromDiscord	<Clyybber> gogolxdong: I mean why you would need to check if its a literal or not
11:27:26	FromGitter	<gogolxdong> Have no idea , it's from nim-chronos, nim-libp2p
11:30:37	FromGitter	<gogolxdong> I can see it uses shallowCopy as long as it's not literal.
11:31:34	FromDiscord	<dom96> lol wtf
11:31:37	FromDiscord	<dom96> that's from chronos?
11:35:20	FromGitter	<gogolxdong> https://github.com/status-im/nim-chronos/blob/master/chronos/transports/common.nim#L515
11:36:18	FromDiscord	<Rika> thats horrible...
11:37:15	narimiran	i recognize that casting! (the same thing is in `nimbus`, IIRC)
11:37:51	narimiran	...and it is the first thing that fails when you try to run it with `--gc:orc`
11:40:07	alehander92	raq
11:40:17	alehander92	Araq * so i talked with the futhark guy
11:41:11	alehander92	and he kinda agrees about the null / option thing, but he made an good point
11:41:40	alehander92	that NaN-s themselves sometimes matter (as carrying additional info different from "nothing")
11:42:37	alehander92	and on the other hand that they are just as dangerous as unchecked null-s indeed
11:43:32	alehander92	i guess that one can indeed have a universal concept of `T not A` (`ref not nil` `float not NaN` etc )
11:44:02	alehander92	back to work now
11:45:13	*	theelous3 joined #nim
11:50:26	*	ehmry joined #nim
11:52:00	ehmry	ldlework: great. with the nim compiler you shouldn't need to override any `buildInputs`
11:52:35	ehmry	and nim-unwrapped *must* only take inputs that are relevant to the system hosting the compiler, not the target system
11:52:59	ldlework	ehmry: it was because I run with nvidia-offload, so I had to run it on the gpu
11:53:03	ldlework	for some reason..
11:54:05	Araq	alehander92, well thanks I guess.
11:58:45	*	JustASlacker joined #nim
11:58:56	alehander92	i was just interested in seeing how fp people actually think of low level details
11:59:16	alehander92	sorry for the offtopic
11:59:19	alehander92	:D
11:59:32	FromDiscord	<Clyybber> no need to be sorry :)
11:59:54	*	mipri left #nim (#nim)
11:59:57	FromDiscord	<Clyybber> this can all be accomplished via a proof system
12:00:12	Araq	clyybber: (ifExpr | complexOrSimpleStmt) now we're getting there
12:00:20	FromDiscord	<Clyybber> yep :D
12:00:43	Araq	complexOrSimpleStmt can start with the 'if' token so you have a reduce-reduce conflict
12:00:59	FromDiscord	<Clyybber> yeah
12:01:11	Araq	so how does it *really* work?
12:01:27	FromDiscord	<Clyybber> we check for if and if its if we parse an ifExpr instead of an ifStmt
12:01:36	Araq	(ifExpr / complexOrSimpleStmt) then
12:01:43	FromDiscord	<Clyybber> eh s/ifStmt/complexOrSimpelStmt
12:01:51	Araq	where '/' is the ordered choice
12:01:56	FromDiscord	<Clyybber> ah!
12:02:04	Araq	(not to be confused with '|' )
12:02:13	FromDiscord	<Clyybber> I figured you used / as an ordered choice, but wasn't sure
12:02:21	alehander92	clyybber i think it can be accomplished in a simpler way
12:02:28	alehander92	but first i have to finish my work
12:02:45	Araq	btw https://github.com/nim-lang/Nim/pull/15281 is here for you to review
12:02:46	disbot	➥ fixes #15280 [backport:1.2]
12:02:57	FromDiscord	<Clyybber> alehander92: The thing is, proof creation/consuming *is* the simplest way
12:03:00	alehander92	i can't wait for nim's latex spec
12:03:38	FromDiscord	<Clyybber> ?
12:03:52	FromDiscord	<Clyybber> oh :D
12:04:27	FromDiscord	<Clyybber> Araq: Looks good to me for what it wants to accomplish, but I don't particularily like the goal :D
12:06:02	*	supakeen quit (Quit: WeeChat 2.9)
12:06:38	*	supakeen joined #nim
12:07:44	FromDiscord	<Clyybber> And this was explicitly introduced as a feature in https://github.com/nim-lang/Nim/pull/13520
12:07:45	disbot	➥ EndsInNoReturn in expressions extension, fixes #13490
12:08:07	FromDiscord	<Clyybber> IMO its better to just allow it. The compiler can handle it anyways
12:08:09	Araq	I figured but 'return' is different from 'raises' and the like
12:08:20	alehander92	clyybber kidding about latex; otherwise possible: but i need to see an exa,plee
12:08:27	Araq	maybe, if you push for it
12:08:37	Araq	distinguish between 'return val' and 'return'
12:09:05	FromDiscord	<Clyybber> Araq: You can see return as just a `result = x; break outerblockIntroducedByTransformation`
12:09:32	FromDiscord	<Clyybber> So IMO it doesn't make sense to specifically forbid return
12:09:34	Araq	that's a fair point, 'break' shouldn't be allowed either
12:10:00	FromDiscord	<Clyybber> No, all control-flow-cancelling/skipping OPs should be allowed
12:10:02	FromDiscord	<Clyybber> IMO
12:11:23	alehander92	guys btw
12:11:47	alehander92	the requiresInit thing
12:12:40	alehander92	nvm i found a test
12:14:29	alehander92	so how does this work with serialization?
12:15:05	alehander92	e.g. i have T which has such a field: does the serialization lib have to know that it has to provide the field?
12:15:10	alehander92	does json handle it?
12:15:38	Araq	yeah it does
12:16:03	Araq	the general solution is pretty much "disable the warning in this section of code"
12:16:30	Araq	json handles it differently but Status's serialization disables the warning
12:17:50	alehander92	ok
12:17:58	alehander92	but `newSeq` breaks it as well
12:18:21	alehander92	so yeah here comes the default warning, right
12:18:23	alehander92	good
12:18:35	FromDiscord	<Clyybber> status has a default(T) variant where they take the addr and then deref it or something
12:18:41	FromDiscord	<Clyybber> but its a workaround IMO
12:18:45	alehander92	yeah i forgot my own rfc
12:19:19	alehander92	the disable thing might work
12:19:45	alehander92	which reminds me another thing: is there an easy way to just list all the "cases" of unsafe usages in my program
12:20:02	alehander92	as we often say "it's simple, just see if you use ptr / cast / etc"
12:20:26	alehander92	but the compiler can simply list them (especially if they are in some third party package)
12:20:33	FromDiscord	<Clyybber> grep "ptr"/"cast"/etc..
12:20:36	alehander92	(and not always instantiated for my code)
12:20:38	alehander92	eh ^
12:20:47	alehander92	grepping doesn't take into account what does my code use
12:21:16	alehander92	and one needs to find somewhere a list of those, grep one by one in the dep tre
12:22:12	alehander92	i am imagining something like `nim --listUnsafe`
12:23:37	alehander92	this can't list all anyway tho, because i can just import posix and memset to nil my totally unrelated location
12:24:06	Araq	*shrug* that's not a hard problem, so also list all .importc procs
12:25:09	FromDiscord	<Rika> unfortunate downside of showing all c wrappers as unsafe
12:25:24	FromDiscord	<Rika> and anything using nimterop wont be caught i assume
12:25:35	FromDiscord	<Clyybber> why?
12:25:43	FromDiscord	<Clyybber> we can do it after sem
12:25:46	Araq	you start with a "list me ALL..." feature and already it's unusable in practice. given no connection to reality, we can make it flag even more cases
12:26:10	Araq	works for me :-)
12:26:14	alehander92	Araq well the thing is not all importc stuff are unsafe in practice :)
12:26:15	*	Zectbumo quit (Remote host closed the connection)
12:26:18	alehander92	so it would have a lot of noise
12:26:28	Araq	neither are my usages of 'addr' unsafe.
12:26:32	FromDiscord	<Clyybber> yeah
12:26:44	FromDiscord	<Clyybber> in practice the only thing thats unsafe is bugs
12:26:55	alehander92	but it's often that you have 2-3 addr-es
12:27:01	alehander92	and 100 importc
12:27:05	alehander92	but i might be wrong here
12:27:31	FromDiscord	<Rika> arrd is not inherently unsafe is the idea
12:27:34	FromDiscord	<Rika> addr*
12:27:45	alehander92	ugh depends on how you define that
12:28:08	FromDiscord	<Rika> well then theres a flaw with listing what's unsafe
12:28:11	FromDiscord	<Rika> define unsafe
12:28:30	FromDiscord	<Rika> everyone probably has a conflicting idea on whats unsafe and now
12:28:31	FromDiscord	<Rika> (edit) 'now' => 'not'
12:28:40	Araq	if you want to find bugs, it ususally helps to do more than just the most naive way you can come up with.
12:28:41	alehander92	basically "warn about everything i can't kinda prove that shouldn't lead to random crashes" :D
12:29:03	alehander92	yes, but until i find this way, the naive way is still > 0
12:29:14	FromDiscord	<Rika> not really
12:29:22	FromDiscord	<Rika> it might lead you to a dead end
12:29:42	alehander92	but it gives you a finite list of stuff
12:29:43	FromDiscord	<Rika> aka misleading warnings or so
12:29:54	alehander92	oh i don't want warnings
12:30:15	Araq	basically the usual "I'm confused and worried..."
12:30:43	alehander92	nah :)
12:30:47	FromDiscord	<Rika> finite list of stuff, but the list has false positives AND negatives...
12:31:06	Araq	the scientific approach is completely different btw, that would be "what bugs usually come up in Nim programs and how to mitigate them"
12:31:15	alehander92	it's just a "ok the lang could force making "unsafe" stuff use `!`" but it doesn't so i need to list them somehow
12:31:43	Araq	once you have '!' you'll soon enough ask for '!!', it's a race to the bottom
12:32:11	alehander92	wait for my `!!?!` hahaha
12:32:12	FromDiscord	<Rika> im tempted to mention the story of rust's actix
12:32:32	alehander92	dont get me wrong i dont really worry about that
12:32:45	FromDiscord	<Rika> other people will
12:33:19	alehander92	it's just annoying one can do
12:34:23	alehander92	if not variable.isNil: purefunc(); echo variable.flag
12:34:53	FromDiscord	<Clyybber> ?
12:35:04	FromDiscord	<Clyybber> thats not even remotely related
12:35:09	alehander92	and the implementor of purefunc to write something like # ignore side effects, memset stuff to zero
12:35:14	FromDiscord	<Clyybber> oh
12:35:39	FromDiscord	<Clyybber> well, should the language really stop you from using code thats badly written?
12:36:12	FromDiscord	<Clyybber> also, you can easily grep for {.noSideEffects.}: blocks
12:36:28	alehander92	no, but i want to be easy to say `nim --listUnsafe` and seeing e.g. "here the programer uses noSideEffect here he casts X"
12:36:29	alehander92	but dude
12:36:36	alehander92	i dont want to grep for 20 things
12:36:48	FromDiscord	<Clyybber> you want that. but it won't get you far
12:36:49	alehander92	:D and nobody even knows the full list of those things
12:36:54	FromDiscord	<Clyybber> you get a bunch of results
12:36:57	FromDiscord	<Clyybber> and then what?
12:37:07	FromDiscord	<Clyybber> you go bother the pkg authors about it?
12:37:14	FromDiscord	<Clyybber> or are you searching for a bug?
12:37:21	Araq	well a .noSideEffect override has no effect on memory safety
12:37:28	alehander92	search for something related to your bug
12:37:54	alehander92	well it can hide a call to a function that reassigns something aliasing to your pointer
12:38:08	Araq	so yeah, the problem gets harder the less you invest into reasoning about possible program behaviours.
12:38:39	Araq	in the worst case, the crash could come from every line of code that you didn't look at and you looked at 0 lines, it's pretty bad
12:39:06	alehander92	yeah, but my point was where is the boundary
12:39:13	FromDiscord	<Varriount> @haxscramper commented on your dedent PR
12:39:16	Araq	better make the compiler list every source code line then
12:39:29	Araq	which you can then ignore further as it's overwhelming
12:39:43	Araq	more to the point though
12:39:44	alehander92	e.g. i can probably construct a case where this can happen, so should I treat `func` calls as stuff that invalidates variable being not nil or not
12:40:03	Araq	say we have --warning[unsafe]:on
12:40:15	alehander92	Araq well, that list can be grouped/sorted somehow
12:40:19	Araq	and then 3rd party libraries grow
12:40:40	Araq	{.push warning[unsafe]:off.}: ... addr(x) ... {.pop.}
12:41:03	Araq	and then you demand to see the unsafe code in your 3rd party deps too
12:41:20	Araq	so we add --warning[unsafeAlways]
12:41:31	alehander92	but isn't the instantiation thing relevant? it's very possible that my code uses just 20% of those things actually
12:41:37	Araq	and then ... well you get the idea
12:41:57	Araq	but you don't declare which ones you use and code evolves
12:42:06	alehander92	but the compiler knows
12:42:16	Araq	maybe later on you more than 20% or a different 20%
12:42:19	alehander92	because it imports/instantiates just the functions and modules
12:42:22	alehander92	that i use
12:42:30	Araq	the compiler doesn't know the history of your security reviews
12:42:51	alehander92	ok, but the advice usually is "grep for X", but there are 20 X-es
12:42:54	alehander92	how is this better
12:43:10	FromDiscord	<Rika> it isnt
12:43:22	Araq	there are 3 X-es. 'addr', 'unsafeAddr' and 'cast'
12:43:40	Araq	'ptr T' itself is a type and cannot be "unsafe", only real code can be unsafe
12:43:50	alehander92	but there are many warning:off etc
12:43:56	alehander92	otherStuff:off
12:44:02	Araq	.noSideEffect overrides are not related to memory safety
12:44:12	alehander92	but they are
12:44:19	Araq	they are not.
12:44:41	FromDiscord	<Rika> not directly
12:44:48	alehander92	i can call a function which sets global stuff to nil
12:44:56	alehander92	and this can be aliased to a local var in my caller
12:45:04	FromDiscord	<Varriount> Hm. Does aliasing (the current aliasing limitations for parameters) influence this discussion at all?
12:45:05	alehander92	local var field*
12:45:05	alehander92	ok
12:45:07	Araq	oh not that again...
12:45:20	alehander92	ok, just asking if this is relevant or not
12:45:27	Araq	that was wrong years ago when you first brought it up
12:45:34	Araq	and it's still wrong.
12:45:38	alehander92	but now we also have fields
12:46:22	alehander92	basically it's exactly your mutation thing but i just wonder
12:46:30	alehander92	if it asssumes that `func` can't mutate
12:46:50	Araq	it doesn't. but it should IMO
12:46:53	FromDiscord	<Clyybber> @Varriount I lifted it
12:47:03	FromDiscord	<Clyybber> If you are talkign about default params
12:47:07	FromDiscord	<Clyybber> and the unindent PR
12:47:13	Araq	but that's a different discussion, currently it doesn't assume deep immutability
12:47:18	FromDiscord	<Clyybber> oh, I don't think you are :D
12:47:19	FromDiscord	<Rika> the issue is not usage but improper usage
12:47:47	alehander92	sorry
12:47:50	alehander92	let me make an example
12:47:58	FromDiscord	<mratsim> btw thoughts on adding __attribute("pure") in the codegen with strictFunc as suggested by snej? (only if there are no "var" I guess)
12:48:32	FromDiscord	<mratsim> might as well add "const * const int"
12:48:58	FromDiscord	<Clyybber> sounds good
12:48:59	Araq	mratsim: yeah :-) first is noalias / restrict though
12:50:30	Araq	these discussions are tiresome, what if your dependency has code like 'exec"rm -rf ~ "' and 'exec' wrapped in a way so that not even Nim's .tag system detects it?
12:50:54	FromDiscord	<mratsim> I need them all, I'm fighting dumb Clang codegen and superdumb GCC codegen when dealing with crypto objects of size ~4608-bit (576 bytes) where they do some aliasing defense in the function prologue and this is costing me 5%~10% perf
12:51:12	Araq	shouldn't you worry about that moreso than about random memory mutations that can only affect the fragile process that is currently running?
12:51:38	Araq	why is it that nobody ever worries about my hard disk contents?
12:51:39	FromDiscord	<mratsim> {.noalias.} is something I can add manually when i don't have attribute pure so i'm fine with the order
12:51:46	alehander92	Araq well memory mutations seem pretty dangerous :D
12:52:33	Araq	"oh no, my browser crashed" vs "oh no, I lost a week of my work"
12:52:34	FromDiscord	<mratsim> btw, someone screenshotted the first malware written in Nim in the wild last Friday (?)
12:52:37	alehander92	i am working on the lang level, os-es / supervisors can deal with the permissions/protection thing
12:52:56	alehander92	but memory bugs can sometimes lead to vulnerabilities
12:53:14	alehander92	maybe not in this case but like in general, they are not always less dangerous
12:53:30	Araq	sure they can. and even worse, they are a pita to debug
12:53:42	FromDiscord	<mratsim> easier than threading bugs
12:54:12	Araq	but if we worry about stuff why stop there, we can worry about so many more things
12:54:44	Araq	there is no shortage of things to worry about.
12:55:15	Araq	I worry I don't get things done, so bye.
12:58:10	FromDiscord	<Varriount> @Clyybber my comment was about the behavior of the implementation
12:58:39	FromDiscord	<Varriount> Of unindent
12:58:52	FromDiscord	<Clyybber> Ah
12:58:56	FromDiscord	<Clyybber> I think the implementation is fine
12:59:21	FromDiscord	<Clyybber> But what you said is correct
12:59:46	FromDiscord	<Clyybber> Although, its power comes through the count parameter
13:00:08	FromDiscord	<Clyybber> the default needs to stay as is, sadly because we break niminaction otherwise
13:00:24	FromDiscord	<Varriount> Which is why I suggest deprecation
13:00:35	FromDiscord	<Varriount> And adopting a new name
13:01:24	FromDiscord	<Rika> unindent -> dedent 😉
13:01:29	FromDiscord	<Clyybber> Fair enough, but the deprecation can be done in a different PR from haxscrampers or mine, because I think it will cause more controversy
13:01:43	FromDiscord	<Clyybber> But now theres not too much difference between our PRs I think
13:02:04	FromDiscord	<Varriount> I would like this resolved before merging.
13:02:17	FromDiscord	<haxscramper> clyyber's implementation has better API than mine (`indentation` function might also be useful)
13:03:04	FromDiscord	<Varriount> "dedent" and "unindent" are, if not grammatically identical, still similar enough to cause confusion.
13:03:23	FromDiscord	<haxscramper> Aside from naming they are pretty much identical, but just adding `indentation` and `dedent` procs to strutils won't be an issue
13:03:24	FromDiscord	<Varriount> One or the other, but not both
13:03:40	FromDiscord	<Clyybber> hmm, theres a small diff; my loop exits when the ident of the current line is bigger than the maximum already established by the previous lines
13:03:54	FromDiscord	<haxscramper> > Aside from naming they are pretty much identical, but just adding `indentation` and `dedent` procs to strutils won't be an issue↵@haxscramper Add explanation in documentation outlining differences? They pretty much identical behaviour-wise too
13:04:00	FromDiscord	<Varriount> I'm not talking about the implementation. I'm taking about the names
13:05:01	FromDiscord	<haxscramper> Well, I don't think we should change behaviour of the `unindent`, and I don't think adding new proc will cause that much of confusion
13:05:18	FromDiscord	<haxscramper> Note differences in documentation, this should be enough
13:05:43	*	SenasOzys joined #nim
13:05:44	FromDiscord	<Varriount> And then people have to remember the differences, and match them up correctly
13:06:04	FromDiscord	<Clyybber> I'm not sure if we should users of unindent to dedent by deprecation
13:06:06	FromDiscord	<Varriount> It's like php and all the MySQL escape variations
13:06:22	FromDiscord	<Clyybber> since """...""".unindent looks fancier than """...""".dedent(int.high)
13:08:08	*	martylake_ joined #nim
13:08:49	*	darwinism-paced quit (Remote host closed the connection)
13:09:03	FromDiscord	<haxscramper> Alright, then (1) mark current implementation of `unindent` as deprecated; (2) Accept PR with `indentation` and `dedent` proc (instead of changing behaviour for existing one). This mostly solves the issue of having users to remember the difference; does not break existing code
13:11:49	*	superbia quit (Quit: WeeChat 2.9)
13:16:40	Tanger	Clyybber: Hey mate, you made cligen yeah?
13:20:45	Araq	iirc it was c-blake
13:21:22	Araq	--experimental:views ?
13:21:32	Araq	good enough for a name? bikeshedding time
13:21:37	Tanger	Ah, you're right, thanks
13:28:05	*	audiofile joined #nim
13:28:48	FromDiscord	<mratsim> views? for first class openarrays?
13:34:24	Araq	yes
13:35:19	*	waleee-cl joined #nim
13:39:28	FromDiscord	<exelotl> :o
13:40:35	FromDiscord	<mratsim> sounds good.
13:40:39	FromDiscord	<Rika> sounds great
13:40:45	FromDiscord	<mratsim> sounds awesome
13:41:43	disruptek	views; what else?
13:42:33	disruptek	Araq: you are anti-block-break now?
13:42:40	FromDiscord	<mratsim> Note, C# is now following C++ lead and calling that Span (instead of Slice): https://github.com/joeduffy/slice.net#important-note
13:42:54	disruptek	or, just because this code is so trivial?
13:43:02	Araq	the latter
13:43:38	disruptek	fair enough.
13:43:53	Araq	but yeah now that you phrase it this way, block and loop structures without 'break' would be sweet
13:44:02	FromDiscord	<Rika> is there a reason theyre called spans?
13:44:05	FromDiscord	<Rika> instead of slices?
13:44:24	FromDiscord	<Rika> in C# as mratsim says i mean
13:44:49	disruptek	you mean, while/else. 😜
13:45:13	Araq	yes while/else but much better
13:45:25	disruptek	such as what?
13:45:59	Araq	unfortunately I haven't had good ideas beyond 'for x in iter while cond'
13:46:18	Araq	(notice the absense of a colon, that's a single loop construct)
13:46:27	disruptek	dude, speaking of crazy...
13:46:46	disruptek	term rewriting macros.
13:46:51	disruptek	completely bananas.
13:47:07	Araq	I don't see the connection
13:47:50	FromDiscord	<mratsim> @Rika, C++ calls them spans
13:48:07	disruptek	i didn't realize there were so many constraints.
13:48:19	FromDiscord	<Rika> yeah and im wondering why they call it spans
13:48:23	disruptek	it's a super powerful part of the language.
13:48:23	FromDiscord	<mratsim> and I expect it's because slice is often used in libraries so introducing a new slice type in the standard library would be a pain for the ecosystem
13:48:33	disruptek	not convinced i can trust it, though.
13:50:11	FromDiscord	<mratsim> apparently they already had "view" as well for something more complex http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2016/p0122r1.pdf
13:50:51	Araq	disruptek: it's bad for compile-times and I have a redesign of this feature in an unwritten RFC
13:51:10	Araq	Nim - The book of unwritten RFCs.
13:51:15	disruptek	so my fears are warranted. 🙁
13:51:42	Araq	well mratsim uses them, they work and when they don't you get missed optimizations
13:51:53	FromDiscord	<lqdev> if an rfc isn't written, is it even an rfc? 🤔
13:52:10	disruptek	i was going to use them for automatic compile-time json.
13:52:18	FromDiscord	<Rika> "you can request for comment without writing anything" (/s)
13:52:21	disruptek	seems ideal, but i don't want to have them break on me.
13:52:44	Araq	doesn't sound like you use them for "optimizations only"
13:53:12	disruptek	well, i haven't found a clean way to make statics slot in naturally and get pre-encoded.
13:53:24	disruptek	i have to specify every typedesc.
13:53:32	FromDiscord	<mratsim> I use them but I'm better served with a DSL + compiler anyway
13:53:46	FromDiscord	<mratsim> ah there is a case where they are needed for correctness
13:54:15	FromDiscord	<mratsim> https://github.com/mratsim/Arraymancer/blob/master/src/arraymancer/math_ops_fusion/math_ops_fusion.nim#L19-L42
13:54:22	disruptek	1500 cycles to 4 cycles is worth it, especially with so much of my generated api stuff.
13:55:11	FromDiscord	<mratsim> to detect ln(1+x) and exp(x) - 1 and rewrites it to proc without catastrophic cancellation
13:55:29	FromDiscord	<mratsim> but I commented those out because somehow I couldn't detect the thing with TRM
13:56:18	alehander92	i think TR had limit on recursion
13:56:30	FromDiscord	<mratsim> I'm also using there as an alignment trick: https://github.com/numforge/laser/blob/master/laser/tensor/datatypes.nim#L90-L102
13:56:33	alehander92	so not sure how good is it for serizalization dsl
13:57:11	*	maier quit (Ping timeout: 258 seconds)
13:57:34	FromDiscord	<mratsim> and this would be nice: https://github.com/nim-lang/Nim/issues/7214#issuecomment-431567894
13:57:36	disbot	➥ Term rewriting macros: not working when declared inside template ; snippet at 12https://play.nim-lang.org/#ix=2wq1
13:57:41	disruptek	hey, our language color got fixed finally.
13:57:56	FromDiscord	<mratsim> when you go in DSL mode, you might as well add your own optimization passes
13:58:47	disruptek	i'm going to have to unignore discord just so i can hear mratsim.
13:59:09	FromDiscord	<Yardanico> Yahoo Disruptek you're right
13:59:15	FromDiscord	<Yardanico> The colour is finally there
13:59:30	disruptek	good job, dude.
14:00:06	audiofile	anyone here develop with juce?
14:01:02	FromDiscord	<Yardanico> @disruptek ironic thing is that they lifted the colour difference restrictions
14:01:14	FromDiscord	<Yardanico> So you don't need to care if you colour conflicts with others
14:01:28	disruptek	eh it's fine. i think it looks good. very golden.
14:01:34	FromDiscord	<Yardanico> But anyway, our colour is more easily distinguishable this way
14:01:36	disruptek	makes javascript look like shit, too.
14:02:54	FromDiscord	<exelotl> Should we consider restructuring the discord server so that irc folks don't feel compelled to mute the bridge?
14:02:58	disruptek	clyybber: now that you're unmuted you can make excuses for breaking criterion.
14:03:12	FromDiscord	<Yardanico> @exelotl it's only Disruptek so far :)
14:03:24	FromDiscord	<Yardanico> And there's not much I can do about word wrapping in code pastes
14:03:37	FromDiscord	<Yardanico> Absence of it
14:03:40	disruptek	it's not word-wrapping, chucklehead.
14:03:57	disruptek	it's putting invalid nim code into the playground.
14:04:05	FromDiscord	<Yardanico> Well yeah I said I'll fix that
14:04:06	disruptek	if you're not going to vet it, just send it to ix.
14:04:14	FromDiscord	<Yardanico> It's sent to ix
14:04:17	disruptek	i've been asking you to fix it for literally months.
14:04:25	FromDiscord	<Yardanico> Nim playground uses ix
14:04:27	FromDiscord	<Yardanico> No you weren't :)
14:04:35	FromDiscord	<Yardanico> I only recently learned what you actually meant
14:04:48	disruptek	well, i thought it was obviously broken. my bad.
14:05:39	FromDiscord	<exelotl> It's also worth fixing cause it would allow servers other than this one to use the bot xD
14:05:55	FromDiscord	<Yardanico> You can do it even now, it's not hard to remove playground
14:05:58	FromDiscord	<Yardanico> Playground uses ix anyway
14:06:12	disruptek	i had to recreate criterion repo because otherwise the bot ignores it as a fork.
14:06:14	FromDiscord	<Yardanico> I just make a playground link to the ix paste I made
14:06:38	Tanger	Can I use nimsuggest to extract the function names from a source file?
14:06:50	disruptek	yep.
14:09:20	Tanger	disruptek, sorry ,is that a yes to me?
14:13:38	*	pietroppeter quit (Quit: Connection closed for inactivity)
14:19:21	Tanger	Also, what does "terse" mode actually do?
14:25:04	*	awe004 joined #nim
14:25:33	*	maier joined #nim
14:27:45	*	awe003 quit (Ping timeout: 240 seconds)
14:35:37	FromDiscord	<Yardanico> What mode?
14:35:48	FromDiscord	<Yardanico> Oh, in nimsuggest
14:35:50	FromDiscord	<Yardanico> No clue
14:46:47	FromDiscord	<Clyybber> Tanger: No, thats c-blake
14:46:55	FromDiscord	<Clyybber> disruptek: I'm already investigating
14:48:23	disruptek	tanger: what are you trying to accomplish? it's more convenient to simply parse the file into ast.
14:50:30	FromDiscord	<Clyybber> Araq: Regarding borrow checking/views. We should unify lent/view/cursor
14:50:47	Araq	I'm using the same logic for them yes
14:51:10	Araq	check out my additions to varpartitions.nim, we know do it "scope" based
14:51:13	Araq	*now
14:51:37	FromDiscord	<Clyybber> disruptek: How come if your usage is so trivial you struggle to give us a test case which we can work with?
14:51:37	Tanger	disruptek, I'm refactoring metacommands in INim. IPython has metacommands like `ls`, but if you define your own ls function, it overrides it. Essentially, I want to see what's available in scope to the buffer file and exclude builtin commands only if they're defined elsewhere
14:52:42	FromDiscord	<Clyybber> theres `defined(symbol)`
14:52:50	disruptek	clyybber: for what? concepts?
14:52:53	FromDiscord	<Clyybber> eh, declared(symbol)
14:52:56	FromDiscord	<Clyybber> disruptek: Yes
14:54:41	disruptek	clyybber: you want me to tell you what's wrong so you can fix it. you're the expert about how it works. how should i know why it doesn't work? have you looked at my usage?
14:55:07	FromDiscord	<Clyybber> No, I want you to tell me *what* doesn't work
14:55:10	disruptek	my car makes a weird noise when i turn left.
14:55:23	Tanger	I'd prefer to get it through nimsuggest as a natural step to autocompletion
14:55:37	disruptek	clyybber: mangling doesn't bootstrap on nim-1.0, iirc.
14:55:50	FromDiscord	<Clyybber> yes, does not tell me what doesn't work
14:56:00	disruptek	my car makes a weird noise when i turn left.
14:56:02	FromDiscord	<Clyybber> what I want is something small
14:56:10	disruptek	do i have to take my car apart?
14:56:13	FromDiscord	<Clyybber> yeah
14:56:18	disruptek	well, i'm not a mechanic.
14:56:24	disruptek	i'm just a driver.
14:56:25	FromDiscord	<Clyybber> use DUST?
14:56:30	disruptek	on the compiler?
14:56:32	FromDiscord	<Clyybber> thats what its made for no?
14:56:34	disruptek	are you mad, man?
14:56:58	disruptek	it's not very good, and not good enough for this.
14:57:07	FromDiscord	<Clyybber> also, its concepts, if *they* are bugged, you should be able to extract a testcase, replace everything with stubs
14:59:28	FromDiscord	<Clyybber> I can't simply "fix" your mangling problems because I don't know what they are and because its too big
14:59:33	FromDiscord	<Clyybber> Like what are they
14:59:38	FromDiscord	<Clyybber> is some concept not matching?
14:59:41	FromDiscord	<Clyybber> is it that they hang?
14:59:43	FromDiscord	<Clyybber> crash?
14:59:52	disruptek	do you have access to nim's ci?
15:01:37	disruptek	gah, /do i have access to nim's ci/
15:01:53	disruptek	0 errors / 0 warnings
15:03:00	disruptek	/home/runner/work/Nim/Nim/compiler/mangler.nim(184, 48) Error: generic instantiation too nested
15:03:24	FromDiscord	<Clyybber> what happens if you increase the nesting limit?
15:06:09	alehander92	hm
15:06:14	alehander92	need to bootstrap as well
15:06:21	disruptek	how do i do that?
15:06:49	disruptek	lemme push this tweak so the line numbers work.
15:07:12	FromDiscord	<Clyybber> compiler/seminst:330
15:07:21	FromDiscord	<Clyybber> increase the 50 to 500
15:13:59	*	maier quit (Ping timeout: 260 seconds)
15:50:14	FromDiscord	<Clyybber> Araq: The fundamental difference between cursors and views is that cursors are introduced as optimization, thus its ok to COW, while views can't be COW
15:50:40	Araq	lol don't lecture me, I'm completely aware
15:51:02	Araq	still it can use the same, smart underlying algorithm for borrow checking
15:51:18	Zevv	Clyybber: one day your day will come. You will be able to lecture Araq and he will be *totally* ignorant and oblivious
15:51:21	FromDiscord	<Clyybber> sorry, didn't mean to lecture you, I wanted to follow up
15:51:22	Zevv	your day will come
15:51:59	FromDiscord	<Clyybber> with this sentence: "So maybe we should encode it into the type. What do we want lent to be?
15:52:09	FromDiscord	<Clyybber> Araq: Can it not?
15:52:20	FromDiscord	<Clyybber> Zevv: lol
15:52:59	Araq	Clyybber: 'lent T's implementation can remain dependent on T's size
15:53:14	Araq	it should work out
15:53:19	FromDiscord	<Clyybber> yeah
15:53:44	FromDiscord	<Clyybber> ok, so lent is COW
15:53:48	FromDiscord	<Clyybber> sounds good
15:53:51	Araq	bah patching the backend for first class openArray is an unhealthy amount of work
15:54:06	Araq	once again an AST to AST rewrite seems simpler
15:54:06	FromDiscord	<Clyybber> I can imagine
15:54:28	FromDiscord	<Clyybber> yeah
15:54:33	Araq	otherwise it bubbles up into everything
15:54:35	Araq	bbl
15:55:05	FromDiscord	<Clyybber> ce ya
15:59:44	FromDiscord	<mratsim> instead of patching, we can have view alongside openarray and then slowly deprecate openarrays
16:18:00	*	martylake_ quit (Ping timeout: 245 seconds)
16:26:29	FromGitter	<StefanSalewski> Salewski may have joined.
16:26:44	FromDiscord	<Clyybber> sup
16:27:32	FromDiscord	<Rika> hello
16:28:19	FromGitter	<StefanSalewski> Admins, I would strongly suggest deleting last post in https://forum.nim-lang.org/t/6766#42238, it is really too unfriendly.
16:29:02	*	awe004 quit (Ping timeout: 246 seconds)
16:30:42	FromDiscord	<Rika> which one
16:31:27	*	awe004 joined #nim
16:33:59	FromDiscord	<Rika> i assume solitude's but its not highlighted when i open the link so
16:35:06	*	hnOsmium0001 joined #nim
16:35:08	narimiran	"really too unfriendly" ...coming from a guy who just the other day called some new nim users morons :D
16:35:51	FromDiscord	<Rika> 👀 yall better not fight
16:37:09	alehander92	hm
16:37:15	alehander92	actually i agree with the lack of democracy
16:37:34	FromGitter	<StefanSalewski> Yes, but a post with only one sentence with "entire post is one of the worst things i've read on this forum" is too extrem I think.
16:38:43	solitudesf	okay okay
16:39:19	FromGitter	<StefanSalewski> Thanks, bye.
16:41:40	FromDiscord	<exelotl> The forums are pretty atrocious these days tbh, just today I saw a 2-year old thread that was necro'd and immediately devolved into people calling each other "triggered"
16:43:49	*	Kaivo quit (Ping timeout: 260 seconds)
16:48:03	Zevv	yeah, I wondered as well. It's not like there is a small group trolling about, it's just the overall tone
16:48:46	FromDiscord	<Rika> why is it like that over there
16:49:09	Zevv	I don't know. here it's usually just fine. The usual name calling from disruptek, but no one takes that seriously
16:49:26	FromDiscord	<Recruit_main707> sent a long message, see http://ix.io/2wqQ
16:49:53	FromDiscord	<Clyybber> because if people write bs in a chat or talk bs during a conversation one can correct the other, and nothing feels permanent
16:50:35	FromDiscord	<Clyybber> but if you make a post, it probably feels more like a postulate (bdumtss)
16:50:47	Zevv	who's in charge of the discord bridge?
16:50:49	FromDiscord	<Rika> post, you late
16:50:55	Zevv	it does funn stuff with these long messages
16:51:01	FromDiscord	<Rika> yardanico
16:51:15	Zevv	ping @Yardanico
16:51:16	FromDiscord	<Recruit_main707> i sent the code so that irc could see it i think
16:51:23	FromDiscord	<Recruit_main707> line by line
16:51:29	Zevv	yeah but your question is also in that message
16:51:31	Zevv	that's a bit silly
16:51:33	FromDiscord	<Yardanico> Wat happen
16:51:43	Zevv	I think the snippet should go to ix, but your text should just be posted
16:51:58	FromDiscord	<Clyybber> disruptek infested Zevv
16:52:03	FromDiscord	<Yardanico> This is not as simple
16:52:07	FromDiscord	<Yardanico> https://media.discordapp.net/attachments/371759389889003532/752571913272033460/Screenshot_2020-09-07-19-51-57-054_com.discord.jpg
16:52:13	FromDiscord	<Yardanico> This is how that message looks
16:52:27	Zevv	if it were simple, anybody could do it. But we have Yardanico!
16:52:43	FromDiscord	<Yardanico> Reverse psychology Huh
16:52:46	FromDiscord	<Yardanico> *huh
16:52:48	Zevv	who, me?
16:52:49	Zevv	what where
16:53:08	FromDiscord	<Yardanico> When
16:53:29	Zevv	I was recently sent to a course "effective communication". I now do everything they didn't tell me, or don't do anything they told me. I forgot
16:54:15	FromDiscord	<Recruit_main707> should i open an issue in the glad repo about this?
16:54:30	FromDiscord	<Recruit_main707> its pretty dumb, how is it that no one has seen it yet
16:54:32	voidpi	Zevv: the earth king has invited you to drink tea at lake laogai
16:55:09	Zevv	I had to google that up
16:55:15	*	Zevv is too old for that shit
16:55:22	FromDiscord	<Rika> there is no war in ba sing se
16:55:45	voidpi	Zevv: me too, just saw the thing a couple of months ago
16:55:54	Zevv	hehe
16:56:14	FromDiscord	<Rika> man im young then
16:56:26	voidpi	it really is good
16:56:42	voidpi	sorry about the OT
16:57:43	Zevv	oh we never do on-topic here. We have #nim-ontopic for that
16:58:05	FromDiscord	<Rika> lol
17:00:08	*	MyMind joined #nim
17:00:49	*	a_chou joined #nim
17:01:23	*	Sembei quit (Ping timeout: 265 seconds)
17:01:32	FromDiscord	<Recruit_main707> there is also this line:↵`GL_TIMEOUT_IGNORED*: GLenum = GLenum(0xFFFFFFFFFFFFFFFF)`↵`Error: -1 can't be converted to GLenum`↵which its because uint32 cant be negative (if im not mistaken)
17:01:48	FromDiscord	<Clyybber> yeah that wrapper is broken then
17:02:50	voidpi	StefanSalewski: I'm reading your book it's helping me a lot thanks
17:03:19	FromDiscord	<Recruit_main707> its only those things, i think opening an issue on github warning about it could fix it
17:03:54	voidpi	the binary sum is wrong btw
17:04:56	alehander92	we need to correct
17:08:21	voidpi	alehander92: since I'm reading it I can ping you if I see any typos
17:09:07	alehander92	i meant to correct ourselves in stuff like forums
17:09:29	alehander92	maybe you want to ping stefan for the book?
17:10:14	voidpi	alehander92: heh, yes
17:10:28	*	maier joined #nim
17:10:56	*	Vladar quit (Remote host closed the connection)
17:15:54	*	maier quit (Ping timeout: 260 seconds)
17:23:08	*	a_chou quit (Quit: a_chou)
17:24:46	*	natrys joined #nim
17:51:35	*	arecacea1 quit (Remote host closed the connection)
17:51:54	*	arecacea1 joined #nim
17:52:20	FromDiscord	<Clyybber> disrupotek
18:09:54	*	gmaggior joined #nim
18:13:23	*	lritter joined #nim
18:17:34	*	supakeen quit (Quit: WeeChat 2.9)
18:18:19	*	supakeen joined #nim
18:24:05	*	JustASlacker quit (Ping timeout: 240 seconds)
18:25:16	disruptek	i live
18:30:22	FromDiscord	<iWonderAboutTuatara> Wow same
18:33:20	supakeen	So much in common.
18:33:56	FromDiscord	<iWonderAboutTuatara> Yeah, it's truly incrediblen
18:41:12	stefantalpalaru	Can I always rely on the Nim compiler generating explicit C++ casts between aliased, but non-distinct, Nim types?
18:41:37	stefantalpalaru	(I may have a special case here, because one of those types is importc-based)
18:42:14	FromDiscord	<Clyybber> I don't think so.
18:46:11	*	pietroppeter joined #nim
18:48:31	disruptek	Araq: .editorconfig ... looks like this:
18:48:37	disruptek	[*]
18:48:37	disruptek	indent_style = space
18:48:37	disruptek	insert_final_newline = true
18:48:38	disruptek	indent_size = 2
18:48:39	disruptek	trim_trailing_whitespace = true
18:48:54	disruptek	shall i PR it into the compiler or PR it into .gitignore? 😉
18:51:28	alehander92	guys
18:51:39	alehander92	is a.isNil a compile time warning/error for not nil ?
18:51:52	disruptek	warning.
18:52:04	disruptek	"condition is always false"
18:53:26	alehander92	do we have that for other builtins?
18:53:40	alehander92	yeah a default error probably would break code
18:53:57	alehander92	it's a bit odd tho
18:54:06	disruptek	eh, see compiler/semtypes for some error message machinery.
18:54:09	alehander92	e.g. it's the same as `[2].len == 0` # always false
18:54:13	alehander92	no no
18:54:20	alehander92	i mean if it's an api precedent
18:54:28	alehander92	to warn for such unreachable cases
18:54:33	*	haxscramper joined #nim
18:54:40	disruptek	well, we have an unreachable code warning, indeed.
18:55:02	alehander92	thanks for semtypes pointer
18:55:05	alehander92	!
18:55:06	alehander92	okk
18:55:54	disruptek	i think unreachableElse is a thing that you might piggyback, too.
18:56:00	haxscramper	How should `nimph fork` be done exactly? When I do `nimph fork gram` I get `unable to find a project; try nimble init?`
18:56:11	haxscramper	In empty directory*
18:56:30	disruptek	ah, it works in a project. the next nimph will let you run in any git repo.
18:56:49	disruptek	ie. we define projects as repos, not directories with .nimble files.
18:57:19	disruptek	!issue nimph 2.0
18:57:20	disbot	no results 😢
18:57:27	disruptek	!pull nimph 2.0
18:57:28	disbot	no results 😢
18:57:31	disruptek	jesus.
18:57:31	haxscramper	So I need to clone gram and then do nimph fork inside clone directory?
18:57:59	disruptek	if you are in any project that has gram in it, or you are in gram itself, you can `nimph fork gram`.
18:58:31	disruptek	technically, you could do it in any project as long as nimph can find gram in your global nimble packages or locally or anywhere.
18:58:42	disruptek	probably we should let you fork with a url, too.
19:00:26	*	NimBot joined #nim
19:02:41	haxscramper	If I do `nimph fork` on globally installed package I will have to deal with git myself because files in `.nimble/pkgs` are stored without any metadata, right?
19:02:41	haxscramper
19:02:41	haxscramper	So I should use `--path` if I want to write my own code & do some minor fixes to forked package as I stumble upon bugs.
19:03:15	haxscramper	When I'm done I commit fixes and PR them
19:03:17	disruptek	honestly, it works really well if you use it like a normal, sane developer.
19:03:38	disruptek	you have a project that uses gram. you discover that i'm an idiot and i have no business tapping on a keyboard.
19:04:06	disruptek	you type, `nimph fork gram`. then cd (nimph path gram); then you make changes. you `git push`. you create a PR.
19:04:10	haxscramper	I do fork-fix-pr and continue with my code
19:04:11	disruptek	it's really that simple.
19:04:34	disruptek	yeah. i do this all day long with my own shit and nimph doesn't know who i am.
19:04:42	disruptek	just behave like me and you can fool nimph, too.
19:04:50	disruptek	or tell me what workflow you want.
19:05:03	disruptek	i have a lot to learn about effective software development.
19:05:50	haxscramper	I actually just worded it badly. What I tried to say in the first message was just 'fork-fix-pr & continue with my own code'
19:06:17	disruptek	yes, but you shouldn't even need to set the --path.
19:06:33	haxscramper	I just have bad habit of writing everything at once, so I tend to do a lot of fixups in my own dependencies; this is probably not a good thing anyway
19:06:46	disruptek	local deps will help you a lot.
19:07:22	*	JustASlacker joined #nim
19:07:32	disruptek	you just make a nim.cfg in your project with `--clearNimblePath` and `--nimblePath="$config/deps/pkgs/" and then you `mkdir deps` and `nimph doctor` will clone them all into the deps and setup all the paths for you.
19:07:54	disruptek	in the future, you won't even do any of that.
19:11:20	*	maier joined #nim
19:11:47	disruptek	and let me know how it works for you. i get precious little feedback on this stuff.
19:13:38	haxscramper	I will
19:16:43	*	maier quit (Ping timeout: 265 seconds)
19:17:48	*	JustASlacker quit (Ping timeout: 260 seconds)
19:22:36	FromDiscord	<iWonderAboutTuatara> Does Nim to JS compiler produce readable js?
19:22:47	stefantalpalaru	"nimStackTraceOverride: enable stack traces in exceptions" - https://github.com/nim-lang/Nim/pull/15284
19:22:48	disruptek	node loves to read it.
19:22:48	disbot	➥ nimStackTraceOverride: enable stack traces in exceptions
19:22:53	FromDiscord	<iWonderAboutTuatara> Ie if I wanted to edit the js directly eeouof that be feasible?
19:23:09	FromDiscord	<iWonderAboutTuatara> Lmao
19:23:11	disruptek	yes but you'd be wise not to.
19:23:15	bozaloshtsh	you _could_ do it but why would you want to?
19:23:28	bozaloshtsh	just emit stuff with the asm statement
19:23:31	FromDiscord	<iWonderAboutTuatara> I imagine the transpiler isn't the best right?
19:23:42	disruptek	what are you comparing it to?
19:23:52	FromDiscord	<iWonderAboutTuatara> Typescript, mostly
19:23:59	disruptek	you can write javascript with other languages, but who cares?
19:24:03	bozaloshtsh	typescript is a superscript of JS, you can't really get betterr than that
19:24:03	FromDiscord	<iWonderAboutTuatara> I'm thinking it will be worse than haxe to is though
19:24:05	disruptek	wouldn't you rather write nim?
19:24:12	FromDiscord	<iWonderAboutTuatara> I would, yeah
19:24:21	disruptek	so does it really matter if haxe has a better transpilation result?
19:24:39	FromDiscord	<iWonderAboutTuatara> But if I'm going to be interfacing with flutter using some dart to js stuff, it might be needed
19:24:50	disruptek	what might be needed?
19:24:52	bozaloshtsh	there's also several angles of comparison. For me, the most important one is how easy is it to interoperate with stuff already written in JS that has no bindings
19:25:00	FromDiscord	<iWonderAboutTuatara> Ie there might be things that are very difficult in Nim which are less difficult in is
19:25:02	FromDiscord	<iWonderAboutTuatara> Yeah
19:25:10	bozaloshtsh	nim wins that in my book over all other transpilers except JS superset ones
19:25:13	FromDiscord	<iWonderAboutTuatara> (edit) 'is' => 'js'
19:25:35	disruptek	Yardanico: can we have the colorized edits back?
19:25:46	FromDiscord	<iWonderAboutTuatara> That's good
19:25:57	Yardanico	disruptek: they're in the last ircord release, but that version apparently had some bugs so I reverted it
19:27:44	bozaloshtsh	there are several issues you'll run into if you expect to be able to use the whole stdlib in the js backend
19:28:08	disruptek	does someone have a cdebug macro that emits debug arguments in c comments?
19:28:15	disruptek	because if so, i want that shit pronto.
19:28:17	FromDiscord	<iWonderAboutTuatara> What's the server address for IRC?
19:28:37	FromDiscord	<iWonderAboutTuatara> Nim irc specifically
19:30:38	bozaloshtsh	irc.freenode.net?
19:31:49	FromDiscord	<iWonderAboutTuatara> I guess
19:32:43	*	Tuatarian joined #nim
19:32:58	Tuatarian	That worked
19:33:15	Tuatarian	So now I have an IRC, incase discord does a bad thing in future
19:33:50	Tuatarian	Interestingly, I have the same pfp as last time I logged on with this Nick
19:34:08	Yardanico	@iWonderAboutTuatara it's just freenode
19:34:14	Yardanico	Tuatarian: you mean the one on discord when you're in IRC?
19:34:23	Tuatarian	Yes
19:34:24	Yardanico	i use gravatar and the avatar is based on the hash of your nickname :)
19:34:34	Yardanico	so yeah, it'll be the same for same nickname
19:34:36	FromDiscord	<iWonderAboutTuatara> I see, makes sense
19:34:38	FromDiscord	<iWonderAboutTuatara> That's good
19:34:50	Yardanico	it's https://robohash.org/
19:35:00	Tuatarian	Messages from irc in discord look cleaner than messages from discord in irc
19:35:04	Yardanico	gravatar supports it
19:35:08	Tuatarian	Probably just a limitation of irc
19:35:10	Yardanico	Tuatarian: well, because discord supports webhooks
19:35:17	Tuatarian	Oh is see
19:35:23	Tuatarian	Yeah nothing to you
19:35:26	*	PMunch quit (Quit: leaving)
19:35:30	Yardanico	there's a way to make a seamless bridge like that for IRC too, but it's much harder and needs to be done on network-level
19:35:33	Tuatarian	The bot is wonderful
19:35:33	Yardanico	like the freenode-matrix bridge
19:35:41	Tuatarian	I see
19:35:54	Tuatarian	The bot does what it needs to incredibly well though, huge props
19:36:00	Tuatarian	I remember I accidentally insulted it lmao
19:36:03	Yardanico	if you see someone with [m] on freenode - they're actually from matrix
19:36:14	Yardanico	I mean in nickname
19:36:15	FromDiscord	<iWonderAboutTuatara> I prefer discord though
19:36:18	FromDiscord	<iWonderAboutTuatara> Oh I see
19:36:21	Yardanico	like leorize[m] for example :P
19:36:23	FromDiscord	<iWonderAboutTuatara> Discord feels cleaner to me
19:36:25	FromDiscord	<iWonderAboutTuatara> Irs a lot faster
19:36:33	Yardanico	irc*
19:36:42	FromDiscord	<iWonderAboutTuatara> Oh my eyes skipped over [m] lol
19:36:45	FromDiscord	<iWonderAboutTuatara> *it's
19:36:50	FromDiscord	<iWonderAboutTuatara> Irc feels sluggish
19:36:54	Yardanico	wat
19:36:54	FromDiscord	<iWonderAboutTuatara> Maybe I'm using a bad client
19:37:00	Yardanico	irc is faster than most stuff out there
19:37:03	FromDiscord	<iWonderAboutTuatara> Maybe I should make a better IRC client for android
19:37:10	FromDiscord	<iWonderAboutTuatara> Probably just a bad irc client then
19:37:23	Tuatarian	I don't know why it feels slower lol
19:37:28	Tuatarian	It just kind of does
19:37:29	bozaloshtsh	on desktop, IRC clients are probably orders of magnitude faster than Discord
19:37:37	Tuatarian	No doubt
19:37:44	Tuatarian	I'm on mobile though
19:37:54	Tuatarian	Don't seem to be very many good IRC clients on mobile
19:37:56	bozaloshtsh	heh, then using Discord suddenly starts to make sense
19:38:06	Tuatarian	Half tempted to take that as a project lol
19:38:07	Tuatarian	Lmao
19:38:13	disruptek	discord never makes sense.
19:38:17	FromDiscord	<Elegant Beef> Could use a matrix client like element, or a native one
19:38:18	Tuatarian	Discord should open source their client
19:38:23	FromDiscord	<Elegant Beef> Lol
19:38:29	Yardanico	I use quassel on desktop/client'
19:38:37	Yardanico	s/client/mobile
19:38:42	Yardanico	and the server is on the vps on hetzner
19:38:42	Tuatarian	I don't like the discord desktop app
19:38:46	FromDiscord	<Elegant Beef> They dont allow third party clients in their TOS it's not their plan to OSS the client
19:38:52	Tuatarian	Irc is a lot better for desktop
19:38:59	Tuatarian	I know they have no intention of doing it
19:39:02	Tuatarian	But they should allow it
19:39:08	Tuatarian	*should open source it
19:39:09	Yardanico	btw we have #nim-offtopic too :P just saying'
19:39:11	FromDiscord	<Elegant Beef> Or just use matrix protocol
19:39:13	Yardanico	occasional offtopic is fine there
19:39:16	*	supakeen quit (Quit: WeeChat 2.9)
19:39:20	Tuatarian	Ohhh
19:39:22	Yardanico	it's bridged to #offtopic on discord
19:39:26	Tuatarian	I see
19:39:32	Tuatarian	How do I add more channels to a server?
19:39:41	*	supakeen joined #nim
19:39:58	Yardanico	on discord? you just create them
19:40:02	Yardanico	or what do you mean
19:40:20	Tuatarian	Says I need to register on freenode
19:40:31	Yardanico	well, you need to register on freenode :)
19:40:42	Yardanico	a lot of channels on freenode are only for registered users
19:40:46	*	Tuatarian quit (Quit: Quit)
19:40:49	Yardanico	https://freenode.net/kb/answer/registration
19:40:52	FromDiscord	<Elegant Beef> Oddly enough the channel you dont need to register for is #nim and not <#371759607934353448>
19:40:58	*	Tuatarian joined #nim
19:41:02	Yardanico	so newbies have an easier time getting into IRC
19:41:07	Tuatarian	Oh I see
19:41:07	Yardanico	to ask a question
19:41:23	Tuatarian	I'll register later, I should be working atm lol
19:49:40	FromDiscord	<Clyybber> Cool people get a blue discord bridge bot avatar
19:49:56	*	clyybber joined #nim
19:50:02	clyybber	amirite Yardanico ;)
19:50:05	*	clyybber quit (Client Quit)
19:50:10	Yardanico	haha
19:51:39	*	supakeen quit (Quit: WeeChat 2.9)
19:53:18	*	supakeen joined #nim
19:54:05	*	uvegbot joined #nim
19:55:37	*	supakeen quit (Client Quit)
19:56:33	*	supakeen joined #nim
19:56:50	*	bra-ket quit (Ping timeout: 265 seconds)
20:03:39	*	awe00 joined #nim
20:06:43	disruptek	global forvars ftl
20:07:23	*	awe004 quit (Ping timeout: 260 seconds)
20:09:53	*	narimiran quit (Ping timeout: 265 seconds)
20:16:12	*	lritter quit (Ping timeout: 256 seconds)
20:18:20	*	snowolf joined #nim
20:21:39	*	lritter joined #nim
20:22:44	*	konkrrrrrr joined #nim
20:25:19	*	kinkinkijkin quit (Ping timeout: 244 seconds)
20:30:08	alehander92	hm
20:30:18	alehander92	can we remove globals
20:30:20	alehander92	from nim
20:30:52	alehander92	just a thought experiment: i dont need to do that
20:32:12	FromDiscord	<Elegant Beef> No it encourages my bad design
20:33:27	haxscramper	disruptek: There is no specific improvements that I would suggest (at least right now), but there are some things related to ergonomics (which is basically nitpicking)
20:33:27	haxscramper
20:33:27	haxscramper	- `nimph fork` does did't change git origin to new fork
20:33:27	haxscramper	- For some reason `nimph fork skiplist` prints wikipedia page about skiplists. It created fork though
20:33:27	Yardanico	alehander92: technically yes, but it'll require to have more arguments in a lot of places
20:33:30	haxscramper	- (not on nimph): maybe I missed something, but test for skiplists uses system version instead of `$config/..` one. I had to temporarily add `switch("excludePath")` + `switch("path", "$config/..")` to actually run tests.
20:33:33	haxscramper
20:33:36	haxscramper	Aside from that there is no comments. I really liked test for skiplists (and other ones) - mainly because there is no endless walls of text about `Verifying ...`. (I know there It can be changed by writing custom `task`, but still).
20:33:40	haxscramper
20:33:43	haxscramper	Maybe nimph could also include some basic nimscript utilities for simplification of common tasks (for example I have couple of helpers for running tests in docker that copies all `develop` packages to container, executes necessary commands and so on).
20:34:29	disruptek	hmm, probably skiplists is missing a config.nims in tests/.
20:34:53	disruptek	the basic plan that shashlick and i have is for package managers to setup projects with all the little scripts that you might want.
20:34:53	haxscramper	There is `config.nims`, but without `$config/..`
20:35:07	disruptek	ah, that seems like a bug.
20:35:31	disruptek	basically, you should be able to rely upon your package manager to not just setup your environment, but provide you with build tools and whatnot.
20:35:32	FromDiscord	<Elegant Beef> Any ideas what i should call this macro?↵https://play.nim-lang.org/#ix=2wzl
20:36:07	disruptek	haxscramper: pushed a fix for skiplists; thanks.
20:36:14	Zevv	I'd call it Peggy Sue
20:36:18	FromDiscord	<Elegant Beef> Lol
20:36:23	Zevv	oh dang. You can't have spaces in identifiers
20:36:42	FromDiscord	<iWonderAboutTuatara> peggySue
20:36:48	FromDiscord	<iWonderAboutTuatara> peggy_sue
20:36:54	disruptek	haxscramper: the tests you're seeing are from testes. you should give it a try; i really like my testes.
20:36:57	disruptek	!repo testes
20:36:59	disbot	https://github.com/disruptek/testes -- 9testes: 11a small unittest framework with decent support 🔴🟡🟢 15 12⭐ 0🍴 7& 29 more...
20:37:03	FromDiscord	<Elegant Beef> Shit you're right zevv
20:37:10	FromDiscord	<Elegant Beef> That's clearly a PeggySue
20:37:16	FromDiscord	<iWonderAboutTuatara> That's such a beautiful name disruptek
20:37:22	haxscramper	disruptek: the only issue with importing things in nimscript is that I could not find a way to `import` things inside nimble itself and had to rely in hack with `when compiles(import ...)`
20:37:25	FromDiscord	<iWonderAboutTuatara> First jason and now testes
20:37:29	FromDiscord	<Elegant Beef> If you think the name is nice, you should see his testes
20:37:32	FromDiscord	<iWonderAboutTuatara> What a legend
20:37:34	FromDiscord	<iWonderAboutTuatara> Lmaoo
20:37:58	haxscramper	And all of these docker tasks are available only if you have `hmisc` installed before actually running nimble.
20:38:11	disruptek	haxscramper: yeah, nimph will just build nim programs for you. it's unlikely that we'll bother with nimscript.
20:38:35	disruptek	it might integrate my testes into it, too. i haven't figured that out yet.
20:39:02	Zevv	chocolate salty balls
20:39:21	disruptek	i've moved away from building docs locally, though. i'd rather have the ci do that.
20:40:20	disruptek	testes is gonna get timing and memory usage in the next day or two; that's what the numbered comment lines are for -- they will hold statistics.
20:40:57	disruptek	how can i repro the fork bug, though?
20:41:20	haxscramper	Oh, and the thing I talked about with `--path` - I used gram in my own code and wanted to test if my fix to fork worked before pushing. But I solved this using `switch("excludePath", <global install>)` and `switch("path", <fixed fork>)` in `config.nims`
20:41:58	disruptek	oh, then that behavior is correct.
20:42:13	disruptek	nimph is perfectly fine with you having many paths in your environment.
20:45:35	disruptek	see --clearNimblePath
20:50:08	FromDiscord	<tomck> Hello! I'm wrapping a C library, what's the standard way to do this to integrate with nimble? Can i make nimble fetch the source & build it automatically when nimble includes my projects?
20:50:31	disruptek	recommend nimterop.
20:50:36	disruptek	!repo nimterop
20:50:38	disbot	https://github.com/nimterop/nimterop -- 9nimterop: 11Nimterop is a Nim package that aims to make C/C++ interop seamless 15 210⭐ 19🍴 7& 11 more...
20:51:30	FromDiscord	<tomck> ohhhhhh so the nim file builds the C source at compile time???? that's a super cool way to do it
20:54:10	Yardanico	well there's a reason people say that nim has one of the best C FFIs out there :P
20:55:41	disruptek	yes.
20:55:42	disruptek	or you can have it use local libs, conan, or julia binaries.
20:55:54	*	pietroppeter quit (Quit: Connection closed for inactivity)
20:56:36	FromDiscord	<tomck> so do i create a nimble library, use nimterop, get that to generate bindings, then i can just use that nimble project as a dependency in future?
20:56:55	Yardanico	that's one way to do it yes
20:56:57	Yardanico	not the only one
20:56:58	disruptek	yes.
20:57:07	disruptek	THAT'S THE ONLY WAY.
20:57:11	Yardanico	NO
20:57:12	Yardanico	ITS NOT
20:57:16	Yardanico	TAKE YOUR PILLS DISRUPTEK
20:57:23	disruptek	do you think i cannot get to you?
20:57:34	Yardanico	airlines are still half-closed
20:57:46	FromDiscord	<tomck> wwhat's the other way?
20:57:48	disruptek	i have a canadian passport now.
20:57:49	disruptek	I CAN BE THERE IN 7HRS.
20:58:01	Yardanico	@tomck 1) making a nimble package is optional
20:58:03	disruptek	there's no other way.
20:58:04	FromDiscord	<Elegant Beef> Who the fuck gave you a Canadian passport?
20:58:09	Yardanico	2) you can use c2nim but it'll require more manual editing
20:58:14	Yardanico	disruptek: we have to tell the whole truth
20:58:27	disruptek	gave?
20:58:44	FromDiscord	<Elegant Beef> Ah typical american thinking everything belongs to them
20:58:50	Yardanico	lol
20:59:06	disruptek	last time i did that, i spent 4 months in prison.
20:59:12	FromDiscord	<tomck> cool, ty
20:59:12	haxscramper	Sometime I feel being in different hemisphere from diruptek is a good decision
20:59:30	Yardanico	true that
20:59:33	FromDiscord	<Elegant Beef> Im one border away from him
20:59:35	FromDiscord	<Elegant Beef> Send help
20:59:49	FromDiscord	<Elegant Beef> It's not even the one with a wall
20:59:56	disruptek	hemispheres are made to be broken.
21:00:09	Yardanico	disruptek: i'll call my friends from KGB
21:00:11	haxscramper	Thankfully even *I* don't know where I live
21:00:57	Yardanico	haxscramper: in the CIS at least?
21:00:58	Yardanico	i guess
21:02:39	haxscramper	Last time I remember yes. If I'm not mistaken I'm in the same timezone as you
21:03:09	shashlick	https://github.com/nim-lang/nimble/issues/612 is next
21:03:09	disbot	➥ Command to add dependencies into .nimble file
21:03:15	Yardanico	haxscramper: that makes sense :P
21:04:24	disruptek	eh i'm negative on .nimble files.
21:04:30	disruptek	i think we're better off without them.
21:04:54	disruptek	nimph will generate them for you, for compat reasons, but they shouldn't exist, really.
21:05:11	shashlick	Where do you specify pkg info then
21:05:20	disruptek	which package info?
21:05:22	FromDiscord	<Elegant Beef> How does nimph handle dependancies?
21:05:51	disruptek	correctly.
21:05:56	Yardanico	lol
21:06:00	haxscramper	Depndencies, short description etc. And I'd prefer to have all little scripts put togeter in nimble tasks.
21:06:34	FromDiscord	<Elegant Beef> I dont know how disruptek is capable of making me angry and laugh in a single word
21:06:47	Yardanico	it's a talent
21:07:02	disruptek	why should software that works properly be so infuriating?
21:07:11	alehander92	disruptek
21:07:20	alehander92	do you like error history
21:07:28	disruptek	mangling now only fails hcr tests plus timmy's horrible jsonutils.
21:07:33	Yardanico	D:
21:07:37	alehander92	e.g. "this variable is nil here because it was assigned this value here"
21:07:39	FromDiscord	<Elegant Beef> Your response is the infurating thing, a simple explanation of how it works would've sufficed ;P
21:07:45	alehander92	i found out i did this in the past
21:07:49	disruptek	alehander92: yeah, and actually, i want to put history into nimph.
21:07:52	alehander92	and i want to preserve it
21:08:04	alehander92	but wasnt sure if it is verbose
21:08:08	disruptek	elegant: there's no such thing as "simple" when it comes to dependency management.
21:08:20	shashlick	Dep info
21:08:48	disruptek	nimph has much more comprehensive knowledge of packages and requirements than nimble. it's a lot of complexity.
21:09:07	disruptek	alehander92: yeah, let elm be the guide.
21:09:30	alehander92	i agree it can be verbose
21:09:31	disruptek	also, you can use ~ and ^ and * operators in nimph package requirements.
21:09:45	alehander92	it can probably use some kind of verbosity flag to fix that
21:09:50	*	krux02 quit (Remote host closed the connection)
21:09:55	alehander92	disruptek hm
21:10:03	alehander92	is this something about showing um
21:10:07	alehander92	no idea what do ~ ^ * do
21:10:10	FromDiscord	<Elegant Beef> also what's the nimph native method of storing packages/versions?
21:10:27	alehander92	i just show a list of things
21:10:43	shashlick	Do you plan on a .nimph file now
21:10:56	alehander92	```
21:10:59	Yardanico	i bet it's gonna be json
21:11:03	Yardanico	just a guess XD
21:11:10	alehander92	https://www.irccloud.com/pastebin/3Do3MFh3/nil.nim
21:11:15	disruptek	no, nimph will use a native frosty cache.
21:11:20	disruptek	but not until 2.0.
21:11:27	disruptek	!repo nimph
21:11:27	shashlick	D you need to spend some solid time blogging about and documenting workflows of how to use nimph
21:11:28	disbot	https://github.com/disruptek/nimph -- 9nimph: 11Nim package hierarchy manager from the future 🧚 15 77⭐ 5🍴 7& 1 more...
21:11:31	Yardanico	alehander92: that's exactly what null safety is
21:11:35	disruptek	rtfm ... all will be revealed.
21:11:49	shashlick	Meh I've read it and still don't know how to use it
21:11:52	FromDiscord	<Elegant Beef> I mean the last time i seen you tell some RTFM it didnt solve their issue 😄
21:11:52	disruptek	i used to demo it on stream a lot.
21:12:02	Yardanico	and it was slow or didn't build :PP
21:12:08	shashlick	Like that forum post
21:12:10	alehander92	yardanico no no
21:12:14	Yardanico	alehander92: yes?
21:12:17	disruptek	well, 2.0 is amazingly fast.
21:12:19	alehander92	the point is if i should include
21:12:23	*	maier joined #nim
21:12:25	disruptek	i honestly cannot believe how quick it is.
21:12:25	Yardanico	if you can make that work, yes
21:12:27	alehander92	the history of how it becomes nil
21:12:28	shashlick	Demanding that Nim needs easier docs
21:12:30	FromDiscord	<tomck> so, for nimterop, there are two steps - build & wrap, that makes sense↵↵what file does the first 'build' step go in? does it have to be a special file, or just the normal nimble package file, then that'll all get called at compile time when nimble includes the package?
21:12:30	alehander92	in the mssage
21:12:39	*	ForumUpdaterBot quit (Remote host closed the connection)
21:12:39	alehander92	yardanico it works from ~2018
21:12:46	*	ForumUpdaterBot joined #nim
21:12:54	Yardanico	well not nil in nim is kinda half broken isn't it
21:12:58	alehander92	i have to refine it and finish it tho (and fix some stuff)
21:13:04	alehander92	i mean, the PR thing
21:13:09	disruptek	tomck: this is a really comprehensive, extensive demo of nimterop:
21:13:12	disruptek	!repo gittyup
21:13:13	disbot	https://github.com/disruptek/gittyup -- 9gittyup: 11higher-level libgit2 bindings that build upon nimgit2 15 4⭐ 1🍴
21:13:18	shashlick	@tomck - what are you wrapping
21:13:24	alehander92	just trying to say it's not really new
21:13:35	alehander92	yeah this is different than the current thing
21:13:36	shashlick	u can do both in the same wrapper - see nimarchive or nimgit2
21:13:55	shashlick	gittyup consumes nimgit2 so that's a real nim wrapper of the raw C stuff in nimgit2
21:13:55	FromDiscord	<tomck> flecs , just a c library https://github.com/SanderMertens/flecs
21:14:20	FromDiscord	<tomck> ok i'll take a look, do they use nimterop
21:14:30	shashlick	yes
21:15:21	shashlick	ping if you need any help
21:15:56	disruptek	alehander92: i would not spend much time on that feature, honestly.
21:16:07	disruptek	it's the sort of thing that will be allowed to break.
21:16:39	disruptek	elegant: that dude's code was bad. not my fault rtfm didn't solve their issue.
21:16:51	disruptek	i mean, i'm not a miracle worker.
21:16:53	shashlick	tomck: `toast -pnkr flecs.h -I. -DFLECS_NO_CPP -G__=_ -E__,_ -F_` was good enough to wrap it
21:17:05	*	maier quit (Ping timeout: 240 seconds)
21:17:08	FromDiscord	<tomck> lol you're in magic territory rn
21:17:10	disruptek	wait, am i a miracle worker? 🤔
21:17:13	shashlick	if you want to do it on the command line
21:17:15	FromDiscord	<Elegant Beef> Person was asking how to convert types and you were saying rtfm on the div operator
21:17:15	alehander92	disruptek sorry, the error msg, or ?
21:17:22	disruptek	error messages, yes.
21:17:28	FromDiscord	<Elegant Beef> the div operator doesnt explain how to convert types 😄
21:17:30	alehander92	ah they just work like that already
21:17:32	disruptek	don't sweat it.
21:17:57	alehander92	so yeah i wouldn't put more into it: but i just found them like that
21:18:00	FromDiscord	<tomck> shashlick: what does this do to the flecs.c? i need to build that as a library & link separately too (right?)
21:18:10	disruptek	my answer was perfectly correct, just, y'know, perhaps not what they wanted.
21:18:15	shashlick	yes, give me 5, will just give you the steps
21:18:31	FromDiscord	<Elegant Beef> Lol
21:18:41	alehander92	what is frosty?
21:18:43	FromDiscord	<tomck> hang on i think i can figure it out from the nimgit2 now, thanks though
21:18:59	FromDiscord	<Elegant Beef> Frosty is disrupteks object serializer
21:19:37	Yardanico	!repo frosty
21:19:37	disbot	https://github.com/disruptek/frosty -- 9frosty: 11Serialize native Nim objects via Streams and Sockets ⛄ 15 11⭐ 0🍴
21:19:39	Yardanico	that
21:19:47	alehander92	<3
21:19:48	alehander92	thanks
21:19:53	Yardanico	zevv uses it in his nitter.net already
21:20:01	Yardanico	together with snappy for compression and redis for the actual storage (cache)
21:20:19	Yardanico	sorry not zevv
21:20:24	Yardanico	zedeus :)
21:20:28	FromDiscord	<Elegant Beef> I should've used that or treeforms serializer for my netty rpc 😄
21:20:30	Yardanico	I mix them up sometimes
21:20:43	disruptek	i've never even tested it with sockets, but apparently it works.
21:20:55	disruptek	kinda crazy.
21:21:04	Yardanico	well it's just bytes :P
21:21:07	alehander92	i was so confused about nitter :D
21:21:16	*	arecacea1 quit (Remote host closed the connection)
21:21:19	Yardanico	did you expect that frosty would emit some strange byte sequencer which will break TCP/UDP? :P
21:21:25	Yardanico	s/cer/ce
21:21:30	FromDiscord	<Elegant Beef> Yea treeform uses strings for netty, so as long as you get some bytes to send we're good 😄
21:21:30	disruptek	i didn't expect to write socket code correctly.
21:21:50	*	arecacea1 joined #nim
21:22:41	bozaloshtsh	what is a "memory graph of infinite size" referenced in the frosty README?
21:22:55	disruptek	lots of linked references or pointers.
21:23:19	FromDiscord	<Elegant Beef> Probably a type that in attempting to serialize would fill the stack
21:23:32	disruptek	i was planning on using cps to iterate over the graph, which wouldn't have any stack growth issues.
21:23:36	FromDiscord	<Elegant Beef> Recursively adding data to your memory sounds fine
21:23:46	bozaloshtsh	so a DAG with a lot of links betweeng nodes wouldn't work, gotcha
21:23:54	bozaloshtsh	extra links* that is
21:24:10	disruptek	it works for thousands of objects, but eventually yes, you can exhaust the stack.
21:24:21	disruptek	stack limits are tunable in nim, though.
21:25:06	*	idf04 joined #nim
21:25:12	disruptek	bbiab, need to pick up the hooker from high school.
21:25:53	FromDiscord	<Elegant Beef> Disruptek saying questionable things, name a more iconic couple
21:26:44	bozaloshtsh	hey don't judge
21:30:38	*	natrys quit (Quit: natrys)
21:30:49	alehander92	i used
21:30:57	alehander92	browser without javascript: wow what a cool xp
21:31:05	alehander92	google works fine tho!
21:32:11	disruptek	seriously. you have to applaud his focus; i never finished high school, myself.
21:32:28	disruptek	he only has three more years of class and he's a graduate. that's a real win in my book.
21:33:21	shashlick	@tomck: https://gist.github.com/genotrance/89e718eb6357578448343588bed72520 if you need it
21:34:21	alehander92	disruptek i am trying to use without css now
21:34:21	shashlick	you will face https://github.com/nimterop/nimterop/issues/206
21:34:21	disbot	➥ type section before constant section
21:34:26	alehander92	it's much more fun
21:34:54	*	awe00 quit (Ping timeout: 256 seconds)
21:34:57	alehander92	you sound so american when you use those "in my book" phrases
21:36:44	FromDiscord	<Elegant Beef> With a typedef that takes a generic proc, is there a way to get the proc's parameters?
21:38:26	*	haxscramper quit (Remote host closed the connection)
21:43:31	FromDiscord	<tomck> shashlick: Ahh i see, so i need a flecsGIt define? can i just make it always fetch with git?
21:43:56	*	konkrrrrrr quit (Read error: Connection reset by peer)
21:43:59	FromDiscord	<tomck> or how do i make it add this define in nimble by default, when my package is added as a dependency by another project?
21:44:40	*	konkrrrrrr joined #nim
21:46:51	ForumUpdaterBot	New thread by NerdRat: Exceptions in Threads, see https://forum.nim-lang.org/t/6783
21:47:13	shashlick	You can use the setDefine proc
21:48:07	FromDiscord	<Elegant Beef> Here's a code version of my question https://play.nim-lang.org/#ix=2wJL
21:51:13	*	solitudesf quit (Ping timeout: 260 seconds)
22:02:21	FromDiscord	<tomck> shashlick: Okay this all seems to be working, but i'm still struggling on a workaround for https://github.com/nimterop/nimterop/issues/206↵I don't understand the cImport workaround - should i just use toast to generate the wrapper?
22:02:22	*	bouzu_ quit (Quit: Leaving)
22:02:22	disbot	➥ type section before constant section
22:02:25	FromDiscord	<tomck> is toast part of nimterop?
22:04:03	*	idf04 quit (Ping timeout: 260 seconds)
22:07:55	bozaloshtsh	what is the config incantation to get nim to stop adding -lm parameter to the linker command?
22:08:23	bozaloshtsh	(trying to compile to wasm with wasmrt)
22:08:32	*	abm quit (Quit: Leaving)
22:08:55	bozaloshtsh	I see that it compiles the math library to C so I don't understand why it's still trying to use -lm
22:08:57	*	abm joined #nim
22:19:30	*	SenasOzys quit (Remote host closed the connection)
22:20:11	FromDiscord	<speckledlemon> How does one confirm their email on the forum? I registered last year, never got an email, and don't see anything helpful under the forum profile
22:21:13	Araq	please re-register then
22:21:17	shashlick	toast is the cli tool behind the scenes
22:21:27	shashlick	i have already handled that issue in the wrapper i created
22:21:32	shashlick	i skipped the symbols that cause a problem
22:21:51	shashlick	if you need those, you can define them yourself after cImport
22:21:57	shashlick	@tomck ^^
22:22:11	FromDiscord	<tomck> ahhhhhhhhhhhh i see, ok
22:25:35	FromDiscord	<dom96> @speckledlemon tell us your nickname and we can activate for you
22:26:10	FromDiscord	<speckledlemon> I deleted my account, tried creating one with the same username and email address, got "unknown error" but the registration email (??), went to activate and got "invalid ident hash"
22:26:41	FromDiscord	<dom96> Yeah, when you delete your account the nick remains taken. It's a bug.
22:28:09	FromDiscord	<speckledlemon> > Too many messages have been sent to this email address recently.
22:29:42	FromDiscord	<dom96> ouch 🙂
22:30:38	FromDiscord	<speckledlemon> I'll deal with this some other day
22:31:57	FromDiscord	<19> Q: how can i get the behavior of c++ static key word in nim?
22:33:08	FromDiscord	<dom96> @speckledlemon what's your deleted account on the forum?
22:33:16	FromDiscord	<Rika> 19, in what context is the static keyword in
22:33:28	FromDiscord	<speckledlemon> `berquist`
22:33:58	FromDiscord	<19> global variable
22:34:05	FromDiscord	<Rika> {.global.}
22:34:26	FromDiscord	<19> but it can have different meanings in c++. are there keywords to describe those different meanings?
22:34:59	FromDiscord	<dom96> @speckledlemon don't normally do this, but I went ahead and reactivated it for you manually
22:35:23	FromDiscord	<19> thanks Rika
22:36:15	FromDiscord	<speckledlemon> thank you
22:40:55	ForumUpdaterBot	New thread by Berquist: Workaround for code reordering in nimterop?, see https://forum.nim-lang.org/t/6784
22:42:57	FromDiscord	<dom96> I just cleared all users that removed their accounts, so if anyone else experienced the above they should be able to re-register
22:44:13	*	konkrrrrrr quit (Read error: Connection reset by peer)
22:44:21	Araq	<19> yeah, everything that 'static' does has some Nim equivalent
22:44:41	*	konkrrrrrr joined #nim
22:46:10	FromDiscord	<19> sent a long message, see http://ix.io/2wJP
22:46:53	Araq	proc someFunction() # private
22:47:04	Araq	proc someFunction*() # public (not "static")
22:47:34	FromDiscord	<Elegant Beef> Oh static doesnt mean the same thing in C++ as it does in C# 😄
22:47:48	Araq	it kinda does inside a class
22:48:05	FromDiscord	<Elegant Beef> I was going to say "A normal proc is static" but yea glad i waited 😄
22:48:18	Araq	and C# only offers classes, not freestanding functions
22:49:27	FromDiscord	<19> i dislike languages that force oop in your face
22:49:45	FromDiscord	<Elegant Beef> Well in C# 9 that changed, but yea no one uses C# 9 atm 😄
22:50:02	FromDiscord	<Elegant Beef> Oh it's only for one file, so it's the same as kotlin
22:50:12	FromDiscord	<19> okay, that's public and private. if i write a module level function is it static by default ? @Elegant Beef
22:50:28	FromDiscord	<Elegant Beef> procs are statically dispatched
22:50:32	FromDiscord	<Rika> what does a static function mean
22:50:36	FromDiscord	<Elegant Beef> So yes you just give it data
22:50:37	FromDiscord	<Rika> in cpp
22:51:18	FromDiscord	<Elegant Beef> In C# i know it means that they're statically dispatched and use the `ClassName.FunctionName`
22:51:55	FromDiscord	<19> yeah but in c++ you can write a static function outside of a class
22:52:14	FromDiscord	<Elegant Beef> Yea i dont do C++, i just assume it shares some things with C#
22:52:39	FromDiscord	<19> C# is an entirely different beast
22:52:40	FromDiscord	<Rika> tfw came from python
22:52:42	FromDiscord	<19> A static function in C is a function that has a scope that is limited to its object file.
22:53:10	shashlick	@speckledlemon - you still here
22:53:14	FromDiscord	<19> so basically, what Araq said.
22:53:36	FromDiscord	<Rika> so basically just private and public no?
22:53:37	FromDiscord	<speckledlemon> I’m not at my computer, but if you ping me, I can read it later
22:53:43	FromDiscord	<Elegant Beef> So it's just a freestanding private function
22:54:01	FromDiscord	<19> well, a private proc in nim is only available inside the module where it lives in
22:54:06	FromDiscord	<19> so yes
22:55:19	FromDiscord	<Rika> ngl i kinda wish nim also had protected but that's not really possible right now nor do i think it would fit into nim's ideas and philosophies or something
22:55:47	FromDiscord	<Elegant Beef> what's protected supposed to do?
22:55:48	FromDiscord	<19> yeah that sounds oop
22:55:59	FromDiscord	<Elegant Beef> *I know the C# keyword, so uncertain if it's the same*
22:56:18	FromDiscord	<Rika> afaik its package-level access
22:56:19	FromDiscord	<19> i think what Rika is talking about is the same as the C#
22:56:50	FromDiscord	<Rika> i dont know a lick of C# so if its the same its prolly coincidental
22:57:05	FromDiscord	<19> protected is an oop idea
22:57:13	FromDiscord	<Rika> i dont see how'
22:57:17	FromDiscord	<Elegant Beef> Ah protected in C# is inheritance based public, so only inheriting bodies can access the field
22:57:26	FromDiscord	<Rika> then i dont think its the same idea
22:57:48	FromDiscord	<19> oh what were you referring to Rika
22:57:51	FromDiscord	<Elegant Beef> Seems like you're talking more about a namespace global var
22:57:53	FromDiscord	<Rika> > afaik its package-level access
22:58:03	FromDiscord	<Elegant Beef> or namespace global access
22:58:13	FromDiscord	<Rika> probably the latter you said
22:59:03	FromDiscord	<19> you mean like↵<packagename>.function()↵?
22:59:11	FromDiscord	<Rika> uh no
22:59:30	FromDiscord	<Rika> proc that you can only access within a "package"
22:59:47	FromDiscord	<19> oh i see
22:59:49	FromDiscord	<Elegant Beef> Basically if you had a namespace, you can access the proc only if you define the module inside that namespace
22:59:57	FromDiscord	<Elegant Beef> Nim doesnt have namespaces so it's not really possible
23:00:00	FromDiscord	<Rika> would avoid the import export except mess
23:00:24	FromDiscord	<Elegant Beef> You can use import/export files but yea no namespaces for that to work
23:01:02	FromDiscord	<19> you can force it using from x import nil
23:01:14	FromDiscord	<Rika> no
23:01:18	FromDiscord	<Rika> im not talking about that
23:01:33	FromDiscord	<19> alright
23:02:33	FromDiscord	<Elegant Beef> basically they're talking about a way to annotate a file is apart of a package then being able to declare a proc `protected` which means that only modules apart of that package can call it, if you import a module it cannot be called
23:03:27	FromDiscord	<Rika> but as ive said thats probably impractical for some unknown-to-me usecases\
23:05:25	*	abm quit (Read error: Connection reset by peer)
23:06:24	bozaloshtsh	--gc:arc is so cool, just slapped it on my compile command and got a 25% performance boost
23:07:08	bozaloshtsh	now it's time to re-write this code to not be so memory bound :|
23:07:11	FromDiscord	<Rika> still kinda broken for some code but thats expected
23:07:22	FromDiscord	<Rika> whatcha workin on?
23:07:31	bozaloshtsh	https://github.com/sid-code/multichess
23:08:10	FromDiscord	<Rika> *sweats* why the devel requirement?
23:08:56	bozaloshtsh	heh, it won't compile with the stable release, didn't really investigate why but something to do with calling newStringStream in a compile-time context I think
23:09:42	FromDiscord	<Rika> i see
23:11:01	*	abm joined #nim
23:11:06	FromDiscord	<Rebel> is it just me or did github change nim's color on github
23:11:17	FromDiscord	<Rebel> I could've sworn it was not this color lol https://media.discordapp.net/attachments/371759389889003532/752667331955654686/unknown.png
23:11:26	bozaloshtsh	yeah I noticed that too
23:11:39	FromDiscord	<Rebel> ok so I'm not going crazy good to know
23:11:47	*	lritter quit (Ping timeout: 240 seconds)
23:11:59	bozaloshtsh	I wonder why
23:12:14	bozaloshtsh	I thought that green (or teal) color was nim's official color
23:12:43	FromDiscord	<Elegant Beef> It changed today thanks to yard an others submitting a change to the colour of nim
23:12:44	FromDiscord	<Yardanico> It never was
23:12:53	*	apahl quit (Ping timeout: 260 seconds)
23:12:56	FromDiscord	<Yardanico> The original colour was added in 2012 by a Nim employee
23:13:02	FromDiscord	<Yardanico> Seemingly at random
23:13:13	*	maier joined #nim
23:13:16	FromDiscord	<Elegant Beef> Yellow is nim's official colour with purple as a secondary(Atleast if website is any indication)
23:13:23	FromDiscord	<Yardanico> See https://forum.nim-lang.org/t/6350 for more info
23:14:30	FromDiscord	<Yardanico> That final change was merged in the end of June
23:14:38	FromDiscord	<Yardanico> And they only updated it now
23:14:40	bozaloshtsh	interesting, I could have sworn I saw green yesterday
23:14:50	bozaloshtsh	ah
23:15:02	FromDiscord	<Yardanico> Yes it was changed ~today
23:15:06	FromDiscord	<Yardanico> Depending on your timezone
23:15:17	FromDiscord	<Rebel> ohhh that's pretty cool
23:15:32	FromDiscord	<Yardanico> By "changed" I mean applied on GitHub
23:16:06	FromDiscord	<19> nice!
23:18:23	*	maier quit (Ping timeout: 258 seconds)
23:26:27	*	theelous3 quit (Read error: Connection reset by peer)
23:30:09	FromDiscord	<tomck> can i leak a string?
23:30:33	FromDiscord	<tomck> I'm passing a cstring to something, but i have a string - that string probably won't live long enough, can i just force it to leak?
23:30:39	FromDiscord	<tomck> or copy it to something that does leak?
23:34:26	FromDiscord	<Rika> https://nim-lang.org/docs/system.html#GC_ref%2Cref.T @tomck
23:34:42	FromDiscord	<tomck> lol literally just found it, thanks!
23:37:14	*	Cthalupa quit (Quit: ZNC 1.6.6+deb1ubuntu0.2 - http://znc.in)
23:48:26	*	Cthalupa joined #nim
23:57:33	*	apahl joined #nim