<< 14-09-2017 >>

---

00:01:14	def-pri-pub	It's in `util.nim`. I think it might be better to go in it's own file and then imported by `util.nim`
00:02:42	FromGitter	<zacharycarter> agreed
00:02:48	FromGitter	<zacharycarter> I don't know if we'll ever get around to using it in zengine
00:05:38	def-pri-pub	Weren't you working on a website for zengine a while back?
00:06:57	FromGitter	<zacharycarter> I was but I stopped I didn't really get anywhere with it I got sidetracked trying to do soemthing with karax :P
00:07:00	FromGitter	<zacharycarter> I should start on one again
00:07:23	jsgrant	Just noticed; There's no nim2nix yet.
00:07:32	FromGitter	<zacharycarter> nim2nix?
00:07:44	def-pri-pub	yeah, I was just about to ask
00:07:58	def-pri-pub	(damn, the IRC logs are already logged in google)
00:08:38	jsgrant	Compat-layer for NixOS; Basically allows a prog-lang's packagemanager to talk to "nix" and install to the system proper to be run.
00:08:49	jsgrant	Because you don't run local-binaries on NixOS.
00:09:10	FromGitter	<zacharycarter> dah fuq is NixOS?
00:09:19	FromGitter	<zacharycarter> ohhh yeah
00:09:21	FromGitter	<zacharycarter> this thing
00:09:35	jsgrant	All but this machine (Fedora) are now on NixOS; Might be a fun-project ... still a year out before I do trivial Nim-hacking it looks. Ordering dom96's book tomorrow though. :^)
00:10:12	*	jsgrant is making a placeholder repo for now, as a reminder, regardless.
00:10:45	FromGitter	<zacharycarter> why are you waiting a year to start hacking on Nim?
00:12:11	*	mahmudov joined #nim
00:13:48	jsgrant	Well, more closely it's going to be a half-year once I start the beginning of the next on my self-imposed "curriculum"; This year has been a weird "reset/rebase" kind-of thing a lot of which is going through actual overly-romantic withdrawls from Lisp.
00:14:36	jsgrant	More-or-less "reteaching" myself to program, non-trivially, outside of Lisp... which I haven't done in like 4-5 years to any notable degree and really not at all, sans the last few months, for like 3 years prior.
00:15:52	FromGitter	<zacharycarter> yeah I feel that's one problem (major) with lisp
00:16:19	FromGitter	<zacharycarter> most lisp programmers I talk to about other programming languages shudder at the thought of switching
00:16:59	FromGitter	<zacharycarter> I think it's the investment one has to make into lisp to become competent in it
00:17:54	jsgrant	Once I finally, after like 2 years of bulling about it ... get my blog up, will eventually write about it; Just such an odd mindset I was in.
00:18:37	jsgrant	The fact that I had any -- let-alone such strong feelings of "selling out" trying to move away from it, is literally crazy to me.
00:19:13	FromGitter	<zacharycarter> :P embrace the new! embrace the Nim!
00:19:43	FromGitter	<zacharycarter> I'm attempting to write a non-recursive flood fill algo implementation and it's not going well
00:21:24	FromGitter	<zacharycarter> is there any way to TCO optimize Nim code?
00:22:01	jsgrant	Was absurdly productive though, during that time; And what get's me is I don't know how much of it was actually due to nice lil quirks in Lisp & how much of it was the almost unshakeable "man on a mission" mindset I had.
00:23:09	jsgrant	The latter, was certainly pushed in the community; And too (I think I've mentioned in here & certainly in other channels generally) I've been pensive to really peruse anything/any community that seems to mirror that ... namely Haskell.
00:23:45	jsgrant	Think there must be a sort-of "brain" that is apt to buy into that kind-of stuff in full-force. :^P
00:24:00	FromGitter	<zacharycarter> meh I gave CL a shot and just floundered for a few days trying to get things working
00:24:15	FromGitter	<zacharycarter> SDL wouldn't work and then I changed from SBCL to some other implementation and things still didn't work
00:24:31	FromGitter	<zacharycarter> then I tried out chicken scheme and the example project there didn't work
00:24:34	jsgrant	zach: One of the reasons Nim actually interested me initally, actually, is the Lispish macro-system.
00:24:59	FromGitter	<zacharycarter> makes sense
00:25:02	jsgrant	Ever play with CEPL?
00:25:27	*	vlad1777d quit (Ping timeout: 240 seconds)
00:26:12	FromGitter	<zacharycarter> nope! is that like a C repl?
00:26:34	jsgrant	*Marco/DSL-Lust is also a big reason why I too started reorienting to Scala.
00:26:49	*	renarc joined #nim
00:26:50	jsgrant	https://github.com/cbaggers/cepl
00:27:30	jsgrant	Baggers also has some pretty nice/decent demo's on his linked Youtube channel.
00:28:19	jsgrant	There was a good half-dozen or so Lisp Gamedev projects about; When I was still about.
00:28:51	jsgrant	Clinch was pretty nice too, really.
00:30:10	FromGitter	<zacharycarter> yeah I talked to a few lips users who were trying to get me to switch and were touting their game projects
00:30:17	FromGitter	<zacharycarter> I still think nim is better suited for gamedev
00:32:44	jsgrant	Yeah, at this point I agree; But your mindset (and mine not-nearly-as-much) doesn't fall into that "Assimilate" modulus.
00:33:34	FromGitter	<zacharycarter> true
00:33:59	jsgrant	I was running Emacs + Stumpwm for like 80% of my environment (dev or otherwise); And that still wasn't enough for me.
00:34:53	jsgrant	You sorta-kinda see that in "RIIR" (Rewrite It In Rust), but even then Lisp is a whole 'nother level because there are parties (myself included) who wanted LispMachines/LispOSes to "be a thing" again.
00:36:35	FromGitter	<zacharycarter> yeah
00:36:44	FromGitter	<zacharycarter> I def see it in the rust community
00:41:58	*	TjYoco joined #nim
00:42:59	jsgrant	But yeah, basically I "have a history" of falling into some of these absolutist traps & once in, it's absurdly hard for me to crawl out of -- so I've more-or-less crafted the loosely aforementioned "curriculum" with the explicit goal if going slow & curbing my enthusiasm.
00:44:48	jsgrant	Trying to go the "slow, but gainful payoffs" game; Which is certainly very counterintuitive to me.
00:47:57	jsgrant	Pretty sure I've functionally got to a point where I know what I'm doing -- but absolutely get why it sounds odd when I say, basically, "Oh yeah, I'm excited to play with Nim; Only going to take me 6+ months to a year to do anything non-trivial in!" :^P
00:55:18	FromGitter	<zacharycarter> hehe
00:55:26	FromGitter	<zacharycarter> makes sense
01:12:02	*	gangstacat quit (Quit: Ĝis!)
01:14:55	*	def-pri-pub quit (Quit: leaving)
01:18:25	*	TjYoco quit (Ping timeout: 248 seconds)
01:19:34	subsetpark	Araq dom96 : is there a handy list anywhere of the C types that Nim uses to represent its base types?
01:23:47	FromGitter	<zacharycarter> subsetpark: how goes the roguelike?
01:25:05	subsetpark	zacharycarter - I converted it to BLT and got roughly working everything already in place - however, I have this Nim/Python talk coming up soon so I have to work on my utility library for that instead.
01:26:08	FromGitter	<zacharycarter> gotcha
01:40:02	*	chemist69_ quit (Ping timeout: 255 seconds)
01:53:56	*	chemist69 joined #nim
01:58:47	*	endragor joined #nim
02:02:56	*	endragor quit (Ping timeout: 240 seconds)
02:56:31	*	Jesin quit (Quit: Leaving)
02:59:02	*	endragor joined #nim
03:01:59	*	Jesin joined #nim
03:03:47	subsetpark	zacharycarter: here's an extremely barebones, no-documentation first sketch: https://pypi.python.org/pypi/nim-types
03:03:54	subsetpark	`import ntypes`
03:26:26	*	onionhammer quit (Ping timeout: 240 seconds)
03:28:36	*	onionhammer joined #nim
03:29:52	*	dddddd quit (Remote host closed the connection)
03:37:05	*	onionhammer quit (Ping timeout: 248 seconds)
03:38:57	*	onionhammer joined #nim
03:41:21	*	onionhammer quit (Client Quit)
04:39:26	*	mahmudov quit (Remote host closed the connection)
04:53:23	FromGitter	<Bennyelg> When I am updating my package I should nimble publish again?
04:53:27	FromGitter	<Bennyelg> how is it working?
05:12:01	*	StarBrilliant quit (Ping timeout: 276 seconds)
05:13:08	*	miran joined #nim
05:19:06	*	StarBrilliant joined #nim
05:24:30	*	Vladar joined #nim
05:30:10	Araq_	benneyelg: no, you don't have to do that. if the URL changed, you need to create a manual PR I think
05:30:24	Araq_	if something else changed (dependencies) you don't need to do anything
05:30:53	Araq_	nimble loads the dependencies from your .nimble file on github and so its knowledge is always up to date
05:35:59	*	gangstacat joined #nim
06:02:27	FromGitter	<Bennyelg> Fantastic
06:02:32	FromGitter	<Bennyelg> https://github.com/Bennyelg/nimPresto ⏎ Is now on nimble
06:07:05	*	miran quit (Ping timeout: 240 seconds)
06:18:52	*	vlad1777d joined #nim
06:20:01	*	ShalokShalom quit (Remote host closed the connection)
06:22:53	FromGitter	<mratsim> @zacharycarter If you manage to implement a fast floodfill ping me. No one will tell you this but Go-playing engine are not compute or memory bound, they are floodfill bound since you have to run flood fill thousand of times per second in a go AI. ⏎ ⏎ Here is my implementation in Rust using functional iterators style: while let, filter_map, fighting the borrow checker: https:/
06:22:53	FromGitter	... /github.com/mratsim/rustygo/blob/master/src/main.rs#L358-L378
06:24:05	*	vlad1777d quit (Ping timeout: 240 seconds)
06:26:10	*	nsf joined #nim
06:39:28	*	jsgrant quit (Ping timeout: 240 seconds)
06:43:39	*	endragor quit (Remote host closed the connection)
06:46:44	*	spica joined #nim
06:51:59	*	ShalokShalom joined #nim
07:08:51	*	endragor joined #nim
07:12:06	*	yglukhov joined #nim
07:13:22	*	Arrrr joined #nim
07:13:22	*	Arrrr quit (Changing host)
07:13:22	*	Arrrr joined #nim
07:14:52	*	solitudesf joined #nim
07:16:52	*	yglukhov quit (Remote host closed the connection)
07:18:39	*	PMunch joined #nim
07:20:05	*	claudiuinberlin joined #nim
07:29:32	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
07:46:20	*	claudiuinberlin joined #nim
07:46:41	*	Sembei quit (Ping timeout: 240 seconds)
07:47:24	*	MyMind joined #nim
07:50:02	*	couven92 joined #nim
08:04:20	*	yglukhov joined #nim
08:07:45	*	spica quit (Ping timeout: 252 seconds)
08:08:31	*	yglukhov quit (Ping timeout: 248 seconds)
08:11:34	*	spica joined #nim
08:23:37	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
08:23:57	FromGitter	<Varriount> @mratsim Your Arraymancer library is really neat!
08:24:02	*	yglukhov joined #nim
08:24:08	*	couven92 quit (Quit: Client disconnecting)
08:27:10	*	claudiuinberlin joined #nim
08:31:46	*	couven92 joined #nim
08:33:11	*	Yardanico joined #nim
08:43:07	FromGitter	<Bennyelg> To whom using presto-db by facebook (sql engine) . ⏎ I created A nim connector :) ⏎ Now I'ts working pretty efficient ⏎ https://github.com/Bennyelg/nimPresto ⏎ ... [https://gitter.im/nim-lang/Nim?at=59ba411bbac826f054b7287f]
08:51:26	Arrrr	If i have a `proc test[T](t: T)` and T is `seq[A]`, is it possible to get the `A`, for example: `proc test[T](t: T) = test2[T.A](t[0])`
08:56:45	Araq_	template elemType[T](a: seq[T]): typedesc[T] = T
08:56:57	Araq_	^ untested though
08:56:59	*	dom96|w joined #nim
09:02:49	Arrrr	Good shot https://play.nim-lang.org/?gist=1a755f0dd718bfd6513ac7bce41fec35
09:03:17	*	Kingsquee quit (Quit: https://i.imgur.com/qicT3GK.gif)
09:04:26	Araq_	for me the button keeps spinning
09:04:35	Araq_	I guess compiler endless loop?
09:04:56	Arrrr	Is a thing playground does sometimes
09:04:56	Araq_	oh now it works, ok
09:05:00	PMunch	It did that for me as well, but then it suddenly worked
09:18:33	FromGitter	<Varriount> @Bennyelg "connect" is a rather ambiguous term - perhaps you should provide an additional procedure called prestoConnect, or connectToPresto?
09:20:11	FromGitter	<Grabli66> Is there "Any" type that i can use in runtime, not in compile time?
09:21:01	dom96|w	Varriount: I don't think that's necessary.
09:21:06	PMunch	You can create a type with a kind field like the JSON module does
09:21:09	dom96|w	As long as the `connect` proc works on a Presto type
09:22:13	FromGitter	<Varriount> dom96: It doesn't. It just takes a hostname, port, etc
09:22:28	Araq_	presto.connect
09:22:50	Araq_	not sure why people pretend we have no disambiguation means
09:23:12	dom96|w	yeah. I agree with Araq.
09:23:26	dom96|w	Araq_ you should kill your other BNC.
09:23:34	dom96|w	Log in to it and /msg *status disconnect
09:23:34	Araq_	I know
09:23:59	*	Araq quit (Quit: Bye!)
09:24:19	*	Araq_ is now known as Araq
09:24:20	FromGitter	<Grabli66> dom96|w, can i use ssl with jester?
09:24:23	FromGitter	<Varriount> Araq: Eh, I guess you have a point
09:24:48	dom96|w	Grabli66: yes, but not using Jester directly. You should put it behind nginx and set up SSL for it.
09:25:37	PMunch	I put mine behind Apache, worked pretty well
09:25:48	FromGitter	<Grabli66> dom96|w, do you plan to add native ssl support to jester?
09:26:09	PMunch	Why? That is just more code to maintain with little benefit
09:26:26	*	Yardanico quit (Remote host closed the connection)
09:26:48	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
09:27:41	FromGitter	<Grabli66> I dont want use nginx or apache. :)
09:28:45	FromGitter	<Varriount> It would benefit webapps, and small applications
09:29:14	FromGitter	<Varriount> I wonder if flask (python) supports ssl
09:31:35	PMunch	Well, someone could probably write a small reverse-proxy in Nim :)
09:33:45	dom96|w	Grabli66: Sure. But it's not a high priority for me.
09:37:11	PMunch	Varriount, it appears that it does: https://blog.miguelgrinberg.com/post/running-your-flask-application-over-https
09:40:36	*	claudiuinberlin joined #nim
09:44:35	couven92	Hmm, @Araq_ I still get `msgs.nim(602, 32) Error: ambiguous call; both tables.[](t: var Table[[].A, [].B], key: A)[declared in lib\pure\collections\tables.nim(167, 5)] and tables.[](t: Table[[].A, [].B], key: A)[declared in lib\pure\collections\tables.nim(161, 5)] match for: (Table[system.string, system.int32], string)` on the 0.17.2 sources
09:45:20	Araq	couven92: you still haven't fixed the vcc bug, so yeah
09:45:47	couven92	I have no idea what is going on here, so how am I to fix it? :O
09:46:30	dom96|w	if you want Jester to support SSL then asynchttpserver needs to support it first
09:46:42	dom96|w	and for that to happen you will likely need to wrestle quite a bit with openssl
09:47:00	couven92	yes, you pointed med at some function in some file, I still have no idea what I am looking at there... (I am reasonably sure it has something to do with the compiler, that's about as far as I got :P )
09:55:10	FromGitter	<Grabli66> dom96|w, asyncnet has ssl support.
09:55:32	dom96|w	Grabli66: Basic SSL support.
09:55:52	dom96|w	But do try to add SSL support to asynchttpserver :)
09:56:05	dom96|w	It might just work without any problems
10:00:08	yglukhov	dom96: btw, even "basic" ssl support needs improvement. cert validation almost never works (
10:00:35	dom96|w	Yeah... SSL is tricky :\
10:00:55	dom96|w	It took me months to get it this far.
10:02:27	yglukhov	Grabli66: https://github.com/yglukhov/variant
10:02:34	dom96|w	I think in the long run we should switch to BearSSL when it's ready https://bearssl.org/
10:03:23	yglukhov	oh, that is interesting. never heard
10:03:42	FromGitter	<Grabli66> yglukhov, thanks
10:04:07	dom96|w	It should be possible for us to link to it statically, which would be awesome.
10:04:27	dom96|w	In fact, it might be a good idea to start using it already.
10:05:31	yglukhov	yeah, i've also been thinking about nimble support for foreign static libs. i think it would be nice to put static libs for different platforms/archs right into the nimble package...
10:06:09	FromGitter	<Bennyelg> @Araq what is the disamiguation when I use connect?
10:06:23	yglukhov	as a matter of fact nimble can already allow that, so nevermind
10:07:12	dom96|w	Yes, via submodules.
10:07:12	yglukhov	but yeah, static ssl lib (with a lot of platform/arch variants) would be great
10:07:21	dom96|w	(git submodules)
10:07:35	dom96|w	It would be awesome for Nimble to download dynamic libs too
10:08:25	FromGitter	<Bennyelg> @dom96 But the signature is unique to the presto, and you should import presto when use connect
10:08:34	FromGitter	<Bennyelg> @dom96 I don't see the problem
10:08:49	FromGitter	<Bennyelg> assuming you connect( with the right signature
10:08:53	dom96|w	Bennyelg: hrm? I'm fine with your code. Varriount isn't.
10:09:23	FromGitter	<Grabli66> Is there plugin system for nim? I want to load from nim process a module with some interface and use it like plugin.
10:09:28	FromGitter	<Bennyelg> Oh,  sorry
10:11:35	FromGitter	<Bennyelg> @dom96 I read your comment on git about the 2 spaces convention. on my opinion is not enough when you read code 4 make a lot of different
10:11:52	*	sdw joined #nim
10:12:29	FromGitter	<Bennyelg> but it's just me I assuming it's an old habit
10:16:01	*	mahsav quit (Ping timeout: 248 seconds)
10:16:43	*	mahsav joined #nim
10:22:21	*	Yardanico joined #nim
10:23:10	*	spica quit (Ping timeout: 264 seconds)
10:24:46	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
10:26:46	*	ShalokShalom_ joined #nim
10:27:40	couven92	Araq, I tried as you suggested to do some GC_rec calls ( https://github.com/nim-lang/Nim/compare/devel...couven92:issue-6229 ) but that didn't help... -.-
10:29:39	*	ShalokShalom quit (Ping timeout: 252 seconds)
10:29:50	*	endragor quit (Remote host closed the connection)
10:34:07	PMunch	BearSSL looks really interesting dom96|w
10:34:17	Araq	couven92: oh ok, hmmm
10:35:28	*	spica joined #nim
10:36:53	FromGitter	<Grabli66> It would be great if someone implements bearssl :)
10:37:17	PMunch	*hint hint*
10:37:52	PMunch	Try to run c2nim on it and see what happens
10:38:34	*	elrood joined #nim
10:38:42	FromGitter	<Grabli66> I am too old for that :)
10:41:16	*	mahsav2 joined #nim
10:41:41	*	mahsav quit (Ping timeout: 240 seconds)
10:45:39	*	Yardanico quit (Remote host closed the connection)
10:46:44	FromGitter	<mratsim> > @mratsim Your Arraymancer library is really neat! ⏎ ⏎ Thanks ! I’m almost done wrapping CUDA C so we can use custom GPU kernel in Arraymancer =)
10:46:54	FromGitter	<mratsim> @Varriount *
10:49:34	*	Ven`` joined #nim
10:49:47	*	dddddd joined #nim
10:50:01	*	vendethiel- quit (Ping timeout: 240 seconds)
10:50:16	*	endragor joined #nim
10:53:14	*	arnetheduck joined #nim
10:55:26	*	endragor quit (Ping timeout: 240 seconds)
11:02:08	yglukhov	Bennyelg: +1, also prefer 4 spaces.
11:03:44	PMunch	Grabli66, saw your question about plugins. Are you talking about dynamic libraries? Because that is available in Nim
11:04:18	Arrrr	2 spaces masterrace
11:04:27	*	dom96|w quit (Quit: My Mac has gone to sleep. ZZZzzz…)
11:04:30	FromGitter	<Grabli66> PMunch, yes. Compile to dll, so, dynlib?
11:05:31	PMunch	https://rosettacode.org/wiki/Call_a_function_in_a_shared_library#Nim
11:05:52	PMunch	That's a pretty bare-bones example of DLLs in Nim
11:05:58	PMunch	It's pretty easy to work with
11:06:38	spica	guys, was the / (string, string) removed from the lib?
11:06:47	Araq	no
11:06:51	Araq	we all love it.
11:06:55	FromGitter	<Grabli66> PMunch, great
11:07:03	Araq	it's in ospaths now
11:07:19	spica	oh
11:07:21	spica	okay
11:07:23	Araq	can play.nim get c2nim support? that would be so cool
11:07:31	PMunch	Grabli66, https://forum.nim-lang.org/t/2051 that's an example showing dynamic loading
11:07:57	Araq	spica: os.nim forwards to ospaths.nim though, so everything should work as before
11:08:02	PMunch	So loading plugins and getting their procedures dynamically
11:08:07	Araq	did we break something?
11:08:49	spica	Araq, I was just reading through dom's book, and my code didn't compile. He only imported "os", so I thought you guys removed the operator
11:09:06	spica	so you broke his book :))
11:09:22	FromGitter	<Grabli66> PMunch, thanks, it's what i want
11:10:26	Araq	huh? as I said, it should still work
11:10:40	Araq	can we get a proper bug report for this please?
11:11:09	spica	you just said you've moved the proc into another module
11:11:24	spica	should it work if just os is imported?
11:11:30	Araq	spica: yes.
11:11:39	Araq	os does 'export ospaths'
11:11:55	spica	I am a dummy
11:12:10	spica	my bad. I had this: "from os import nil" :)
11:12:16	spica	so it's ok, sry
11:12:39	FromGitter	<BigEpsilon> @Araq , talking about c2nim: https://github.com/nim-lang/c2nim/pull/97
11:12:41	FromGitter	<ephja> are enum values limited to <int.high?
11:12:58	Araq	ephja: I think so
11:14:13	FromGitter	<ephja> `int.high` raises OverflowError in semEnum. I'm gonna have a look
11:28:26	*	claudiuinberlin joined #nim
11:29:10	*	claudiuinberlin quit (Client Quit)
11:29:22	*	Snircle joined #nim
11:31:00	*	dom96|w joined #nim
11:42:05	*	Sentreen quit (Ping timeout: 240 seconds)
11:43:40	*	endragor joined #nim
11:43:43	*	claudiuinberlin joined #nim
11:48:41	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
11:50:27	*	Yardanico joined #nim
11:50:39	*	vendethiel- joined #nim
11:51:03	*	ShalokShalom_ quit (Ping timeout: 252 seconds)
11:51:27	couven92	Araq, I just realized: vcc does not seem to care about `vcc.options.debug`, what do I need to do to trigger the use of that configuration? I tried `nim c koch` and `nim c -d:debug koch`
11:51:53	*	ShalokShalom joined #nim
11:52:08	*	Ven`` quit (Ping timeout: 240 seconds)
11:55:02	*	Sentreen joined #nim
11:57:37	Araq	couven92: huh, good question, --debugger:native ?
12:03:28	*	MyMind quit (Ping timeout: 240 seconds)
12:03:29	*	Pisuke joined #nim
12:05:56	*	claudiuinberlin joined #nim
12:06:56	*	sz0 joined #nim
12:09:36	spica	guys, why does this fail? https://pastebin.com/HbeLkH7q unicode issues
12:10:26	FromGitter	<Grabli66> Is it possible to fix that https://play.nim-lang.org/?gist=b63a811237ba256c7821c174ed456553 ?
12:10:27	Araq	spica: are you on windows?
12:10:30	spica	i am
12:10:37	spica	dayum
12:10:43	Araq	run start.bat that came with Nim
12:10:47	spica	ok
12:10:55	Araq	and retry your program
12:11:10	Araq	in that terminal/shell/console
12:12:23	Araq	Grabli66: that's impossible, you need to duplicate the valueType field
12:12:35	FromGitter	<Grabli66> ok
12:13:01	Araq	you also might want to use JSonNode instead
12:14:41	PMunch	I think he's just doing it for fun Araq
12:14:52	spica	Araq, that works. Is it just a displaying issue?
12:14:57	FromGitter	<ephja> does anyone actually use 'items' for enum types to iterate over enums with holes?
12:15:10	Araq	spica: yes. it's a Windows bug really.
12:15:35	FromGitter	<Grabli66> yglukhov, Your variant type is slow? :)
12:15:46	spica	yep, it works fine withot start.bat in mingw terminal
12:16:33	Araq	spica: look at start.bat's code to see the workaround
12:16:49	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
12:17:58	spica	small inconsistency issue in the api: unicode.runSubStr vs system.substr
12:19:18	Yardanico	well it's runeSubStr
12:19:31	*	claudiuinberlin joined #nim
12:19:42	FromGitter	<ephja> should 'items' be left as it is and then have an iterator that accounts for holes?
12:20:25	spica	Yardanico, it is. but the Str is title cased, in system.substr it's not
12:20:36	Yardanico	spica, it doesn't matter
12:20:53	Yardanico	you can call it like "runesubstr" or "runeSubStr"
12:20:55	spica	Well, it doesn't but it's a bit inconsistent
12:21:28	Yardanico	echo "asdasd".runesubstr(5, 3)
12:21:31	PMunch	ephja, what would be the purpose of an items iterator that didn't account for holes? When would you want to use that instead of one that accounted for them?
12:21:32	Yardanico	if you like it
12:21:36	*	dom96|w quit (Quit: My Mac has gone to sleep. ZZZzzz…)
12:22:03	PMunch	runE_suB_strinG
12:22:14	PMunch	Reverse camel snake-case
12:22:21	Yardanico	yeah
12:22:29	spica	lol
12:22:34	Yardanico	runE_suB_stR
12:22:45	spica	that stll doesn't change the facts
12:22:49	Yardanico	why?
12:22:55	Yardanico	you can call it like you want
12:23:08	spica	lol
12:23:09	Yardanico	it's not a big deal in nim
12:23:11	spica	I do know that
12:23:30	PMunch	That should be implemented as well, shorthand insensitivity :P proc sub{str|srting}
12:23:35	spica	but the api is inconsistent
12:23:41	spica	whatever
12:23:50	spica	not a big deal, just saying
12:23:57	FromGitter	<ephja> PMunch: I dunno if it was done for reasons of simplicity or for some other reason
12:24:02	spica	look at php, you don;t want to end up like that
12:24:09	Yardanico	like what?
12:24:10	spica	absolute rubbish consistency
12:24:13	PMunch	I agree with spica, even though it doesn't matter in Nim the core modules should be consistent with the style guide
12:24:36	Yardanico	well then you can say "Nim compiler should be consistent with the style guide" :P
12:24:56	PMunch	ephja, ah I get you. Do a git blame on it and see who did it :P
12:25:06	FromGitter	<ephja> though it's not that difficult to account for holes in the compiler, but `e.low .. e.high` is very simple
12:25:41	*	TjYoco joined #nim
12:32:00	Araq	the Nim compiler predates the style guide
12:32:31	Yardanico	yeah, I know
12:32:39	Yardanico	but maybe in the future
12:37:00	*	endragor quit (Remote host closed the connection)
12:37:08	*	ShalokShalom quit (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.)
12:38:40	*	endragor joined #nim
12:38:55	*	ShalokShalom joined #nim
12:42:13	couven92	Araq, sry, suddenly had to attend a meeting... Uhm, yes: `--debugger:native` did work...
12:42:59	*	ShalokShalom quit (Remote host closed the connection)
12:43:21	*	ShalokShalom joined #nim
12:43:34	couven92	Is there a way to set vcc.options for the usual case without using .always (since that also would affect `--opt:size` and `--opt:speed`?
12:45:50	FromGitter	<zacharycarter> subsetpark: nice!
12:49:21	*	kunev quit (Ping timeout: 252 seconds)
12:50:06	*	relax joined #nim
12:50:59	*	kunev joined #nim
12:51:29	*	dom96|w joined #nim
12:51:39	spica	how does one convert a string into an array or seq or hashset? Is there any api for that?
12:52:34	PMunch	Turning "Hello, world" into ["H","e","l","l",...]?
12:52:54	spica	yes
12:53:51	PMunch	First init the seq to str.len then just iterate and set the fields
12:54:04	PMunch	Not sure if there is any code for that anywhere
12:54:10	PMunch	But it would be like 5 lines of code
12:54:17	*	pilne joined #nim
12:55:00	spica	Yeah, I was just curious if there's a proc or something for this task.
12:55:44	spica	so if you want to sort the chars of a string, make a seq first via loops, then sort. In python we can sort strings, so that's what I was after
12:56:11	Araq	you can sort strings in Nim
12:56:18	Araq	no idea why you think you can't
12:57:15	spica	how
12:57:26	*	relax quit (Ping timeout: 240 seconds)
12:57:28	PMunch	https://forum.nim-lang.org/t/3075
12:58:23	*	renarcii joined #nim
12:58:55	spica	Dayum. I was using sort(cmp[string]), that's why it failed
13:02:39	*	renarc quit (Remote host closed the connection)
13:02:50	*	renarcii is now known as renarc
13:03:18	*	Ven`` joined #nim
13:04:13	*	Sembei joined #nim
13:04:47	*	renarc quit (Quit: Leaving)
13:05:01	*	vendethiel- quit (Ping timeout: 255 seconds)
13:05:12	FromGitter	<zacharycarter> I think you can just do @"Hello World" spica
13:05:13	*	renarc joined #nim
13:05:40	Yardanico	yeah
13:05:51	*	Pisuke quit (Ping timeout: 252 seconds)
13:06:11	spica	it wokrs
13:08:05	yglukhov	Grabli66: its superfast if wrapped type size is <= sizeof(pointer). otherwise, it allocates a copy, and still remains super fast ;)
13:08:37	FromGitter	<Grabli66> yglukhov, good :)
13:09:00	yglukhov	by "superfast" i mean it is as fast as it can be given its semantics =)
13:10:56	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
13:10:59	FromGitter	<Grabli66> Ok. I'll try to use it :)
13:13:52	yglukhov	Grabli66: it is worth mentioning though that such functionality is rarely needed. in most cases you can get away without losing compile time type tracking, or use nim obj variants. so i would suggest to revisit your program's concept before using variants.
13:13:56	*	onionhammer joined #nim
13:14:16	*	vlad1777d joined #nim
13:14:30	FromGitter	<Grabli66> Maybe later
13:15:34	*	renarc quit (Remote host closed the connection)
13:23:06	subsetpark	Araq: here's something interesting - the same code fails in the VM
13:23:12	*	claudiuinberlin joined #nim
13:23:37	subsetpark	https://www.irccloud.com/pastebin/jGgQombE/
13:29:57	spica	works fine on my system. nim32 on windows
13:30:03	spica	with gcc
13:30:24	spica	but what kind of shell is that?
13:30:54	TjYoco	works on mine too
13:33:08	Yardanico	guys, wrap this code in "static" block
13:33:10	Yardanico	and try again :)
13:33:30	Yardanico	VM - compile-time VM for executing Nim macros and other stuff which is computed at compile-time
13:34:06	FromGitter	<krux02> spice: try calling "nim secret" on the command line
13:34:26	FromGitter	<krux02> on linux I recommend to use "rlwrap nim secret"
13:35:22	spica	yeah, sorting fails that way
13:36:40	FromGitter	<krux02> subsetpark: please create a bug report
13:39:30	spica	lol
13:40:05	spica	:((
13:40:22	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
13:40:40	*	relax joined #nim
13:41:29	spica	>>> var c=[1,2,3,4,-1]
13:41:29	spica	>>> sort(c, cmp)
13:41:37	spica	try that one guys
13:41:40	spica	in nim secret
13:41:47	spica	Error: unhandled exception: value out of range
13:43:28	TjYoco	where are you accessing the array
13:43:48	*	claudiuinberlin joined #nim
13:44:15	spica	what do you mean?
13:44:50	spica	sortin fails, not arr access
13:45:54	FromGitter	<krux02> yes the nim vm was my disillision to nim as well
13:46:22	TjYoco	It worked for me in nim secret
13:46:51	spica	what nim ver?
13:46:57	TjYoco	0.17.2
13:47:01	spica	same
13:47:06	spica	windows ?
13:47:10	TjYoco	yeah
13:47:17	spica	nim 32 or 64?
13:47:55	TjYoco	64
13:48:16	spica	fails on 64 and 32...
13:48:46	spica	https://pastebin.com/QdehtiPz
13:51:20	Yardanico	arch doesn't matter in VM
13:51:25	Yardanico	AFAIK
13:51:43	Yardanico	just do a bugreport already :)
13:52:51	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
13:54:35	*	vlad1777d quit (Ping timeout: 240 seconds)
13:54:56	FromGitter	<Grabli66> Yardanico, can nimble watch changes in files and build?
13:55:11	Yardanico	what do you mean?
13:55:13	Yardanico	hot reloading?
13:55:21	FromGitter	<Grabli66> yes
13:55:37	FromGitter	<Grabli66> Build on changes in source files
13:56:10	Yardanico	well I mean you want your app to be stopped and started again?
13:56:40	Yardanico	if you want to do this without restarting the program itself (with a script or by hand) - your only way is dynamic libraries
13:57:08	Yardanico	but no, I think nimble doesn't have a feature like that
13:58:22	*	claudiuinberlin joined #nim
13:58:22	FromGitter	<Grabli66> Vscode plugin can do that. But it not use nimble, just nim compiler
13:58:50	Yardanico	well can you explain more precisely what do you want to do?
13:59:23	*	PMunch quit (Quit: Leaving)
14:00:19	TjYoco	Grabli66 do you mean you're making a nimble package and importing it elsewhere and want to know if it detects source code changes automatically?
14:01:12	FromGitter	<Grabli66> I want "nimble" looked for changes in the project and automatically began to build as if they were
14:01:32	FromGitter	<Grabli66> I use nimble for project managment
14:01:39	Yardanico	well you would make a script for that
14:01:47	Yardanico	ah
14:01:52	FromGitter	<Grabli66> And build my project like "nimble build"
14:01:54	Yardanico	or simple nim program (if you're on linux)
14:02:11	FromGitter	<Grabli66> And i want "nimble build watch" :)
14:02:27	Yardanico	well there's no cross-platform way to do that in Nim currently
14:02:37	Yardanico	and I don't think it would be a good feature for a package manager :)
14:02:53	FromGitter	<Grabli66> But it also a project manager
14:02:58	Yardanico	yeah
14:03:00	Yardanico	manager
14:03:00	FromGitter	<Grabli66> Build system, i mean
14:03:14	Araq	yup, I think it would be a cool nimble feature
14:03:38	Yardanico	Araq, currently we don't have cross-platform way to watch file changes, do we?
14:03:57	Araq	we have fsmonitor
14:04:02	Araq	only supports linux
14:04:14	Araq	apparently somebody has a Win implementation too
14:04:24	Araq	check the forum
14:04:40	Yardanico	ah, https://gist.github.com/mashingan/4a31d3a630e22cf9274eebf475ee60e6
14:05:05	Yardanico	winim is this - https://github.com/khchen/winim
14:05:09	*	dom96|w quit (Quit: My Mac has gone to sleep. ZZZzzz…)
14:07:07	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
14:08:44	*	claudiuinberlin joined #nim
14:10:33	Araq	ok, I see you volunteered.
14:10:37	Araq	nice
14:11:01	Yardanico	Araq, what ? :)
14:11:09	FromGitter	<Grabli66> Cool. We'll be waiting for results :)
14:11:21	Araq	give fsmonitor a Windows implementation.
14:11:29	Araq	and then an OSX one.
14:11:53	Araq	and then a (Free)[BL]SD one
14:15:10	*	ShalokShalom quit (Remote host closed the connection)
14:15:21	*	endragor quit (Remote host closed the connection)
14:16:08	*	sz0 quit (Quit: Connection closed for inactivity)
14:16:42	FromGitter	<ephja> `errSetTooBig: "set is too large",` big or large? ;)
14:16:45	*	endragor joined #nim
14:17:14	Araq	sometimes it shows I'm no native speaker :P
14:17:20	Araq	PRs welcome
14:17:34	*	dom96|w joined #nim
14:18:13	FromGitter	<ephja> I'm not so sure myself, but "large" might be more formal
14:19:51	Araq	"set is too <insert dick joke here>"
14:19:59	Araq	^^ problem solved
14:20:23	*	endragor quit (Remote host closed the connection)
14:20:30	*	endragor joined #nim
14:20:55	*	endragor quit (Remote host closed the connection)
14:21:19	TjYoco	"Overflow Error: Hint: Try uinch8 instead of uinch4"
14:21:50	Araq	"your mother's set ..."
14:22:52	FromGitter	<ephja> roughly what is needed in order to assign a run time variable to the value that a compile time variable holds at the very end of the compile time scope?
14:23:20	Yardanico	well I create new variable in a proc to do this
14:23:29	*	dom96|w quit (Quit: My Mac has gone to sleep. ZZZzzz…)
14:23:29	Yardanico	e.g. in a proc: "var myvar = SomeConst"
14:23:34	*	endragor joined #nim
14:24:06	FromGitter	<k0pernicus> Hello!
14:24:16	Yardanico	hi
14:24:17	Araq	"Uncountable set detected, ask Cantor for help"
14:24:22	FromGitter	<k0pernicus> Is anybody here has trouble with example 4.26 from Nim in Action?
14:24:33	FromGitter	<krux02> ¡olleH
14:24:37	FromGitter	<k0pernicus> The code is ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=59ba9125210ac269202edb38]
14:24:47	FromGitter	<ephja> Yardanico: yes but sometimes it's not good enough
14:24:49	*	dom96|w joined #nim
14:25:01	FromGitter	<k0pernicus> I compiled the code with `nim c -d:ssl`
14:25:24	FromGitter	<Yardanico> @k0pernicus works fine for me
14:25:25	FromGitter	<Yardanico> with -d:ssl
14:25:38	FromGitter	<k0pernicus> But I got this error running it: `Error: unhandled exception: SSL support is not available. Cannot connect over SSL.`
14:25:56	FromGitter	<Yardanico> well maybe you've compiled it to another directory? :)
14:26:25	FromGitter	<k0pernicus> Oh well - sorry…
14:26:30	FromGitter	<k0pernicus> The error is: `Error: unhandled exception: error:14077410:SSL routines:SSL23_GET_SERVER_HELLO:sslv3 alert handshake failure`
14:26:43	FromGitter	<k0pernicus> Different from the previous one :-]
14:26:45	FromGitter	<Yardanico> oh
14:26:54	FromGitter	<Yardanico> something bad with your openssl?
14:27:03	FromGitter	<k0pernicus> Mmmmh...
14:27:33	FromGitter	<k0pernicus> I will update `brew`… just in case… but I don’t think so
14:27:38	FromGitter	<Yardanico> https://bugs.launchpad.net/ubuntu/+source/openssl/+bug/861137
14:27:42	FromGitter	<k0pernicus> `Warning: openssl 1.0.2l is already installed`
14:28:15	*	endragor quit (Ping timeout: 248 seconds)
14:28:26	FromGitter	<ephja> @Yardanico one example is implementing opengl.loadExtensions without compiler support. lexicographic order shouldn't matter
14:29:00	FromGitter	<Yardanico> @k0pernicus are you sure nim uses this new ssl?
14:29:04	FromGitter	<ephja> well, I should figure it out myself eventually
14:29:23	FromGitter	<k0pernicus> @Yardanico It’s in the book
14:29:37	FromGitter	<k0pernicus> That’s why I asked if anybody has got troubles with the book’s code
14:29:44	Yardanico	It's the error in OpenSSL
14:29:47	Yardanico	what's why I'm asking
14:29:59	Yardanico	and I mean openssl by "ssl"
14:30:16	FromGitter	<k0pernicus> Oh ok, sorry
14:31:29	Yardanico	well I don't really know, you can google this error
14:31:53	FromGitter	<k0pernicus> Yep, I tried
14:31:57	dom96|w	Mac OS X's older openssl doesn't support cloudflare's SSL IIRC
14:32:48	Yardanico	brew update && brew install openssl && brew link --force openssl && openssl version -a
14:32:54	Yardanico	but this can break your system :)
14:33:08	Yardanico	because every program will use newer OpenSSL
14:33:09	FromGitter	<k0pernicus> @Yardanico openssl from brew is already installed ;-0
14:33:31	Yardanico	did you do "brew link --force openssl" ?
14:33:35	FromGitter	<k0pernicus> Yep
14:33:46	Yardanico	what's the output of "openssl version -a" ?
14:33:54	*	relax quit (Ping timeout: 246 seconds)
14:34:00	Yardanico	and if you did this right now, try to relogin or reboot
14:35:13	*	Calinou quit (Ping timeout: 248 seconds)
14:36:07	FromGitter	<k0pernicus> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=59ba93d7bc4647297432edd0]
14:36:08	FromGitter	<edubart> How to enable gcc warnings when compiling with nim? (disable the -w argument that nim passes to gcc to ignore all warnings)
14:36:45	Yardanico	edit nim.cfg ?
14:37:05	Yardanico	yourNimCompilerDir/config/nim.cfg
14:37:15	Yardanico	maybe there's a better way
14:37:22	FromGitter	<edubart> I've tried adding -passC:-Wall but gcc still ignores all warnings
14:37:36	Yardanico	you should try to edit nim config as I said
14:37:47	Yardanico	but why do you want gcc warnings? just curious :)
14:38:15	TjYoco	Im trying to fix a dependency issue by using Generics, how should I fix this 'Field' is not a concrete type? https://gist.github.com/Tyler-Yocolano/ebe83c4b1051ec31d15b8cbb57ac3ad3
14:38:53	FromGitter	<edubart> why I need is explained in my comment on https://github.com/nim-lang/Nim/issues/6324
14:38:59	Yardanico	you can't have a seq of different types
14:39:09	Yardanico	generics are compile-time thing
14:39:12	FromGitter	<edubart> what line I have to add/edit in nim.cfg?
14:39:54	Yardanico	try to edit line 128
14:39:58	Yardanico	and remove "-w" here
14:40:09	Yardanico	(if you're on linux)
14:40:09	TjYoco	Yardanico alright, thought they were all technically Field objects so that wouldn't matter
14:40:19	Yardanico	TjYoco, no, that would matter
14:40:26	Yardanico	https://forum.nim-lang.org/t/3167
14:41:06	Yardanico	generics are not a runtime thing
14:42:28	TjYoco	ohhh gotcha
14:43:36	FromGitter	<k0pernicus> @Yardanico Does the code run well on your computer?
14:43:39	FromGitter	<k0pernicus> No exception?
14:43:59	Yardanico	yes. I'm on linux (arch linux)
14:44:12	FromGitter	<k0pernicus> And what is your version of OpenSSL ?
14:44:41	Yardanico	OpenSSL 1.1.0f 25 May 2017 :)
14:45:05	FromGitter	<k0pernicus> Mmmmh ok
14:45:06	Yardanico	well it's arch, here's everything has latest versions
14:45:24	Yardanico	dom96|w himself uses mac IIRC
14:45:44	*	relax joined #nim
14:45:45	FromGitter	<k0pernicus> @dom96 Are you using MacOS? Can you paste your OpenSSL please?
14:45:50	*	sz0 joined #nim
14:45:50	dom96|w	what URL are you trying to request?
14:45:58	dom96|w	I'm not on my MBP right now
14:45:59	Yardanico	dom96|w, he's trying to request nim-lang.org
14:46:07	FromGitter	<k0pernicus> Yep
14:51:55	FromGitter	<Grabli66> How to create an interface type in nim?
14:52:32	Yardanico	there's no "interface" in Nim
14:52:50	FromGitter	<Grabli66> I know :)
14:53:02	Yardanico	what are you trying to do?
14:53:04	FromGitter	<Grabli66> But. I want something like interface
14:53:26	Yardanico	well there should be a reason for that
14:53:40	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
14:53:54	*	nsf quit (Quit: WeeChat 1.9)
14:54:05	FromGitter	<Grabli66> Multiple inheritance
14:54:27	Yardanico	well there should be a reason for that :D
14:54:37	FromGitter	<Grabli66> :)
14:54:50	FromGitter	<edubart> I've been absent from nim related stuff for about one year, how is the progress towards 1.0 since then? closer or farther?
14:54:53	Yardanico	well, really, you can post your code (which you want to do)
14:55:01	Yardanico	edubart: closer indeed
14:55:08	FromGitter	<Grabli66> I want object that has field from other two objects
14:55:15	FromGitter	<Grabli66> fields
14:55:51	Yardanico	ehh, you don't get me :D I mean why do you want to have an object like this? can you post a code which you're trying to make work?
14:56:32	TjYoco	I thought I read somewhere about composition instead of inheritance, something like the new object would have fields of the other ones instead of being a child of them... is that anywhere close to correct?
14:56:32	Yardanico	also this: http://openmymind.net/Interfaces-In-Nim/
14:57:01	*	arnetheduck quit (Ping timeout: 240 seconds)
14:57:17	*	claudiuinberlin joined #nim
14:57:47	FromGitter	<edubart> but I saw that the github issues is larger (I remember that were ~900 a year ago) and some looks scary, this gives me the feel that is farther
14:57:56	FromGitter	<Grabli66> Yardanico, Good trick :)
14:57:57	*	pilne quit (Quit: Quitting!)
14:58:15	Yardanico	edubart: well because issues are closed and new ones are opened? :)
14:58:29	FromGitter	<Grabli66> Yardanico, Maybe there is a mixins?
14:58:35	FromGitter	<edubart> I am talking about open ones
14:58:35	*	Ven`` quit (Ping timeout: 240 seconds)
14:58:52	Yardanico	look at rust or swift then :P
14:59:11	Yardanico	ah, swift doesn't have issues in github repo
15:00:37	FromGitter	<edubart> rust 2770 much more haha, however 19382 closed ones, they have a better ratio
15:00:50	Yardanico	because nim community is smaller
15:01:31	Yardanico	https://github.com/nim-lang/Nim/issues?utf8=%E2%9C%93&q=is%3Aissue%20is%3Aclosed%20updated%3A%3E2016-09-14
15:01:32	*	Ven`` joined #nim
15:01:45	Yardanico	not the best search, but still
15:01:48	TjYoco	how many issues though are "Q: I can't get this to work; A: You're doing it wrong, here's how"
15:01:56	FromGitter	<k0pernicus> @edubart The first release of Nim will be ok when `high-priority` issues will be fixed
15:02:06	FromGitter	<k0pernicus> (I think…)
15:02:08	Yardanico	well Araq wanted to release 1.0 years ago IIRC
15:02:14	Yardanico	but community stopped him
15:02:20	*	Jesin quit (Quit: Leaving)
15:02:20	FromGitter	<k0pernicus> Why?
15:02:38	FromGitter	<k0pernicus> (11 high-priority issues for the moment I think)
15:02:48	FromGitter	<edubart> we all I hope so, rhoughtly in how many years maybe 1.0 will be out?
15:03:14	Yardanico	no one can ask questions like this, there's no schedule
15:03:22	FromGitter	<k0pernicus> Maybe it’s too late for 2017… 2018?
15:03:32	FromGitter	<zacharycarter> these kind of questions are pointless
15:03:41	FromGitter	<edubart> is nim in feature freeze or it keeps adding new features often?
15:03:51	Araq	I'm still shooting for 2017 fwiw
15:04:00	Araq	every year is Nim v1 year.
15:04:01	FromGitter	<zacharycarter> please look at the release notes if you want to see when features are added @edubart
15:04:09	Yardanico	edubart: it's not feature freeze
15:04:50	Yardanico	nimble became much smarter in this year AFAIK
15:05:11	Yardanico	or for example "The C/C++ code generator has been rewritten to use stable name mangling rules. This means that compile times for edit-compile-run cycles are much reduced."
15:05:13	FromGitter	<edubart> aside from bugs, is thread a roadmap or list of new features that will be added?
15:05:24	FromGitter	<edubart> is there*
15:05:47	Yardanico	well, https://github.com/nim-lang/Nim/blob/devel/todo.txt, but i don't know if it's a good source
15:05:56	Araq	edubart: I'm working on a new one
15:06:05	Yardanico	Araq, yay
15:07:29	Yardanico	but AFAIK not nil is not complete now
15:08:10	Araq	in a nutshell: make 'not nil' the default for 'ref', make 'nil' an invalid state for strings/seqs, make 'case' do pattern matching for objects
15:09:04	*	Jesin joined #nim
15:10:17	FromGitter	<edubart> other thing that I'm afraid a little is that the project itself is much dependent on Araq, I hope Araq keeps rocking and improving nim forever, however what happens in other scenario?
15:10:55	Araq	it's a valid concern, we hired "zahary" who knows pretty much everything too
15:11:29	Araq	making the bus factor 2 for Nim's core. Jehan too knows almost everything afaict but he is not very active either
15:11:41	FromGitter	<edubart> Araq, do you feel more tired of improving and fixing Nim compared with yourself years ago? will you keep rocking as much as you can?
15:12:50	TjYoco	Do you expect him to say No he's giving up soon?
15:12:52	FromGitter	<edubart> hmm, good to know that you are hiring others to help, I hope the active core developers grows
15:13:08	Araq	I'm improving Nim's ecosystem more than the compiler these days (sorry), we got "karax" for browser based frontends (SPAs), we're getting a unique ORM
15:13:09	Yardanico	TjYoco, well I think Araq wouldn't give up :)
15:13:21	Yardanico	at least because there's large code bases written in Nim AFAIK :)
15:14:00	Araq	that means two things:
15:14:11	Araq	- progress is slower than I would like
15:14:29	Araq	- I am having lots of fun with no signs of growing tired :D
15:14:39	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
15:14:53	*	vivus joined #nim
15:15:08	TjYoco	Couldn't someone just fork Nim and upgrade it their own way anyhow?
15:15:21	Yardanico	TjYoco, yes, if he knows Nim compiler internals
15:15:25	Yardanico	nim compiler is MIT
15:15:34	FromGitter	<edubart> improving the ecosystem is a good strategy, a good ecosystem will attract more people to help out, thus improving more the ecosystem in a recursive manner
15:15:42	Yardanico	but I don't know any reason to fork Nim
15:15:54	Araq	I know most of you don't agree
15:16:19	Araq	but the compiler is really not hard, complex algorithms have been avoided for the most parts
15:16:53	FromGitter	<ephja> I'm probably just slow
15:17:11	*	haha_ joined #nim
15:17:19	TjYoco	Yardanico, I agree from what I've seen in the past month or so. Just saying if there were major concerns in the community or whatever then anyone can just make OtherNim
15:17:41	Araq	once you get over my eccentric style of writing things ("hi, I am a single char variable. guess what it means") it's not all that hard
15:18:46	Araq	there is the tendency though that hard bugs stay around for long. that's quite common in open source land.+
15:20:25	Yardanico	well there's 160 issues with "feature"label
15:21:39	*	claudiuinberlin joined #nim
15:23:05	Yardanico	oh, we have 80 open PRs
15:26:50	*	jsgrant joined #nim
15:30:13	*	vlad1777d joined #nim
15:31:51	FromGitter	<krux02> Sorry I knew I am annoying, but it is annoying that by default the debug build has no debug symbols
15:33:04	Araq	krux02: edit your nim.cfg and add --debugger:native
15:33:22	FromGitter	<krux02> nim.cfg is confusing
15:33:26	FromGitter	<krux02> and I did
15:34:13	FromGitter	<krux02> the result was the output "using config file ..." broke my code
15:34:30	couven92	@krux02 on Windows?
15:34:39	couven92	using vcc?
15:35:57	FromGitter	<krux02> https://github.com/nim-lang/nimble/issues/397
15:36:13	FromGitter	<krux02> couven92 I am on Linux only
15:36:34	couven92	okay... because i just found out some things about vcc :P
15:37:27	FromGitter	<krux02> Araq: so just editing nim.cfg actually breaks things
15:38:17	Araq	seems a nimble bug to me
15:38:35	FromGitter	<krux02> yea it is in the nimble bug tracker alread
15:39:56	Araq	can't we fix the real issue here?
15:40:14	FromGitter	<krux02> The real issue to me here is that the debug build has no debug symbols
15:40:19	Araq	this staticRead"nimble path" hack needs to die
15:40:31	FromGitter	<krux02> ah yea that one
15:41:38	FromGitter	<krux02> the compile command should accept arguments that are relative to the file it is in
15:41:50	FromGitter	<krux02> I thit that was the reason I did that hack in the first place
15:42:12	FromGitter	<krux02> I *think*
15:45:35	FromGitter	<Grabli66> What is wrong? https://play.nim-lang.org/?gist=7cdb8206df4b0f94d59ca883a336672a
15:45:44	FromGitter	<krux02> Araq: I think a pragma ``{.addCIncludePath: "some/path/".}`` with a relative path the the current file would fix this.
15:46:56	Araq	grabli66: use 'of' instead of 'is' here.
15:47:26	Araq	krux02: we have no way to ship C code with Nim code
15:47:46	Araq	I mean no official way.
15:47:56	couven92	we don't have custom GC finalizers (destructors) yes, do we?
15:47:59	Araq	just a bunch of hacks people came up with
15:48:24	Yardanico	couven92, I think we do
15:48:42	FromGitter	<krux02> Araq: well no official way, but this way here works pretty well: https://github.com/krux02/nimAntTweakBar/blob/master/AntTweakBar.nimble
15:48:47	*	haha_ quit (Quit: haha_)
15:48:54	Yardanico	couven92, https://github.com/nim-lang/Nim/blob/master/lib/system.nim#L220
15:49:19	couven92	oh! nice!
15:51:01	FromGitter	<krux02> couven92: I would be careful with the finalizers, test if they actually can do what you want
15:51:33	FromGitter	<krux02> I remember that I tried destructors in Nim and they ware pretty far away from being reliable
15:51:33	*	haha_ joined #nim
15:51:34	Araq	btw I invented a new GC algorithm
15:51:41	Araq	how can I get the fame?
15:51:51	couven92	Araq, WHAAAAT? :O
15:52:01	FromGitter	<krux02> whoohoo, party
15:52:11	Araq	it's just a variant of existing algorithms really
15:52:14	Araq	but still.
15:52:18	Yardanico	Araq, ahh
15:52:22	Yardanico	I remember this link
15:52:29	FromGitter	<krux02> well isn't a variant of existing algorithm what we all do
15:52:38	Araq	yeah indeed.
15:52:46	Yardanico	you've said that you was experimenting with a new GC
15:55:12	Araq	I read about something similar in a children's book once.
15:56:33	Yardanico	it seems I found it :P https://people.cs.umass.edu/~emery/pubs/berger-oopsla2002.pdf
15:56:38	FromGitter	<k0pernicus> In “Garbage Collection for kids” ?
15:56:46	FromGitter	<krux02> random information of the day: in gdb the command ``rbreak`` matches with regular expressions on function names and allows setting breakpoints in Nim functions pretty neatly
15:56:50	Yardanico	(found it again)
15:57:26	FromGitter	<krux02> and ``tui enable`` in gdb shows the sourcecode interactively
15:57:34	Yardanico	Araq, so new GC will arrive to Nim? or it's just an algorithm, not an implementation?
15:57:52	*	dom96|w quit (Quit: My Mac has gone to sleep. ZZZzzz…)
15:58:03	FromGitter	<krux02> I think it is already in Nim, we just never noticed it :D
15:58:41	FromGitter	<Grabli66> Araq, why there is "is" and "of" to check types?
15:58:42	FromGitter	<k0pernicus> @Araq Did you benchmarked it yet?
15:59:23	FromGitter	<krux02> @Grabli66 well I guess it has to do with inheritance and type aliases
15:59:47	FromGitter	<krux02> is just checks for type aliases and of also matches with super types
16:00:09	*	couven92 quit (Read error: Connection reset by peer)
16:00:16	FromGitter	<krux02> Araq: I broke the GC
16:00:28	FromGitter	<krux02> I made a change and now my test fails the GC initialization
16:04:30	*	solitudesf quit (Remote host closed the connection)
16:07:30	*	Trustable joined #nim
16:07:34	*	dom96|w joined #nim
16:08:43	*	miran joined #nim
16:08:47	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
16:09:56	*	claudiuinberlin joined #nim
16:11:56	*	couven92 joined #nim
16:12:37	*	spica left #nim ("Leaving")
16:13:17	*	vlad1777d quit (Remote host closed the connection)
16:15:08	*	vlad1777d joined #nim
16:15:42	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
16:21:01	Araq	will look at it later, krux02
16:21:02	Araq	bbl
16:23:52	FromGitter	<ephja> hmm ⏎ ⏎ ```type E = enum a = <int.high ⏎ echo({a})``` [https://gitter.im/nim-lang/Nim?at=59baad18c101bc4e3ac882a7]
16:30:58	*	relax quit (Ping timeout: 264 seconds)
16:34:50	Arrrr	try `type E = enum c = 0, a = <int.high`
16:36:07	Arrrr	mm what i'd like to have is a len(enum) proc
16:36:49	*	relax joined #nim
16:40:51	*	jolik69 joined #nim
16:41:34	*	haha_ quit (Quit: haha_)
16:44:17	*	Yardanico_ joined #nim
16:44:27	*	Yardanico_ quit (Client Quit)
16:44:43	*	Yardanico_ joined #nim
16:46:08	*	sz0 quit (Quit: Connection closed for inactivity)
16:46:15	*	salewski joined #nim
16:46:22	*	haha_ joined #nim
16:46:35	*	Yardanico quit (Ping timeout: 240 seconds)
16:48:52	*	Yardanico_ quit (Client Quit)
16:49:06	*	Yardanico joined #nim
16:49:12	salewski	gtype.h:#if GLIB_SIZEOF_SIZE_T != GLIB_SIZEOF_LONG || !defined __cplusplus
16:49:29	*	jolik69 quit (Quit: Leaving)
16:49:58	salewski	How do I test for cplusplus backend in Nim code? when defined(cplusplus)?
16:56:03	FromGitter	<k0pernicus> Hey
16:56:11	FromGitter	<k0pernicus> I am experimenting with `concept`
16:56:12	salewski	My guess was not bad, seems to be when defined (cpp). Bye.
16:56:30	FromGitter	<k0pernicus> This is the code I wrote: ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=59bab4beb59d55b823f79373]
16:57:12	Yardanico	ok?
16:57:18	Yardanico	ah, I see
16:57:32	Yardanico	you can't use concept as a concrete type
16:57:34	FromGitter	<k0pernicus> As I implement `<` for the `Nothing` type, it should be satified the `Comparable` concept, right?
16:57:45	Yardanico	yes
16:57:52	Yardanico	but your myMax is wrong
16:57:55	Yardanico	you can't return Comparable
16:57:58	FromGitter	<k0pernicus> Ow
16:58:26	FromGitter	<Yardanico> use something like this
16:58:29	FromGitter	<Yardanico> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=59bab5357b7d98d30df5c81e]
16:59:43	FromGitter	<k0pernicus> Cool, it works!
16:59:46	FromGitter	<k0pernicus> Thanks a lot!
17:00:16	FromGitter	<k0pernicus> So, instead of returning a concept, we return a concrete type that is a “similar type” (or alias) of the concept type, right?
17:00:27	FromGitter	<Yardanico> nonono
17:00:54	FromGitter	<Yardanico> myMax will accept any type which satisfies "Comparable" concept, and myMax will return a result with this type
17:01:10	FromGitter	<k0pernicus> Ow ok, I see
17:01:13	FromGitter	<k0pernicus> Wrong deduction
17:01:15	FromGitter	<k0pernicus> Thanks :-)
17:01:32	FromGitter	<k0pernicus> (Actually, the `myMax` procedure is from the book)
17:02:07	FromGitter	<k0pernicus> (page 254)
17:02:12	*	couven92 quit (Quit: Client disconnecting)
17:02:19	FromGitter	<Yardanico> also you wouldn't have any concepts or generics at runtime, they all resolved at compile-time
17:02:43	FromGitter	<Yardanico> so for each type pair new C (or C++) function is generated
17:02:49	FromGitter	<k0pernicus> Yep
17:03:09	FromGitter	<Yardanico> Here it would be this: ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=59bab64d1081499f1f301e50]
17:03:56	FromGitter	<k0pernicus> (beautiful)
17:04:08	FromGitter	<k0pernicus> Like generics, right?
17:04:12	FromGitter	<k0pernicus> All in compile time
17:04:20	FromGitter	<Yardanico> yes
17:04:28	FromGitter	<Yardanico> also I've used generics in my code
17:04:31	FromGitter	<k0pernicus> (just saw your “generics at runtime” - sorry for that)
17:04:34	FromGitter	<Yardanico> myMax is a generic proc
17:04:40	FromGitter	<k0pernicus> I see
17:07:16	*	Jesin quit (Quit: Leaving)
17:07:44	*	yglukhov quit (Remote host closed the connection)
17:08:40	*	haha_ quit (Quit: haha_)
17:12:20	*	mahmudov joined #nim
17:16:12	*	relax quit (Ping timeout: 260 seconds)
17:18:01	*	Sentreen quit (Ping timeout: 240 seconds)
17:18:12	*	ShalokShalom joined #nim
17:18:35	*	dom96|w quit (Quit: My Mac has gone to sleep. ZZZzzz…)
17:21:41	*	adeohluwa joined #nim
17:23:36	FromGitter	<krux02> args
17:24:01	FromGitter	<krux02> spending enough time with gdb I start to like it
17:24:09	FromGitter	<krux02> even though I never thought I would ever do that
17:24:30	FromGitter	<krux02> it is all so complicated and needs a lot of knowledge to actually do anything
17:24:33	*	dom96|w joined #nim
17:26:30	FromGitter	<Grabli66> Can i create recursive iterator? Nim says "recursive dependency"
17:26:56	FromGitter	<krux02> well I tried that, too
17:27:33	FromGitter	<krux02> aparently, it doesn't seem so
17:27:40	FromGitter	<krux02> I even triey clusure iterators
17:27:44	FromGitter	<krux02> tried
17:28:06	FromGitter	<Grabli66> Success?
17:28:11	FromGitter	<krux02> nope
17:28:29	FromGitter	<krux02> my solution was to not use an iterator, but to write out the items into a seq
17:28:51	FromGitter	<krux02> for my use case that worked, too, but I would have preferred iterators
17:29:17	*	joby joined #nim
17:29:24	FromGitter	<Grabli66> Ohh. I know what to do
17:32:20	*	Sentreen joined #nim
17:33:17	Araq	seriously guys. iterators/generators vs coroutines
17:33:30	Araq	without a stack you can't have recursion
17:33:57	FromGitter	<krux02> Araq: yes, but aren't clusure iterators with a Stack?
17:34:10	FromGitter	<Grabli66> But we want :)
17:36:22	FromGitter	<krux02> It not that we all implemented iterators with the knowledge that recoursion would not be possible with it.
17:36:34	FromGitter	<krux02> We just try it and then realized that it doesn't work
17:36:40	*	dom96|w quit (Quit: My Mac has gone to sleep. ZZZzzz…)
17:36:45	ehmry	assuming that the stack can grow indefinitely is also not portable
17:38:32	*	haha_ joined #nim
17:39:03	FromGitter	<krux02> ehmry: sure
17:39:19	*	joby quit (Quit: Page closed)
17:40:14	FromGitter	<Grabli66> Ok. I can that understand. But why where is no interfaces? :(
17:44:52	FromGitter	<mratsim> Is there a way to tell the compiler not to optimise a proc? I’m trying this multiply by two on GPU but the resulting C is “optimized” via a CPU function. ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=59bac0147b7d98d30df604a1]
17:45:07	Yardanico	yeah
17:45:21	FromGitter	<Yardanico> ah
17:45:25	FromGitter	<Yardanico> you mean N_INLINE?
17:45:35	FromGitter	<mratsim> yes
17:45:53	FromGitter	<mratsim> I don’t want the *= to be replaced by stareq ….
17:45:56	Yardanico	IDK then
17:46:32	FromGitter	<Bennyelg> how do I push an update of package in nimble
17:46:47	FromGitter	<mratsim> create a new tag on github
17:46:54	Yardanico	make a new tag and update version in .nimble file
17:46:58	FromGitter	<mratsim> and update your nimble
17:47:19	FromGitter	<Bennyelg> so new nimble publish ?
17:47:23	Yardanico	nooo
17:47:30	Yardanico	new version tag on GitHub
17:47:41	Yardanico	make a new commit which updates version in .nimble file
17:47:48	Yardanico	and then create a tag which refers this commit
17:48:25	FromGitter	<Bennyelg> how do I create new tag
17:48:44	FromGitter	<krux02> what do you guys see in this gif https://camo.githubusercontent.com/bbec54f58f9f10beb07a20d5fe0465586d888763/68747470733a2f2f6d656469612e67697068792e636f6d2f6d656469612f6c314a3973363553576a486979614a656f2f67697068792e676966
17:49:00	Yardanico	bad performance? :)
17:49:21	FromGitter	<Bennyelg> Ok, i got the tag
17:49:24	Yardanico	Bennyelg: https://github.com/VKBots/VBot/releases/new
17:49:26	FromGitter	<Bennyelg> and nimble version update
17:49:27	FromGitter	<Bennyelg> what now
17:49:43	*	yglukhov joined #nim
17:49:57	FromGitter	<krux02> bad performance?! why that?
17:49:58	*	couven92 joined #nim
17:50:00	*	haha_ quit (Quit: haha_)
17:51:04	FromGitter	<Bennyelg> :|
17:51:37	FromGitter	<Bennyelg> oh great
17:51:48	FromGitter	<Bennyelg> it's automatically after tag, thanks
17:54:05	*	yglukhov quit (Ping timeout: 240 seconds)
17:54:26	*	haha_ joined #nim
18:01:31	FromGitter	<mratsim> I see "Non-Image content-type returned"
18:03:54	Araq	mratsim: need more help with .emit?
18:05:19	FromGitter	<mratsim> Mmm, that may be a solution yes
18:06:09	*	haha_ quit (Quit: haha_)
18:06:54	*	relax joined #nim
18:08:51	FromGitter	<mratsim> got to go. ⏎ ⏎ I will start with emit before trying something else. ⏎ I was actually thinking of nim-templates: https://github.com/onionhammer/nim-templates ⏎ ... [https://gitter.im/nim-lang/Nim?at=59bac5b3b59d55b823f7f824]
18:10:23	*	Snircle quit (Quit: Textual IRC Client: www.textualapp.com)
18:11:08	*	yglukhov joined #nim
18:16:27	FromGitter	<adamrezich> can anyone tell me what I'm doing wrong such that this whole thing works except for the last test? https://play.nim-lang.org/?gist=b839e751c5b638b3fd339ac87fc9b16a still figuring out macros :/
18:18:59	FromGitter	<krux02> @adamrezich maybe you should use quote in macros
18:19:12	FromGitter	<krux02> it lets you create code blocks with actually using code
18:19:23	FromGitter	<krux02> ``````
18:20:13	Arrrr	do not pass evproc directly to add. Use a variable first
18:20:27	*	Snircle joined #nim
18:20:43	FromGitter	<krux02> @adamrezich well that doesn't solve anything, but it helps to keep things simpler
18:20:48	FromGitter	<adamrezich> @krux02 I didn't know about that, I'll look into it though
18:21:08	FromGitter	<krux02> you can use packticks to insert nodes
18:21:14	TjYoco	adamrezich, this works too a.add(evproc(Tick, echo "hi"))
18:21:43	FromGitter	<adamrezich> @Arrrr the point is to make it as DSL-like as possible, so I'll probably make the actual proc that takes in an evproc into a macro too
18:22:06	FromGitter	<krux02> ```let node: NimNode = ... ⏎ result = quote do: ⏎ foo.bar(`node`)``` [https://gitter.im/nim-lang/Nim?at=59bac8cebac826f054ba5296]
18:23:03	FromGitter	<krux02> to my experience it works best, when you try to be minimal in what your dsl supports
18:23:29	Arrrr	Yep, i know prefer to do everything with templates. Easier to maintain and understand
18:23:33	Arrrr	*now
18:23:33	FromGitter	<krux02> each feature needs to be maintained, documented
18:24:08	FromGitter	<krux02> Arrrr: I agree, when templates allow that, then I totally agree
18:24:19	FromGitter	<krux02> but not everything is possible
18:24:31	Yardanico	well this example will be a bit wordy with quote do
18:24:44	FromGitter	<krux02> but if you try you will realize that a lot is actually possible with templates
18:24:55	FromGitter	<krux02> doesn't matter
18:25:02	FromGitter	<krux02> maybe it does
18:25:35	FromGitter	<krux02> I just don't like super long ``nnkXYP.newTree(...)`` expressions
18:25:47	FromGitter	<krux02> and not everything can be created with quote do
18:26:03	FromGitter	<krux02> quote do always creates nnkStmtList
18:26:09	FromGitter	<krux02> not always what you want
18:27:48	FromGitter	<adamrezich> the whole point of this is to cut down on code duplication and make defining game logic and stuff as simple and DSL-like as possible. I want to make it so users of this game engine can quickly define "event types": a tuple containing the event, and a proc type that takes in an Elem (game object) and said tuple type. this part works fine, but I want to take it a step further and allow users to easily create lambdas
18:27:48	FromGitter	... of that proc type and pass them to the event system. I wouldn't be worrying about this last part if Nim had the ability to use proc types when declaring lambdas like ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=59baca24c101bc4e3ac92324]
18:28:36	FromGitter	<adamrezich> but you can't do that without metaprogramming in nim for reasons I more or less understand (but I still wish that was a feature, though!)
18:29:23	Arrrr	Well is good to learn how to use macros and report bugs.
18:29:49	FromGitter	<adamrezich> yeah this was definitely a good excuse to take a couple hours and learn the macro/AST side of Nim c:
18:29:55	LyndsySimon	Question - is there a Nim library out there yet specifically for developing RESTful APIs?
18:30:26	Yardanico	LyndsySimon, no
18:30:31	Arrrr	you can search it: `nimble search rest`
18:30:33	Yardanico	specifically for restful apis - no
18:33:28	FromGitter	<Araq> @adamrezich isn't that trivially solved with a helper template?
18:34:05	Araq	after(getUpComp[TimestepMgr](elem).evTick, OnTick =
18:34:05	Araq	# body here
18:34:05	Araq	)
18:34:47	LyndsySimon	Yardanico: Interesting. Is that something that would be useful, or is it easy enough that a library isn't needed?
18:35:18	Araq	is still full of boilerplate anyway
18:35:30	FromGitter	<adamrezich> @Araq: i-is it? like I said, I'm still new to nim's metaprogramming faculties. also can you paste that somewhere? gitter botched it
18:36:02	FromGitter	<adamrezich> oh it turned the () syntax into a markdown link lol
18:36:40	Araq	still reads like a C++ "inspired" construct
18:37:31	Araq	LyndsySimon: does "produces JSON from SQL queries" count as "rest" library?
18:38:12	FromGitter	<adamrezich> which part? it's the best system I could come up with for making an event system with user-defined events attached to objects in the scene graph
18:38:35	FromGitter	<adamrezich> though I have been trying to be more nim-like, cutting down on ptrs and using refs where it makes sense haha
18:39:40	Yardanico	can't you just do something like myObject.on(Tick): body ?
18:39:48	Yardanico	and on would be some macro or template
18:42:46	Yardanico	and generate all boilerplate yourself :)
18:43:31	*	haha_ joined #nim
18:59:24	FromGitter	<ephja> @krux02 I sometimes discard the statement list if necessary
19:03:08	*	couven92 quit (Ping timeout: 246 seconds)
19:08:47	FromGitter	<adamrezich> @Yardanico that would require the event system to be even more generic than it is but I think that's what I'm going to move towards
19:08:56	Yardanico	well why? no
19:09:13	Yardanico	you would store some const variables and read them from macros to generate all the stuff you have now
19:09:40	FromGitter	<adamrezich> well also elements don't have events, components do, because i went down that crazy architectural route lol
19:10:15	FromGitter	<adamrezich> this all sounds really complicated but it's actually working really well and i'm enjoying writing it in nim
19:10:23	Araq	yeah don't do that. events are hard to debug. ;-)
19:10:58	FromGitter	<adamrezich> true!
19:11:04	Araq	why? because they turn control flow into data flow.
19:11:34	Araq	all the debugging tools attack control flow.
19:12:26	LyndsySimon	Araq: I'd say that's part of it (i.e., serializing/deserializing from SQL). There are other parts though, like dealing with pagination, enveloping, and error handling. Maybe even auth[n|r], depending on how thin you're trying to keep the library.
19:12:57	Yardanico	jester + ornim ? :P
19:13:01	Yardanico	+ some stuff
19:13:11	*	haha_ quit (Quit: haha_)
19:13:19	Araq	ormin is vaporware but I'm testing it now...
19:13:31	Yardanico	well would it be possible to add async to it?
19:14:10	Araq	oh yeah that's planned too but requires more changes -.-
19:14:32	Yardanico	well you're free to do everything you want to do :)
19:15:36	Araq	LyndsySimon: pagination is just 'order by foo desc; limit 10 offset 20' :P
19:15:55	*	claudiuinberlin joined #nim
19:15:59	Araq	error handling is try: ... except: sendError(getCurrentExceptionMsg)
19:16:09	Araq	auth is hard...
19:17:00	Yardanico	https://github.com/dom96/jester/issues/26
19:17:28	Yardanico	well but you'll still have to implement auth yourself :P
19:18:34	Araq	we need a jsongen library
19:18:55	Araq	sucks to produce a JSON in memory representation only to render it to strings anyway
19:19:20	Yardanico	Araq, also, some question about ornim - will it always parse .sql file, or maybe in some future you would be able to declare your DB tables from Nim?
19:19:46	*	haha_ joined #nim
19:19:57	Yardanico	ah, soorry, ormin, not ornim
19:20:00	Araq	Yardanico: good question, I'm not too happy with the helper exe that parses the SQL
19:20:39	Araq	it's a weird setup that will cause problems for nimble package management etc
19:22:26	*	couven92 joined #nim
19:23:16	Araq	I think yeah, we should have a DSL for table creation too
19:24:06	FromGitter	<ephja> a DSL for DSLs?
19:24:19	Yardanico	lol
19:24:25	Yardanico	hmm, actually SQL is a DSL?
19:27:11	Araq	ephja: SQL is easy to improve upon.
19:27:41	Araq	for example, make 'not null' the default for every field declaration... *cough*
19:28:15	Yardanico	xDDD
19:28:20	Araq	standardize on a "ID primary key auto increment" syntax.
19:28:37	Araq	standardize the 'references table(column)' syntax
19:29:31	FromGitter	<ephja> does standard SQL not cover these aspects or do engines often diverge from it anyway?
19:29:53	Araq	standard SQL is pretty weak and the db engines shit on it
19:30:12	Araq	that means "both".
19:32:49	*	Matthias247 quit (Read error: Connection reset by peer)
19:40:59	LyndsySimon	Araq: One way to do pagination is through limit/offset, yes - but not the only way. For instance, it may be necessary to ensure that subsequent pages do not return records that were sent on previous pages.
19:41:24	LyndsySimon	Araq: To be clear, I'm not saying that's a requirement for anything I'm working on right now - but it is something that I've had to do in the past, and something that other languages' libraries support.
19:42:02	Araq	sure but I pagination can get tricky so I'm not sure how far it can be abstracted
19:42:07	*	LyndsySimon nods
19:42:29	Araq	it's so hard DBs support it out of the box with limit/offset ;-)
19:42:37	LyndsySimon	Yeah, I'm not trying to nail down the details. I'm really just looking for a project I can sink my teeth into so I can come up to speed in Nim.
19:42:38	*	Jesin joined #nim
19:43:12	*	LyndsySimon remembers dealing with it manually in MS SQL Server.
19:43:25	LyndsySimon	MS SQL didn't get limit/offset support until 2012, I want to say.
19:43:46	LyndsySimon	It's still not part of ANSI SQL, is it?
19:44:35	Araq	not sure, ANSI SQL is like ANSI C. yes there is a standard but there also only a handful of implementations so you can safely ignore it
19:46:15	Araq	actually it's like most standards. who cares if it adheres to the HTML spec when all the browser agree on it. the spec will adapt soon enough to the real world.
19:46:51	*	Araq doesn't believe in standards.
19:53:35	*	ipjk joined #nim
19:54:15	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
19:55:48	*	nsf joined #nim
19:59:08	*	chemist69 quit (Ping timeout: 246 seconds)
20:00:12	*	Matthias247 joined #nim
20:03:30	*	claudiuinberlin joined #nim
20:11:29	*	adeohluwa quit (Quit: Connection closed for inactivity)
20:11:38	*	claudiuinberlin quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
20:13:18	*	miran quit (Remote host closed the connection)
20:16:14	*	Jesin quit (Quit: Leaving)
20:17:48	*	chemist69 joined #nim
20:17:59	*	haha_ quit (Quit: haha_)
20:19:51	*	Yardanico quit (Remote host closed the connection)
20:21:08	LyndsySimon	I'm struggling to understand how this works: https://github.com/nim-lang/Nim/blob/devel/lib/impure/db_postgres.nim#L503-L523
20:21:27	LyndsySimon	The "connection" parameter doesn't seem to be used for anything as far as I can tell.
20:22:01	TjYoco	the same proc is used for like 4 different db's
20:22:25	TjYoco	so if you switch to another one you still just call open
20:22:43	LyndsySimon	Right, that makes sense. So it's just a case of maintaining the same functional signature?
20:23:03	*	claudiuinberlin joined #nim
20:23:27	TjYoco	Yeah, it was in Nim in Action, I'm pretty sure thats what it said.
20:24:05	LyndsySimon	Gah. I didn't even see the note in the docstring :(
20:24:58	Araq	in retrospect I would do it differently. it's not really useful to have the same API for 'open'
20:26:12	LyndsySimon	Fair. Do you take PRs on GitHub?
20:36:10	Araq	always.
20:36:28	LyndsySimon	https://github.com/nim-lang/Nim/pull/6381
20:36:54	LyndsySimon	It's a tiny, ridiculously small change - but I figure it's better to submit it as a PR than to demand someone else fix it.
20:37:20	Araq	er ... this is needs to be fixed differently then
20:37:30	LyndsySimon	That works too :)
20:37:54	Araq	"host=localhost dbname=dbname" # you need to construct this string in 'open'
20:38:15	Araq	I take it back, a unified open proc is awesome for reasons like this
20:38:17	LyndsySimon	Hmm? I tested it locally and it works.
20:38:28	Araq	yes it works, but it's not correct
20:38:44	Araq	that's nuking the abstraction
20:38:48	LyndsySimon	OK. Let me know if I can correct it, or if I should leave it alone.
20:39:01	Araq	try to understand what I mean
20:39:18	Araq	let db = open("localhost", "user", "password", "dbname") # this is how it is supposed to work
20:39:28	Araq	# currently the 'localhost' is IGNORED
20:39:37	Araq	that's the bug
20:40:04	Araq	within open you need to construct the
20:40:07	LyndsySimon	Gotchya. I understand the concept of what needs to be done, then. I lack a few pieces of information but i can get them.
20:40:22	Araq	'host=localhost dbname=dbname' part
20:41:07	*	salewski quit (Quit: WeeChat 1.4)
20:41:10	*	LyndsySimon nods
20:42:55	LyndsySimon	The connection param is intended to be in the format `<hostname>:<port>`, correct? E.g. `localhost:5432` in the default case?
20:43:08	Araq	yes
20:43:53	*	gokr joined #nim
20:44:10	LyndsySimon	Great. Outside the scope of this discussion, have you considered supporting URI-style configuration in the standard library?
20:44:26	Araq	ugh, the doc comment says
20:44:37	Araq	## Example:
20:44:37	LyndsySimon	E.g.: `open("postgresql://username:password@localhost:5432/dbname")`
20:44:37	Araq	## con = open("", "", "", "host=localhost port=5432 dbname=mydb")
20:45:18	Araq	no, we want less string parsing, not more
20:45:47	Araq	the wrappers do not support runtime switching between DBs anyway
20:46:01	*	LyndsySimon nods
20:46:30	*	mahmudov quit (Remote host closed the connection)
20:46:33	Araq	there is a nimble package for that though. then your app depends on sqlite.dll, postgre.dll and mysql.dll at the same time, a hallmark of good engineering
20:47:16	LyndsySimon	If I build the string in `open`, should I also check to ensure that the host and port are not already specified in the `database` parameter? Not doing so would break backward compatibility.
20:47:59	Araq	you break compat by checking the 'database' parameter, I think
20:48:04	LyndsySimon	Let me know if I'm asking too many questions, by the way. I'm definitely off on a tangent, but it is helping me understand things.
20:48:41	Araq	you need to continue to support what the docs show, aka connection empty, stuff cooked up in the dbname section
20:49:08	FromGitter	<zacharycarter> hoping someone can help me figure out why this code is not exiting the while loop - https://gist.github.com/zacharycarter/dbb8a2561f64fe67ada0edb0e1d538d6#file-nimrl-nim-L53-L95
20:51:48	TjYoco	echo your queue.len in the while loop. Maybe checkValidity isnt working and you keep adding 4 for every 1 you pop
20:52:12	Arrrr	I have seen something strange:
20:52:14	FromGitter	<zacharycarter> yeah that is probably the case
20:52:24	FromGitter	<zacharycarter> as the queue seems to be growing like crazy
20:52:28	Arrrr	This inside your while: dungeon[coord.column, coord.row] = CellKind.Floor
20:52:37	Arrrr	And this inside your check: dungeon[column, row].kind != CellKind.Floor
20:52:54	Arrrr	i don't understand why is there a kind but not in the other place
20:53:01	FromGitter	<zacharycarter> ooo good catch
20:53:08	FromGitter	<zacharycarter> oh wait
20:53:14	FromGitter	<zacharycarter> because one is an assignment and one is checking the property
20:53:17	FromGitter	<zacharycarter> if you look further up in the code
20:53:28	FromGitter	<zacharycarter> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=59baec48177fb9fe7ec02fc4]
20:53:57	FromGitter	<zacharycarter> TjYoco: yeah I'm thinking that's what's happening I'm just not sure why...
20:54:34	Araq	zacharycarter: it's an interesting bug, add a duplicate check
20:54:49	FromGitter	<Bennyelg> @zacharycarter What does `[]` and `[]=` means
20:54:58	FromGitter	<Bennyelg> and how do you call them
20:55:06	Araq	var traversed = initHashSet[(int, int)]()
20:55:07	FromGitter	<zacharycarter> Araq: what do you mean a duplicate check?
20:55:22	Araq	if traversed.containsOrIncl((x, y)):
20:55:29	Araq	echo "cycle? ", x, y
20:55:48	FromGitter	<zacharycarter> I'll try that thanks Araq
20:56:17	Araq	btw I love Nim's containers
20:57:05	Arrrr	I have written code like this before, move the CellKind.Floor assignment inside your ifs to gain speed.
20:57:23	FromGitter	<zacharycarter> 
20:57:27	FromGitter	<zacharycarter> thanks for the protip
20:57:31	Araq	if checkValidity(dungeon, coord.column, coord.row - 1):
20:57:31	Araq	queue.addLast((coord.column, coord.row - 1))
20:57:44	Araq	4 times copied and pasted with minor modifications?
20:57:51	FromGitter	<zacharycarter> yeah
20:58:01	Araq	use a template here
20:58:09	Araq	eliminates typos
20:58:17	FromGitter	<zacharycarter> good idea
20:59:20	FromGitter	<zacharycarter> @Bennyelg they are operator overrides
21:01:16	Arrrr	maybe is a bug in queue
21:05:13	FromGitter	<zacharycarter> Maybe
21:05:18	Arrrr	I compiled it in playground with the change i suggested and it works https://play.nim-lang.org/?gist=340671337ddf5a97bd252e014be5b8fb
21:05:32	Arrrr	less checks i suppose
21:06:25	FromGitter	<zacharycarter> Awesome thank you!
21:08:03	*	Vladar quit (Quit: Leaving)
21:12:03	LyndsySimon	I'll be glad when the staticly-typed nature of Nim is more natural to me :(
21:12:30	LyndsySimon	I'm calling a proc, and I want to pass an argument only if that argument is not an empty string; if it is an empty string I want to pass that argument as nil.
21:12:51	*	xet7 quit (Remote host closed the connection)
21:15:38	Araq	proc p(param = "")
21:15:45	Arrrr	(if str == "": string(nil) else: str)
21:16:56	*	claudiuinberlin quit (Quit: Textual IRC Client: www.textualapp.com)
21:17:27	Araq	Arrrr: string(nil) will cease to exist
21:17:40	Arrrr	why is that
21:17:46	*	vlad1777d quit (Ping timeout: 255 seconds)
21:19:15	*	MyMind joined #nim
21:20:20	Arrrr	I suppose you are going to init seqs and strings by default. Will {.noinit.} work for them?
21:20:23	*	gokr left #nim (#nim)
21:21:05	*	Sembei quit (Ping timeout: 240 seconds)
21:27:46	LyndsySimon	Araq: Here you go, round 2 - https://github.com/nim-lang/Nim/pull/6381
21:27:53	Araq	"" will be mapped to 'nil' internally, Arrrr
21:27:53	*	vlad1777d joined #nim
21:28:11	Araq	btw len(s) for s.isNil returns 0 already
21:28:18	Araq	and does so since quite some time
21:28:35	*	relax quit (Ping timeout: 240 seconds)
21:28:42	Araq	maybe later we'll get a different string implementation
21:29:34	Araq	but having both nil and "" sucks, always makes for subtle differences nobody should care about. if you do care, use an Option type
21:30:25	Araq	LyndsySimon: looks good
21:30:35	LyndsySimon	Woot!
21:30:35	Arrrr	So, will "" consume memory or not, that's the question.
21:30:41	*	LyndsySimon celebrates
21:30:43	Arrrr	other than pointer that is
21:31:09	LyndsySimon	The code is mostly copied from db_mysql, but still. I had to figure out what it was doing, I didn't just copypasta.
21:32:48	Araq	LyndsySimon: I suppose 'port' should be a parameter of open too :-)
21:33:01	Araq	but that's a PR for another time
21:33:27	LyndsySimon	Araq: Yes, I agree. If we take the approach of "less string parsing is better", then that's what I would do to be consistent.
21:33:32	Araq	Arrrr: no, we map it to 'nil', newString(0) will continue to allocate though
21:34:57	LyndsySimon	You mentioned earlier that the package that lets you use URIs required your app to be dependent upon multiple DBs' DLLs. Is that necessarily the case, though? Couldn't you have compile-time code that parses the connection string and conditionally includes only the library that's appropriate? Assuming of course that the DB URI is not configurable at runtime.
21:36:26	Araq	then the type would be static[string], not 'string'
21:38:36	Araq	but string based DSLs are suspicious, I know the Nim syntax, I can construct a call like connect(database = "x") # use defaults for the other parameters
21:38:57	Araq	but how do I the same in the string literal? it is not obvious
21:39:56	Araq	the error messages are also likely worse when I get it wrong, the compiler's error messages are good, ymmv
21:40:41	*	couven92 quit (Ping timeout: 240 seconds)
21:47:34	*	couven92 joined #nim
21:51:10	*	vivus quit (Quit: Leaving)
22:00:59	*	elrood quit (Quit: Leaving)
22:12:44	FromGitter	<zacharycarter> Arrrr: I'd still like to know why my first post didn't work
22:12:54	FromGitter	<zacharycarter> because I don't think your solution gives me the results I'm expecting
22:13:02	FromGitter	<zacharycarter> unless there's another bug somewhere else in my code
22:14:45	Araq	zacharycarter: replace the queue with a seq to rule out a stdlib bug
22:15:22	FromGitter	<zacharycarter> will do
22:17:12	*	Trustable quit (Remote host closed the connection)
22:22:02	FromGitter	<zacharycarter> must be a bug in my implementation thanks for the suggestion Araq, doesn't work with a seq either
22:22:30	Araq	trace it and it should become obvious
22:22:47	Araq	there is something we're all missing :-)
22:28:10	ipjk	zacharycarter: mind pasting a link so I can have a go at it?
22:28:17	FromGitter	<zacharycarter> sure
22:28:36	FromGitter	<zacharycarter> I think I know the problem though - Araq hinted at it earlier, the same neighbors are getting added to the queue over and over again
22:28:59	FromGitter	<zacharycarter> ipjk: https://play.nim-lang.org/?gist=13389d1a61a89cbaceedc48f66b01b55
22:29:26	Araq	play.nim-lang is so awesome :D
22:29:32	FromGitter	<zacharycarter> thank you! :D
22:29:40	*	nsf quit (Quit: WeeChat 1.9)
22:29:40	Araq	oh yeah
22:29:53	Araq	it's your work, good job
22:30:00	Araq	can you add c2nim support?
22:30:14	FromGitter	<zacharycarter> probably - I'll take a stab at it over the weekend
22:30:21	Araq	that would be so cool
22:30:31	Araq	often I only need to translate C snippets
22:30:42	Araq	and creating a file, running c2nim on it etc is overkill
22:30:54	FromGitter	<zacharycarter> yeah that would be super handy I'm often doing the same thing
22:32:39	Araq	huh, what's this? word wrap for the editor? is it 2017 already?
22:34:00	FromGitter	<zacharycarter> I wish we had a better editor for the playground
22:36:28	FromGitter	<zacharycarter> alright finally fully connected caves :D
22:37:38	FromGitter	<zacharycarter> https://imgur.com/a/HlLll
22:47:17	Araq	I still don't get why your cycle detection doesn't work :-)
22:47:35	*	Arrrr quit (Ping timeout: 240 seconds)
22:51:12	FromGitter	<zacharycarter> ¯\_(ツ)_/¯
22:51:37	FromGitter	<zacharycarter> time for some other type of dungeon
22:51:57	FromGitter	<zacharycarter> BSP I guess
23:03:09	*	yglukhov quit (Remote host closed the connection)
23:08:18	TjYoco	When I try nimble publish it tells me 'Host key verification failed. fatal: Could not read from remote repository.'
23:08:39	TjYoco	I have public_repo checked for my access token, is there anything else I need?
23:09:02	Araq	sorry, no idea
23:09:22	*	renarc joined #nim
23:11:42	couven92	Araq, any new ideas for the vcc regression bug?
23:12:03	Araq	couven92: does it work without -d:release?
23:12:08	couven92	yes
23:12:33	dom96	TjYoco: That sounds like a git error
23:12:41	dom96	Is it setup correctly?
23:12:53	*	jsgrant quit (Ping timeout: 255 seconds)
23:13:30	TjYoco	dom96 I have setup through VSCode and I can push and pull from remote
23:14:33	dom96	TjYoco: Can you git clone from command line?
23:14:38	TjYoco	although... it does say 'Command: git clone [email protected]:Tyler-Yocolano/packages' and my thing isnt called packages, but idk if thats a nimble thing
23:15:29	TjYoco	dom96 yeah I just cloned my repo
23:15:30	Araq	couven92: try to compile semcalls.c without optimization settings
23:15:48	Araq	it used to be easy to do that but I forgot
23:16:03	couven92	{.opt: none.}
23:16:06	couven92	?
23:20:21	Araq	no.
23:21:21	couven92	Yeah, that would have been too easy right? :P Well, would've been nice :P
23:22:12	couven92	Araq, can I run `nim compile -c compiler/nim` and then edit `nimcache/nim.json`?
23:22:52	Araq	no
23:23:35	couven92	what is the projectname.json for anyways? just randomly useful metadata or does nim actually use it?
23:26:53	Araq	metadata for other tools
23:27:03	Araq	nim only produces it, it doesn't consume it
23:27:22	couven92	Araq, I can compile nim with a non-optimized semcalls.c manually from the command-line and see what happens
23:27:29	couven92	but that's really messy :(
23:28:07	couven92	since nim.json tells what I need to do... :)
23:28:19	FromGitter	<mratsim> @zacharycarter Is there a bug tracker/feature request on play.nim-lang.org? I would like to see ASM output similar to https://play.rust-lang.org/ or https://gcc.godbolt.org/ to understand where the compiler uses AVX and where I should give him a small nudge.
23:29:11	FromGitter	<mratsim> oh, and loop unrolling
23:30:37	Araq	couven92: add this to nim.cfg
23:30:52	Araq	semcalls.speed = ""
23:31:02	Araq	and pray it works :-)
23:31:15	couven92	WOW! okay, let's see :)
23:31:42	couven92	that would be in `compiler/nim.cfg`, right?
23:37:21	couven92	Araq, nope same error
23:37:41	Araq	check these options are used please
23:37:51	Araq	maybe you need
23:38:05	Araq	semcalls.speed = "-O0" or vcc's equivalent
23:38:50	TjYoco	It says verified as Tyler-Yocolano, so that should mean my access token worked right? And my repo is public as seen https://github.com/Tyler-Yocolano/nimbomb and my token says "public_repo" not sure what else to try
23:38:56	FromGitter	<zacharycarter> @mratsim please just file an issue on github https://github.com/zacharycarter/nim-playground
23:40:34	couven92	Araq, uhm... that's weird... nim does not generate `nimcache/compiler_semcalls.c` at all... is semcalls included somewhere instead of imported?
23:41:24	couven92	which would also mean, that `semcalls.option` can't really work, right? since nim doesn't actually invoke the compiler for semcalls.nim
23:41:33	Araq	yeah
23:41:53	Araq	nimgrep for it
23:44:52	couven92	Araq, no match?
23:44:58	dom96	TjYoco: Afraid you'll need to create a PR manually
23:46:45	couven92	ah, semcall not semcalls and it's in sem.nim
23:47:11	couven92	so `sem.speed = "/Od"` in the config then? let's see...
23:49:06	TjYoco	dom96 could you confirm that it should output: 'Command: git clone [email protected]:Tyler-Yocolano/packages ...' and not '[email protected]:Tyler-Yocolano/nimbomb', my actual package name? As well, I enable every tick mark under the access token and instead of outputting: 'Host key verification failed.', it read: `Permission denied (publickey).' That seems strange...
23:49:31	dom96	yes, it should
23:50:10	TjYoco	dom96 whoops thats the nimble packages thing, duh
23:51:55	couven92	Araq, no setting `sem.speed = "/Od"` (with `/Od` being vcc's don't optimize) didn't do it... same error and nim.json shows `/O2` (optimize for speed)
23:53:06	Araq	sem.always = ?
23:56:59	couven92	Araq, nope and even with both configs there it doesn't affect anything... (maybe need `compiler` prefix, since it's in compiler?
23:57:46	couven92	AH! maybe do `sem.options.always`...
23:57:57	couven92	like with `vcc.options.always`