<< 09-01-2015 >>

---

00:00:13	Araq	(ref Foo)(field: 4) also works
00:00:22	onionhammer	it's less efficient
00:00:54	Araq	it's not. and if it were, we could optimize it in the compiler more
00:01:09	onionhammer	ah nevermind, its the same
00:01:12	onionhammer	just more writing
00:01:35	onionhammer	and looks weird
00:01:35	Araq	according to ldlework nobody ever requested getting rid of () though
00:01:48	*	Sphax quit (Quit: CYA!!)
00:01:48	onionhammer	ref Foo(field: 4) ?
00:01:55	onionhammer	then you're treating ref like new in other languages
00:02:01	Araq	(which must be the reason why everybody loves Lisp)
00:03:26	onionhammer	the fact that people have found so many ways to do this indicates they're looking for an easier way :P
00:03:50	Araq	to me it indicates they are using too many different types
00:04:21	onionhammer	in what way
00:06:17	ldlework	Araq: for expressions, no I haven't heard the majority of this field complaining about ()
00:06:26	ldlework	talk about a feature no one's asked for (strong spaces) ;)
00:06:50	*	ajhager joined #nim
00:07:06	Araq	yeah, everybody stumbles upon 'echo $foo' but "nobody asked for the feature". sure.
00:08:00	onionhammer	the echo $foo thing does suck
00:08:53	ldlework	Araq: yeah, in the most simple cases people do it
00:09:29	ldlework	The only time I use that is for new, so it stands out as a memory allocation
00:12:31	*	z1y joined #nim
00:12:32	dtscode	aporia is still actively developed right?
00:13:12	onionhammer	theoretically
00:13:52	dtscode	i wonder how much trouble i would get in for suggesting gedit like features
00:16:06	*	Sergio965 joined #nim
00:16:25	Araq	dtscode: just go ahead
00:16:47	ldlework	dtscode: like what
00:27:32	dtscode	ldlework, like being able to close all of the tabs
00:29:31	Var|Mobile	Gah. Anyone seen the latest post in the constructors thread on the forum?
00:29:48	Araq	Var|Mobile: just read it
00:30:36	Araq	dtscode: I never have more than 10 tabs open, it's unmanagable with more. so I don't need "close all tabs", but ymmv
00:30:48	Var|Mobile	It made me feel... annoyed
00:31:53	Araq	I can answer
00:32:32	*	irrequietus quit ()
00:32:44	dtscode	Araq, i mean like ill be working on nimbus, then switch to bill bot, but i have to open billbot before i can close out nimbus
00:33:28	Araq	you can disable the "already running" check and start 2 instances of the process
00:33:43	*	Trustable quit (Remote host closed the connection)
00:33:43	dtscode	but i only need the one open
00:34:11	Araq	you just described you need 2. :P
00:34:33	dtscode	no i didnt
00:34:36	*	gmpreussner_ joined #nim
00:34:45	dtscode	that would mean i need two projects open at once
00:34:46	*	gmpreussner quit (Ping timeout: 255 seconds)
00:34:55	dtscode	i want to completely close out a project and start a new one
00:35:48	*	casci joined #nim
00:36:06	*	arekzb_ joined #nim
00:36:43	*	arekzb_ quit (Client Quit)
00:37:12	*	areckizb joined #nim
00:37:16	*	enquora quit (Quit: enquora)
00:49:09	*	Sergio965 quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
00:52:25	*	no_name quit (Remote host closed the connection)
00:55:22	*	casci left #nim (#nim)
00:55:47	*	EXetoC joined #nim
01:07:32	*	no_name joined #nim
01:08:19	Varriount	Araq: Here are the installers: https://drive.google.com/file/d/0B077nrrf63xtNTc1UkhubExIUGs/view?usp=sharing
01:10:12	Varriount	Araq: Now what should I work on? Buildbot? IDETools?
01:10:48	dtscode	what are the ide tools?
01:11:04	Araq	buildbot
01:11:09	Varriount	dtscode: Features of the compiler meant to help augment support for text editors.
01:11:12	jpoirier	Are untyped proc parameters a part of the language? e.g. this compiles proc tst(a, b, c) = echo(a); echo(b); echo(c) and runs tst("Hello", "World", 42.0)
01:11:18	dtscode	ah
01:11:21	Varriount	jpoirier: Yes.
01:11:26	jpoirier	roger
01:11:49	dtscode	that would be under varargs wouldnt it be?
01:12:09	Araq	jpoirier: I might reconsider this though
01:12:39	Araq	I don't like it too much tbh
01:13:21	Araq	I think I'll put this to .experimental
01:13:31	flaviu	> undeclared identifier: 'NEVER_UTF'
01:13:54	flaviu	But right next to it is declared another identifier that does work
01:14:59	Varriount	Araq: Could you clarify what you said about generic parameter lists in the nim for scientific application thread?
01:15:39	jpoirier	Araq: I don't really care for it in Nim either; it doesn't seem orthogonal with the rest of the language
01:16:07	Araq	jpoirier: it is. it has been introduced for reasons of orthogonality
01:16:28	Araq	template t(a, b, c) = ... # works and should work
01:16:33	Araq	it's a shortcut for:
01:16:41	Araq	template t(a, b, c: expr)
01:16:57	flaviu	Oh, it's my fault after all.
01:17:39	Araq	Varriount: generics can always be in [. .] but also in [] for convenience and backwards compatibility
01:18:14	Araq	but for [. .] we can allow obj.m[.int, float.](a, b)
01:18:32	Araq	which is m[.int, float.](obj, a, b)
01:19:17	Araq	in contrast to that obj.m[int, float](a, b) is (obj.m[int, float])(a, b) which fails
01:19:21	Varriount	Araq: Hm, but isn't the ambiguity resolvable? The compiler just has to peek and see if 'm' is a generic, rather than a field with `[]` access
01:19:29	flaviu	I'm not sure I like that. There are already a thousand ways to do everything, two thousand it too many.
01:19:34	flaviu	*is
01:19:52	flaviu	Varriount: No, that's a bad idea
01:20:05	Varriount	flaviu: Why?
01:20:22	flaviu	The parser should be totally decoupled from the rest of the language
01:20:28	Araq	Varriount: what about generic [] accessors though?
01:20:54	Varriount	flaviu: This wouldn't be coupled with the parser? It would be done in the semantic checking phase.
01:21:01	Araq	flaviu: the parser doesn't do the precedence rewriting. but I don't like it either.
01:22:28	Varriount	Araq: What generic procedure also has a generic accessor?
01:22:52	renesac	Araq: I was trying to apply c2nim in the first code sample from here: http://www-unix.mcs.anl.gov/~kazutomo/rdtsc.html
01:24:38	Varriount	if `m` is not a field in `obj`, and `m` is a generic procedure or macro, rewrite the ast
01:27:42	*	Var|Mobile quit (Quit: AndroIRC - Android IRC Client ( http://www.androirc.com ))
01:27:52	jpoirier	Araq: orthogonal from the point of view of "does what you expect" and not from the AST point of view
01:28:45	Araq	jpoirier: do you mean the proc p(a, b, c) feature or the obj.m[T](x) stuff?
01:29:52	*	willwillson quit (Ping timeout: 240 seconds)
01:32:14	Araq	good night
01:32:28	*	yglukhov______ joined #nim
01:34:03	*	gmpreussner_ quit (Remote host closed the connection)
01:34:35	*	gmpreussner_ joined #nim
01:35:40	*	saml_ joined #nim
01:36:07	*	leavengood quit (Ping timeout: 244 seconds)
01:36:50	*	yglukhov______ quit (Ping timeout: 244 seconds)
01:47:40	gmpreussner_	hmm... stdin.readChar still requires Return to be pressed. is there a simple way to read just key presses?
01:48:03	*	jpoirier quit (Quit: Leaving)
01:50:45	*	jpoirier joined #nim
01:56:38	*	jpoirier quit (Quit: Leaving)
01:57:16	renesac	gmpreussner_: unix specific, but maybe fits: http://nim-lang.org/readline.html
01:58:24	gmpreussner_	that looks good, but some crossplatform solution would be better
01:58:56	gmpreussner_	oh well, maybe i'll just leave that out for now
02:00:54	*	gsingh93 quit (Quit: Connection closed for inactivity)
02:03:38	*	jpoirier joined #nim
02:03:56	*	jpoirier quit (Client Quit)
02:05:30	*	jpoirier joined #nim
02:06:06	*	superfunc joined #nim
02:10:16	*	gokr why do I let myself get sucked into discussions .... oh well.
02:10:49	EXetoC	/me? :p
02:11:45	*	JinShil quit (Quit: Konversation terminated!)
02:15:18	gokr	Someone called Nikki on forum?
02:16:52	*	jpoirier quit (Quit: Leaving)
02:17:01	gokr	gnite
02:17:11	*	jpoirier joined #nim
02:19:06	*	jpoirier quit (Client Quit)
02:19:27	*	jpoirier joined #nim
02:20:13	*	EXetoC quit (Ping timeout: 264 seconds)
02:23:13	*	ajhager quit (Ping timeout: 264 seconds)
02:27:16	*	shevy joined #nim
02:31:41	*	gokr quit (Quit: Leaving.)
02:31:45	flaviu	I copied the code from <https://github.com/Araq/Nim/blob/devel/lib/wrappers/pcre.nim#L38-L49> into one of my own files
02:31:55	*	ajhager joined #nim
02:32:00	flaviu	but "Error: invalid pragma: dynlib: pcreDll"
02:32:40	flaviu	Which I don't understand, if it works for Nim, why doesn't it work for me?
02:35:15	flaviu	dom96: ping, I'm getting missing css on http://irclogs.nim-lang.org/16-06-2012.html
02:37:42	Varriount	flaviu, dom96: As am I.
02:38:09	*	kapil__ joined #nim
02:38:10	Varriount	Although, whether missing CSS impacts usability greatly is debatable.
02:38:16	Varriount	Hello kapil__
02:38:27	Varriount	Hello ajhager
02:39:06	def-	flaviu: you defined const pcreDll?
02:39:20	Varriount	def-: Yeah, it's defined
02:39:24	def-	strange
02:39:33	Varriount	def-: He's using Nim's PCRE wrapper
02:39:39	flaviu	Varriount: No
02:39:52	flaviu	I'm using my own PCRE wrapper with Nim's dynlib code.
02:40:16	flaviu	https://gist.github.com/e4a30004d0bb8e90fb8c
02:41:01	renesac	Varriount:: importing windows module is scary
02:41:11	renesac	only the docs already make my browser pause
02:41:20	def-	flaviu: this is funny indeed
02:41:30	Varriount	renesac: Yeah. Windows has a *huge* api
02:41:49	renesac	I ended up copying just what I wanted...
02:42:00	renesac	about 6 lines
02:42:03	Varriount	renesac: Are you sure what you need isn't in winlean?
02:42:12	renesac	maybe
02:42:24	renesac	where is it?
02:42:35	Varriount	renesac: Same directory as the windows module
02:42:53	Varriount	renesac: It's the module most of the Nimrod stdlib uses when interfacing with Windows.
02:43:03	renesac	http://nim-lang.org/lib.html <-- I'm not seeing it here
02:43:30	Varriount	renesac: *shrug*, it exists.
02:43:35	Varriount	I mean, as a file.
02:43:47	flaviu	http://nim-lang.org/winlean.html
02:43:56	flaviu	renesac: The good stuff is hidden from that page :P
02:44:52	renesac	why?
02:45:21	flaviu	renesac: I was just joking, it was probably overlooked.
02:45:42	renesac	well, it got the types I need, just the GetCurrentProcess() I need to copy from windows.nim
02:46:07	def-	flaviu: remove the pcreImport inside the type definitions, then it works
02:46:10	renesac	BOOL WINAPI == WINBOOL?
02:46:23	def-	the error message should say something about being triggered there =/
02:49:44	renesac	Varriount:: the pascal case in the windows API is annoying now that Nimrod is first case sensitive
03:01:49	*	SplinterOfChaos joined #nim
03:04:13	*	gmpreussner__ joined #nim
03:04:57	*	gmpreussner_ quit (Ping timeout: 265 seconds)
03:06:01	*	BitPuffin quit (Ping timeout: 244 seconds)
03:07:11	jpoirier	what's the func keyword for/do? I assumed s/proc/func but that gives an error.
03:08:00	def-	jpoirier: it's reserved for potential use as procs without side effects (I think)
03:08:38	jpoirier	roger
03:09:47	*	Sergio965 joined #nim
03:11:00	def-	there it is: http://nim-lang.org/manual.html#nosideeffect-pragma
03:12:59	jpoirier	def-: thanks. was just reading that...
03:15:48	*	VinceAddons quit (Read error: Connection reset by peer)
03:16:12	Varriount	renesac: Feel free to change it, using the deprecated pragma
03:18:06	def-	anyone who can remove forum threads online?
03:20:12	*	pwernersbach joined #nim
03:20:13	superfunc	hey man
03:20:18	superfunc	fifa coins are important :3
03:20:36	pwernersbach	hey, just dropping in to alert everyone that we've been spammed on the forums
03:21:00	*	darkf joined #nim
03:21:06	superfunc	pwernersbach: yeah def- is looking for someone to pull it
03:21:20	pwernersbach	Cool, thanks for being on it
03:21:21	def-	Probably no one with forum moderation access awake
03:21:23	*	yglukhov______ joined #nim
03:21:34	pwernersbach	I g2g though goodbye
03:21:39	*	pwernersbach left #nim (#nim)
03:23:04	*	brson quit (Ping timeout: 264 seconds)
03:25:53	*	yglukhov______ quit (Ping timeout: 240 seconds)
03:27:55	*	superfunc quit (Quit: Page closed)
03:33:35	*	yglukhov______ joined #nim
03:37:53	renesac	Varriount: : I though there would be discussion
03:38:11	renesac	so all that is lacking is someone making a script to do it?
03:38:18	*	yglukhov______ quit (Ping timeout: 265 seconds)
03:47:01	Varriount	renesac: Or something. What happens when you run it through nimfix?
03:55:19	*	BitPuffin joined #nim
03:57:56	renesac	it probably don't create the "deprecated" pragmas
03:58:02	renesac	but i don't know
03:58:06	renesac	I will try tomorro
03:58:07	renesac	w
03:58:49	renesac	Varriount: :BOOL WINAPI == WINBOOL ?
04:00:24	*	gsingh93 joined #nim
04:01:45	*	renesac is now known as renesac|away
04:04:22	*	gmpreussner__ quit (Remote host closed the connection)
04:04:49	*	gmpreussner__ joined #nim
04:14:48	Varriount	renesac|away: Huh?
04:14:49	*	BitPuffin quit (Ping timeout: 264 seconds)
04:16:46	*	BitPuffin joined #nim
04:21:39	*	BitPuffin quit (Ping timeout: 264 seconds)
04:34:45	*	Sergio965 quit (Read error: Connection reset by peer)
04:38:53	*	jefus_ joined #nim
04:42:35	*	jefus quit (Ping timeout: 265 seconds)
04:47:44	*	Sergio965 joined #nim
04:47:52	onionhammer	spammers on the forum.....
04:48:21	*	Sergio965 quit (Client Quit)
04:48:55	*	Sergio965 joined #nim
04:51:40	*	z1y quit (Ping timeout: 264 seconds)
04:56:19	*	hjjyuuu joined #nim
04:56:56	*	hjjyuuu quit (Client Quit)
05:01:45	*	saml_ quit (Quit: Leaving)
05:02:12	Varriount	?
05:02:16	Varriount	onionhammer: Hm?
05:02:49	Varriount	onionhammer: Do you mean Nikki?
05:10:57	*	brson joined #nim
05:14:20	*	z1y joined #nim
05:17:35	*	BitPuffin joined #nim
05:17:54	*	vbtt joined #nim
05:19:52	*	z1y quit (Ping timeout: 240 seconds)
05:22:44	*	BitPuffin quit (Ping timeout: 245 seconds)
05:34:28	*	Epic|gmpreussner joined #nim
05:35:16	*	gmpreussner__ quit (Ping timeout: 265 seconds)
05:54:33	*	nRike joined #nim
05:55:11	nRike	hey guys, I'm having trouble when trying to make a httpclient request from a URL
05:55:33	nRike	every time I want to look up for a website, it throws 400 error
06:01:17	*	bjz joined #nim
06:27:49	*	z1y joined #nim
06:34:21	*	gmpreussner joined #nim
06:34:49	*	Epic|gmpreussner quit (Ping timeout: 245 seconds)
06:45:58	*	nRike quit (Ping timeout: 246 seconds)
06:52:37	*	ajhager quit (Ping timeout: 264 seconds)
07:00:09	*	kniteli quit (Ping timeout: 252 seconds)
07:01:35	*	Demon_Fox quit (Quit: Leaving)
07:06:38	*	BitPuffin joined #nim
07:11:29	*	BitPuffin quit (Ping timeout: 245 seconds)
07:18:04	*	jefus joined #nim
07:20:56	*	jefus_ quit (Ping timeout: 244 seconds)
07:23:51	*	nande quit (Remote host closed the connection)
07:26:06	*	bjz quit (Ping timeout: 244 seconds)
07:29:42	dtscode	does jester work reasonably well with apache?
07:30:36	*	gokr_ joined #nim
07:32:06	*	vbtt quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client)
07:35:10	*	bjz joined #nim
07:39:34	*	bjz quit (Ping timeout: 255 seconds)
07:46:50	vegai	dtscode: don't see why it wouldn't
07:47:02	vegai	apache might be a bottleneck in such a setup, though :P
07:47:12	dtscode	what do you mean?
07:47:38	vegai	half-joking there. I meant that jester will be churning out stuff faster than apache can serve it
07:47:45	vegai	but I'm not certain if that's true
07:47:55	dtscode	oh well that i dont care about so much
07:48:03	dtscode	its just for my personal site
07:48:05	vegai	yeah, it should be fast enough anyway
07:48:13	dtscode	awesome
07:48:22	vegai	you'll know when you try it :)
07:51:09	dtscode	is dom96's fork of jester the current one?
07:52:59	vegai	he started it and I believe it's the most current one
07:53:10	dtscode	awesoem
07:53:11	dtscode	thanks
07:53:32	*	gokr joined #nim
07:53:39	gokr	Good morning!
07:54:06	gokr	dom96: A spammer account has appeared on the forum, chou66
07:54:09	dtscode	good morning gokr
07:54:30	gokr	Morning to ya!
07:54:46	gokr	Where are you btw?
07:55:05	dtscode	cali
07:55:07	*	gour joined #nim
07:55:11	dtscode	i just keep weird hours
07:56:42	gokr	Cali?
07:56:52	dtscode	california
07:57:00	gokr	Ah :)
07:57:05	dtscode	you?
07:57:19	gokr	Sweden, north of Stockholm
07:57:29	gokr	So its... 9 in the morning here
07:57:49	dtscode	are the bank accounts as good as they say?
07:57:54	Varriount	Hey gokr
07:58:02	gokr	Bank accounts?
07:58:07	gokr	Varriount: hey :)
07:58:09	Varriount	gokr: I admire your civil responses on the forum.
07:58:09	dtscode	swedish bank accounts
07:58:20	dtscode	i should start being on the forum more
07:58:20	gokr	Varriount: hehe... yeah.
07:58:42	dtscode	i can get input on my impending tutorial there :}
07:58:45	gokr	I am spoiled by the extremely polite Smalltalk community I guess.
07:58:48	Varriount	gokr: I also liked that article link you posted.
07:58:58	gokr	You mean by Gilad?
07:59:20	gokr	dtscode: You saw my Nim articles, right?
07:59:23	Varriount	gokr: Well, I'd like to think that the Nim community, overall, is polite. Though, a few... sour apples are bound to pop up.
07:59:35	dtscode	gokr, i did, but havent had time to read them
07:59:59	gokr	Varriount: I agree, the community has been very friendly to an "outsider" like me for sure.
08:00:24	gokr	dtscode: Feel free to pillage for a tutorial
08:00:33	Varriount	gokr: Did you see my earlier response? The last one on page 2?
08:00:37	dtscode	i probably will. you write great stuff
08:01:36	gokr	Varriount: Yes, i read it
08:02:15	dtscode	> Constructors are supposed to guarantee that the object is initialized fully and properly.
08:02:16	dtscode	heh
08:02:18	dtscode	n00b
08:02:41	Varriount	Well, *supposed* is the key term.
08:03:08	gokr	Now... there are two things Smalltalk has that Nim doesn't - that may be a factor here.
08:03:19	gokr	Smalltalk has "method categories".
08:03:25	gokr	Which is IMHO a brilliant thing. :)
08:03:59	gokr	Its basically "named groups" of methods, but have no impact - so they are like annotations.
08:04:17	Varriount	"All methods are public.1 Methods are grouped into protocols that indicate their intent. Some common protocol names have been established by convention, for example, accessing for all accessor methods, and initialization for establishing a consistent initial state for the object."
08:04:22	gokr	And very convention driven - so you have say "accessing" which always contains pure getters/setters etc.
08:04:32	gokr	Yes, right.
08:04:45	gokr	And same goes for the class side.
08:04:51	*	gmpreussner quit (Ping timeout: 264 seconds)
08:04:54	*	gmpreussner_ joined #nim
08:05:02	gokr	Now... this is where most non Smalltalkers get confused - classes are objects.
08:05:08	Varriount	gokr: Which makes me wonder if smalltalk wouldn't have been a better choice for Android. :P
08:05:20	gokr	Sure it would.
08:05:40	gokr	But the reasons Smalltalk is not more popular - its a whole book.
08:05:58	gokr	But it's a brilliant language - similar to how Lisp is brilliant.
08:06:13	dtscode	Varriount, im sure lots of things woul dhave been better
08:06:15	gokr	Anyway... so classes have methods, and they also are categorized of course.
08:06:35	gokr	So... my point: there is a class side method category called "instance creation"
08:06:54	gokr	By convention it always has all the "constructor methods".
08:07:12	gokr	So you can easily find them - and you know they can be used to create instances.
08:07:33	gokr	Example code: Date today
08:07:53	gokr	That's Smalltalk for ... `Date.today()"
08:08:07	gokr	So "Date" is a class.
08:08:18	gokr	And today is a method on the class side.
08:08:41	Varriount	Neat.
08:08:50	gokr	It doesn't follow any naming convention - but since there are method categories - you find them anyway.
08:09:25	gokr	And since classes are objects - and thus are instances of meta classes - they all inherit one such method. Its called "new".
08:10:28	gokr	The second thing different in Smalltalk - is that since we send a message to the class - its very easy to find examples of "instance creations."
08:10:28	Varriount	So you can create new classes on-the-fly
08:10:42	gokr	Yes, in Smalltalk you are always in runtime.
08:10:59	gokr	As I have said a few times - but I understand its "foreign".
08:11:15	Varriount	gokr: Not for someone whose first language was python.
08:11:50	gokr	Well, Python is close - agree.
08:11:59	gokr	Ruby is even closer.
08:12:37	gokr	Matz has even said he basically took the Smalltalk object model and put a more "traditional" syntax on it, and made it script friendly.
08:13:04	*	dts|pokeball joined #nim
08:13:10	Varriount	gokr: Matz?
08:13:24	gokr	So the second thing there - since 95% of all messages sent to classes - are in fact these instance creation messages - then its very easy to find all places where "Date" is instantiated.
08:13:55	gokr	http://en.wikipedia.org/wiki/Yukihiro_Matsumoto
08:14:11	gokr	Met him once at OOPSLA
08:15:20	gokr	To summarize - these two things (method categories and the fact that we easily can find all places where Date is the receiver) - make "constructor methods" a non problem in Smalltalk.
08:16:00	gokr	Nim doesn't have those two things though, but it could easily be improved if we annotated them using pragmas.
08:16:09	gokr	Follow?
08:16:34	Varriount	gokr: I think.
08:16:59	Varriount	gokr: Certainly makes me want to read up on smalltalk. It sounds like a fascinating language.
08:17:12	gokr	It is very fascinating.
08:17:23	gokr	6 keywords :)
08:17:30	Varriount	O_o
08:17:40	gokr	Or wait, 5.
08:17:45	Varriount	o_o
08:18:05	gokr	And they are probably not what you think.
08:18:17	gokr	Most people would guess "if", "while" etc
08:18:31	gokr	But nah, those are all methods in Smalltalk.
08:18:58	gokr	They are: nil, true, false, self, super, thisContext
08:19:03	gokr	Oh, 6 it is.
08:19:13	gokr	thisContext is often forgotten :)
08:19:30	Varriount	gokr: Unfortunately, I don't think Nim can be quite so dynamic as Smalltalk.
08:19:52	gokr	No, it can't. The languages are quite far from each other.
08:21:04	gokr	I have programmed in about 25-30 languages since I was 12 yo. I am 45 now. I discovered Smalltalk in 1993 I think - addict ever since.
08:22:23	gokr	Nim is actually the first language since 1993 that I am giving more time than just "playing a while and then going back home to Smalltalk".
08:22:42	Varriount	:D
08:22:51	gokr	So that's quite telling for Nim.
08:24:20	gokr	My influence on Nim is from that extensive Smalltalk experience. I know when things in Smalltalk "don't fit" - but I also know when its worth raising a hand and saying that... "hey, guys, this language from the 1980s does it like this...".
08:25:11	gokr	Btw, I used to teach Smalltalk too - the language fits comfortably on 5 slides.
08:26:22	Varriount	gokr: Well, annoying users on the forum aside, know that I value your feedback. I think Araq does too (or maybe he just enjoys discussing/arguing with you. Maybe it's the same thing? :P )
08:27:30	gokr	I haven't used the pragmas much yet - are they also used for purely annotation purposes?
08:27:53	Varriount	gokr: Well, they have a couple of uses (like everything else)
08:27:56	gokr	I am cool with Araq.
08:28:25	*	dtscode quit (Quit: ZNC - http://znc.in)
08:29:23	Varriount	You can use macros as pragmas to pass various AST's to the macro. You can use them to control how the compiler treats the target (calling convention, load from dll, etc)
08:30:09	gokr	Damn Nim is extensive :)
08:30:19	Varriount	gokr: There also macros which describe the procedure, such as what kinds of side-effects it has, if it requires locks, etc
08:30:38	gokr	Yeah, I know several of them etc, just curious if there are any that "just" do annotations.
08:30:54	Varriount	gokr: Well, I think you can have an empty user pragma.
08:30:57	gokr	So anyway - we could mark "constructor methods" using a pragma.
08:31:09	Varriount	*user-defined
08:31:43	gokr	Or in fact... we could even create a "proc category pragma".
08:31:58	gokr	So we could categorize behaviors - like in Smalltalk.
08:32:22	Varriount	http://nim-lang.org/manual.html#pragma-pragma
08:32:34	gokr	Ah, right :)
08:32:51	*	gmpreussner__ joined #nim
08:33:10	gokr	It may be overkill in your eyes - but the method categories are very, very nice in Smalltalk.
08:33:29	gokr	Now, if you look at say http://nim-lang.org/queues.html - you do get the iterators separately, which is nice.
08:33:37	*	gmpreussner_ quit (Ping timeout: 265 seconds)
08:33:47	gokr	But the "construction procs" are not easily found.
08:34:30	Varriount	gokr: Yep, you can create an empty user-defined macro, and place it on a procedure.
08:35:04	*	BlaXpirit joined #nim
08:35:11	Varriount	gokr: https://gist.github.com/Varriount/8442ead784268640ab41
08:35:36	gokr	yeah
08:36:14	gokr	So I think it would be a worthwhile thing - especially for even better auto completion I guess.
08:37:12	Varriount	Anyway, I need to get back to bed. It was nice talking with you gokr.
08:37:16	Varriount	Goodnight.
08:37:19	*	Sergio965 quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
08:37:19	gokr	Gnite!
08:37:55	BlaXpirit	i kinda like it without constructors
08:38:25	gokr	BlaXpirit: We do too. We are just talking about annotating the "construction procs".
08:38:39	BlaXpirit	kinda pointless
08:38:44	gokr	Eh, no?
08:38:55	BlaXpirit	yes
08:39:09	BlaXpirit	if a function returns something of that type, you could call it construction proc
08:39:20	BlaXpirit	expecially if it doesn't take an argument that is the type itself
08:39:29	BlaXpirit	s*
08:40:00	gokr	So you think we could skip the newXXX, initXXX convention too?
08:40:16	BlaXpirit	uh now that i think about it, yes
08:40:28	BlaXpirit	i've been skipping "init" myself
08:40:36	BlaXpirit	new may be good though
08:41:44	gokr	I don't agree, although I agree the heuristic would catch many of them. But it would also catch "false" ones.
08:42:10	BlaXpirit	i don't know what would even be "false"
08:42:40	BlaXpirit	you can't have a strict definition of a constructor in context of nim anyway
08:42:46	gokr	So, say you have an object holding lots of Bananas, and you call "find" to get one of them. Construction? No.
08:43:01	BlaXpirit	what if the heuristic is actually the sanes definition of them
08:43:19	BlaXpirit	gokr, that is true and renders a lot of my points invalid :p
08:44:49	gokr	One could however possibly argue - that if the proc returns a single T, and it also actually allocated it... but you know, I do think an annotation may be more intention revealing :)
08:45:08	*	brson quit (Quit: leaving)
08:45:33	BlaXpirit	nah... using proc body is not an option
08:45:39	BlaXpirit	especially in the context of FFI
08:46:08	gokr	Either way, if you read what I wrote above - my experience from Smalltalk indicates this "annotation" is very useful.
08:46:27	BlaXpirit	was gonna say i can't... but chat log
08:46:49	gokr	Ah, you just entered - sorry, didn't see.
08:48:48	BlaXpirit	another problem is, nobody's gonna use those
08:49:03	gokr	Why not?
08:49:19	BlaXpirit	i dunno. it's too late?
08:49:33	gokr	To add annotations like that?
08:49:50	gokr	It would be trivial to scan the stdlib and add them I think.
08:50:10	BlaXpirit	if someone is writing code, they don't wanna bother with something that has no effect
08:51:00	gokr	I disagree - that's invalidating all of documentation too.
08:51:01	BlaXpirit	now, if special convenient syntax for constructors is introduced, that may be a different story
08:51:33	gokr	Comments don't have effects either, yet we write them, don't we?
08:51:43	BlaXpirit	sometimes :p
08:52:36	gokr	Its of course a library maturity thing, you are not forced to use it in your little script etc, but nice to have it in stdlibs.
08:53:44	*	infrashortfoo joined #nim
08:55:25	*	BitPuffin joined #nim
08:55:29	BlaXpirit	ok, all things considered, it's not bad
08:55:52	gokr	Ha! :) I will treasure this moment
08:56:06	*	gokr taking a snapshot picture of my screen...
08:56:16	gokr	Just kidding.
08:56:49	BlaXpirit	no change in nim required. might as well start a movement right now
08:58:32	gokr	A sidenote: In Smalltalk when I want to see how a class is used, I tend to hit ctrt-N (or whatever) to have a browser on all methods referencing the class.
08:59:02	*	z1y quit (Remote host closed the connection)
08:59:18	gokr	95% of those methods would be methods calling an instance creation method - so I get lots of examples of usage that way.
08:59:23	BlaXpirit	"all methods referencing the class" is indeed a must
08:59:48	BlaXpirit	generated documentation is barely usable right now because if its absence
08:59:53	*	BitPuffin quit (Ping timeout: 240 seconds)
09:00:09	BlaXpirit	someone has been working on documentation full of sorting, filtering and other goodies
09:00:09	gokr	Having this annotation would make it easy to find all places where a type is "created" - and thus easy to find sample code.
09:00:12	BlaXpirit	no time now
09:00:51	BlaXpirit	gokr, but what do you mean "in smalltalk"
09:01:06	gokr	In the Smalltalk IDE
09:01:11	BlaXpirit	ok
09:01:30	BlaXpirit	vim users laugh at that
09:01:35	gokr	Smalltalk is 50% language, 50% IDE.
09:01:39	BlaXpirit	even though i'm not a vim user, i don't use no IDEs
09:02:28	gokr	Well... the Smalltalk "IDE" is not like other IDEs.
09:02:37	gokr	But "IDE" is the only name fitting.
09:03:09	gokr	Well, I guess Emacs is similar.
09:03:43	gokr	Similar in a theoretical POV. As in being immersed inside a Lisp "world".
09:04:43	gokr	Hard to explain, but its very cool stuff and you should try it sometime :)
09:04:46	*	Epic|gmpreussner joined #nim
09:05:02	BlaXpirit	easy to dismiss as unpopular
09:05:03	*	gmpreussner__ quit (Ping timeout: 256 seconds)
09:05:10	gokr	You mean like Nim?
09:05:12	BlaXpirit	but then again, we are in #nim
09:05:26	gokr	The really cool things are never mainstream.
09:05:55	gokr	Smalltalk is in this regard just like Nim. A gem going under the radar for most people.
09:06:11	BlaXpirit	for how many years has it been evading radars now?
09:06:26	gokr	Smalltalk was quite big you know
09:07:21	gokr	In say 1990-95 C++ had about 60% and Smalltalk had 33% of the "OOP" market.
09:10:10	gokr	In fact, lots of the big companies using Smalltalk didn't want to tell - because it was considered a competitive advantage. This was true especially in the stock market domain.
09:11:12	gokr	Sorry for the offtopic - but most seem to be asleep anyway ;)
09:15:37	*	minciue joined #nim
09:19:17	*	dts|pokeball quit (Remote host closed the connection)
09:25:57	*	dts|pokeball joined #nim
09:29:24	*	zahary1 quit (Ping timeout: 245 seconds)
09:45:26	*	dts|pokeball quit (Quit: Leaving)
09:47:44	*	dts|pokeball joined #nim
09:51:02	*	johnsoft quit (Read error: Connection reset by peer)
09:54:28	*	dts|pokeball quit (Quit: Leaving)
09:56:27	*	dts|pokeball joined #nim
09:57:29	*	dts|pokeball quit (Remote host closed the connection)
10:04:44	*	yglukhov______ joined #nim
10:05:12	*	gmpreussner joined #nim
10:05:14	*	Epic|gmpreussner quit (Ping timeout: 245 seconds)
10:24:02	*	MajorTennis joined #nim
10:24:33	*	MajorTennis quit (Changing host)
10:24:33	*	MajorTennis joined #nim
10:28:37	*	JinShil joined #nim
10:34:38	*	Sphax joined #nim
10:48:03	*	yglukhov_______ joined #nim
10:48:03	*	yglukhov______ quit (Ping timeout: 264 seconds)
10:55:14	*	BitPuffin joined #nim
10:57:53	*	JinShil quit (Ping timeout: 240 seconds)
11:00:44	*	bjz joined #nim
11:00:54	*	gsingh93 quit (Quit: Connection closed for inactivity)
11:11:31	*	yglukhov_______ quit (Quit: Lingo: www.lingoirc.com)
11:12:51	*	VinceAddons joined #nim
11:16:49	*	gour_ joined #nim
11:19:34	*	gour quit (Disconnected by services)
11:19:38	*	gour_ is now known as gour
11:22:39	*	Sphax quit (Quit: ZZZZZzzzzz)
11:23:05	*	BitPuffin quit (Ping timeout: 264 seconds)
11:28:02	*	JinShil joined #nim
11:29:14	*	BitPuffin joined #nim
11:32:42	*	bjz quit (Read error: Connection reset by peer)
11:34:41	*	gmpreussner_ joined #nim
11:36:04	*	gmpreussner quit (Ping timeout: 264 seconds)
11:39:40	JinShil	What does the .magic prama do? Does it just forward to a C function?
11:40:24	JinShil	Let me rephrase that. What does the magic pragma do?
11:42:10	Varriount	JinShil: It's used by the compiler to detect special procedures.
11:42:16	*	zielmicha joined #nim
11:42:31	Varriount	Usually procedures implemented by the compiler.
11:43:03	JinShil	Varriount, ok that makes sense. Thanks.
11:44:24	*	keyle joined #nim
11:44:34	JinShil	May I ask, why aren't those special procedures implemented in the runtime?
11:45:31	keyle	hi
11:46:24	Araq	JinShil: depends.
11:46:49	Araq	often the magic is transformed into a .compilerProc which is implemented in the runtime
11:49:48	*	bjz joined #nim
11:50:23	Araq	lol @ "forward to a C function". We don't use C. We generate it. I know it's hard to imagine but Nim doesn't require a core written in C.
11:53:59	*	johnsoft joined #nim
11:56:46	*	EXetoC joined #nim
12:00:14	JinShil	I guess I keep expecting the man behind the curtain written in C.
12:03:35	*	repax joined #nim
12:04:37	*	bjz quit (Ping timeout: 264 seconds)
12:06:15	*	leavengood joined #nim
12:06:38	*	leavengood is now known as Guest7667
12:12:55	*	bjz joined #nim
12:21:17	*	willwillson joined #nim
12:21:26	*	darkf quit (Ping timeout: 244 seconds)
12:22:40	*	areckizb quit (Ping timeout: 246 seconds)
12:25:44	*	bjz quit (Read error: Connection reset by peer)
12:28:56	*	Trustable joined #nim
12:34:55	*	gmpreussner__ joined #nim
12:35:39	*	gmpreussner_ quit (Ping timeout: 245 seconds)
12:37:29	*	bjz joined #nim
12:37:38	Araq	Varriount: I was quicker :P
12:38:34	Varriount	Araq: Hm?
12:38:56	Araq	I replied too in the IUP thread
12:40:22	keyle	is there a tutorial with a real world spawn example? ... I'm trying to hookup ircclient with readLine and I'm not having much luck :>
12:40:47	Varriount	keyle: Could you show us your code?
12:40:49	Araq	keyle: there are a couple of tests in tests/parallel
12:41:59	keyle	Varriount: note that I'm a nim noob - https://gist.github.com/keyle/46337a707c23492c7e11
12:43:25	*	Epic|gmpreussner joined #nim
12:43:37	*	gmpreussner__ quit (Ping timeout: 245 seconds)
12:44:02	keyle	I guess I have to use spawn here
12:44:17	Araq	keyle: no, don't use it here
12:44:32	Araq	oh wait
12:44:56	*	Sphax joined #nim
12:45:12	keyle	the goal here is to make a very basic irssi inspired client
12:45:35	Araq	well it doesn't work that way
12:46:00	Araq	launchClient shouldn't be .async
12:46:39	Araq	and should call runForever I guess which is not .gcsafe
12:47:06	Varriount	keyle: Async is not quite the same as spawn
12:47:29	Araq	but I need to talk with dom96 about this, currently async + spawn is simply not supported
12:48:04	keyle	yeah I admit I'm poking in the dark with this code
12:48:21	Varriount	keyle: For now, stick to async only
12:48:46	Varriount	Araq: There's also the problem that 'await' is used by both async and thread procedures. Is that intended?
12:48:51	keyle	ok, how do I make my readLine work with irc events?
12:49:48	Araq	hrm nope, asyncdispatch is already threadsafe
12:50:03	Araq	so it should work. do this:
12:50:27	Araq	don't make launchClient .async and call runForever in it
12:51:06	Araq	spawn launchClient() # note that this doesn't return anything
12:51:54	keyle	to be clear, I do runForever in launchClient or not?
12:52:07	Araq	in launchClient
12:53:10	Araq	but if it doesn't work we have to ask dom96 ;-)
12:53:43	Araq	Varriount: that's a problem I guess
12:53:51	Araq	the awaits are completely different
12:54:02	keyle	yep, it compiles fine, but I get this at start - http://pastebin.com/7e71xw4k
12:55:29	Araq	well it looks like it fails to 'connect'
12:56:31	Varriount	Araq: Did the irc module get removed from the stdlib?
12:57:00	Araq	dunno
12:58:05	Varriount	Araq: Yep, it got removed...
12:58:20	Varriount	Despite the fact that several tests and modules use it.
12:58:24	Varriount	>_>
12:59:01	renesac|away	Varriount: : I meant this type of BOOL WINAPI: http://msdn.microsoft.com/en-us/library/windows/desktop/dd627165%28v=vs.85%29.aspx
12:59:13	renesac|away	how do I translate as a nimrod type?
12:59:17	*	renesac|away is now known as renesac
13:00:26	Varriount	renesac: WinBool will work
13:00:38	Varriount	renesac: This page will help you: http://msdn.microsoft.com/en-us/library/windows/desktop/aa383751%28v=vs.85%29.aspx
13:01:18	keyle	thanks guys
13:01:37	Varriount	keyle: Sorry for the trouble.
13:01:48	renesac	right, thanks
13:02:01	Varriount	keyle: We'd be happy for any help with documentation or tutorials
13:03:13	Varriount	Araq: I'm working on the patch for PR 1896, and I have a design question
13:04:32	Varriount	The incrSeq procedure in sysstr.nim needs to somehow have the part where it modifies the actual length of the sequence disabled or deferred for a single call.
13:04:50	keyle	Varriount: thanks, I guess I'm learning as well and trying to do something non-trivial here... constantly readLine while another process outputs
13:05:30	keyle	I am having difficulty with the runtime errors though... they can be really good or really puzzling. And google is no help.
13:07:06	keyle	I think those connect fails are due to the await going on in the irc lib, that I spawned
13:07:07	Varriount	Araq: What would you prefer, that I add an argument to the incrSeq procedure to enable/disable the incrementing of the sequences length field, or remove the length increment entirely, and patch each usage of incrSeq to modify the length field right after the call?
13:08:15	Varriount	keyle: If you must use another thread, put the readLine in it's own thread, rather than the irc client
13:08:59	keyle	I'm used to GUI development, I would have thought that readline needs to be in the UI thread? ie. the terminal in this case?
13:09:15	keyle	can I readline from another thread?
13:09:27	*	BitPuffin quit (Ping timeout: 245 seconds)
13:09:41	Varriount	keyle: I don't see why not.
13:09:56	renesac	and going on the windows folder I discovered one more wrapper not listed on the libs page...
13:10:02	renesac	the psapi.nim....
13:10:10	keyle	Varriount: also, is it possible to do this without another thread? how would a readline work within an evented environment?
13:10:18	Varriount	renesac: Perhaps you could add those to the documentation index?
13:10:23	renesac	yeah
13:10:32	Varriount	keyle: Optimally, readLine would not be a blocking operation.
13:10:32	renesac	a bit frustrating because that is what I was wrapping
13:10:34	renesac	^^'
13:11:02	keyle	but it _is_, right?
13:11:05	renesac	and nimfix psapinimfix.nim --styleCheck:auto
13:11:16	renesac	does nothing to those PascalCase windows names
13:11:43	Varriount	keyle: You could poll stdin to see if a read() on it would be non-blocking
13:12:01	Varriount	keyle: However keep in mind that this is not cross-platform. It won't work on Windows.
13:12:22	keyle	ah.
13:12:48	keyle	will try spawn readline, thanks for your time
13:13:09	*	bjz quit (Ping timeout: 245 seconds)
13:14:08	Varriount	Araq: Also, do compilerProcs accept default arguments?
13:14:21	*	darkf joined #nim
13:14:33	*	darkf quit (Changing host)
13:14:34	*	darkf joined #nim
13:16:23	*	infrashortfoo quit (Read error: Connection reset by peer)
13:16:33	*	infrashortfoo joined #nim
13:20:42	*	bjz joined #nim
13:28:52	*	Demos joined #nim
13:34:52	*	Epic|gmpreussner quit (Ping timeout: 245 seconds)
13:34:59	*	Epic|gmpreussner joined #nim
13:37:16	Demos	Has anyone looked at the c++ variadic folding feature, seems crazy to me
13:37:36	Varriount	Demos: Huh?
13:37:46	EXetoC	that's a surprise
13:38:26	Demos	n4295
13:38:43	Demos	allows stuff like
13:39:11	Demos	template<typename... Args> bool f(Args... args) { return (true + ... + args); }
13:45:08	keyle	Varriount: what's poll() ?
13:48:54	Araq	Varriount: patch every usage of incrSeq
13:49:43	Araq	but yes, these can have default arguments but you need to pass every argument when it's used by the codegen
13:58:34	*	bjz quit (Ping timeout: 244 seconds)
14:05:35	*	BitPuffin joined #nim
14:08:54	*	bjz joined #nim
14:09:03	*	Demos quit (Ping timeout: 264 seconds)
14:10:08	*	BitPuffin quit (Ping timeout: 244 seconds)
14:11:34	*	kapil__ quit (Quit: Connection closed for inactivity)
14:14:52	*	z1y joined #nim
14:17:46	keyle	Is there a way to get a onkeydown instead of a readline ?
14:20:14	BlaXpirit	probably not
14:20:17	BlaXpirit	unless... curses?
14:20:28	ekarlso-	anyone that wanna help up with packages crud ?
14:20:45	keyle	I figured as much
14:22:32	keyle	afaik just importing and compiling threadpool within an app causes anything async to crash
14:22:43	ekarlso-	hmmms, noone seems to have a clue for that ;P
14:22:56	keyle	at runtime that is. No need to even use spawn.
14:24:03	keyle	night
14:24:07	*	keyle quit (Quit: Page closed)
14:34:15	*	ajhager joined #nim
14:36:24	*	gokr quit (Quit: Leaving.)
14:43:53	*	bjz quit (Ping timeout: 240 seconds)
15:01:57	*	bjz joined #nim
15:02:49	*	z1y quit (Remote host closed the connection)
15:03:47	*	z1y joined #nim
15:09:57	*	willwillson quit (Ping timeout: 265 seconds)
15:11:37	*	Sergio965 joined #nim
15:13:17	*	bjz quit (Read error: Connection reset by peer)
15:15:05	*	JinShil quit (Quit: Konversation terminated!)
15:38:00	*	bjz joined #nim
15:42:48	*	Demos joined #nim
15:52:15	*	gokr joined #nim
15:52:20	Araq	Demos: I'm about to try your VS plugin
15:52:29	Demos	uh oh
15:52:51	Araq	does it use the "server mode" of idetools?
15:53:18	*	Epic|gmpreussner is now known as gmpreussner
15:53:51	Demos	nope
15:53:55	Demos	server mode does not work
15:54:09	Demos	or it did not when I wrote it. There is code to do it if I remember though
15:55:21	Araq	well it's a nice chicken and egg
15:55:39	Araq	I can't fix the server mode since no editor supports since it doesn't work
15:57:22	*	repax quit (Ping timeout: 245 seconds)
16:02:28	*	renesac quit (Ping timeout: 246 seconds)
16:03:05	jpoirier	The compiler is okay with using using a data type as a proc parameter name, e.g. proc tst(int, uint) = echo(int); echo(uint), and, proc tst(int: char) = echo(int)
16:05:40	*	gmpreussner quit (Ping timeout: 244 seconds)
16:05:49	*	gmpreussner joined #nim
16:05:52	*	z1y quit (Ping timeout: 244 seconds)
16:11:59	onionhammer	Araq NimLime has servermode implemented but disabled iirc
16:12:22	onionhammer	(since it doesnt work:))
16:12:58	onionhammer	I would love to use a VS plugin though Demos
16:12:59	onionhammer	:)
16:13:48	Demos	The biggest thing with the VS plugin is testing and fixing the crash when you exit the IDE
16:14:26	*	bjz quit (Read error: Connection reset by peer)
16:14:35	onionhammer	and just tidying things up imo
16:14:47	onionhammer	getting real projects working, if thye dont
16:14:58	Demos	my wishlist features include a better project tree view that does not rely on a XML project file, either using some kind of IDEtools output or mirroring the filesystem, and a way to build the project that does not involve a bash script embedded in an XML file
16:16:06	Demos	real projects should work, and I am willing to help people get them to work, but the amount of time I have to devote to doing these wishlist features is very small
16:16:06	onionhammer	you should also rename it ;)
16:16:18	onionhammer	yeah
16:16:21	onionhammer	understandable
16:16:24	Demos	yeah, and make sure it works in VS 2015
16:16:25	onionhammer	also inline debugging
16:16:29	onionhammer	lol
16:16:31	Demos	hm? debugging works
16:16:37	onionhammer	i wasnt able to get it to work
16:17:58	Demos	it might have been a bug in the project templates. I really want to figure out how to deal with those deployment wise. I think I will put the project generator in nimble, that should make things easier. When I use the VS plugin to write code I add the solutions and project files to .gitignore and generate them with that tool when needed
16:18:49	onionhammer	yeah
16:18:51	*	z1y joined #nim
16:18:51	onionhammer	good idea
16:20:24	Demos	the tools is really basic at the moment (and I think it should remain so). you invoke it like vsgen mainfile.nim and it traverses the filesystem and adds all sibling and child .nim files to the project, it then sets mainfile.nim as the "startup object" and writes all the bash build script stuff
16:21:06	Demos	this is all stuff that I should be able to do in code that overrides the VS build and project systems, but the API to actually do that is a total and complete fustercluck
16:21:11	onionhammer	mm, and you cant do that without the command line tool?
16:21:19	onionhammer	oh ok
16:23:21	*	z1y quit (Ping timeout: 252 seconds)
16:27:32	*	darkf quit (Quit: Leaving)
16:29:44	*	bjz joined #nim
16:31:01	*	infrashortfoo quit (Ping timeout: 264 seconds)
16:31:34	Demos	the problem is not only that the API is really bad but also that there is little documentation
16:38:01	*	johnsoft quit (Ping timeout: 252 seconds)
16:38:17	*	johnsoft joined #nim
16:38:56	*	jefus_ joined #nim
16:39:51	onionhammer	Demos you should also test it w/ the free version of vs
16:40:07	Demos	The community edition?
16:42:46	*	bjz quit (Read error: Connection reset by peer)
16:43:01	*	jefus quit (Ping timeout: 264 seconds)
16:43:06	onionhammer	yea
16:43:12	onionhammer	its supposed to support plugins, right?
16:46:11	BlaXpirit	does anyone else find the code examples on home page cryptic?
16:46:29	BlaXpirit	i can imagine them scaring me off if i just came to that page
16:46:32	onionhammer	the parallel one yah
16:47:13	*	gokr_ quit (Ping timeout: 264 seconds)
16:48:08	*	BitPuffin joined #nim
16:50:06	Demos	onionhammer, yeah I think it is actually just a license for VS 2013 pro
16:57:35	*	gsingh93 joined #nim
16:57:39	*	Varriount|Busy joined #nim
16:58:19	*	MajorTennis quit (Quit: Leaving)
16:58:20	*	bjz joined #nim
16:59:39	Varriount|Busy	Gah. I think I'm just going to disable the timeout for testament for the buildbots
17:04:31	Araq	BlaXpirit: well the parallel one needs to outline the basic idea
17:04:39	Araq	it's not code you can run
17:04:49	BlaXpirit	i'd just like to see more simple examples of typical things
17:05:12	Araq	we have the other snippets for that
17:05:30	*	gmpreussner quit (Remote host closed the connection)
17:05:46	Araq	the parallel example is there to say "shut up, Nim can do things that Rust cannot"
17:06:03	*	gmpreussner joined #nim
17:14:14	Varriount|Busy	onionhammer: NimLime also needs to be cleaned up. (Why is there assignment aligning code? How is that related to Nim?)
17:14:35	*	Demos quit (Ping timeout: 244 seconds)
17:15:17	*	bjz quit (Ping timeout: 264 seconds)
17:17:53	*	gokr testing QtCreator...
17:18:54	*	Matthias247 joined #nim
17:19:10	BlaXpirit	gokr, how exactly......????
17:19:20	Varriount|Busy	gokr: QtCreator is nice...
17:19:43	BlaXpirit	it sure is, but how does it relate to Nim?
17:19:54	*	Demos joined #nim
17:19:58	gokr	I got syntax highlighting working using NimKate definitions
17:20:03	onionhammer	assignment aligning is awesome, thats why
17:20:13	BlaXpirit	gokr, whatever...
17:20:18	gokr	I have also read that QtCreator is good at debugging.
17:20:24	gokr	So wanted to try that.
17:20:32	gokr	And filcuc's QML stuff would be nice to test.
17:20:34	BlaXpirit	it's OK but how can you make it work with nim?
17:20:37	BlaXpirit	how?
17:20:41	gokr	Debugging?
17:20:51	BlaXpirit	let's start with that for example
17:20:52	gokr	It uses gdb AFAIK
17:20:58	onionhammer	but agree, it can be taken out of the plugin..........
17:21:07	gokr	KDevelop works fine, so I think QtCreator also will work
17:21:17	BlaXpirit	gokr, :o where is info on that?
17:22:14	*	dtscode joined #nim
17:22:14	gokr	Info on... debugging?
17:22:31	BlaXpirit	info on debugging Nim in a C++ IDE
17:22:33	BlaXpirit	,yes
17:22:55	gokr	I intended to write an article - but its nothing magical. Just compile with -debuginfo and -lineDir:on and then debug away. Works great in KDevelop.
17:23:11	gokr	I have said so here several times ;)
17:23:22	BlaXpirit	I have no idea
17:23:36	BlaXpirit	I barely was able to get KDevelop to work with C++
17:24:01	BlaXpirit	and while I'm quite confident with Qt Creator, I still have no idea how to do anything else but C++ in it
17:24:16	BlaXpirit	please do make an article...
17:24:36	onionhammer	Varriount|Busy okay, im removing it
17:24:40	*	Sphax quit (Quit: CYA!!)
17:25:18	*	dtscode quit (Quit: ZNC - http://znc.in)
17:25:49	*	dtscode joined #nim
17:25:55	gokr	BlaXpirit: I will.
17:25:57	Varriount|Busy	onionhammer: And the favorite file thing?
17:26:05	onionhammer	idk what that is
17:27:16	onionhammer	oh i think that's the license for the 'favorite file' code, which we used to set the 'main' nim file
17:27:21	onionhammer	(I think)
17:27:49	onionhammer	https://github.com/facelessuser/FavoriteFiles/
17:28:46	onionhammer	Varriount|Busy are you working on the idetools stuff?
17:29:17	*	gokr quit (Quit: Leaving.)
17:30:25	*	vendethiel quit (Quit: q+)
17:31:20	Varriount|Busy	onionhammer: No, I'm busy at work.
17:31:36	*	brson joined #nim
17:32:03	Varriount|Busy	After that, I've been tasked with getting more of the buildbot done.
17:32:21	onionhammer	the thing I think nimlime needs is probably to refactor the naming a bit in our tmLanguage file
17:32:31	onionhammer	make sure stuff goes into the correct scope
17:32:54	onionhammer	i.e. misc vs store vs entity vs keyword etc
17:33:25	onionhammer	I just cleaned up the proc highlighting stuff
17:33:27	onionhammer	yesterday
17:33:31	*	bjz joined #nim
17:33:40	onionhammer	and added highlighting for call syntax
17:35:41	onionhammer	probably highlight pragmas finally as well.. not sure how to classify them yet though, i'll probably follow C++'s plugin
17:35:55	Varriount|Busy	onionhammer: Put in some issues, so I have a reminder
17:36:59	Varriount|Busy	Araq: Where is the proc definition for the `[]` in 'addr(someString[0])'
17:38:35	Araq	Varriount|Busy: it's builtin in a way that's not visible in system.nim
17:39:20	Varriount|Busy	Araq: Oh, so there's no way to replicate it in arbitrary types?
17:39:44	Araq	Varriount|Busy: sure there is
17:40:00	Araq	proc `[]`(x: Foo: index: int): var Bar
17:40:11	Araq	proc `[]`(x: Foo; index: int): var Bar
17:41:17	Varriount|Busy	proc `[]`(ls: LinkedString, n: int): var char = ls.value[n]
17:41:26	Varriount|Busy	doesn't work
17:41:59	Araq	maybe you need 'var LinkedString'
17:43:11	*	bjz quit (Read error: Connection reset by peer)
17:43:42	ekarlso-	hmm
17:44:10	ekarlso-	if you have two procs that take a string but returns different types how does it call the right one ?
17:45:57	Araq	ekarlso-: it complains it doesn't know what to call
17:46:29	ekarlso-	Araq: how you get it to call the right one :|
17:46:31	Varriount|Busy	ekarlso-: You can resolve the ambiguity by specifying the module the desired proc is in.
17:46:34	*	willwillson joined #nim
17:47:01	ekarlso-	Varriount|Busy: it's in the same module
17:47:09	ekarlso-	like a fromRow proc
17:47:39	Varriount|Busy	ekarlso-: So... you have two procedures, named the same thing, taking the same arguments, but returning a different return type?
17:47:51	*	kniteli joined #nim
17:48:05	ekarlso-	lemme make a example instead ;P
17:51:13	ekarlso-	https://bpaste.net/show/804f39e6ae08
17:51:45	*	superfunc|busy joined #nim
17:51:59	Varriount|Busy	ekarlso-: There's no way around that ambiguity at the moment. Consider changing the procedure names
17:52:09	ekarlso-	darn :(
17:52:26	Varriount|Busy	eg: getPackageFrom(...) and getStringFrom(...)
17:52:39	ekarlso-	rowToPackage
17:52:41	ekarlso-	I guess
17:53:08	Varriount|Busy	ekarlso-: The other alternative is to use a type variant to encapsulate the two types.
17:53:27	ekarlso-	type variant ? sounds beyond the tutorials
17:56:05	Araq	don't use type variants if you don't need them
17:56:32	ekarlso-	why can't the compiler infer the type though ?
17:57:20	Araq	why can't you come up with 2 different names though?
17:58:42	ekarlso-	Araq: I did now
17:58:47	ekarlso-	packageFromRow or rowToPackage ?
17:58:49	ekarlso-	:P
18:00:03	Varriount|Busy	ekarlso-: This is type variants: https://bpaste.net/show/3e5b80c7e2e0
18:00:28	Varriount|Busy	(That example may not work exactly, I can't remember if variant fields can share names)
18:02:09	Araq	(they cannot, why would they?)
18:02:42	onionhammer	Varriount|Busy i forgot the issue i was having with ##'s, but i dont see it anymore
18:07:33	*	bjz joined #nim
18:11:48	EXetoC	not packageToRow and rowToPackage?
18:12:33	EXetoC	I misread
18:17:23	*	kniteli quit (Remote host closed the connection)
18:18:15	Varriount|Busy	onionhammer: I'm gonna try something ... interesting this afternoon that might help with the plugin organization.
18:18:47	onionhammer	oh?
18:18:56	onionhammer	(i just checked in pragma highlighting 5 secs ago)
18:19:49	*	kniteli joined #nim
18:20:10	onionhammer	Varriount|Busy what, pray tell? :)
18:21:14	Varriount|Busy	onionhammer: Using __all__ or eval()
18:21:29	onionhammer	come again?
18:22:11	Varriount|Busy	onionhammer: One is a hidden module variable that controls what is exported in a python module. The other is a function that evaluates a string as python code, in context of the current scope.
18:22:14	onionhammer	you're reorganizing the python?
18:22:43	Varriount|Busy	onionhammer: Yeah.
18:22:48	*	BitPuffin quit (Ping timeout: 265 seconds)
18:23:17	Varriount|Busy	onionhammer: We can't really move the other files. :/
18:26:14	*	AMorpork is now known as ShovelingMorpork
18:27:48	EXetoC	the python? sounds like some kind of mutant
18:28:56	BlaXpirit	the programming language..
18:30:57	*	Demos quit (Ping timeout: 252 seconds)
18:35:34	*	gmpreussner_ joined #nim
18:36:13	*	gmpreussner quit (Ping timeout: 244 seconds)
18:38:34	*	Demon_Fox joined #nim
18:42:06	*	BitPuffin joined #nim
18:45:39	*	bjz quit (Ping timeout: 264 seconds)
18:47:41	*	bjz joined #nim
18:48:09	*	superfunc|busy quit (Ping timeout: 256 seconds)
18:50:41	*	Varriount|Busy quit (Quit: Page closed)
19:00:30	*	bjz quit (Ping timeout: 265 seconds)
19:02:16	*	ShovelingMorpork is now known as AMorpork
19:09:47	*	Lordovos joined #nim
19:11:54	*	brson quit (Quit: leaving)
19:17:39	*	bjz joined #nim
19:25:13	*	repax joined #nim
19:27:48	*	renesac joined #nim
19:28:19	*	bjz quit (Read error: Connection reset by peer)
19:36:04	*	gmpreussner_ quit (Ping timeout: 245 seconds)
19:36:09	*	gmpreussner__ joined #nim
19:38:09	jpoirier	Can spawn be used as a cheap thread or is it not advisable?
19:41:04	jpoirier	...to create a cheap thread that is.
19:45:53	*	rpag joined #nim
19:46:02	*	rpag quit (Client Quit)
19:49:25	*	Guest7667 is now known as leavengood
19:49:33	*	dloss_ joined #nim
19:50:40	*	rpag joined #nim
19:51:18	Varriount	jpoirier: Spawn is used for threadpools, so in a way, yes
19:52:48	dloss_	Hi. Running "nimble install c2nim" gives me "Unsatisfied dependency: nim (>= 0.10.3)." Any hints?
19:53:49	dloss_	I'm on nimble v0.6.0 and nim installed from github
19:55:30	willwillson	dloss_: yeah, you need to use latest devel code
19:57:07	*	bjz joined #nim
19:57:19	dloss_	willwillson: git pull says I'm up-to-date. Freshly installed from 647aadc1
19:57:59	willwillson	what does nim -version say?
19:58:16	willwillson	nim --version
19:58:23	dloss_	Nim Compiler Version 0.10.2 (2015-01-09) [MacOSX: amd64]
19:58:28	*	BlaXpirit_ joined #nim
19:58:57	willwillson	there is your problem, you need to compile nim from the latest devel sources
19:59:27	willwillson	or step back a few commits from c2nim HEAD
19:59:37	*	rpag quit (Quit: Leaving)
19:59:50	dloss_	thx. will try that
20:01:37	*	BlaXpirit quit (Ping timeout: 264 seconds)
20:07:33	*	nande joined #nim
20:09:26	dloss_	using the nim devel branch helped. thx.
20:12:51	*	bjz quit (Read error: Connection reset by peer)
20:13:53	*	EXetoC quit (Ping timeout: 240 seconds)
20:14:27	*	bjz joined #nim
20:16:49	*	repax quit (Quit: repax)
20:25:35	*	billp joined #nim
20:28:45	*	bjz quit (Read error: Connection reset by peer)
20:33:24	dtscode	BlaXpirit_, was it you that wrote the article on writing jester apps?
20:33:32	BlaXpirit_	definitely not
20:33:36	BlaXpirit_	why
20:33:56	dtscode	im trying to figure out how to get jester to work with apache
20:36:19	*	gour quit (Quit: Leaving)
20:38:08	*	shodan45 joined #nim
20:38:39	*	gokr joined #nim
20:42:21	dtscode	Araq, is that build status https://github.com/Araq/Nim here current?
20:43:08	*	bjz joined #nim
20:45:42	gokr	Varriount: You need nodejs on the build slaves?
20:45:44	gokr	I presume
20:51:36	def-	dtscode: you would just run the jester app locally and on apache setup a reverse proxy
20:52:04	dtscode	def-, well as it turns out i need apache
20:52:09	dtscode	dtscode.io:5000
20:52:14	dtscode	thats just jester afaik
20:52:17	def-	yes
20:52:39	dtscode	so now i just need jester to work on port 80, and i can cut apache out of the equation
20:52:57	def-	yes, that would work
20:53:09	def-	may be insecure, but I'm not sure
20:53:27	dtscode	i think thats all the nim forums use
20:54:05	def-	I bet the nim forum is behind an nginx reverse proxy for the "forum.nim-lang.org" domain
20:54:27	dtscode	how would i setup a reverse proxy
20:54:49	def-	with apache you need a line like "ProxyPass /foo http://localhost:5000/"
20:55:31	def-	nginx should be faster, there it's called proxy_pass
20:56:02	dtscode	meh, speed i dont care about so much right now
20:56:12	dtscode	ive already got apache installed
20:56:56	ldlework	if you use Docker its easy peasy
20:56:58	ldlework	:)
20:57:05	ldlework	https://registry.hub.docker.com/u/jwilder/nginx-proxy/
20:57:06	dtscode	for a webserver?
20:57:07	ldlework	magic
20:59:26	*	bjz quit (Read error: Connection reset by peer)
21:01:19	*	bjz joined #nim
21:05:50	*	Epic|gmpreussner joined #nim
21:06:53	*	gmpreussner__ quit (Ping timeout: 264 seconds)
21:07:57	*	Demos joined #nim
21:12:46	dtscode	def-, do i need to write that in my apache2.conf file?
21:13:36	jsudlow	there you go again sharing the magical image ldlework
21:14:03	ldlework	jsudlow: worked for you right?
21:14:20	jsudlow	ldlework: It worked surprisingly well - the first time
21:14:34	def-	dtscode: yes, something like that
21:14:40	jsudlow	if your patient enough to read a few paragraphs you can have it
21:14:42	gokr	Can report that debugging in QtCreator works fine
21:16:19	*	bjz quit (Ping timeout: 252 seconds)
21:16:20	*	Matthias247 quit (Quit: Matthias247)
21:18:28	flaviu	https://gist.github.com/flaviut/e4a30004d0bb8e90fb8c
21:18:40	flaviu	Can anyone tell what "pcre.nim(10, 30) Error: invalid pragma: dynlib: pcreDll" is all about?
21:18:50	def-	flaviu: yes, i told you last time
21:19:03	def-	flaviu: in the type definitions, remove all pcreImports
21:19:45	flaviu	def-: Oh, sorry. I had gone to bed by that time.
21:19:50	def-	no problem
21:27:27	*	gokr1 joined #nim
21:27:57	*	arekzb quit (Quit: Page closed)
21:28:02	*	Kal_ joined #nim
21:29:06	*	gokr quit (Ping timeout: 244 seconds)
21:30:43	*	irrequietus joined #nim
21:33:51	*	billp left #nim (#nim)
21:38:44	gokr1	Its slick that debugging goes fine into templates etc
21:43:27	*	dyu joined #nim
21:45:15	ldlework	Is devel still broken?
21:45:31	*	Matthias247 joined #nim
21:47:45	*	bjz joined #nim
21:48:42	*	brson joined #nim
21:49:15	*	minciue quit (Quit: minciue)
21:52:15	*	Demos quit (Ping timeout: 252 seconds)
21:54:38	dtscode	devel compiles fine for me
21:55:15	ldlework	yeah
21:55:20	ldlework	dtscode: how the hell do you use the irc module
21:55:26	ldlework	is the example in the docs wrong?
21:55:37	dtscode	ldlework, no idea. i mostly tore apart nimbot
21:56:00	ldlework	sure it instantiates a PAsyncIRC directly
21:56:05	ldlework	but the docs say to use irc.irc
21:56:19	dom96	The docs are likely outdated.
21:56:19	ldlework	which I'm getting: talmud.nim(3, 16) Error: expression 'irc(IRC_NETWORK)' cannot be called
21:56:24	ldlework	ok
21:56:40	dtscode	i might make a deal with dom96 where he writes proper asyncio code for nimbus and i write eval stuff for nimbot
21:56:45	dtscode	anyways, i can show you my code
21:57:02	ldlework	I'm just checking that the docs are wrong and not me
21:57:04	dom96	sure, how much will you pay me? :P
21:57:06	ldlework	I can read nimbot
21:57:29	dtscode	dom96, ummmm
21:57:36	dtscode	how much do you cost?
21:58:09	dom96	$100 an hour :P
21:58:17	dtscode	fuck that
21:58:25	ldlework	lol
21:58:28	onionhammer	$100/hour for a student? :P
21:58:34	onionhammer	i'll give u $10/hour
21:58:45	dtscode	it would just be cheaper for me to learn the asyncnet module
21:59:06	dtscode	im broke right now anyways so its a mute point
22:00:52	*	bjz quit (Ping timeout: 255 seconds)
22:02:33	ldlework	moot*
22:02:54	dtscode	woot
22:03:14	dom96	onionhammer: deal.
22:06:06	*	gmpreussner joined #nim
22:06:11	*	adam_s joined #nim
22:07:04	*	Epic|gmpreussner quit (Ping timeout: 265 seconds)
22:07:57	ldlework	dom96: why can't I call irc.irc
22:08:28	dom96	dunno
22:08:50	dom96	what are you trying to do?
22:09:04	dtscode	ldlework, have you installed irc?
22:09:14	ldlework	dtscode: oh I thought it was stdlib
22:09:25	dtscode	ldlework, nope :/ gave me some trouble
22:09:46	dtscode	nimble install irc
22:12:55	ldlework	dtscode: I already have it installed
22:12:58	ldlework	irc.irc not callable
22:13:08	ldlework	dom96: I'm trying to run the code in the documentation
22:13:20	dtscode	ah
22:14:07	dom96	you want newIrc
22:14:15	dom96	or something like that
22:20:37	*	bjz joined #nim
22:22:48	*	jefus_ is now known as jefus
22:31:41	*	Lordovos quit (Quit: Leaving)
22:31:59	*	bjz quit (Ping timeout: 256 seconds)
22:43:44	*	dloss_ quit (Quit: Leaving)
22:45:53	*	shevy left #nim ("I'll be back ... maybe")
22:46:35	*	rpag joined #nim
22:56:49	ldlework	that worked
23:01:35	*	ajhager quit (Ping timeout: 244 seconds)
23:03:17	*	EXetoC joined #nim
23:09:35	ldlework	Does anyone know where theindex.html is generated?
23:12:55	ldlework	I can't figure it out
23:13:13	EXetoC	have yu tried t grep fr it?
23:13:23	EXetoC	damn keybard
23:14:43	ldlework	EXetoC: yeah
23:14:51	ldlework	I've found various related things
23:14:57	ldlework	but not where the actual html is generated for the index
23:15:32	EXetoC	tools/nimweb shows up when I grep for it
23:15:57	*	bjz joined #nim
23:16:37	ldlework	yes I see that too, but that's not really the actual work
23:17:16	ldlework	(it shells out back to the compiler)
23:18:47	ldlework	i have been trying to trackdown what 'buildIndex' does but not really getting lucky
23:18:52	ldlework	all I find is shit in the csources
23:19:05	EXetoC	docgen.commandBuildIndex?
23:19:10	BlaXpirit_	don't forget the possibility that it is not in repo
23:19:22	BlaXpirit_	index might be an external idx file or some stuff
23:19:48	ldlework	this is turning out to be a time sink
23:20:03	flaviu	ldlework: ack?
23:20:11	ldlework	whyyyy don't we just use a normal template system for generating html like every other project in the world
23:20:16	flaviu	--ignore-dir=nimcache --ignore-dir=csources
23:20:17	ldlework	flaviu: ack?
23:20:17	*	saml quit (Quit: Leaving)
23:20:24	flaviu	ack
23:20:25	flaviu	http://beyondgrep.com/
23:20:30	EXetoC	he was using grep already
23:20:31	ldlework	flaviu: try it?
23:20:47	EXetoC	commandBuildIndex doesn't lead you anywhere?
23:21:02	ldlework	EXetoC: and where did you find 'commandBuildIndex'?
23:21:29	EXetoC	compiler/docgen.nim
23:21:37	flaviu	ldlework: compiler/docgen.nim:636
23:21:43	ldlework	right but how did you /find that/
23:21:57	*	saml joined #nim
23:22:15	flaviu	EXetoC: Yeah, but with ack it doesn't look at .git, and with --ignore-dir=nimcache --ignore-dir=csources, it's even more useful
23:22:50	ldlework	guys do you have any idea where the actual logic of generating html actually is?
23:22:56	ldlework	where is the template for the index that this is using?
23:23:01	EXetoC	ok I never got a match in csources
23:23:14	BlaXpirit_	search for css?
23:23:29	ldlework	BlaXpirit_: what
23:23:45	BlaXpirit_	search all files for "text/css" literally
23:24:01	BlaXpirit_	ah you're talking about index specifically
23:24:23	flaviu	ldlework: config/nimdoc.cfg
23:24:39	*	bjz_ joined #nim
23:25:38	*	bjz quit (Read error: Connection reset by peer)
23:25:47	ldlework	lib/packages/docutils/rstgen.nim
23:25:50	ldlework	jesus
23:25:55	ldlework	this isn't complicated at all :)
23:26:12	ldlework	getting closer :P
23:26:19	*	saml quit (Client Quit)
23:26:42	ldlework	yay line 416
23:26:44	BlaXpirit_	ah, so many reinvented wheels
23:26:45	ldlework	thanks guys
23:26:49	ldlework	BlaXpirit_: no kiddin
23:27:06	*	juhani joined #nim
23:27:12	EXetoC	round ones or square wheels?
23:27:23	ldlework	EXetoC: they are parametric wheels at least
23:28:26	ldlework	I wonder how I will go about testing my changes
23:29:10	*	Sergio965 quit (Remote host closed the connection)
23:30:01	*	Sergio965 joined #nim
23:30:14	*	Sergio965 quit (Client Quit)
23:30:33	*	Sergio965 joined #nim
23:30:35	Araq	yeah "reinvented wheels". lets invoke some external python script instead which didn't even exist when the code was written
23:31:07	Araq	the real question here is why the docutils people never used Nim's improved RST parser and called it a day
23:31:24	Araq	it's like 20 times faster too
23:31:39	Trixar_za	Because that would be too easy and people never use the easy or obvious solution?
23:32:42	EXetoC	must have round wheels down to the molecular level
23:33:12	flaviu	Araq: There are markdown parsers written in C available.
23:33:48	Araq	flaviu: in what way is that even relevant?
23:34:28	Araq	RST has been designed for documentation of programming stuff. markdown ... well I don't even know if that's been designed in the first place
23:34:48	flaviu	Because "C library" sounds a lot better than "external python script"
23:34:55	ldlework	Araq: haha
23:35:14	ldlework	Araq: where should invoke buildIndex to get a full theindex.html
23:35:23	flaviu	It hasn't been designed by committee, but it's been designed over time by people
23:35:38	flaviu	It's supposed to be similar to plain-text email formatting.
23:35:42	Araq	by that logic everything has been "designed"
23:36:21	flaviu	I wouldn't consider that a flaw in the argument.
23:36:38	*	gmpreussner_ joined #nim
23:36:43	*	gmpreussner quit (Ping timeout: 255 seconds)
23:36:53	Araq	the only reason why everything now uses markdown is that RST is a PITA to implement and the reference impleemntation is unreadable
23:37:13	ldlework	Does anyone know where to invoke `nim buildIndex`
23:37:19	flaviu	RST isn't fun to use as a user either, IMO
23:37:33	Araq	yes, much worse than Nim's implementation, spare me, I am the one who actually *tried* to port it
23:38:14	ldlework	:(
23:38:42	Araq	ldlework: 'koch web' does that
23:38:47	ldlework	thanks
23:40:07	flaviu	Araq: Nothing wrong with just wrapping a library
23:41:02	Araq	flaviu: nothing wrong with continuing our own stuff which actually works fine already
23:41:18	*	juhani left #nim ("WeeChat 1.0.1")
23:41:38	ldlework	I think its wierd that we piece together html via string concatenation
23:41:49	flaviu	As I said, I feel like the RST syntax is inferior. IMO, it's not near as nice as markdown.
23:42:16	ldlework	nm
23:43:27	Trixar_za	It's not that Markdown is superior to RST, nor that it's easier to use. It's down to to what people are used to or familiar with. Markdown has been spoon fed to people by places like github - that's the only reason people prefer it.
23:43:47	BlaXpirit_	umm
23:44:07	BlaXpirit_	no, the reason is rst is ugly, markdown is not
23:44:11	Araq	RST is superior because it uses indentation based parsing for things like ..code-block avoiding escape hell
23:44:22	Triplefox	So do we need an RST editor or something
23:44:44	flaviu	Triplefox: That may be part of the reason, but my reasoning is that `.. image:: gnu.png` is too long and annyoing
23:44:45	BlaXpirit_	Araq, please elaborate on the last sentence
23:44:57	BlaXpirit_	what do you mean by escape hell
23:45:10	Trixar_za	BlaXpirit_: Markdown can get ugly too...
23:45:29	BlaXpirit_	maybe, but rst is always ugly
23:45:36	flaviu	Araq: GHFMD uses indentation too, but it also allows you to avoid indentation
23:45:42	flaviu	with ```lang
23:45:46	Triplefox	I don't think md would help much once you get into actually complex things like images
23:45:47	Trixar_za	Try converting a wiki page over to Markdown including tables and tell me again that Markdown is pretty.
23:45:49	*	tinAndi joined #nim
23:46:02	flaviu	Trixar_za: Tables are a pain in every markdown language.
23:46:18	Araq	BlaXpirit_: I write RST for any .txt file I write. it's not ugly by any means.
23:46:30	flaviu	Trixar_za: RST doesn't really do better: http://sphinx-doc.org/rest.html#tables
23:46:56	BlaXpirit_	Araq, but what were you saying
23:46:59	reactormonk	flaviu, except org-mode ;-)
23:47:14	BlaXpirit_	[:44:11] <Araq> RST is superior because it uses indentation based parsing for things like ..code-block avoiding escape hell
23:47:35	flaviu	reactormonk: I've never touched org-mode, but it seems nice. It's more a property of the tooling than of the table.
23:48:18	reactormonk	flaviu, yes, indeed. And I really like org-mode to just patch my experimental data somewhere in a section with :noexport: and generating the fancy graphs somewhere in a source-code block
23:49:24	*	shodan45 quit (Read error: Connection reset by peer)
23:49:32	*	shodan45_ joined #nim
23:49:43	*	shodan45_ quit (Client Quit)
23:50:52	*	shodan45 joined #nim
23:51:59	Araq	flaviu: RST has been designed so that it reads very well as text on its own as a .txt file
23:52:12	ldlework	org-mode is the amaze
23:52:20	Trixar_za	flaviu: True. However that table style is closer to the one I use out of habit.
23:52:35	Araq	it has not been optimized for saving keystrokes
23:52:46	ldlework	RST is totally fine
23:53:12	ldlework	before markdown, I remember advocating for RST all the time
23:53:20	flaviu	Araq: I have two complaints: ".." and requiring more than 3 characters for underlining.
23:53:35	flaviu	That's my only arguments.
23:53:41	flaviu	I don't want to type all that.
23:53:48	BlaXpirit_	Araq, no, no, no, rst definitely doesn't read as normal text
23:53:52	BlaXpirit_	while markdown does
23:54:02	Araq	#Header
23:54:04	Araq	Header
23:54:07	Araq	======
23:54:13	BlaXpirit_	what is this
23:54:17	Araq	case closed.
23:54:23	BlaXpirit_	which case
23:54:30	BlaXpirit_	you are not explaining anything
23:54:48	flaviu	Araq: # Header is not part of the spec, as far as I can tell.
23:55:03	BlaXpirit_	what are you all talking about
23:55:16	flaviu	RST may be technically superior, but markdown is incredibly easy to use.
23:55:22	BlaXpirit_	if you think markdown doesn't have ====== header, you are wrong
23:55:38	flaviu	BlaXpirit_: It does, but it only requires 3 underline characters.
23:55:48	BlaXpirit_	3 or more
23:55:52	flaviu	And I can use # anyway.
23:56:07	Araq	BlaXpirit_: ok, I forgot about that syntax :P
23:56:09	flaviu	BlaXpirit_: I never gave an upper bound :)
23:56:15	BlaXpirit_	"only"
23:56:20	BlaXpirit_	whatever
23:57:06	BlaXpirit_	Araq, please just actually look at some markdown documents
23:57:38	*	z1y joined #nim
23:58:09	BlaXpirit_	rst is not technically superior
23:58:12	*	bjz_ quit (Ping timeout: 245 seconds)
23:58:12	Triplefox	Well...which markdown
23:58:15	Araq	like our readme.md?
23:58:20	BlaXpirit_	most of the things that you could call ugly in markdown, are not even possible in rst
23:58:52	*	ajhager joined #nim
23:58:56	flaviu	BlaXpirit_: It is technically superior. It has an official method of creating extensions in a non-breaking way.
23:59:02	BlaXpirit_	that readme.md is definitely not written in the best practices
23:59:12	BlaXpirit_	yeah but you can't have bold links
23:59:23	BlaXpirit_	can't have bold links
23:59:30	Triplefox	I don't know how anyone could consider md for lengthy documentation
23:59:32	BlaXpirit_	or part of a link, for that matter