<< 22-09-2013 >>

---

00:20:31	*	q66 quit (Quit: Leaving)
00:24:24	*	q66 joined #nimrod
00:25:54	*	q66 quit (Remote host closed the connection)
00:29:22	*	ltbarcly joined #nimrod
00:34:01	*	ltbarcly quit (Client Quit)
00:44:14	*	ltbarcly joined #nimrod
01:17:48	*	brson quit (Quit: leaving)
01:23:02	*	Guest14108 quit (Ping timeout: 256 seconds)
01:26:38	reactormonk	how do I map over $? .map($) tells me it expects an expression, .map(`$`) Error: '$' cannot be passed to a procvar
01:26:48	reactormonk	or, how do I convert a string to a cstring?
01:37:52	*	ltbarcly quit (Quit: Computer has gone to sleep.)
02:24:31	*	ltbarcly joined #nimrod
03:00:57	*	ltbarcly quit (Quit: Computer has gone to sleep.)
03:11:14	*	OrionPK quit (Read error: Connection reset by peer)
03:21:24	*	ltbarcly joined #nimrod
04:02:05	*	ltbarcly quit (Quit: Computer has gone to sleep.)
05:08:19	*	Associat0r joined #nimrod
05:35:59	*	zhtx joined #nimrod
05:51:25	*	ponce quit (Disconnected by services)
05:51:34	*	ponce joined #nimrod
05:51:51	*	ponce is now known as Guest41546
06:04:32	*	zhtx quit (Quit: Leaving.)
06:33:09	*	zahary joined #nimrod
07:27:46	*	Associat0r quit (Quit: Associat0r)
07:43:56	*	Associat0r joined #nimrod
07:43:56	*	Associat0r quit (Changing host)
07:43:56	*	Associat0r joined #nimrod
08:15:42	*	Guest14108 joined #nimrod
09:07:41	*	Guest14108 quit (Read error: Connection reset by peer)
09:14:00	*	q66 joined #nimrod
10:02:33	*	Guest41546 is now known as p0nce
10:46:27	*	girvo joined #nimrod
10:47:02	girvo	hey all
10:48:40	dom96	hello girvo
11:00:57	girvo	out of curiosity, do many use nim on Mac? I know Aporia has problems with it
11:01:52	dom96	I wouldn't say many. But some do yes. Araq got a Mac recently so he may be able to fix the issues with Aporia on it.
11:02:20	girvo	ah cool, cheers. seems to have great Windows support?
11:02:36	girvo	trying to decide what VM i should set-up for nimrod dev :)
11:02:54	dom96	Yeah. I'd say it works best on Linux though.
11:03:28	girvo	Oh true? Neat. I couldn't get Aporia compiling on Ubuntu oddly enough, but I'll fire up a new vm and give it a shot!
11:03:31	girvo	cheers dom96 :)
11:03:59	dom96	Yeah, I heard that people are having the same issues as on Mac. Weird.
11:04:04	dom96	It works fine for me on Arch Linux.
11:04:10	girvo	yeah, pretty much. x64?
11:04:18	dom96	yes
11:04:36	girvo	huh weird. must be Canonical's fault (like every other damn problem I have with Ubuntu :P)
11:04:52	girvo	one of these days I should just deal with it and swap to arch
11:06:55	dom96	lol
11:38:17	*	girvo quit (Remote host closed the connection)
12:22:04	*	shevy joined #nimrod
12:42:40	*	OrionPK joined #nimrod
13:34:23	*	gs joined #nimrod
13:34:46	*	gs is now known as Guest58253
13:51:28	*	girvo joined #nimrod
13:51:32	girvo	hi again!
13:52:12	girvo	very quick question, nimforum: how is it setup? is there an nginx reverse proxy pointing to a nimrod-compiled app that has a http server?
13:52:28	*	zhtx joined #nimrod
13:57:42	dom96	girvo: not really. It's done using scgi.
13:58:21	girvo	dom96: ah cool. I thought about that, but I haven't touched scgi in a few years now. performance is okay?
13:59:51	dom96	Guest58253: You should really change your nick to something which isn't already taken.
14:00:36	*	io2 joined #nimrod
14:01:28	*	Guest58253 quit (Quit: Leaving)
14:01:34	*	Guest58253 joined #nimrod
14:01:35	*	Guest58253 quit (Client Quit)
14:02:13	*	gsp joined #nimrod
14:03:00	gsp	fair enough
14:20:43	*	dyu joined #nimrod
14:27:05	*	zhtx quit (Quit: Leaving.)
14:44:45	girvo	how stable is the httpserver module currently? (like, api wise)
14:45:03	girvo	the docs say unstable, so I figure thats still the case?
15:01:10	*	girvo quit (Remote host closed the connection)
15:04:55	*	sdw joined #nimrod
15:15:19	Araq	I'm back
15:15:30	Araq	but not for long ...
15:22:31	*	XAMPP quit (Read error: Connection reset by peer)
15:51:23	*	io2 quit (Read error: Connection reset by peer)
15:52:04	*	Associat0r quit (Quit: Associat0r)
15:52:31	*	io2 joined #nimrod
16:01:51	*	Associat0r joined #nimrod
16:01:51	*	Associat0r quit (Changing host)
16:01:51	*	Associat0r joined #nimrod
16:34:36	*	gsp quit (Ping timeout: 245 seconds)
17:08:51	*	filwit joined #nimrod
17:09:06	filwit	hey dom96
17:09:37	filwit	what APIs, if any, are you guys using on the forums?
17:09:42	filwit	also, who rights it?
17:10:39	filwit	i forgot, but did i hear that you wrote it? or maybe i just mis-remember that?
17:10:57	filwit	i was going to say good job to the person who wrote it :)
17:13:20	filwit	i just joined and it's much more functionally complete that D's.. it's kinda silly that D uses a newsgroup with the editing capabilities of Notepad and no after-post editing abilities.
17:52:16	dom96	hi filwit
17:53:05	dom96	Take a look at the contributors stats on github, it was me, Araq and gradha who wrote it.
17:53:30	dom96	The forum just uses jester.
17:53:38	dom96	and the compiler's rst parser
17:57:13	filwit	nice, i'm taking a look at all this, one sec
17:57:20	Araq	hi filwit, I wrote the forum and then dom96 rewrote essentially every line ;-)
17:57:22	filwit	good work then, btw
17:57:32	filwit	hi Araq
17:57:41	filwit	k, i'll look at your response
17:58:05	Araq	what?
17:58:10	Araq	what response?
17:58:20	filwit	nevermind
17:58:39	filwit	i thought you where talking about the forum post i made
17:58:54	filwit	one sec
17:59:05	Araq	already read it
17:59:21	filwit	yeah i figured, besides it's just what i showed you before
17:59:54	filwit	just wanted to see what other's said about all that, and figured the forum was better than IRC for it
18:00:24	filwit	i really like the Nimrod syntax integration in the forums, btw
18:00:42	Araq	thanks
18:00:51	Araq	coming from you I know this means something ;-)
18:01:16	Araq	you surely know about design, everybody likes our new homepage
18:01:35	filwit	glad to hear it :)
18:04:27	filwit	btw, Araq, maybe you know
18:04:53	filwit	is there a way to do this: var x {.nodecl.}: SomeType ??
18:05:12	Araq	yeah? just like the way you wrote it?
18:05:31	filwit	it caused some stack-trace error message
18:05:43	filwit	i think, i was trying a bunch of stuff
18:05:54	Araq	JS backend?
18:06:02	filwit	i will try again later, but you're saying it should work like that?
18:06:08	filwit	yes, for the JS backend
18:06:11	filwit	forgot to mention that
18:06:37	Araq	yeah that should work but it's not useful, use "importc" instead
18:07:11	Araq	otherwise the mangled symbol name is used but not declared anywhere
18:07:28	Araq	or maybe nodecl implies importc ... can't remember
18:07:55	filwit	okay, i will test it later and ask you
18:08:12	filwit	also, there's no {.forceinline.} ability anywhere is there?
18:08:36	Araq	no but there is .inline
18:08:46	filwit	i'm assuming not, since you probably just pass __inline down to the C compiler
18:08:48	Araq	for forceinline a template should do the job
18:09:02	filwit	oh right, right, dom told me that too
18:09:07	filwit	i forgot, haha
18:09:19	Araq	yeah for now, but I have a new optimizer in the works
18:09:33	Araq	well it's in the "design phase"
18:09:58	Araq	turns out that you get a global CSE for free when you do the control flow dependent typing
18:10:20	filwit	CSE?
18:12:11	filwit	jester looks pretty cool as well, dom. What does the '@' do to strings again?
18:16:07	Araq	cse = common subexpression elimination
18:18:14	dom96	filwit: it grabs a piece of form data, or the corresponding @name in the url pattern, or the value of a query string.
18:21:24	filwit	aw, i see. didn't do a full read through. Thought it was a built-in string operator
18:22:27	filwit	i was looking at Adrian's code here: https://github.com/AdrianV/nimrod-tools/blob/master/oopExample2.nim
18:22:34	filwit	and it looks pretty cool
18:22:51	filwit	that's the sort of thing i was trying to do before, but couldn't get it to work
18:23:39	filwit	is it possible to do something similar to his 'declClass' macro, but for procs ?
18:24:09	filwit	like... declFunc foo(bar:int): ...
18:24:21	filwit	?
18:25:18	Araq	kind of yeah
18:26:51	Araq	declFunc do (bar: int) -> gah:
18:26:53	Araq	...
18:28:04	Araq	I dislike that we don't have 'specialRoutine foo(bar: int) = ...' but it's hard to do with nimrod's parsing technology
18:34:53	filwit	yeah that would be interesting
18:35:06	filwit	so could i do something like:
18:35:23	filwit	declType Foo from Bar: ...
18:35:51	filwit	declFunc baz gives int: ...
18:36:12	filwit	where 'from' and 'gives' are the same as your '->' operator?
18:36:49	Araq	no
18:37:05	filwit	hmm, okay
18:37:05	Araq	'do' is a keyword introducing a lambda
18:37:27	filwit	'do'?
18:37:34	filwit	you mean the ->?
18:37:45	Araq	no I mean 'do'
18:38:29	filwit	okay. well i will take a deeper look at it later
18:45:02	Araq	filwit: what time is it for you?
18:46:07	filwit	2:45
18:46:21	filwit	well, 14:45
18:47:33	Araq	hmm that's why it all feels wrong
18:57:04	dom96	It's called jetlag :P
19:20:02	*	algoban joined #nimrod
19:20:06	*	algoban left #nimrod (#nimrod)
19:29:25	Araq	filwit: you never answered how 'var Foo' would be handled with your scheme
19:29:38	Araq	proc var Foo.abc(x: int) ?
19:29:42	Araq	for
19:29:52	Araq	proc abc(this: var Foo, x: int)
19:57:57	*	dyu quit (Quit: Leaving)
19:59:30	filwit	sorry, was gone for a sec
20:00:15	filwit	and yes, just like that i think
20:00:23	filwit	unless there's some issue with that
20:01:21	filwit	or.. hmm.. wouldn't it default to that?
20:01:26	filwit	so:
20:01:42	Araq	it would work but it's ugly imo
20:01:43	filwit	proc Foo.abc()
20:01:44	filwit	for
20:01:51	filwit	proc abc(this: var Foo)
20:02:02	Araq	meh
20:02:04	filwit	since you normally want that
20:02:15	Araq	no
20:02:18	Araq	if you have
20:02:25	Araq	type Foo = ref object
20:02:28	Araq	you don't
20:02:36	filwit	ah, right right
20:02:51	filwit	yeah, so just the 'proc var Foo()' then
20:03:11	filwit	but hey, if there's some way i could define my own through macros, i am not care
20:03:33	filwit	like that: declFunc foo(): ..
20:03:48	filwit	if that can work like i want, then i'll work around that
20:04:09	filwit	if not, i'm not gunna try and fight for a feature no one wants
20:04:12	Araq	there needs to be a 'do' in there though
20:04:32	filwit	ahh.. i see what you meant now
20:04:40	filwit	so '->' == 'do'?
20:04:49	Araq	but I'd just use the declClass or something macro
20:05:04	Araq	that does want you want even more nicely anyway
20:05:10	dom96	I think it's better if we have a 'class' macro or something: class (x: Foo): proc blah() = ... # we get the x param added to our proc's params
20:05:24	Araq	dom96: well have that
20:05:32	Araq	it's not in the stdlib though
20:05:33	filwit	yes, i agree with dom96
20:05:59	filwit	but i have a couple of ideas and questions about what's possible
20:06:09	filwit	which will take some time for me to digg through things again
20:06:20	filwit	and i don't really have the time to do that just yet
20:06:41	Araq	good I don't feel like thinking anyway
20:06:51	Araq	my stomach seeks to explode for some reason
20:07:13	filwit	ate too much :)
20:07:17	filwit	are you back in Germany?
20:08:05	filwit	oh, one question though. is there a way to make a "static object"?
20:08:17	filwit	type Foo = global object ...
20:08:25	filwit	so then i can just use 'Foo'
20:08:45	*	Guest_1134 joined #nimrod
20:08:46	filwit	without needed to do: type TFoo = object ... var Foo: TFoo
20:09:13	*	Guest_1134 quit (Client Quit)
20:09:22	filwit	just wondering
20:10:49	dom96	I don't think so.
20:11:20	Araq	no and it's misguided
20:11:26	dom96	What would you use it for?
20:11:35	filwit	sanity :P
20:11:54	Araq	objects use nominal typing so ... the type name is important
20:12:14	*	dom96 was looking for an example
20:13:03	filwit	yeah, Araq.. i think something to 'grab' the TypeInfo of a def is a better system
20:13:19	filwit	typeof(instance).name # const data
20:13:31	filwit	type of thing
20:13:45	filwit	and it's not so rigid i think
20:13:46	Araq	type(instance).name should do exactly that
20:14:08	Araq	but then you prefer to write type(Foo) everywhere instead of TFoo ?
20:14:18	filwit	right, so why does there need to be a unique name for the Type of a global instance?
20:14:18	Araq	you're a strange guy ... :P
20:14:35	filwit	yes, i prefer write type(Foo) everywhere
20:14:54	filwit	i think it helps with sanity on large code bases
20:15:04	filwit	less symbols to remember
20:15:41	Araq	'T' is a common prefix though, no need to "remember" it if you *use* the language
20:15:51	filwit	so i read: type Foo = gobal object.. and I know there will never be but one instance of it, and how to reference the type data from it
20:16:07	Araq	how so?
20:16:14	Araq	var otherFoo: type(Foo)
20:16:29	filwit	true, it's a common prefix, but it helps developers to understand structure if they see how the type is used, at definition
20:17:05	filwit	good point
20:17:38	filwit	still prefer it for naming, but that's just a personal preference i guess
20:17:44	Araq	devs are more helped by a proper CS education instead of your constant fight to care the biggest dumbass around
20:17:51	Araq	*to care for
20:18:16	filwit	there are reasons verbose syntax help sanity i think
20:19:20	Araq	and yet you propose a shortcut for singletons
20:19:24	filwit	it helps you focus on important data, how often is type(Foo) really used?
20:19:39	Araq	it could be everywhere
20:19:52	Araq	proc a(this: var TFoo) ...
20:19:57	dom96	i'm confused, are we still discussing static objects or 'T' prefixes now.
20:20:03	dom96	?
20:20:15	Araq	dom96: both I think
20:20:35	filwit	they are related
20:22:41	Araq	filwit: lets flip this:
20:23:02	Araq	if a guy comes along and says "waaah, I can't remember the T prefix for types"
20:23:10	Araq	would really like to work with him?
20:23:27	Araq	*would you
20:23:53	filwit	that has little objective relevance to the pros or cons of my arguments
20:23:57	Araq	chances are high he can't script your cool game anyway, right?
20:24:22	filwit	that's a horrible outlook, lol
20:25:02	Araq	but it's relevant in the sense that I think you think way too much about trivialities
20:25:13	dom96	I guess you could think of a module as a static object, if I understand static objects correctly.
20:25:24	Araq	dom96: yes that's correct
20:25:41	filwit	we are talking about language syntax design, Araq.. it's all trivial details at this level
20:25:49	*	Associat0r quit (Quit: Associat0r)
20:26:11	Araq	no. there are objectively bad syntaxes simply because they are ambiguous or very hard to parse
20:26:49	filwit	yes, true.
20:27:11	filwit	but i don't see "it's trivial to me" as a convincing argument
20:27:31	filwit	i'm not sitting here saying "you have to use do it or NOONE will use the language"
20:27:37	Araq	sure
20:28:19	Araq	but then you're the only guy that constantly comes up with strange things like that ;-) ok, daddye does too
20:28:59	Araq	the real thing is: you're not grateful :P
20:29:13	filwit	haha
20:29:20	Araq	program in Pascal and Ada for a few years and then tell me Nimrod's syntax isn't pure joy
20:29:27	filwit	nah, i just like language design, even the petty things about it
20:29:53	filwit	and i would never program in Pascal or Ada, nor am I claiming Nimrod is comparable to those languages
20:30:44	filwit	but I'm used to features from the C#/D languages, and I think some have valid, objective arguments in there favor (OOP-style i mean)
20:31:17	filwit	so i raise those arguments here sometimes... really it was just my question that sparked this argument anyways, lol
20:31:40	filwit	you're the one who originally claimed it was misguided thinking
20:31:56	Araq	knowing the history of the C#/D syntax that's a hard sell here ...
20:32:08	filwit	sorry, that sounded more accusatory than i meant it
20:32:47	filwit	well i wasn't advocating C#/D get everything right, there are some really nice features of Nimrod, even in the syntax, lol
20:33:30	filwit	i don't mind the Python style (expect the object decl syntax), and I like the macro/template/proc definitions and stuff
20:33:37	filwit	more than C#/D i mean
20:34:19	filwit	the only really issue i have with Nimrod, is categorizing things bye objects, like i'm used to in those languages
20:34:31	filwit	the line between types and modules is very blurred in Nimrod
20:34:39	Araq	lol
20:34:46	Araq	in fact it's the opposite
20:34:48	filwit	and while that's good, some people like thinking that way. it's also confusing to some
20:35:11	Araq	it's 'class' that completely blurs things
20:35:41	filwit	well, actually, you might be right about that, and it brings me to another question i had
20:35:54	filwit	can i do this: Console.log("...")
20:36:01	Araq	I know it's useful for you but *objectively* it really does, go read a book about it ;-)
20:36:03	filwit	where 'Console' is module?
20:36:09	dom96	yes
20:36:09	Araq	yeah
20:36:39	dom96	I still would like to see an example of where static classes really are useful.
20:36:42	*	dom96 is curious
20:36:47	filwit	so basically i don't need to use types for global objects? i can just use modules
20:37:26	dom96	I guess in C# you can't just have global functions, you need an enclosing class. So you would create a static class.
20:37:30	filwit	also, what i meant bye "Console.log()" is that i don't import it where "log" is available without the prefix
20:37:34	filwit	is that possible?
20:37:59	filwit	so basically you could either "import Console" then just use 'log', or you could just use "Console.log".. ?
20:38:06	filwit	guess i should just try it out...
20:38:08	filwit	one sec
20:38:43	dom96	no, you need to import it.
20:38:53	dom96	Same way you need a reference to something in C#.
20:39:01	dom96	(and a 'use'?)
20:39:21	dom96	hrm, nah I guess the 'use' is not needed.
20:41:02	filwit	that's the problem with modules then
20:41:07	filwit	compared to static objects
20:41:28	filwit	cause i don't want to "import" everything inside of Console into my module
20:41:42	dom96	You can do: from console import nil
20:41:57	filwit	hmm... that's an interesting idea
20:42:28	filwit	maybe I could make a template that makes that easier.. using Console
20:42:44	filwit	where 'using' is template that just spits out: from Console import nil
20:42:49	dom96	But I don't see why you would, unless you want to get rid of a lot of ambiguities
20:44:27	filwit	^ yes that
20:44:46	Araq	filwit: you need the import; ultimately the IDE can produce it for you though
20:44:47	filwit	and so i can think of modules like static objects... at least to a degree...
20:45:09	filwit	they still can't have hierarchy like static objects can
20:45:43	dom96	what if you put a global var in your module? :P
20:45:58	filwit	Araq: it's not about typing the import at the top, it's about not being able to use 'log' without the prefix
20:46:12	dom96	hrm, nah, that's probably a bad idea.
20:46:21	Araq	dom96: I'm making the compiler 'guess' the package name from the directory the module resides in
20:46:26	dom96	filwit: why is that so important to you?
20:46:28	filwit	dom96: i don't follow
20:46:40	dom96	Araq: what?
20:46:56	Araq	like jester/src/foo.nim --> jester_foo.c
20:47:31	Araq	the rule I have in mind is: parentDir(filename) unless that is in ["src", "source"
20:47:39	dom96	oh, right. Well. 'guess' sounds risky.
20:47:45	Araq	, "package", "pckg", "lib", "library" ]
20:47:52	Araq	then we go further up
20:48:11	dom96	I think it should go from the directory that is in Nimrod's PATH
20:48:32	Araq	I don't see how that changes anything
20:49:14	dom96	why are you ignoring the 'src' etc?
20:49:22	Araq	and a heuristic seems to be good enough, we only need the prefix for disambiguation
20:49:26	Araq	because
20:49:37	Araq	jester/src/utils.nim
20:49:46	Araq	fooPackage/src/utils.nim
20:49:49	Araq	might be common
20:50:43	Araq	dunno what Babel says about this
20:50:47	dom96	huh
20:51:07	dom96	What if we have: jester/src/utils.nim and jester/source/utils.nim?
20:51:15	Araq	2 jesters?
20:51:35	Araq	Babel can't handle it this way
20:51:39	Araq	it would be
20:51:43	Araq	jester-1.0/src
20:51:50	Araq	jester-1.1/source
20:52:03	Araq	and then it's fine again :P
20:52:14	dom96	I don't see the problem in just making it: jester_src_utils.c
20:52:34	Araq	max command line length is 4K on linux iirc
20:52:41	Araq	so long names suck
20:52:55	dom96	well, don't assume that people are going to be using babel
20:53:18	filwit	okay, great, i guess i can live with this, haha
20:53:22	dom96	4k should be enough
20:53:36	filwit	template using(e:expr) = from e import nil
20:53:41	filwit	works out very well :)
20:54:39	filwit	is there a way to make every compiled file implicitly import another nim right?
20:54:43	Araq	yup, it's superb, making me write Console.write like I'm some idiot
20:54:58	Araq	I appreciate these things
20:55:10	Araq	;-)
20:55:21	filwit	haha, well it's good for script kiddies dude
20:55:27	Araq	so you say
20:55:36	filwit	and if you write a game engine, you want to make limitations that way
20:55:46	filwit	and if you dont like it.. just use 'import Console"
20:55:49	filwit	fixed
20:56:01	filwit	so both are still supported
20:56:03	filwit	i like it
20:56:28	Araq	why? what's wrong with using your brain instead of trying to switch it off during coding? and try to rely on some minor things that don't guarantee anything?
20:56:34	dom96	yay, Nimrod will become C#
20:56:38	filwit	i would actually advocate to make a built-in 'using' template to the object.nim
20:56:59	*	Araq doesn't get it
20:57:15	filwit	that's not an argument Araq, it's all just ad hominem
20:57:36	Araq	yeah it's 'log' and that's still the Console.log. The *opposite* is perhaps confusing. Like having 'log' and it finally means something differentt
20:57:39	dom96	filwit: why do you prefer to waste hours of your life typing 'console'?
20:57:59	filwit	the point is to limit the brain power required so the programmer can focus on other tasks
20:58:04	Araq	filwit: I don't mean it personal
20:58:16	Araq	like what other tasks?
20:58:21	filwit	dome96: yes, i prefer to type Console.write than echo for large projects
20:58:32	dom96	huh?
20:58:37	Araq	like what can I do apart from *reading*?
20:58:47	dom96	If anything you need to use more brain power to remember the extra piece of info, which is 'console'.
20:58:48	Araq	play some game at the same time?
20:59:11	filwit	Araq: reading other's code in a large code base is very important
20:59:39	Araq	reading 'Console.' all over place doesn't help with reading
20:59:57	filwit	dom96: no, because then you catagorize everything under 'console'. there's less symbols to remember, especially with inteli-sense
21:00:11	dom96	ahh, so the point is to be able to find the function quickly. Because you know it's part of some 'Console' class.
21:00:33	filwit	Araq: Console is just an example. the point is to use if for other things, like Physics.rayCast(...)
21:00:39	Araq	and nobody actually really *reads* code, it's too large anyway, instead you skim around and fix what's wrong
21:00:58	Araq	in my experience anyway
21:01:11	Araq	but sure
21:01:23	Araq	go ahead and read the 100K of Quake's engine
21:01:24	filwit	Araq: that's my point.. you want to read exactly what the structure is rather than constantly guess where the procs are coming from and what they effect in global space
21:02:12	filwit	Quakes engine is very old, so IDK if you're saying it's good code or bad, but i'm sure things have changed in game-dev practice since then
21:02:40	filwit	i also have never read the Quake code, so idk what it's like
21:03:12	filwit	you know guys.. this is really hard to talk to both you over text..
21:03:18	Araq	it was just an example
21:03:22	filwit	we should just talk sometime over skype
21:03:38	filwit	i think we can do conference calls there right?
21:03:49	Araq	sure thing
21:04:05	Araq	but it's bad for the people who read logs
21:04:34	Araq	filwit: the real problem that I have is that you optimize for the *casual* reader
21:04:41	filwit	yes, but it's hard to give eachother time to be clear, and talk to multiple people at the same time
21:04:57	Araq	not for the guy who already *knows* quite a bit about your code base
21:05:06	filwit	Araq: i just want the *ability* to optimize for the casual user
21:05:25	filwit	Araq: and i'm glad to see that my using template makes that much easier
21:05:30	Araq	yeah but no need to argue for that, nimrod in this case gives you want
21:05:40	Araq	*gives you want you want
21:05:43	filwit	Araq: i really would suggest adding that to the object.nim as an option for people
21:06:12	filwit	Araq: the 'using' template i mean
21:07:07	Araq	well zahary doesn't like it, we wants to make 'using' a keyword for Pascal-like 'with' functionality
21:07:50	filwit	yes you told me that
21:07:50	Araq	and I don't like it because it encourages pettiness which I increasingly cannot stand anymore
21:08:04	filwit	if that's actually going to happen, i would argue 'with' is a better name for it
21:08:42	Araq	in the near future I will demand "show me in a very concrete way how it avoids bugs" ;-)
21:08:54	filwit	'pettiness' is just another way of saying "things i dont like that arn't major features" lol
21:09:32	filwit	i see, well as long as i can make my own modules that are included everywhere, i would be happy
21:09:36	filwit	is that possible?
21:09:46	Araq	kind of yes
21:10:01	Araq	make an "import inlcude" and include that
21:10:02	filwit	so something like: nimrod c -implictInclude:global.nim main.nim
21:10:13	Araq	yeah we have taht too iirc
21:10:22	Araq	never used it so I don't remember its name
21:10:33	filwit	okay, i will look it up and try it
21:12:13	Araq	dom96: I'm wrong max command line is more like 2MB these days
21:12:17	filwit	what's the difference between --import and --include again?
21:12:29	Araq	--import doesn't help you in this case
21:12:44	Araq	import imports, include includes
21:12:55	Araq	yeah I know that's not a useful explanation
21:13:06	Araq	include is like C's #include
21:13:19	Araq	and import like a module system that C lacks
21:14:03	filwit	well import imports exposed things right?
21:14:11	Araq	right
21:14:25	dom96	Araq: there ya go
21:14:31	filwit	yeah, okay so --import works just fine
21:14:32	filwit	great
21:14:40	filwit	awesome, i can live happy with this :)
21:14:43	dom96	Araq: You know, you still have not tested babel :P
21:14:57	Araq	dom96: babel doesn't work
21:15:07	Araq	it can't work
21:15:16	dom96	what!? why?
21:15:17	Araq	everybody now has an "utils" module
21:15:30	Araq	I need to fix the compiler to make it work
21:17:23	dom96	you can still test it :P
21:17:30	Araq	dom96: but jester_src_utils.c is much uglier and what's worse I only have absolute paths here
21:18:08	Araq	so it would be usr_ugly_unix_babel_jester-1.0_src_utils.c
21:18:20	dom96	that's why I told you
21:18:27	dom96	Base it on the PATH
21:18:30	dom96	that the compiler gets
21:18:49	Araq	what makes you think the PATH is not absolute?
21:19:14	dom96	ok, let me explain.
21:22:54	dom96	So the PATH to the jester package is: ~/.babel/pkgs/jester-0.1.0/, I import "jester/private/utils.nim". I guess you could just use the "jester/private/utils.nim"?
21:23:09	dom96	So jester_private_utils.c
21:24:01	Araq	ok but this doesn't work if you use --path:jester/private and import utils but I guess packages can't do that
21:24:29	Araq	but then I can't assume Babel is used either
21:24:41	Araq	so ... this doesn't work
21:24:54	dom96	I guess you can't trivially check whether the path will conflict.
21:25:09	Araq	and apart from that I don't store the "import path" anywhere anyway
21:25:21	Araq	so that would be lots of work
21:25:58	*	q66_ joined #nimrod
21:26:00	Araq	lets just try my way
21:26:32	*	q66 quit (Read error: Operation timed out)
21:26:39	dom96	also, why not just create directories :P
21:26:43	filwit	so wait, back to my declFunc thing.. could it work like this: on update do: ...
21:26:51	filwit	where 'on' and 'do' are macros
21:26:52	dom96	in the nimcache dir?
21:27:04	filwit	'update' is the name of the proc or whatever
21:28:10	filwit	i'm thinking about how you could apply it for a game engine script areas, where every nimrod file under the 'scripts' folder had implicit 'on' procs to handle game-engine callbacks (like Unity3D).. and that syntax would actually be really nice
21:28:33	dom96	Araq: Please think about it carefully, I fear these rules will end up with cases where the paths clash.
21:29:33	Araq	dom96: lots of directories in nimcache/ are annoying for debugging generated C code
21:29:47	Araq	in general directories are annoying
21:29:57	Araq	hierarchies are
21:30:36	dom96	I think a good file browser can help with that
21:30:42	Araq	you want a banana but you have to write jungle.fruits.banana
21:30:48	Araq	lol yeah
21:30:56	Araq	a "good" file browser
21:31:00	Araq	where do I find that?
21:31:07	Araq	and how can I make Aporia use it?
21:31:17	*	dom96 adds another project to his todo list
21:31:27	Araq	ever used google btw?
21:31:40	Araq	does it make you write jungle.fruits.banana?
21:31:52	dom96	The thing is, you're turning hierarchies into single files.
21:31:55	Araq	ever wondered why it doesn't? now you know :P
21:32:07	dom96	it will certainly annoy people :P
21:32:20	dom96	To be honest I don't care that much.
21:32:25	Araq	good
21:32:29	dom96	as long as it works
21:33:40	Araq	I wouldn't mind banana.fruits.jungle that much though because that's in the order how the brain works
21:34:05	Araq	forum.nimrod-code.org # good design
21:34:05	dom96	hrm, perhaps you should add the path to the nim file which was compiled to the .c sources
21:40:07	Araq	what do you mean?
21:41:19	filwit	https://gist.github.com/PhilipWitte/6664132
21:41:23	dom96	In the generated .c files: "Generated from: path/blah/file.nim"
21:41:24	filwit	any way to get that to work?
21:41:37	filwit	or something similar
21:42:13	dom96	argh, github. Why do you mess up the indentation!?
21:42:29	dom96	Do I need to write my own pastebin now too?
21:43:09	filwit	gist has always worked well for me
21:47:30	Araq	you can make it work with a macro for this case I think
21:47:41	Araq	problem is that (a: b) is a tuple construction
21:47:53	Araq	and you want to turn it into a param declaration
21:48:15	filwit	well i just want it to work, i don't care if it's a expr or tuple
21:48:19	filwit	let me try with tuple
21:48:32	Araq	you don't understand I think
21:48:53	filwit	yeah i see
21:49:01	filwit	cause it's not the same thing
21:51:38	Araq	onUpdate do (y: int): ...
21:51:47	Araq	that's what we had in mind
21:52:11	Araq	but that passes a lambda to 'onUpdate'
21:53:37	Araq	you could try a source code filter
21:54:42	filwit	is passing a lamba bad for performance?
21:55:24	Araq	it's not too bad but also not too good
21:55:32	filwit	crap
21:55:39	filwit	cause i would be fine with the syntax
21:55:45	filwit	but i don't want any performance hit
21:56:08	filwit	hmm.. you know what, it might not matter maybe
21:57:12	Araq	er
21:57:22	filwit	but i'll figure it out later i guess.
21:57:31	filwit	i need to take off, actually
21:57:43	Araq	wait a sec
21:57:46	filwit	okay
21:58:01	Araq	#! replace(sub="on ", by = "proc ")
21:58:15	Araq	try that in some.tmpl file please
21:58:21	Araq	and include some.tmpl
21:58:33	filwit	well that's actually not what i'm trying to do. but that's interesting
21:58:45	Araq	and in some.tmpl you can then use 'on ' instead of 'proc '
21:58:57	Araq	but yeah likely not what you want
21:59:34	Araq	you can easily make 'onUpdate' a macro though and eliminate the lambda this way
21:59:39	filwit	i don't want just another name for proc, i want to do some bookkeeping on when and where that proc is used... (wire it up to the game engine, for example)
21:59:51	Araq	yeah that's what I thought
22:01:10	filwit	for the most part, things don't need incoming parameters... so just: on update: ... would be fine
22:01:29	filwit	but sometimes, like with collisions, you want: on collision(c:CollisionPoint): ...
22:02:15	filwit	i'm thinking though, that i might be able to just pre-define these 'CollisionPoint' params
22:03:22	filwit	so.. template on(name:expr, body:stmt) = when name == collision: proc name = var c: CollisionPoint .. body
22:03:26	filwit	something like that
22:03:38	filwit	kinda hard to understand strung out like that
22:04:09	*	comex quit (Ping timeout: 252 seconds)
22:07:27	filwit	hmm... how do i do the "when name == 'collide':" line exactly?
22:07:48	*	comex joined #nimrod
22:07:55	Araq	when astToStr(name) == "collide" I guess
22:08:45	Araq	## This is often the same as the ``head`` result of ``splitPath``.
22:08:47	Araq	## If there is no parent, ``path`` is returned.
22:08:52	filwit	k thanks i'll try
22:08:58	Araq	from 'parentDir's docs
22:09:04	Araq	gah, that is nastry
22:09:07	Araq	*nasty
22:09:12	Araq	it should return "" otherwise
22:09:30	Araq	while (let parent = parentDir(x); parent != ""): ...
22:10:15	dom96	just make it 'parent != x'?
22:10:28	Araq	but that's stupid, no other API behaves this way
22:10:33	dom96	it should raise an error
22:10:39	Araq	no way
22:10:43	dom96	I swear we've discussed this before.
22:10:45	Araq	look at my example code
22:10:49	filwit	lol
22:10:56	Araq	yes we did
22:11:01	Araq	but I didn't push it enough
22:11:13	Araq	and now gradha went on and made it worse
22:11:16	dom96	it seems as if we never arrive at a solution
22:11:27	filwit	astToStr(name) == "collide" is not working :-/
22:11:37	Araq	why not?
22:12:00	dom96	somebody must have pulled his pull request *cough*
22:12:02	filwit	https://gist.github.com/PhilipWitte/6664132
22:12:20	filwit	says it can't find 'x' in the 'on collide:'
22:12:41	Araq	welcome to the world of hygienic macros
22:12:57	Araq	you need 'var x {.inject.} = 324
22:13:09	filwit	ahh, okay
22:13:16	dom96	Well as I said before, I would rather get a good runtime error. Than for some code to later fail because parentDir returns ""
22:13:24	Araq	or just make the template a dirty, immediate
22:13:26	filwit	i thought when didn't introduce scope like D's static if
22:13:32	Araq	it doesn't
22:13:38	Araq	just like in D
22:13:49	Araq	it has nothing to do with 'when' but with 'template'
22:13:51	*	io2 quit ()
22:14:06	Araq	just make the template a {.immediate, dirty.}
22:14:15	Araq	that's what everybody else does anyway ;-)
22:14:21	filwit	okay
22:14:21	Araq	and then it should work
22:14:55	Araq	dom96: but but but ... I like to loop without try ... except
22:15:22	Araq	and in any case, returning path unchanged is the worst of all worlds
22:15:30	Araq	return "" or raise
22:15:44	Araq	but don't just return the old path. when is THAT useful?
22:16:11	dom96	does 'normalise' return "" when there is nothing to normalise? :P
22:16:36	Araq	normalize cannot fail so it has nothing to do with the matter
22:16:42	filwit	well it fixed one problem, but not the other
22:16:44	filwit	one sec
22:18:51	filwit	https://gist.github.com/PhilipWitte/6664132
22:19:30	Araq	get rid of the 'params'
22:19:49	Araq	and of the 'inject' too
22:20:08	filwit	that causes a different problem, one sec
22:20:52	dom96	Araq: I don't get what is so bad about returning the path unchanged.
22:21:01	filwit	updated, just refresh the gist page
22:21:29	Araq	1. get rid of params
22:21:35	Araq	2. invoke like:
22:21:37	Araq	update()
22:21:57	filwit	oh, duh, completely missed the params..
22:22:17	filwit	great, it worked
22:22:18	filwit	thanks
22:22:53	Araq	dom96: oh well, I don't care
22:23:02	filwit	awesome.. so, that's not bad syntax for a script engine i think
22:23:20	filwit	just as easy as Unity3D's, and you have much more power if you want
22:23:37	Araq	wow
22:23:45	Araq	can't believe you like it :P
22:24:25	Araq	finally you and Nimrod are getting warm ;-)
22:25:00	dom96	they're getting what now?
22:25:02	filwit	i'm trying to narrow down my options. I'll probably be working on my current app (the one i'm going to port to Nimrod) for another year or more, depending on how the marketers want to evolve it
22:25:30	filwit	so i have a lot of animation rendering stuff i've made in JS for it, and I would like to structure that into Nimrod in a more friendly way
22:26:09	Araq	dom96: I guess you can't say that in English this way :P
22:26:18	Araq	filwit: nice
22:26:22	filwit	and i'm thinking, since i'll need a HTML5 canvas backend & an OpenGL backend, might be nice to start a little Nimrod 2D game sdk that targets both
22:26:39	Araq	excellent, I have to sleep now
22:26:46	Araq	good night
22:26:47	filwit	alright, night
22:27:30	dom96	lol, good night
22:29:31	dom96	filwit: You must make sure to convert some people on the D forums to the Nimrod side :P
22:31:45	filwit	the connection might be worth something, but i don't actually post much on the D forums
22:32:15	filwit	and i may end up arguing against some of Nimrod's designs, lol.. but then, I'm liking that i can do some things like what i just figured out
22:32:21	dom96	ok, then you need to earn their trust first.
22:32:41	dom96	And then convert them from the inside :P
22:32:47	filwit	maybe with a little game sdk example of how to use Nimrod like i want, more people there will agree
22:33:03	filwit	yes, haha, infiltration
22:33:15	Araq	you know
22:33:27	Araq	they don't come here and try to convert us either
22:33:38	Araq	that's not polite
22:33:48	Araq	*would be
22:33:53	dom96	Araq: shhhhh, you should be sleeping.
22:34:06	filwit	dom just has more ambition then they do :P
22:34:45	dom96	indeed
22:35:15	dom96	I remember a certain someone encouraging people to convert us on the D forums :P
22:35:29	Araq	well I guess they might try it if we had a decently sized community ...
22:35:49	dom96	yea-- ... :(
22:35:51	Araq	what?
22:36:04	Araq	do you mean me?
22:36:11	dom96	no
22:36:25	filwit	i'm sure things like presentations at StrangeLoop conferences and better documentation will help the community grow
22:36:26	dom96	Andrei or whatchma call him
22:37:00	dom96	oh yeah, in fact. We need to get to work on writing blog articles.
22:37:13	filwit	yes ^ that too
22:37:22	filwit	gotta advertise a bit more i think
22:37:41	filwit	keep new content coming to the web-page and get some "google juice"
22:37:59	dom96	This one really gathered a fair bit of attention: http://forum.nimrod-code.org/t/220
22:38:10	dom96	if we could sustain the momentum with articles like that
22:38:22	dom96	it would be great.
22:39:36	Araq	I will release my slides this week
22:39:54	Araq	with an article that comes with it
22:40:02	dom96	good.
22:40:18	dom96	When will the video be public?
22:40:21	filwit	make sure to talk about how much you love kittens... the internet loves kittens
22:40:31	Araq	no idea, dom96
22:40:47	Araq	filwit: well I don't
22:41:15	Araq	but you're right
22:41:20	dom96	puppies ftw
22:41:48	Araq	pussies?
22:42:36	filwit	haha, that might detour all the current "women in tech" activist groups if you write about that..
22:42:45	Araq	lol yeah, will talk about how much I love pussies instead
22:42:58	filwit	LOL
22:43:25	filwit	Nimrod tech talk with a side of pornographic oversharing
22:45:12	Araq	well here in germany we have a different problem
22:45:26	Araq	too few women are into waste management
22:45:33	Araq	and yet nobody talks about it!
22:45:44	dom96	haha
22:45:49	filwit	lol wut..
22:46:08	Araq	I call it a big discrimination
22:46:22	filwit	oh, haha
22:46:39	Araq	we need to do something about it
22:46:52	Araq	some free day where girls can practice to collect waste
22:47:36	filwit	i thought i knew what you where talking about.. but now i'm back to being confused ;P
22:47:41	dom96	oh yeah, I could use a maid.
22:47:50	filwit	^
22:48:22	Araq	filwit: I think you understand me
22:48:43	Araq	just forget the last sentence if it confuses you
22:49:35	filwit	okay folks, i'm taking off
22:49:49	dom96	bye filwit!
22:49:55	filwit	good to talk with you both again, i will be around more
22:49:58	filwit	bye
22:50:00	*	filwit quit (Quit: Leaving)
22:50:54	dom96	well I might as well go to sleep then