<< 21-07-2013 >>

---

01:18:31	DAddYE_	http://forum.nimrod-code.org/t/182
01:41:30	*	q66 quit (Quit: Leaving)
02:59:22	*	XAMPP joined #nimrod
03:32:29	*	DAddYE_ quit (Remote host closed the connection)
03:47:40	*	DAddYE joined #nimrod
03:48:44	*	DAddYE quit (Remote host closed the connection)
03:49:18	*	DAddYE joined #nimrod
03:53:48	*	DAddYE quit (Ping timeout: 260 seconds)
04:39:11	*	DAddYE joined #nimrod
04:41:49	*	DAddYE quit (Remote host closed the connection)
05:02:10	*	DAddYE joined #nimrod
05:18:23	DAddYE	Araq: how hard will be if I use from foo import nil be able to use foo.Type.proc_that_match_it()
05:18:38	DAddYE	aka preserve chain-ability
05:29:44	DAddYE	also what is the nim way to map hashes/tables https://github.com/joyent/node/blob/master/lib/_http_server.js#L40-L59 ?
05:30:00	DAddYE	I know there are table but I don't know if there is a better way
05:40:42	DAddYE	haaaa here we are http://build.nimrod-code.org/docs/manual.html#table-constructor
05:40:45	DAddYE	nvm
07:55:48	Araq	DAddYE: I don't know how hard it'll be and frankly I don't care, you want explicit 'module.' prefixes everywhere except that you don't
07:57:03	DAddYE	two reasons
07:57:09	DAddYE	in the module where I'm working
07:57:19	DAddYE	I will not need to prefix with module name
07:57:54	DAddYE	in the module that import it I'll need and it's idiomatic since I don't need to know where a thing come frome
07:57:57	DAddYE	*from
07:58:10	DAddYE	and I also don't have (not sure if allowed by compiler)
07:58:28	DAddYE	redefinition of types/enum/procs
07:59:14	Araq	'from module import nil' is not idiomatic, it's very new and nobody uses it
08:00:26	DAddYE	yea in nimrod and usually before meet other languages also in ruby
08:00:44	DAddYE	is the same (except you have more fine grain control)
08:01:23	DAddYE	btw the big thing is enclose in a namespace
08:01:33	Araq	use .pure enums then
08:01:55	DAddYE	I had to
08:02:04	Araq	I don't like if x.callingConvention in {CallingConvention.Cdecl, CallingConvention.Stdcall}
08:02:36	Araq	it's verbose and only helps people who wouldn't understand the compiler's source code anyway
08:03:08	DAddYE	yea so you have the option to do: use my_module
08:03:41	DAddYE	yea so you have the option to do: use CallingConventions
08:03:47	DAddYE	haha
08:04:03	Araq	yeah but with my way I dont need it
08:04:26	Araq	look
08:04:41	Araq	what you perceive as asthetic is well known to me
08:04:52	Araq	I called the 'dot' disease
08:04:55	Araq	*call it
08:05:02	Araq	GC_disable() # bad, ugly
08:05:08	Araq	GC.disable() # nice and clean
08:05:14	DAddYE	exactly
08:05:15	Araq	ccDefault # bad, ugly
08:05:24	Araq	CallingConvention.Defautl # nice and clean
08:05:34	DAddYE	you know why?
08:05:48	DAddYE	you put here your code
08:05:58	DAddYE	but here I (will) be able to say
08:06:04	DAddYE	that disable is in gc.nim
08:06:12	DAddYE	Default is in cc.nim
08:06:20	Araq	but you can't and it's not important anyway
08:06:38	Araq	*why* do you read code?
08:06:58	DAddYE	usually I don't use anything where I'm not able to read the code
08:07:02	Araq	you seem to imply it's always about reading *unknown new code*
08:07:04	DAddYE	(now you know)
08:07:11	DAddYE	but also for me
08:07:21	DAddYE	I'd like to have every thing separated
08:07:22	DAddYE	in files
08:07:23	Araq	whereas in practice it's always *known* and old code
08:07:38	DAddYE	small portion of code
08:07:41	DAddYE	in many files
08:07:42	Araq	it's always "lets add a feature"
08:07:47	Araq	or "fix the bug"
08:07:59	Araq	"where does this symbol come from?" is not an important question
08:08:14	DAddYE	for me that I'm reading your code YES
08:08:19	DAddYE	what it does?
08:08:25	DAddYE	where is the proc?
08:08:40	DAddYE	and so on
08:08:46	Araq	and btw the navigability you seek DOESN'T exist in Ruby at all
08:08:52	Araq	and yet you don't mind
08:09:05	Araq	because Ruby's very *semantics* prevent it
08:09:43	DAddYE	mmm /dir/lib/module_name/foo.rb => Module::Foo
08:10:18	DAddYE	mmm /dir/lib/module_name/foo.rb => ModuleName::Foo
08:10:23	DAddYE	is pretty standard
08:11:17	DAddYE	also I can write modules named with the same name of nimrod/lib ?
08:11:58	DAddYE	and btw I prefer much more Python module system
08:12:07	DAddYE	or clojure
08:12:31	DAddYE	than ruby
08:13:53	Araq	obj.method() # which 'method' is it?
08:13:56	Araq	you don't know
08:14:07	Araq	and it's a feature in these languages
08:14:30	Araq	it depends on 'obj's *runtime* type
08:19:27	Araq	and fyi I wasn't desperate to write my own RST parser, I looked at the Python implementation and wanted to re-use it or re-code it in Nimrod. It was faster to start from scratch because as soon as something gets more complicated Python and other scripting languages break very quickly.
08:19:44	Araq	It doesn't scale at all.
08:19:56	Araq	ymmv of course
08:20:40	DAddYE	yea
08:20:48	DAddYE	I know a bit better clojure
08:21:03	DAddYE	and recently they got some problems with the module system
08:21:16	DAddYE	for the reloader
08:21:27	DAddYE	hard to track changes
08:21:49	DAddYE	and in jvm you know that respawn a jvm is slow like compile 40K lines of nimcode
08:36:03	DAddYE	is a thing var buf = "".setLen(suggested_size)
08:36:05	DAddYE	valid?
08:36:17	Araq	it was invalid yesterday
08:36:31	Araq	what makes you think it's valid now?
08:38:23	DAddYE	????????
08:38:40	Araq	in other words: no it's not valid
08:38:53	DAddYE	in term of syntax or
08:39:00	DAddYE	code?
08:39:21	DAddYE	*terms
08:39:22	Araq	syntax is fine bug setLen returns void
08:40:13	DAddYE	haaaaa
08:40:44	DAddYE	sorry I mean "".setLen(xx) => (char*) malloc(size)
09:14:40	reactormonk	Araq, remember the times when I trolled around?
09:34:20	*	DAddYE quit (Remote host closed the connection)
09:34:47	*	DAddYE joined #nimrod
09:39:22	dom96	hello
09:39:31	*	DAddYE quit (Ping timeout: 276 seconds)
09:39:36	dom96	ping Araq
09:53:49	*	EXetoC joined #nimrod
09:54:00	dom96	yo EXetoC
09:55:11	EXetoC	ohai
10:08:25	NimBot	Araq/Nimrod master 838c35e Dominik Picheta [+0 ±1 -0]: Fixes OSError deprecation warning.
10:09:01	NimBot	Araq/Nimrod asyncmacro 838c35e Dominik Picheta [+0 ±1 -0]: Fixes OSError deprecation warning.
10:09:01	NimBot	Araq/Nimrod asyncmacro 406edb0 Dominik Picheta [+0 ±1 -0]: Merge branch 'master' into asyncmacro
10:13:00	*	BitPuffin joined #nimrod
10:17:32	EXetoC	daddye really does whatever he can to avoid those prefixes :>
10:20:10	EXetoC	BitPuffin: lo
10:20:18	BitPuffin	EXetoC: howdy!
10:20:31	BitPuffin	EXetoC: considering using Nimrod to do some magic
10:21:23	EXetoC	ok
10:21:53	BitPuffin	EXetoC: what are you up to?
10:22:44	dom96	hey BitPuffin
10:22:50	BitPuffin	hey dom96!
10:23:03	EXetoC	writing some UI code
10:23:29	BitPuffin	coolio
10:26:28	EXetoC	c(:)-<
10:38:12	BitPuffin	EXetoC: shouldn't you bump the revision number of glfw3 to 1?
10:41:16	*	q66 joined #nimrod
10:41:18	EXetoC	no one other than me has even tested it I think. it's not exactly a complicated API though, so I guess I'll bump it in a couple of days
10:42:26	BitPuffin	I will use it
10:42:40	*	DAddYE joined #nimrod
10:43:56	BitPuffin	I thought of another project though that might have the possibility to become quite popular, not the kind of game I thought I'd be making but I kind of like the genre, so I might use Allegro to get desktop, android and iOS support and then compile to js for webgl support
10:44:38	BitPuffin	However I might just write my own Nimrod library that does this for easier portability to additional platforms (like Sailfish, Ubuntu Phone etc)
10:46:06	BitPuffin	But I guess it would also work to contribute additional platforms to allegro, but then I have to write C
10:49:07	*	DAddYE quit (Ping timeout: 260 seconds)
10:55:36	BitPuffin	EXetoC: did you use c2nim to generate the glfw bindings?
11:00:47	EXetoC	I probably did. takes so much time otherwise
11:08:41	EXetoC	there are about 3-4 things I want to change, and then I should add documentation for at least newWindow and a couple of other things
11:44:06	BitPuffin	EXetoC: did you use the --header option?
11:44:13	BitPuffin	apparently that's discouraged
11:45:55	*	DAddYE joined #nimrod
11:48:50	EXetoC	BitPuffin: nah
11:49:01	BitPuffin	EXetoC: but you ran it on the header?
11:51:38	EXetoC	BitPuffin: yes
11:51:43	BitPuffin	I see
11:52:29	*	DAddYE quit (Ping timeout: 245 seconds)
12:00:17	EXetoC	shouldn't be too hard in general, assuming that said header doesn't resemble, say... FreeImage.h in any way :>
12:04:57	*	Araq0 joined #nimrod
12:05:18	Araq0	Hey dom96
12:05:30	dom96	oh no it's Araq's evil twin!
12:06:23	Araq0	Wup?
12:06:48	dom96	When are you going to be back at your desktop?
12:07:13	Araq0	Tonight
12:07:51	dom96	I'm having a barbecue at 5pm my time.
12:08:16	Araq0	I dont
12:08:22	dom96	Not sure how long it will take.
12:08:35	EXetoC	oh, I thought it was an invitation
12:08:54	dom96	EXetoC: sure, if you can get to Northern Ireland in 4 hours.
12:09:12	EXetoC	:>
12:10:35	Araq0	Well you said yoi need me...
12:10:49	dom96	I did?
12:10:53	dom96	I pinged you.
12:10:58	Araq0	Yeah
12:11:04	dom96	I wish you would read the logs...
12:11:15	dom96	<dom96> [23:49:20] Araq: This is how I do the arg passing: https://github.com/Araq/Nimrod/blob/9f297120d1622dd9f6b08ec7cab582419bb811d2/tests/compile/tasynciterraw.nim#L35
12:11:15	dom96	<dom96> [23:49:23] Thoughts?
12:11:47	Araq0	I did read the logs
12:14:10	dom96	well, what are your thoughts?
12:15:37	Araq0	Dunno why you need the cast
12:15:50	dom96	maybe it's a bug?
12:16:15	dom96	I'm not sure how the safe casting is done at runtime.
12:16:21	Araq0	Ordinary conversion shoild be fine
12:16:50	Araq0	But you need to declare ptest once
12:17:15	Araq0	Not twice for it to work
12:17:32	dom96	yes, but how does processServer know the type then?
12:19:39	Araq0	Processserver knows the type already
12:19:57	Araq0	No idea what you mean
12:20:40	dom96	it only knows it because I define it there
12:20:44	Araq0	How does redeclaring the type avoid knowimg it?
12:21:13	dom96	how else am I suppose to cast it if I don't have the concrete type?
12:21:24	dom96	the PTest below is inaccessible
12:22:32	dom96	hrm, unless I generate the PTest above the proc definition.
12:27:34	EXetoC	the order of declaration is important, by default anyway, if that's what you mean
12:33:01	Araq0	Bbl
12:43:02	dom96	yay an infinite loop...
12:43:18	*	zahary__ quit (Ping timeout: 240 seconds)
12:47:03	dom96	yay, it works
12:49:01	*	DAddYE joined #nimrod
12:55:24	*	DAddYE quit (Ping timeout: 245 seconds)
13:16:49	BitPuffin	EXetoC: so you basically just did c2nim glfw.h ?
13:26:17	EXetoC	BitPuffin: I might've had to add one or more #def. chances are you need to if an error message says something about a missing ';'
13:26:43	BitPuffin	EXetoC: it says expected ')' where there is one :s
13:26:56	BitPuffin	local/include/allegro5/allegro_color.h(6, 29) Error: ')' expected
13:28:13	NimBot	Araq/Nimrod asyncmacro 426fbdb Dominik Picheta [+0 ±2 -0]: Better param passing. Implemented send, readLine and support for... 1 more lines
13:28:13	NimBot	Araq/Nimrod asyncmacro b3673c2 Dominik Picheta [+1 ±2 -0]: Calling an async proc with invalid arguments will now fail at compile-time.
13:30:06	BitPuffin	EXetoC: man I'm not sure if this will work as expected, it's adding a bunch of when not defined stuff etc
13:31:11	EXetoC	you can always remove the conditionals if you think they aren't needed
13:31:28	EXetoC	and replace them with c2nim-specific stuff of course, if necessary
13:31:55	BitPuffin	EXetoC: I suppose, but I'm not sure what I'm gonna do with this error I got where it expects a ) where there is a ) :(
13:33:37	BitPuffin	EXetoC: I get it in almost every file
13:34:09	BitPuffin	Araq0: is this a known issue?
13:36:51	EXetoC	Araq0: http://subversion.assembla.com/svn/procedural-tree-generation/trunk/project/ProceduralTreeGenerator/ProceduralTreeGenerator/lib/allegro-5.0.5/allegro/include/allegro5/allegro_color.h
13:37:30	EXetoC	gotta love this stuff. too bad they aren't using begin/end headers
13:37:58	EXetoC	anyway, it's not present in that many headers, is it?
13:41:33	BitPuffin	EXetoC: I don't see what's wrong with it
13:45:57	EXetoC	c2nim might not be able to handle it for whatever reason. I removed most such conditionals, but I don't know if that's necessary
13:46:18	BitPuffin	EXetoC: so you think I should ditch the importing stuff?
13:47:52	BitPuffin	or the conditionals, guess I can't skip importing
13:48:33	BitPuffin	EXetoC: if allegro.nim imports a bunch of other stuff like allegro.h does, is this stuff available to me if I only import allegro.nim?
13:50:30	BitPuffin	EXetoC: doesn't seem like there is a flag to skip the conditionals :(
13:50:35	BitPuffin	only includes
13:51:23	NimBot	Araq/Nimrod asyncmacro 2e3ef1f Dominik Picheta [+0 ±1 -0]: Fixes problems with async docs.
13:52:31	BitPuffin	is it Araq0 or Araq or Araq__ one should ping
13:53:48	dom96	Araq0 I think
13:53:52	EXetoC	I haven't dealt with this crap much at all unfortunately, but I'm pretty sure that much of it can just be removed :>
13:54:29	BitPuffin	EXetoC: well that doesn't really help when the conversion fails because of it
13:54:38	BitPuffin	EXetoC: it doesn't keep going or anything
13:59:42	BitPuffin	EXetoC: maybe I need to do: import allegro_blabla \n export allegro_blabla
14:44:02	BitPuffin	EXetoC: guess I'm gonna have to write the binding manually huh
14:44:22	fowl	I thought someone else did allegro already
14:44:55	fowl	I tried once, it was too much work
14:45:19	fowl	It was gradha IIRC
14:45:56	BitPuffin	fowl: really? latest allegro?
14:46:04	BitPuffin	I searched for it on google and nothing showed up
14:46:40	EXetoC	were you thinking of this https://github.com/nimrod-code/packages/blob/master/packages.json#L38 ?
14:47:03	EXetoC	I couldn't find anything more relevant when googling
14:48:41	fowl	Allegro sux compared to sdl2
14:48:53	BitPuffin	nope
14:49:02	fowl	They require hijacking main() iirc
14:49:04	BitPuffin	why?
14:49:32	BitPuffin	mmmmmm nah only on OSX I think, and that's not their fault, it's OSX's iirc
14:50:57	fowl	Any lib with convoluted headers sucks IMO.. ncurses may be useful but it still sux
14:52:40	*	DAddYE joined #nimrod
14:57:05	BitPuffin	fowl: convoluted?
14:57:41	dom96	maybe try grabbing a pascal wrapper and using pas2nim :P
14:59:10	fowl	BitPuffin the c is ugly and hard for c2nim to get through
14:59:13	*	DAddYE quit (Ping timeout: 246 seconds)
14:59:23	fowl	Parsing c is not trivial
14:59:25	BitPuffin	fowl: I think it looks alright
15:00:28	BitPuffin	dom96: no pascal bindings for allegro either haha
15:00:39	BitPuffin	only for allegro4
15:00:51	*	Araq__ quit (Read error: Connection timed out)
15:00:59	BitPuffin	fowl: more trivial than C++ :P
15:01:27	BitPuffin	are there bindings to SDL2 available though?
15:01:36	EXetoC	yup
15:01:52	EXetoC	https://github.com/fowlmouth/nimlibs/tree/master/fowltek
15:01:55	*	Araq_ joined #nimrod
15:02:11	BitPuffin	of course he did D:<!
15:02:13	BitPuffin	haha
15:02:25	BitPuffin	Well
15:02:39	BitPuffin	the only reason I'd consider it would be to not have to mess around with writing bindings
15:02:45	BitPuffin	time is precious
15:02:46	BitPuffin	hrm
15:03:21	dom96	why not sfml?
15:03:56	BitPuffin	Well
15:03:58	BitPuffin	why sfml?
15:04:04	fowl	Sdl2 is niiice compared to sdl1
15:04:18	BitPuffin	Oh, I guess it's worth mentioning I'm not even making anything 2d here
15:04:41	dom96	oh, well then sdl2 is fine I guess.
15:05:40	BitPuffin	I'm just looking to write OpenGL ES 2 code and have touch and audio API stuff cross platform and have desktop version with audio
15:05:44	Araq0	Bitpuffin the header looks trivial to me
15:05:45	fowl	BitPuffin I have horde3d wrapped and assimp
15:05:57	BitPuffin	Araq0: yeah same
15:06:08	BitPuffin	Araq0: but c2nim can't handle the truth
15:06:16	Araq0	For c2nim to handle
15:06:35	fowl	Ah also I started on window management for a 3d engine, that's in fowltek too
15:06:58	Araq0	Just make the define a def
15:07:28	BitPuffin	Araq0: Well, I wasn't lying when it was expecting a ) where there was a ) and for some reason that's an error
15:07:50	BitPuffin	fowl: window management?
15:08:39	Araq0	Well i am not lying when i say i can easily get it to work
15:08:57	fowl	BitPuffin well i was going to do more there's a scene graph and WM just the basics
15:09:22	BitPuffin	Araq0: well then go for it :P
15:09:48	Araq0	Whats the error msg
15:09:56	BitPuffin	fowl: well what kind of window management are we talking
15:10:23	BitPuffin	Araq0: local/include/allegro5/allegro_color.h(6, 29) Error: ')' expected
15:10:33	EXetoC	creating, destroying, handling events?
15:10:46	fowl	^
15:12:07	BitPuffin	yeah that's what I was thinking
15:12:22	BitPuffin	I was thinking maybe he ment OpenGL windows (GUI stuff)
15:13:11	fowl	U could use it as a starting point or something, I likely won't work on it further
15:13:14	Araq0	Make it defined(x) instead of (defined x) then
15:14:06	fowl	BitPuffin the window stuff is xorg, gl context is created though
15:14:22	BitPuffin	Araq0: so you mean I need to go in to every header and change it manually?
15:14:44	Araq0	God is there a single header out there without arcane bs...
15:14:45	BitPuffin	fowl: well if I needed that I'd use glfw
15:15:01	BitPuffin	fowl: but since I also want touch I'll use either allegro or sdl2
15:15:20	Araq0	Bitpuffin yes
15:15:40	BitPuffin	Well then no thanks, I think c2nim should be able to understand it as it is
15:15:41	Araq0	Headers always need tweaking for c2nim
15:16:15	fowl	BitPuffin sure use whatever, I thought u were talking about writing ur own tho
15:16:43	BitPuffin	fowl: hmm maybe I mentioned that earlier, but even then I was thinking putting libs in the background
15:17:05	Araq0	Bitpuffin its not c2nims fault c is bs
15:17:45	BitPuffin	Well
15:17:46	Araq0	And hostile to interop
15:17:47	BitPuffin	yes
15:17:50	BitPuffin	C is total bs
15:18:06	BitPuffin	But c2nim is trying to handle C, so it is c2nims fault
15:18:06	BitPuffin	or rather both
15:19:02	*	Araq_ quit (Read error: Connection timed out)
15:19:33	Araq0	C2nim is still more useful than wrapping by hand
15:19:55	*	Araq_ joined #nimrod
15:21:00	BitPuffin	Yeah it is
15:21:19	BitPuffin	But it needs more work
15:22:22	BitPuffin	fowl: I guess I'll go with SDL2 then, I don't really love how blitting and all that stuff is in the same header as window creating and all that jazz
15:22:44	EXetoC	what doesn't?
15:23:19	fowl	You don't blit in sdl2 unless u work with software renderer
15:24:16	fowl	Also if ur doing 3d all u use from sdl is event handling right
15:25:27	BitPuffin	fowl: exactly so why would I want it to clutter my namespace
15:25:38	BitPuffin	I guess I'm gonna have to from bla import bla
15:26:01	fowl	What lol clutter ur namespace
15:27:57	fowl	I don't understand
15:27:59	EXetoC	very few things are going to collide, if any
15:28:48	BitPuffin	Well, I know
15:28:59	BitPuffin	but still! :P
15:29:08	BitPuffin	they shouldn't keep them in the same header
15:29:13	BitPuffin	Imo
15:29:51	dom96	BitPuffin: fix c2nim then
15:29:51	BitPuffin	fowl: sdl2 runs on android and ios right?
15:30:36	fowl	That's what I hear
15:30:49	BitPuffin	dom96: well I'm no expert at parsing C and don't have time to become one :P
15:43:01	EXetoC	well, there's this time thing that you mentioned earlier :>
15:44:31	BitPuffin	EXetoC: the lack of? :D
15:53:00	fowl	I'll wrap a library if u need me to and I can have it done in a day, probably, price based on complexity
15:55:46	*	DAddYE joined #nimrod
15:59:59	*	DAddYE quit (Ping timeout: 245 seconds)
16:00:26	BitPuffin	fowl: ah, sorry, can't afford to pay anything right now :)
16:02:34	BitPuffin	fowl: but I'll keep that in mind :D
16:02:58	fowl	Ok
16:09:10	dom96	see you later guys
16:14:44	BitPuffin	dom96: don't leave us!
16:14:47	BitPuffin	the world needs you
16:16:34	BitPuffin	Okay so help me out
16:16:56	BitPuffin	When I install a package with babel
16:17:10	BitPuffin	how do I import it in my project later
16:17:25	BitPuffin	like with fowl's fowltek binding to sdl2
16:17:27	fowl	Import modulename
16:17:28	BitPuffin	how do I import it
16:17:52	fowl	The compiler knows about babel
16:18:31	fowl	Import fowltek/sdl2
16:19:07	fowl	That's just because all modules in fowltek are in that folder, I want my name in ur source
16:19:52	Araq	credit where credit is due, fowl ;-)
16:21:50	fowl	My commercial stuff will be import com/fowl/is/awesome/libraryname lol
16:23:12	*	BitPuffin_ joined #nimrod
16:23:44	*	BitPuffin quit (Ping timeout: 245 seconds)
16:24:22	BitPuffin_	kinda strange syntax to have / in the import statement
16:24:27	BitPuffin_	most languages use .
16:25:35	Araq	yeah I'll support '.' too I think
16:25:44	*	BitPuffin_ is now known as BitPuffin
16:25:47	Araq	import fowltek.sdl2
16:26:03	Araq	that import fowltek doesn't work doesn't matter, does it?
16:26:21	BitPuffin	Araq: but it's not supported currently then?
16:26:26	fowl	Fowltek isn't a module it's a directory
16:26:48	BitPuffin	Would be pretty rocking if we could import fowltek.*
16:27:05	BitPuffin	(says the guy who was earlier bitching about cluttered namespace :$)
16:27:21	Araq	fowl: people decided '.' is good enough for everything and I feel like supporting it
16:27:43	Araq	if only to tell them later "wie man sich bettet so ruht man"
16:28:00	BitPuffin	enguarish prease
16:28:32	fowl	4 years of German in hs and I learned nothing
16:28:57	BitPuffin	same here with spanish
16:29:26	EXetoC	me too, but I passed anyway!
16:29:40	fowl	Oh I know: du bits ein verdammter idiot
16:29:49	fowl	Du bist*
16:29:50	BitPuffin	EXetoC: Yeah I got a VG :s
16:30:00	EXetoC	:>
16:30:35	BitPuffin	Is there any OpenGL ES wrappers for nimrod?
16:30:43	BitPuffin	just use the regular openGL stuff or what
16:31:19	fowl	Bp u should try to work.it into the.opengl module
16:32:13	Araq	"As one makes his bed, so he must lie."
16:32:30	Araq	not sure if it conveys the same meaning ...
16:32:48	BitPuffin	fowl: huh? Does that mean I should use what's in the opengl module or contribute it? :s
16:34:10	BitPuffin	Araq: how far back does the opengl stuff go in terms of compability, I know it stops at 4.2, but does it support everything back to OpenGL 1.0?
16:34:32	Araq	I think so
16:34:37	BitPuffin	I see
16:34:42	BitPuffin	well then it probably already works
16:34:48	Araq	but I'm not sure nor do I see why it is important
16:34:48	EXetoC	it has glBegin. does that answer your question? :p
16:34:49	fowl	Use what u can, contribute if u can
16:35:15	BitPuffin	Since OpenGL ES 2 is OpenGL 2(.1?) without the fixed function bs
16:53:19	EXetoC	~2.0
16:56:54	*	DAddYE joined #nimrod
16:57:24	*	DAddYE quit (Remote host closed the connection)
16:57:31	*	DAddYE joined #nimrod
16:58:38	*	xilo joined #nimrod
16:59:16	*	Mat2 joined #nimrod
16:59:20	Mat2	hi @ all
16:59:43	EXetoC	hi @ you
17:00:16	Mat2	how's work going ?
17:00:55	BitPuffin	hi Mat2!
17:01:16	Mat2	hi BitPuffin
17:01:37	*	DAddYE quit (Remote host closed the connection)
17:04:14	EXetoC	work? well, I'm not getting paid for this :( but it's going well
17:04:33	Mat2	Araq; It was needed to mark the reserved memory-areas for machine-code as executable for calling without generating an exception but otherwise it worked as you suggested
17:06:01	Araq	what? I did suggest something?
17:07:03	Mat2	EXetoC: Well, I assumed the word 'work' includes unpaid programming
17:07:43	Mat2	Araq: Sorry, wrong translation :(
17:08:06	Araq	say it in german then
17:08:53	Mat2	ansonsten funktionierte alles so wie Du beschrieben hattest
17:12:09	Araq	well I don't remember describing anything but alright
17:16:25	Mat2	anyhow, I can now compile vm-code sequences at run-time. Now I'm exploring the thread and process interfaces Nimrod offer
17:16:54	Araq	er ... does that mean your VM is now in Nimrod?
17:17:18	Mat2	yes
17:17:30	Araq	o.o wow aweseom
17:17:51	Araq	how come I hear no complaints from you?
17:18:46	Mat2	because it works at this point
17:19:46	EXetoC	operations on two float32's gives me a float. is that intended?
17:21:28	Araq	no
17:21:33	EXetoC	ok
17:49:56	Araq	Mat2: one more thing
17:51:06	Araq	if you need to put your VM object types on the heap a register based VM has another advantage
17:51:40	Araq	python for instance caches the integer objects < 255 or something to avoid excessive allocations
17:52:06	Araq	but with a register based VM you always have a destination slot that's likely to already point to appropriate memory
17:53:04	Araq	because you can easily assign a fixed type to a temporary slot
17:53:18	Araq	in a register based vm
17:54:26	Mat2	Araq: That's a problem by design - the language design. You think in terms of languages which compile static code sequences (even if this may happen embedded in a dynamic context)
17:55:53	Araq	well I surely don't think in Forth :-)
17:57:10	Mat2	yes. Look how expensive the Java VM for example handles dynamic code-generation
17:59:56	Mat2	I would say, if a VM need to cache object references for performance, this is led by result of flawed language-design decisions
18:00:17	Araq	maybe but I dunno
18:00:30	Araq	it's very common for smalltalk like object systems
18:01:39	BitPuffin	Araq: can I load a dynamic library at runtime manually with some proc?
18:02:00	Araq	BitPuffin: yeah but I forgot the name
18:02:33	Araq	dynlib.nim
18:02:50	*	DAddYE joined #nimrod
18:02:56	Mat2	I think the IO language is a nice example how to avoid such problems by design
18:03:00	BitPuffin	Araq: do you think that will work on platforms like android (with NDK) and iOS?
18:03:35	Araq	whenever you ask me "will it work on iOS" my default answer is "No"
18:03:50	Araq	iOS is the new unoperating system
18:03:55	BitPuffin	lol
18:03:58	BitPuffin	But it can run C!
18:04:05	Araq	it should work in theory
18:04:12	Araq	it's been ported to mac os x
18:04:26	Mat2	have here some one experiences with iOS ?
18:04:44	Araq	Mat2: ask gradha when he's around
18:05:05	Mat2	thanks
18:05:17	Mat2	brb
18:05:28	BitPuffin	it appears that it works with the operating system
18:05:34	BitPuffin	but is not permitted on the appstore..
18:05:46	BitPuffin	hrm
18:05:52	BitPuffin	I hate apple
18:06:04	BitPuffin	How am I supposed to make this then
18:08:28	BitPuffin	I was hoping to package the worlds in to packages with assets and a dynamic library for the compiled nimrod "scripts"
18:09:11	*	DAddYE quit (Ping timeout: 260 seconds)
18:09:40	BitPuffin	I guess I could make some workaround for iOS so that it comes with all the worlds by default, and if you log in or buy or whatever it unlocks the other content somehow
18:26:24	Araq	why does a "world" reside in a dll anyway, BitPuffin? (I use DLL for shared object because it makes more sense)
18:27:03	BitPuffin	Araq: well if you think of a game like Diablo or Torchlight 2 which has Acts
18:27:34	Araq	Diablo, Torchlight 2? never heard of them
18:27:38	BitPuffin	Araq: Each act has special scripts for each acts that has nothing to do with the other acts, and each act can be compiled to a library
18:28:25	BitPuffin	Araq: well okay then think of acts as worlds then, but basically each act has it's own environment and its own objectives and events in those objectives, so why not package it all in a neat little package and load it at runtime
18:28:36	Araq	BitPuffin: I'm kidding
18:28:56	BitPuffin	Araq: lol, I was like, are you serious :P
18:29:45	BitPuffin	Araq: I kind of want to make it so that you only have to buy the game once via the website, so you download a free client that contains the demo data and then when you've bought it you download the other acts
18:29:55	BitPuffin	and I can't think of any other way to achieve this
18:30:03	Araq	well most games include a scripting engine for these things
18:30:17	Araq	and most of the game is then "scripted"
18:30:32	BitPuffin	Araq: yeah but I plan to write the scripts in nimrod that way
18:31:20	Araq	well I'm working on a nimrod VM but it'll run nimrod slower of course
18:31:44	BitPuffin	Araq: hmm, but I'm not even sure iOS let's you use scripting engines even lol
18:32:01	Araq	it can't prevent it properly
18:32:16	BitPuffin	Araq: either way, I was thinking this would be nice, because it gives full performance
18:32:37	EXetoC	can't you just load everything and make it accessible based on some flag? or is cracking common on mobile platforms?
18:32:54	BitPuffin	EXetoC: well with jailbroken devices they sure could crack away
18:33:13	BitPuffin	EXetoC: but I might have to do that exclusively for iOS :(
18:33:34	Araq	BitPuffin: well you have "pure data" like maps that is not code anyway
18:33:43	Araq	it suffices to exclude those
18:34:30	BitPuffin	Araq: hm? Yeah
18:34:34	Araq	you can keep the logic monolithic and it'll crash when it can't load level 10
18:34:46	BitPuffin	But I probably can't download those either
18:34:54	Araq	s/level/map
18:35:19	Araq	why not? it's just some binary blob or even a simple text file
18:35:33	BitPuffin	Araq: well it's obviously technically possible
18:35:40	BitPuffin	but I'm wondering about app store terms
18:35:49	BitPuffin	but actually
18:35:57	BitPuffin	I think there is a game that does that (pocket legends)
18:36:09	Araq	I can't follow
18:36:16	BitPuffin	on iOS
18:36:20	Araq	either downloading files is problematic or it's not
18:36:31	BitPuffin	there is a game that I think even updates itself inside the client
18:36:36	Araq	if it is downloading code is as problematic as downloading data
18:36:40	BitPuffin	called pocket legends
18:48:47	*	NimBot joined #nimrod
18:49:33	BitPuffin	Araq: well it's not like it's shipping today lol
18:49:52	*	dom96 joined #nimrod
18:49:58	BitPuffin	Araq: okay so dylib is not allowed, but scripting engines totally fine, what bullshit
18:50:16	Araq	well that's what I'm saying, create the game and then worry about iOS
18:51:05	BitPuffin	Araq: well I don't want to reach a point where I suddenly have to redo all the scripts in order to comply with some stupid shit
18:51:17	Araq	well I would buy it for PC
18:52:01	*	mario-goulart quit (Ping timeout: 273 seconds)
18:52:01	BitPuffin	Araq: I intend to Run it on Phones and Desktop and Web :) ES 2 will give me all of that with no pain
18:52:12	BitPuffin	"no pain"
18:52:14	BitPuffin	some pain
18:52:26	BitPuffin	but not a whole lot
18:52:46	Araq	we'll see about that
18:52:51	BitPuffin	I guess
18:52:57	Araq	the controls are vastly different for a start
18:53:05	Araq	and can't be abstracted over
18:53:09	BitPuffin	For sure
18:53:13	Araq	well I don't believe they can
18:53:30	BitPuffin	Well I was thinking on desktop and web use traditional diablo style control scheme
18:54:12	BitPuffin	and on handhelds, if touch only use an on screen joystick with buttons for attack and special attacks etc and some targeting assistance
18:54:48	Araq	I always thought Diablo would be much more fun with keyboard input
18:55:05	BitPuffin	Araq: How do you mean?
18:55:16	Araq	pressing the arrow keys for walking is much more natural than clicking a single unit around
18:55:21	BitPuffin	Hmm
18:55:23	BitPuffin	Well
18:55:33	BitPuffin	if I do that then it's closer to the touch counterpart
18:58:49	BitPuffin	Araq: then add gamepad support, and I've got myself a console version :P
18:58:55	Mat2	hmm, that depends on habituation I think. People which grows with smartphones and tablets may tend to touch on everything, so clicking a single unit would be more natural for them
18:59:41	Mat2	(or there try to touch on there displays in use of some mouse emulation)
18:59:45	BitPuffin	Mat2: I did consider that
19:00:10	BitPuffin	Mat2: But I always feel it's a shame when the fingers get in the way of the graphics
19:00:14	Araq	Mat2: Diablo predates smartphones by decades
19:01:01	Mat2	how old is this game ?
19:01:05	BitPuffin	Diablo?
19:01:10	Mat2	yes
19:01:17	BitPuffin	1996
19:01:48	BitPuffin	17 years then?
19:02:14	Mat2	ok, I think there play it inside an emulator :D
19:02:46	BitPuffin	Mat2?
19:03:50	BitPuffin	Araq: I don't think it seems like too much work to support the various control schemes
19:05:20	*	mario-goulart joined #nimrod
19:05:46	Mat2	as historical remark, the virtual-device-interface (VDI) of GEM was specially designed for abstracting all kinds of input *and* output devices (ther is always a way)
19:05:49	*	DAddYE joined #nimrod
19:07:03	Mat2	(GEM was also unfortunately notorious for its performance)
19:08:27	BitPuffin	haha :)
19:08:29	Araq	dunno abstracting away all kinds of input looks like abstracting away the user experience to me
19:08:46	BitPuffin	not really looking to abstract it so it's fine
19:08:53	BitPuffin	I'm looking to provide options
19:08:59	BitPuffin	they differ from each other
19:09:04	Araq	sure
19:09:21	BitPuffin	if you use a mouse to click on things you don't need help targeting mobs
19:09:30	BitPuffin	but if you are using a gamepad it could be useful
19:09:41	BitPuffin	(or on screen joystick)
19:12:45	*	DAddYE quit (Ping timeout: 264 seconds)
19:13:04	BitPuffin	So I wouldn't call it abstraction
19:13:15	BitPuffin	I think torchlight has an xbox version actually
19:14:05	BitPuffin	http://youtu.be/4fIjmKFCTpQ
19:14:55	*	apotheon_ joined #nimrod
19:17:51	EXetoC	I played an FPS once on a console once. that was rubbish
19:18:17	BitPuffin	well
19:18:30	BitPuffin	some people swear by using a game pad for fps games
19:18:31	BitPuffin	I don't but
19:18:33	BitPuffin	yeah
19:19:39	*	Raynes_ joined #nimrod
19:20:08	*	xilo quit (*.net *.split)
19:20:10	*	apotheon quit (*.net *.split)
19:20:10	*	Raynes quit (*.net *.split)
19:20:15	EXetoC	those people can sod off :p
19:21:00	BitPuffin	EXetoC: they'll play the game they want and you won't care since you only have to write a few lines of code for those additional customers
19:21:31	*	Raynes_ is now known as Raynes
19:22:01	*	Raynes quit (Changing host)
19:22:01	*	Raynes joined #nimrod
19:22:20	*	xilo joined #nimrod
19:22:59	EXetoC	sure $$
19:23:15	BitPuffin	EXetoC: and hapiness all over the place
19:23:38	BitPuffin	Like prince Daniel's feelings
19:25:56	EXetoC	ål åver!
19:27:03	BitPuffin	ål åver de plejs
19:28:31	*	NimBot joined #nimrod
19:29:59	BitPuffin	Mat2: Förhoppningsvis lättare att läsa än att prata
19:32:17	*	NimBot joined #nimrod
19:32:31	BitPuffin	Mat2: säkert? Isf är det kanske bara jag hahaha
19:33:07	*	DAddYE joined #nimrod
19:33:23	EXetoC	English or German! Don't upset the general!
19:33:46	DAddYE	hello world
19:34:15	EXetoC	hello.. human?
19:34:23	BitPuffin	EXetoC: I don't see why they get to speak German if we don't get to speak Swedish :(
19:34:27	Mat2	BitPuffin: ;)
19:35:51	Mat2	how much swedish speakers are around here ?
19:35:54	Araq	BitPuffin: quod licet iovi non licet bovi ;-)
19:36:13	BitPuffin	Mat2: from what I know me, you and EXetoC
19:37:03	BitPuffin	Araq: now I don't even know what that is
19:37:14	Araq	Latin
19:38:15	EXetoC	it means, stop watching those cat videos
19:38:36	BitPuffin	"allowed Jupiter not allowed"
19:38:51	Araq	just start a #nimrod-swedish channel or whatever
19:39:04	Araq	in fact German is not really fine here either
19:39:07	BitPuffin	Factus a universus sors of sensu
19:39:25	BitPuffin	Araq: well then stop speaking it :)
19:39:34	Araq	I always provide a translation afterwards
19:39:48	Araq	often it's hard to translate :P
19:39:48	BitPuffin	Araq: what if we do the same?
19:40:03	Araq	then you're allowed single swedish sentences
19:40:17	Araq	like I'm allowed single German sentences
19:41:10	BitPuffin	dealio
19:41:24	BitPuffin	now help me find the definition of PWindow in fowl's lib
19:41:32	BitPuffin	for SDL2
19:42:09	Mat2	BitPuffin: I'm a german speaker (that mean I can read and understand swedish and better danish to some extend but you would likely not understand me if I try to speak it)
19:42:29	BitPuffin	Mat2: but then why doesn't Araq understand swedish
19:42:37	*	DAddYE_ joined #nimrod
19:44:08	Araq	BitPuffin: maybe I should try
19:45:12	BitPuffin	If you can, doesn't that mean we can speak more swedish than you can speak german? :P
19:45:43	*	apotheon_ is now known as apotheon
19:45:43	*	apotheon quit (Changing host)
19:45:43	*	apotheon joined #nimrod
19:46:04	BitPuffin	Seriously where the heck did he get PWindow from :s
19:46:24	EXetoC	does it compile?
19:47:10	BitPuffin	yeah
19:47:19	EXetoC	I see some header imports, but that's only evaluated when SDL_Static is defined
19:47:43	BitPuffin	EXetoC: are you looking at this? https://github.com/fowlmouth/nimlibs/blob/2511e6fa0f965669d26f815f78167603003446ae/fowltek/sdl2.nim
19:47:49	*	XAMPP quit (Read error: Connection reset by peer)
19:48:27	BitPuffin	EXetoC: and in those headers it's not even called PWindow most likely
19:49:13	EXetoC	"t_type.add parseExpr("type P$1* = ptr T$1" % structName_s)[0]" ?
19:49:33	*	Araq__ joined #nimrod
19:49:35	EXetoC	in the sdl_struct macro, which is used a couple of times
19:51:15	BitPuffin	EXetoC: what does it do?
19:51:37	BitPuffin	I haven't learned nimrod100%, I thought I'd learn by doing from now on
19:52:03	EXetoC	I don't know exactly, but that must be the relevant part. I guess it generates type declarations automatically
19:52:49	*	DAddYE quit (Ping timeout: 264 seconds)
19:52:57	BitPuffin	EXetoC: you might be right
19:52:59	BitPuffin	oh well
19:53:12	*	Araq_ quit (*.net *.split)
19:53:39	EXetoC	maybe it shows up in the generated documentation
19:54:21	BitPuffin	woop there is generated documentation?
19:55:25	BitPuffin	okay so since nimrod "hates" oop, is it better to just have procs for creating and destroying and manipulating windows rather than an object?
19:56:14	EXetoC	oo is such a broad time
19:56:27	BitPuffin	term*?
19:56:33	EXetoC	oops
19:56:49	EXetoC	and I don't even use auto-correction
19:56:56	EXetoC	and yes, you'd use regular functions for ctors, and there's a destructor pragma
19:57:15	BitPuffin	hmm
19:57:39	EXetoC	proc newFoo*(x: int): TFoo = result.x = x
19:58:45	EXetoC	proc destruct(o: var TFoo) {.destructor.}
19:58:54	BitPuffin	aha
19:59:10	*	q66 quit (Excess Flood)
19:59:19	*	q66 joined #nimrod
19:59:56	BitPuffin	EXetoC: does it have to be called newFoo?
20:00:02	EXetoC	no
20:00:56	Araq	in fact the convention is initFoo for TFoo, newFoo for PFoo
20:02:03	BitPuffin	aha
20:02:05	BitPuffin	I see
20:02:13	BitPuffin	well, I decided to not do a new type
20:02:30	BitPuffin	just createContext and destroyContext, and have the module manage a window variable
20:06:10	Mat2	that's the classical handler approach
20:06:43	BitPuffin	yep, seemed appropriate
20:10:36	*	gradha joined #nimrod
20:11:53	*	Araq__ quit (Read error: Connection timed out)
20:12:50	*	Araq_ joined #nimrod
20:14:02	BitPuffin	so in nimrod, where you'd traditionally do FLAG | FLAG2 etc how does that work out in nimrod? FLAG xor FLAG2?
20:14:13	BitPuffin	or is it just or
20:14:30	BitPuffin	ah no it's not xor
20:14:32	BitPuffin	is it or?
20:14:43	EXetoC	I think so
20:15:57	EXetoC	pretty much all of that is in system.nim
20:16:32	Araq	it's 'or' but the idiomatic way is to use a bitset
20:16:52	Araq	thankfully C doesn't know about bit sets or type safety
20:17:06	Araq	so you often have no choice when you're wrapping
20:18:20	BitPuffin	Araq: but then what is logical or instead of bitwise?
20:18:33	BitPuffin	also or?
20:18:36	EXetoC	"/ proc `or` n n..." :-)
20:19:12	*	XAMPP joined #nimrod
20:19:12	*	XAMPP quit (Changing host)
20:19:12	*	XAMPP joined #nimrod
20:20:27	dom96	hello
20:20:42	Mat2	hi dom96
20:20:52	Araq	BitPuffin: it's 'or' indeed
20:20:54	EXetoC	so it's logical for bools only as is indicated by the argument passed to the 'magic' pragma. '.magic: "Or"' for the bool variant, '.magic: "BitorI"' for all the int types (except int64, where it is "BitorI64")
20:20:58	BitPuffin	Araq: neato
20:21:16	Araq	EXetoC: the 'magic' is an implementation detail
20:21:19	EXetoC	I might've misinterpreted what you say, but whatever
20:21:31	EXetoC	Araq: right
20:22:51	EXetoC	but the arguments give you a clue. it's all documented though
20:24:45	dom96	Araq: I generate a proc blah() = nil and call it, does the compiler simply get rid of the proc definition and call?
20:24:50	dom96	*if I
20:25:11	Araq	not yet
20:25:18	Araq	it does it for a template though
20:27:46	dom96	what if I make the proc {.inline.}
20:27:48	dom96	?
20:29:58	Araq	then the C compiler will do that for you, but not Nimrod
20:30:04	Araq	but sure, that's fine too
20:30:57	dom96	cool
20:31:44	dom96	Not curious as to why i'm asking? :P
20:32:06	Araq	not really ...
20:32:27	*	Araq_ quit (Read error: Connection timed out)
20:34:51	*	Araq_ joined #nimrod
20:36:33	dom96	pity
20:36:43	dom96	It's pretty clever.
20:37:22	BitPuffin	dom96: tell uss!!
20:38:12	dom96	It's how I get the compiler to make sure that when the user calls an async proc they pass it the correct params.
20:39:24	dom96	I do it by generating a dummy proc which does nothing when an async proc is defined (remember that the async proc is actually converted into an iterator which has very different params)
20:40:06	dom96	Then when the user calls the async proc I just generate a dummy call to it so that the compiler can check the params.
20:41:12	*	dom96 wonders if his explanation is clear
20:41:13	Araq	that works with clean templates too, no need for inline procs
20:41:25	dom96	but then the error message isn't as nice :P
20:41:32	Mat2	I am right to interpret this as some kind of dynamic type-checking ?
20:41:33	Araq	how so?
20:41:48	Araq	error: got (int, char) but expected one of:
20:41:58	Araq	template typecheck(x: int, b: bool)
20:42:01	Araq	vs.
20:42:02	dom96	yeah, expected one of 'template ...' instead of 'proc ...'
20:42:05	Araq	proc typecheck ...
20:42:08	Araq	lol
20:44:51	dom96	I'm pretty proud of that though :P
20:45:30	Araq	C++ does the same for type checking in its generics
20:45:51	dom96	cool
20:46:03	Araq	well it's considered to be good style by Stroustrup
20:46:15	Mat2	ehm, ok
20:48:38	dom96	Araq: btw I noticed that with the old doc gen I can get nice docs for my {.async.} procs generated, but not with the new one.
20:48:57	dom96	but i'm sure we can figure something out
20:49:23	Araq	well the new one expands macros ...
20:49:28	dom96	indeed
20:50:11	Araq	doc generation is hard with meta programming
20:50:25	Araq	sometimes you want to expand what's generated, sometimes you don't
20:50:52	Araq	I'm afraid you need to live with doc1 for now
20:50:56	dom96	I managed to change what is generated if 'nimdoc' is defined though IIRC
20:51:24	dom96	This was only for documenting the built in async functions.
20:51:36	dom96	Haven't tried the actual custom ones yet.
20:55:06	Mat2	ciao
20:55:22	*	Mat2 quit (Quit: Verlassend)
20:56:45	Araq	hmm
20:56:53	Araq	cool that can work
20:56:59	Araq	nice work, dom96
20:57:31	dom96	thanks :D
21:08:10	*	DAddYE_ quit (Remote host closed the connection)
21:52:26	*	gradha quit (Quit: bbl, need to watch https://www.youtube.com/watch?v=1ZZC82dgJr8 again)
21:54:53	*	DAddYE joined #nimrod
22:25:21	dom96	noooo, did github get rid of the pink unicorn for their 500 page? :(
22:25:56	EXetoC	I've never encountered that
22:26:35	dom96	Never saw this? https://github.com/images/error/angry_unicorn.png
22:26:51	EXetoC	ok maybe
22:27:36	EXetoC	maybe my brain is better at dealing with images
22:27:45	EXetoC	I should make a programming language with that in mind
22:27:56	dom96	try scratch :P
22:34:34	EXetoC	:>
23:01:42	NimBot	Araq/Nimrod asyncmacro 95dcabc Dominik Picheta [+0 ±1 -0]: Another docs fix for asyncio.
23:02:19	*	Araq__ joined #nimrod
23:03:01	*	Araq_ quit (Ping timeout: 248 seconds)
23:03:09	BitPuffin	dom96: I've seen it :) It was epic, I think they had it because they're using the Unicorn web server
23:03:20	dom96	indeed
23:07:37	BitPuffin	okay
23:07:42	BitPuffin	time to work on linagl
23:13:22	BitPuffin	can parameterized types have more than one parameter?
23:14:27	dom96	yes
23:21:25	BitPuffin	good
23:21:34	BitPuffin	otherwise this vector would kind of suck :)
23:25:01	EXetoC	type parameters can't yet hold values though
23:25:14	BitPuffin	EXetoC: hmm?
23:25:32	BitPuffin	EXetoC: So I can't have a dimension parameter?
23:26:01	BitPuffin	What do you guys think, should elements of a vector start being counted from 0 or 1
23:26:39	BitPuffin	so the first element is 1 or 0
23:28:37	dom96	good night
23:28:42	EXetoC	cya
23:28:52	EXetoC	I prefer 0. are you also going with arrays?
23:29:02	BitPuffin	type
23:29:02	BitPuffin	TVector[T,D]* = object of TObject
23:29:02	BitPuffin	elements: array[1..D, T]
23:29:09	BitPuffin	nighty night dom96
23:29:16	BitPuffin	EXetoC: hmm yeah perhaps 0
23:29:18	BitPuffin	I dunno
23:29:30	EXetoC	that implies inheritance
23:29:31	BitPuffin	people who like math will probably prefer 1..D
23:29:33	EXetoC	" of..."
23:29:38	BitPuffin	EXetoC: sure
23:29:55	BitPuffin	EXetoC: I was thinking why not give the option :P
23:30:05	BitPuffin	but maybe that's ridiculous
23:30:38	*	Araq__ quit (Read error: Connection timed out)
23:31:14	BitPuffin	EXetoC: but will the D parameter not work now or is that what you are saying?
23:32:54	*	Araq_ joined #nimrod
23:34:03	EXetoC	it won't. you can use type classes though, but it's a little buggy for me
23:34:30	BitPuffin	EXetoC: how would I do it with type classes?
23:34:32	EXetoC	unless you want to allow arbitrary dimensions of course, because then you can't have a fixed set of types
23:35:22	EXetoC	TVT*[T:TReal] = TV2|TV3|TV4
23:35:53	BitPuffin	EXetoC: I want arbitrary dimensions
23:36:20	BitPuffin	So now the final workaround I can think of is using sequences..
23:36:22	BitPuffin	but that sucks
23:36:31	BitPuffin	because this should really work
23:39:45	*	DAddYE quit (Remote host closed the connection)
23:40:15	*	DAddYE joined #nimrod
23:40:43	BitPuffin	Araq, Araq_ or Araq0 Is this true?
23:41:48	EXetoC	they haven't gotten around to it yet
23:42:18	EXetoC	"But then the compiler doesn't support values as generic parameters for now ... It's better to wait until the compiler is ready for it."
23:42:45	BitPuffin	:(
23:42:46	BitPuffin	when is that
23:42:52	BitPuffin	because I need it like, now lol
23:44:17	BitPuffin	does this mean that I'm gonna have to go in to the compiler? because I don't really know how to code a compiler :'(
23:44:37	*	DAddYE quit (Ping timeout: 248 seconds)
23:44:51	BitPuffin	EXetoC: wait a sec, what does it mean when the compiler whines about implementation of blah expected
23:46:37	EXetoC	exactly that. if you forget = and the stuff after, then all you have is a forward declaration
23:46:45	*	DAddYE joined #nimrod
23:47:09	EXetoC	is that the issue? I got it just now, for that very reason
23:47:22	EXetoC	type Ohai
23:47:55	BitPuffin	EXetoC: I was doing this
23:48:09	BitPuffin	type Vector = object (and stuff)
23:48:14	BitPuffin	with params
23:48:20	BitPuffin	so Vector[T, D]
23:48:24	BitPuffin	and then:
23:48:44	BitPuffin	Vec3 = Vector[float, 3] for example
23:49:56	BitPuffin	EXetoC: is it because I haven't yet created a constructor for TVector?
23:50:28	EXetoC	constructors aren't required. let me try that
23:51:17	EXetoC	that literal can't be there. I just get "type expected"
23:51:50	EXetoC	and your error if I do "Vector[T, D]" without specifying an impl
23:52:42	BitPuffin	EXetoC: literal?
23:52:47	BitPuffin	what literal?
23:53:31	EXetoC	or value (the number)
23:54:44	BitPuffin	EXetoC: I'm not sure I get it, you mean the dimension?
23:56:20	EXetoC	yes, it simply doesn't work
23:56:36	EXetoC	I'm sure it's possible with some macro magic
23:56:47	*	DAddYE quit (Remote host closed the connection)
23:57:03	*	Araq_ quit (Read error: Connection timed out)
23:57:20	*	DAddYE joined #nimrod
23:57:27	BitPuffin	EXetoC: But is it valid if the dimension wasn't there?
23:58:39	EXetoC	not without an implementation, unless it's found at a later stage
23:58:53	*	Araq_ joined #nimrod
23:59:15	BitPuffin	EXetoC: well but what is implementation?