<< 21-03-2015 >>

---

00:00:15	fowl	.eval import macros; macro x(s:stmt):stmt = echo(toStrLit(x)); x(1)
00:00:16	Mimbus	fowl: eval.nim(4, 37) Error: recursive dependency: 'x'
00:00:21	fowl	.eval import macros; macro x(s:stmt):stmt = echo(toStrLit(x));x(1)
00:00:22	Mimbus	fowl: eval.nim(4, 37) Error: recursive dependency: 'x'
00:00:25	dom96	zipR4ND: great. let me know if you have any other questions.
00:00:54	*	davidhq joined #nim
00:01:44	EXetoC	fowl: I just wanted to fix this code https://github.com/Araq/Nim/issues/2373
00:01:54	EXetoC	ok :p
00:05:14	*	davidhq quit (Ping timeout: 246 seconds)
00:06:56	fowl	EXetoC, he can fix it by making it take string{lit}
00:08:00	EXetoC	yeah
00:10:09	onionhammer	hola
00:10:11	onionhammer	whats up
00:11:46	*	mwbrown joined #nim
00:12:24	*	TEttinger joined #nim
00:17:51	brson	Araq: what are you refering to?
00:20:27	*	davidhq joined #nim
00:25:06	fowl	CryptoToad, hey try changing the first if clause in that function to use == 0 intead of "not x > 0"
00:27:25	CryptoToad	lol just realized how dirty that was, but it's still doing the same thing
00:28:00	CryptoToad	i'm wondering if ms code is just outdated
00:28:01	*	brson quit (Ping timeout: 255 seconds)
00:28:03	fowl	what is it doing
00:28:08	CryptoToad	i added the console output
00:28:11	CryptoToad	in the hastebin
00:28:20	CryptoToad	see "Err!: @["
00:28:36	CryptoToad	where @[ is szProcessName
00:28:47	CryptoToad	the getModuleBaseName function
00:28:57	CryptoToad	should set szProcessName to the process name
00:29:05	CryptoToad	but i'm messing it up somehow
00:29:14	CryptoToad	cause it's returning false
00:29:20	fowl	CryptoToad, get rid of cast[bool], cast[dword]
00:29:36	CryptoToad	which ones? i use a lot of em
00:29:38	fowl	use bool(), dword() type conversions
00:29:41	CryptoToad	ooohhh
00:30:40	fowl	Error: undeclared identifier: 'closeHandle'
00:30:55	CryptoToad	yeah i commented that for now
00:31:12	CryptoToad	oh
00:31:13	CryptoToad	wow
00:31:14	CryptoToad	it works
00:31:14	CryptoToad	now
00:31:19	CryptoToad	just that type conversion issue
00:31:21	CryptoToad	:O
00:31:47	fowl	i got it working
00:31:48	CryptoToad	thanks fowl
00:31:52	fowl	oh i see you did too
00:32:03	CryptoToad	:)
00:32:06	fowl	CryptoToad, so only use cast[] when ptr is involved
00:32:14	CryptoToad	or if it's a weird type?
00:32:20	CryptoToad	or should all types have
00:32:29	fowl	weird how?
00:32:29	CryptoToad	the type() conversion proc
00:32:46	CryptoToad	like imported from c or something
00:33:16	fowl	ah um with c we mainly run into this with the billions of different int types involved
00:33:23	CryptoToad	lol
00:33:34	CryptoToad	yeah, seems a bit silly to complicate it
00:33:39	CryptoToad	just let me pass an int xD
00:33:47	EXetoC	you might accidentally cast a 32-bit value to a 64-bit value for example
00:35:14	fowl	CryptoToad, should be CloseHandle() btw
00:35:43	*	rzzz joined #nim
00:36:02	fowl	CryptoToad, and to convert that array of characters to a string its just $szProcessName
00:38:41	*	pregressive quit (Remote host closed the connection)
00:39:28	CryptoToad	yeah lol i know it is, i think i was just trying to do things until something worked
00:39:49	CryptoToad	prints fine without the $ though
00:40:02	CryptoToad	does echo convert to a string?
00:40:21	fowl	yea
00:40:34	fowl	echo(varargs[string, `$`]) means it applies $ to every arg
00:40:35	EXetoC	echo(varargs[`$`, T])
00:40:39	*	Matthias247 quit (Read error: Connection reset by peer)
00:40:43	CryptoToad	ahh
00:40:46	EXetoC	damn, so close!
00:40:49	CryptoToad	lol nice try
00:40:50	CryptoToad	:P
00:40:52	EXetoC	wait
00:41:05	EXetoC	details
00:41:43	fowl	bbl food
00:45:49	CryptoToad	exetoc how long have you been coding nim?
00:48:18	EXetoC	since early 2013
00:51:57	CryptoToad	ahh nice
00:52:56	*	CryptoToad quit (Quit: Leaving)
00:56:46	Trustable	Hi all, does anyone know, if someone tried to make Geany compatible to Nim?
01:07:10	*	saml_ joined #nim
01:09:24	*	zipR4ND quit (Ping timeout: 245 seconds)
01:10:56	EXetoC	Trustable: I only know of that PR from 2013 mentioned on the wiki, which remains open
01:11:41	Trustable	EXetoC: What was that PR about?
01:15:28	EXetoC	Trustable: for adding nim support to geany
01:16:27	Trustable	EXetoC: ok, I guess it is about the filename:line:column format of the compiler output
01:17:14	Trustable	EXetoC: But I would like to add a lexer for Nim to Scintilla (which is used by Geany)
01:19:23	Trustable	EXetoC: Sorry, I missread you.
01:20:27	Trustable	I found something interesting here: https://github.com/geany/geany/pull/193
01:20:45	*	akiradeveloper quit (Remote host closed the connection)
01:20:58	Trustable	EXetoC: thanks for the hint
01:23:51	EXetoC	yes that one
01:26:52	*	davidhq quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
01:32:29	*	akiradeveloper joined #nim
01:35:03	*	davidhq joined #nim
01:36:08	*	nuew joined #nim
01:36:32	nuew	Hello - I'm trying to use the events module, but can't figure out how to pass arguments.
01:36:58	nuew	I've sub-classed EventArgs
01:37:29	nuew	but taking EventArgs as the argument results in nothing
01:38:12	nuew	switching the argument to StringArgs (the name of the subclass) makes the proc say it can't be called
01:39:03	nuew	and switching the proc to a method says that it expected an expression.
01:39:22	nuew	Thanks in advance for the help.
01:53:36	*	darkf joined #nim
01:53:58	EXetoC	care to upload the code to gist or something?
01:54:29	EXetoC	nuew: I haven't used that module yet so I don't know what you're referring to
01:54:40	nuew	sure
01:55:46	EXetoC	what do you intend to use it for?
01:55:49	*	Trustable quit (Remote host closed the connection)
02:01:03	nuew	Here are some examples: http://pastebin.com/swtZLvWW
02:01:16	nuew	Using it for a basic event/callback system in a game
02:05:25	EXetoC	EventArgs does not have such a field, and so you must cast it (cast[Type](var))
02:06:02	nuew	ok thanks
02:06:53	*	davidhq quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
02:10:13	EXetoC	nuew: and you can't substitute the class type with the type of one of the children
02:10:35	nuew	i knew that, but I was hoping it would work ;)
02:12:34	EXetoC	and it takes a proc so a method can't be used there. not that it works. you could implement your own methods on top
02:13:02	nuew	huh... gcc is now complaining: error: conversion to non-scalar type requested pushstring_159662((*HEX3Aenvp_176530).Lua176271, ((stringargs172009) ((*e))).X);
02:13:20	*	gokr quit (Quit: Leaving.)
02:13:39	EXetoC	you can extend EventArgs if you want, or you could have your own "base" class with methods
02:13:45	nuew	ok
02:14:00	EXetoC	that's a bug. I'll see if there are bug reports that might be relevant
02:14:29	nuew	i guess i'm just going to reimplement the thing
02:14:51	EXetoC	nope. please report that.
02:15:45	nuew	I mean as well
02:24:44	nuew	reported: https://github.com/Araq/Nim/issues/2375
02:25:26	EXetoC	thanks
02:26:34	EXetoC	-d:release rarely needs to be used
02:28:23	EXetoC	you can use ```nim\n...``` to highlight nim code
02:28:25	*	reem quit (Remote host closed the connection)
02:29:18	nuew	that's on the compiler
02:29:22	nuew	not the build
02:29:59	*	reem joined #nim
02:31:23	EXetoC	right
02:38:27	fowl	dont cast
02:38:43	fowl	type conversion
02:39:01	fowl	StringArgs(e).x
02:40:50	*	Varriount quit (Read error: Connection reset by peer)
02:41:30	fowl	nuew, it works for me with the type conversion ^
02:41:48	fowl	EXetoC, stop telling people to use cast>_>
02:42:23	nuew	thansk
02:42:31	nuew	I swear I tried that.
02:43:01	nuew	but it woks
02:43:05	fowl	mhm :p
02:43:05	nuew	*works
02:44:16	*	reem quit (Remote host closed the connection)
02:45:10	EXetoC	fowl: I recently adviced it needlessly?
02:46:26	EXetoC	I know I did advice against it today
02:46:35	fowl	yes you drunk
02:46:39	*	mwbrown quit (Ping timeout: 252 seconds)
02:46:41	fowl	lol
02:46:42	fowl	<EXetoC> EventArgs does not have such a field, and so you must cast it (cast[Type](var))
02:47:03	EXetoC	other than then obviously :p
02:51:18	EXetoC	I have done it before regardless. I'll write it down
02:53:27	*	bcinman joined #nim
02:55:21	fowl	good write it on the back of your hand D:
02:58:11	EXetoC	yeah why not. nothing else is written there
03:07:28	*	reem joined #nim
03:07:29	*	akiradeveloper quit (Read error: Connection reset by peer)
03:13:31	*	reem quit (Remote host closed the connection)
03:31:23	*	akiradeveloper joined #nim
03:44:31	*	brson joined #nim
03:45:14	*	reem joined #nim
03:55:32	*	vendethiel quit (Ping timeout: 246 seconds)
03:58:48	*	vendethiel joined #nim
04:02:49	*	xtagon joined #nim
04:04:38	*	kjo quit (Quit: Leaving.)
04:05:07	*	kjo joined #nim
04:07:20	*	kjo quit (Client Quit)
04:10:38	*	xtagon quit (Quit: Leaving)
04:13:54	*	xtagon joined #nim
04:17:02	*	epichero quit (Remote host closed the connection)
04:17:18	*	epichero joined #nim
04:21:12	*	vendethiel quit (Ping timeout: 272 seconds)
04:23:55	*	brson quit (Quit: leaving)
04:25:15	*	vendethiel joined #nim
04:33:19	*	johnsoft quit (Ping timeout: 256 seconds)
04:50:55	*	saml_ quit (Quit: Leaving)
05:11:08	*	pleiosaur joined #nim
05:18:16	*	Woflox quit (Ping timeout: 255 seconds)
05:31:17	*	vendethiel quit (Ping timeout: 252 seconds)
05:39:21	*	vendethiel joined #nim
05:42:28	*	filwit quit (Quit: Leaving)
06:19:46	*	xtagon quit (Read error: Connection reset by peer)
06:25:33	*	vendethiel quit (Ping timeout: 252 seconds)
06:26:14	*	vendethiel joined #nim
06:44:23	*	reem quit (Remote host closed the connection)
06:45:46	*	reem_ joined #nim
06:47:05	*	vendethiel quit (Ping timeout: 244 seconds)
06:48:22	*	brononymous joined #nim
06:50:50	*	brononymous left #nim (#nim)
06:51:15	*	vendethiel joined #nim
06:52:02	*	fluoride joined #nim
07:08:23	*	fluoride quit (Ping timeout: 246 seconds)
07:49:46	*	johnsoft joined #nim
07:56:23	*	akiradeveloper quit (Remote host closed the connection)
07:57:03	*	akiradeveloper joined #nim
08:13:22	*	akiradeveloper quit (Remote host closed the connection)
08:13:55	*	akiradeveloper joined #nim
08:18:01	*	akiradeveloper quit (Ping timeout: 244 seconds)
08:19:15	*	rzzz quit (Ping timeout: 250 seconds)
08:23:46	*	vendethiel quit (Ping timeout: 272 seconds)
08:26:20	*	reem_ quit (Read error: Connection reset by peer)
08:26:49	*	vendethiel joined #nim
08:30:06	*	reem joined #nim
08:35:49	*	milosn quit (Ping timeout: 264 seconds)
08:37:39	*	milosn joined #nim
08:38:09	*	nuew quit (Remote host closed the connection)
08:45:20	*	rzzz joined #nim
08:49:39	*	rzzz quit (Ping timeout: 252 seconds)
08:58:26	*	reem quit (Remote host closed the connection)
08:58:31	*	dtscode quit (Ping timeout: 256 seconds)
08:59:00	*	reem joined #nim
09:03:34	*	reem quit (Ping timeout: 265 seconds)
09:06:00	*	BlaXpirit joined #nim
09:07:33	*	gokr joined #nim
09:07:33	*	gokr quit (Read error: Connection reset by peer)
09:07:57	*	gokr joined #nim
09:09:26	*	akiradeveloper joined #nim
09:10:41	*	akiradeveloper quit (Remote host closed the connection)
09:11:49	*	vendethiel quit (Ping timeout: 255 seconds)
09:13:49	*	akiradeveloper joined #nim
09:15:22	*	rzzz joined #nim
09:16:01	*	TEttinger quit (Ping timeout: 250 seconds)
09:19:57	*	Matthias247 joined #nim
09:21:27	*	rzzz quit (Ping timeout: 265 seconds)
09:23:04	*	reem joined #nim
09:27:37	*	vendethiel joined #nim
09:30:33	*	BlaXpirit-UA joined #nim
09:33:17	*	BlaXpirit quit (Ping timeout: 246 seconds)
09:39:48	*	epichero quit (Remote host closed the connection)
09:47:08	*	epichero joined #nim
09:50:30	*	vendethiel quit (Ping timeout: 244 seconds)
09:52:26	*	epichero quit (Ping timeout: 272 seconds)
09:55:18	*	vendethiel joined #nim
10:12:15	*	epichero joined #nim
10:13:15	*	HakanD joined #nim
10:15:22	*	rzzz joined #nim
10:20:25	*	rzzz quit (Ping timeout: 265 seconds)
10:21:49	*	akiradeveloper quit (Remote host closed the connection)
10:22:47	*	bcinman quit (Quit: My Mac has gone to sleep. ZZZzzz…)
10:24:18	*	akiradeveloper joined #nim
10:30:36	*	HakanD quit (Quit: Be back later ...)
10:34:08	*	HakanD joined #nim
10:36:11	*	TEttinger joined #nim
10:53:00	*	johnsoft quit (Ping timeout: 252 seconds)
10:53:52	*	johnsoft joined #nim
10:55:46	flyx	morning folks. I have a macro taking a "params: varargs[expr]". now the compiler complains that it cannot evaluate len(params) at compile time. I don't really see why not.
10:57:25	*	TEttinger quit (Ping timeout: 264 seconds)
10:58:27	flyx	ah, forget it, I see the problem
11:01:52	flyx	I wish I could use openarray as non-parameter type
11:10:05	*	Ven joined #nim
11:15:21	*	rzzz joined #nim
11:20:08	*	rzzz quit (Ping timeout: 264 seconds)
11:21:48	flyx	so, when I have a "param: varargs[…]", how would I construct an array that has a length of len(param) + 1?
11:22:29	flyx	array[len(param) + 1, …] doesn't work because len(param) is not static, and I don't see another method to do it
11:40:04	*	BlaXpirit-UA quit (Remote host closed the connection)
11:41:53	*	BlaXpirit joined #nim
11:42:40	*	epichero quit ()
11:43:04	*	HakanD quit (Quit: Be back later ...)
11:45:01	BlaXpirit	flyx, exactly, it is not static
11:45:05	BlaXpirit	so it is impossible
11:45:20	BlaXpirit	how do you expect to create a type based on a runtime value?
11:45:35	flyx	well, it works in Ada
11:45:42	flyx	so it's possible
11:45:52	BlaXpirit	-_-
11:45:59	BlaXpirit	you're definitely confusing something
11:46:05	BlaXpirit	how about u explain what u want
11:46:57	flyx	I have a proc taking a varargs[…] param, and inside it, I want to call another proc taking a varargs[…] param, but I want to add one value to the varargs when calling it
11:48:33	flyx	I would use a seq, but the varargs param doesn't take a seq
11:49:21	BlaXpirit	.eval proc f(a: varargs[int]) =; var t = @a; t.add(5); f(t)
11:49:24	Mimbus	BlaXpirit: <no output>
12:02:37	*	Gonzih quit (Quit: WeeChat 1.0.1)
12:12:23	*	akiradeveloper quit (Remote host closed the connection)
12:12:56	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
12:15:20	*	rzzz joined #nim
12:20:13	*	irrequietus joined #nim
12:21:03	*	rzzz quit (Ping timeout: 250 seconds)
12:21:07	*	Trustable joined #nim
12:31:22	*	HakanD joined #nim
12:32:17	*	HakanD_ joined #nim
12:35:50	*	HakanD quit (Ping timeout: 244 seconds)
12:44:38	*	akiradeveloper joined #nim
12:52:35	*	Ven joined #nim
12:55:58	*	akiradeveloper quit ()
12:56:03	*	davidhq joined #nim
12:59:06	flyx	hmm okay, I misinterpreted the error message
12:59:54	*	saml_ joined #nim
13:00:47	flyx	is it possible to pass a parameter by value to a macro? so that if I pass a string, it's still a string and not an strlitnode?
13:11:25	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
13:15:22	*	rzzz joined #nim
13:17:32	*	irrequietus quit ()
13:19:48	*	reem quit (Remote host closed the connection)
13:34:16	*	Ven joined #nim
14:04:44	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
14:13:17	*	irrequietus joined #nim
14:14:42	*	Ven joined #nim
14:26:05	*	fluoride joined #nim
14:27:25	*	fluoride quit (Client Quit)
14:29:11	*	Jehan_ joined #nim
14:34:07	*	Woflox joined #nim
14:35:24	*	mwbrown joined #nim
14:37:06	*	untitaker joined #nim
15:09:40	*	darkf quit (Quit: Leaving)
15:23:25	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
15:28:11	*	Matthias247 quit (Read error: Connection reset by peer)
15:33:43	tmku	http://pastebin.com/k5Zp9ge0 is there a way to make s_w and s_h recognized by the rest of object attributes in the scope?
15:33:51	tmku	something akin to python's __init__ I guess
15:34:32	*	Jehan_ quit (Ping timeout: 246 seconds)
15:36:29	flaviu	tmku: array length is a compile-time value, not a runtime value.
15:37:21	tmku	ah. so I should use open arrays instead?
15:38:01	flaviu	I don't think openarray works in that context, but I never figured out what it was.
15:38:21	flaviu	What I'd do is create a type that maps a 2d index to a sequence.
15:40:34	tmku	hm, I see. kind of like pointer arithmetic?
15:41:46	flaviu	Well, it's not pointer arithmetic, this sort of thing is done is languages like javascript.
15:42:14	flaviu	really just (x, y) -> (y * width) + x
15:45:20	tmku	aha, got it. How would I go about overloading the [] operator?
15:47:03	*	Jehan_ joined #nim
15:47:56	flaviu	.eval type Test = object; proc `[]`(self: Test, x, y: int) = echo((y * 10) + x); Test()[1,2]
15:47:57	Mimbus	flaviu: 21
15:49:14	*	Jehan_ quit (Client Quit)
15:49:42	*	Jehan_ joined #nim
15:49:45	*	Ven joined #nim
15:54:14	tmku	thanks, flaviu.
15:57:14	*	Jehan_ quit (Ping timeout: 272 seconds)
15:58:16	*	Jehan_ joined #nim
16:02:18	*	rzzz quit (Ping timeout: 272 seconds)
16:06:46	*	dhasenan joined #nim
16:09:00	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
16:12:57	*	kjo joined #nim
16:20:48	*	devzerp quit (Remote host closed the connection)
16:20:48	*	no_name quit (Remote host closed the connection)
16:22:37	*	devzerp joined #nim
16:22:37	*	no_name joined #nim
16:26:10	*	Demos joined #nim
16:26:23	*	Senketsu quit (Quit: Leaving)
16:28:05	*	no_name quit (Remote host closed the connection)
16:28:05	*	devzerp quit (Remote host closed the connection)
16:28:27	*	Senketsu joined #nim
16:29:18	*	no_name joined #nim
16:29:18	*	devzerp joined #nim
16:30:47	*	bcinman joined #nim
16:32:12	*	bcinman quit (Read error: Connection reset by peer)
16:33:41	dhasenan	This seems like a dumb question, but: how do I cast BiggestInt to float?
16:34:00	dhasenan	If I do cast[float](someBiggestInt), that does a reinterpret cast using a union.
16:34:09	dhasenan	I want the same numeric value instead.
16:38:40	dhasenan	Ah, toFloat() in system. Though I have to cast to an int for that to work, which means losing a large chunk of the allowable range.
16:38:54	def-	dhasenan: float(someBiggestInt)
16:39:12	def-	cast is just reinterpretation of the bits, usually you want a type conversion
16:39:30	dhasenan	Thanks, def- .
17:05:26	*	NimBot joined #nim
17:06:33	tmku	aha, I see
17:07:53	*	gsingh93 joined #nim
17:12:11	*	banister joined #nim
17:12:15	*	banister quit (Max SendQ exceeded)
17:12:58	*	banister joined #nim
17:13:06	*	wink-s joined #nim
17:47:10	*	JehanII joined #nim
17:49:20	*	Jehan_ quit (Ping timeout: 272 seconds)
17:53:22	*	elbow_jason left #nim ("Leaving")
18:01:28	*	matkuki joined #nim
18:09:58	*	Jehan_ joined #nim
18:11:32	*	JehanII quit (Ping timeout: 252 seconds)
18:12:46	*	dhasenan_ quit (Ping timeout: 272 seconds)
18:13:35	*	Matthias247 joined #nim
18:14:45	*	HakanD_ quit (Quit: Be back later ...)
18:14:48	matkuki	Is there something builtin for removing duplicates from a sequence?
18:15:32	Jehan_	sequtils.deduplicate
18:15:44	matkuki	Jehan_: Thanks.
18:16:22	Jehan_	Note however that it has worst case O(n^2) performance, since it doesn't assume anything about the type except that elements can be compared for equality.
18:16:45	matkuki	Got it.
18:16:50	Jehan_	You can write faster implementations for types that support hashing or ordered comparison.
18:17:35	matkuki	For now I need convenience, so deduplicate will do.
18:32:49	*	JehanII joined #nim
18:36:28	*	Jehan_ quit (Ping timeout: 252 seconds)
18:37:28	*	Woflox quit (Ping timeout: 272 seconds)
18:50:57	*	brononymous joined #nim
18:54:30	*	Ven joined #nim
18:55:23	*	Jehan_ joined #nim
18:58:18	*	JehanII quit (Ping timeout: 276 seconds)
18:59:31	*	mwbrown quit (Read error: Connection reset by peer)
19:03:08	*	MagusOTB quit (Remote host closed the connection)
19:03:29	*	wink-s_ joined #nim
19:03:53	*	wink-s_ quit (Client Quit)
19:07:00	*	wink-s quit (Quit: Bye)
19:10:33	*	mwbrown joined #nim
19:19:12	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
19:25:35	*	brononymous quit (Quit: Page closed)
19:30:31	*	Matthias247 quit (Quit: Matthias247)
19:31:40	*	Ven joined #nim
19:37:16	*	wink-s joined #nim
19:48:40	*	banister quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
19:52:53	*	Matthias247 joined #nim
19:57:32	*	TEttinger joined #nim
20:03:24	*	JehanII joined #nim
20:07:50	*	Jehan_ quit (Ping timeout: 276 seconds)
20:22:39	*	matkuki quit (Quit: ChatZilla 0.9.91.1 [Firefox 36.0.3/20150319201009])
20:26:04	*	Jehan_ joined #nim
20:29:17	*	JehanII quit (Ping timeout: 276 seconds)
20:34:10	*	reem joined #nim
20:34:19	BlaXpirit	can i prevent copying of my object?
20:34:34	def-	BlaXpirit: ref or ptr
20:34:50	BlaXpirit	object
20:35:01	def-	I mean: use refs or ptrs
20:35:02	BlaXpirit	there is probably some pragma
20:35:07	BlaXpirit	but no, that is not worth it
20:36:07	def-	what would happen when the original object gets removed from the stack?
20:37:17	BlaXpirit	i want to prevent var b = a
20:38:17	def-	ah, I misunderstood. I guess with the planned overloading of `=` you could do that
20:42:56	*	filwit joined #nim
20:45:56	*	dhasenan_ joined #nim
20:59:27	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
21:10:12	*	reem quit (Remote host closed the connection)
21:12:40	*	reem joined #nim
21:13:10	*	reem quit (Remote host closed the connection)
21:14:26	*	reem joined #nim
21:15:26	*	HakanD_ joined #nim
21:19:44	*	HakanD_ quit (Ping timeout: 244 seconds)
21:35:09	*	dhasenan_ quit (Ping timeout: 265 seconds)
22:07:42	*	rzzz joined #nim
22:08:48	*	Jesin quit (Quit: Leaving)
22:13:43	*	dhasenan_ joined #nim
22:21:03	BlaXpirit	why is intset not sorted?
22:21:14	*	CryptoToad joined #nim
22:23:10	BlaXpirit	IntSet with numbers 1..2000 .items gives 1536..2000 & 1024..1535 & 512..1023 & 1..511
22:24:38	BlaXpirit	https://github.com/Araq/Nim/blob/master/lib/pure/collections/intsets.nim#L199
22:25:02	def-	BlaXpirit: where does it say that they should be sorted?
22:25:10	BlaXpirit	this gives 1056 1 2 7 1 2 7 1056
22:25:12	BlaXpirit	it doesn't say
22:25:18	BlaXpirit	but why wouldn't it be sorted?
22:25:54	def-	Haven't looked at the implementation, maybe it's just more efficient this way
22:26:03	EXetoC	is that more important than what you get if it isn't? but maybe there should be two different types then
22:26:19	BlaXpirit	well of course it is important
22:29:48	def-	In this case I guess you could traverse r.bits in reverse in items
22:32:14	EXetoC	is it a more common requirement? sometimes you need ordering and other times not
22:32:52	*	bcinman joined #nim
22:35:05	BlaXpirit	geez
22:35:29	BlaXpirit	so my lib HAS been horribly broken the whole time
22:36:19	EXetoC	yeah?
22:36:23	BlaXpirit	i don't know what to do
22:37:32	BlaXpirit	https://github.com/BlaXpirit/nim-random/blob/7c434e3cf698f0e9a979c503cdc742ac66403065/src/random/common.nim#L158
22:37:32	*	gsingh93 quit (Ping timeout: 264 seconds)
22:38:24	BlaXpirit	if someone wants 900 randomly chosen numbers from 1..1000 i actually only choose 100
22:38:44	def-	it's broken because of the intsets?
22:38:46	BlaXpirit	and yield numbers that are not in the intset
22:39:00	BlaXpirit	it's broken because i rely on intset to be sorted
22:39:23	def-	I guess you can try to make intset sorted, document it, and make sur performance is still about the same
22:39:29	def-	or change your code
22:39:30	BlaXpirit	lol i have no idea
22:39:38	BlaXpirit	and there is no good way to change the code
22:39:46	def-	let me check
22:39:52	def-	i think it should be easy to make the intset sorted
22:40:04	BlaXpirit	the only way to change the code is
22:40:14	*	johnsoft quit (Ping timeout: 245 seconds)
22:40:23	BlaXpirit	drop any optimization and just generate 900 numbers and return intset
22:40:29	*	johnsoft joined #nim
22:41:14	flaviu	correctness is better than performance.
22:43:28	BlaXpirit	def-, check out intsets.nim line 199 that i linked to
22:43:53	BlaXpirit	whoever wrote that (that reminds me of git blame) knew that it is not sorted
22:44:38	def-	well, it's a set, why would it be sorted?
22:45:08	BlaXpirit	C++'s sets are sorted :|
22:45:10	EXetoC	def-: don't you think there should be two types? or maybe the current type could be made more flexible
22:45:13	BlaXpirit	again, why wouldn't it be sorted?
22:45:39	EXetoC	and there's also unordered_set for a reason
22:45:51	CryptoToad	does nim have a command to get the last X characters of a string?
22:47:42	BlaXpirit	s[0 .. s.high-2] is a safe bet
22:47:58	BlaXpirit	there is a neater way but it's subject to change
22:48:32	flaviu	don't you mean s[s.high-n .. s.high]?
22:48:44	BlaXpirit	what a dummy :|
22:49:25	BlaXpirit	it's actually s[s.len-2 .. s.high]
22:49:40	BlaXpirit	and by 2 I mean n
22:50:07	BlaXpirit	s.substr(s.len - n)
22:50:51	def-	EXetoC: I haven't needed sorted sets yet, don't know
22:52:29	*	brson joined #nim
22:56:28	*	MagusOTB joined #nim
22:59:06	MagusOTB	How do I read an int32 from a file?
23:01:23	def-	BlaXpirit: i don't see a simple way to make Nim's intsets sorted
23:02:03	def-	MagusOTB: as raw data or as a text?
23:03:08	MagusOTB	raw data.
23:03:14	BlaXpirit	what should i do then?
23:03:22	MagusOTB	(I'm trying to write a program that loads MD3 model files)
23:04:40	flaviu	MagusOTB: That's a little harder, is it stored as big or little endian?
23:04:43	def-	MagusOTB: http://nim-lang.org/streams.html#readInt32,Stream
23:05:59	MagusOTB	flaviu: native (in my case, small)
23:06:05	MagusOTB	def-: oh hey perfect
23:06:08	MagusOTB	thanks
23:08:21	def-	MagusOTB: You can always search through http://nim-lang.org/theindex.html if you don't know where to look
23:09:06	flaviu	From what I can tell, the streams module uses pointer punning to read bytes into an int, so it should work here.
23:09:08	MagusOTB	def-: Yeeeaaaah, the problem with that is knowing what to look for. I've been writin this language for like, 2 days :)
23:10:12	MagusOTB	also streams don't seem to support seeking, which sort of makes sense, but also sort of isn't what I need.
23:10:24	MagusOTB	Files appear to
23:12:01	def-	they have getPosition and setPosition at least
23:12:06	MagusOTB	oh
23:12:29	MagusOTB	that'll do
23:13:24	*	MagusOTB was looking for the wrong thing
23:14:24	MagusOTB	What's the difference between 'pointer' and 'ptr <thing>'?
23:14:40	MagusOTB	it seems like in theory I should be able to convert between them, but for some reason I can't.
23:14:40	BlaXpirit	untyped pointer, maybe you're familiar with void*\
23:14:45	MagusOTB	yea
23:14:55	MagusOTB	I know C/Python pretty well
23:15:02	BlaXpirit	cast should convert between them
23:15:15	MagusOTB	but in C you can go from void* to basically anything really easy (and hilariously unsafely)
23:15:23	MagusOTB	hmm... lemme try again
23:15:24	BlaXpirit	that's cast
23:15:40	MagusOTB	ok wait
23:15:40	BlaXpirit	and by cast i mean cast[Type](thing)
23:15:47	MagusOTB	ah
23:16:08	MagusOTB	I was sing thing.Type
23:16:11	MagusOTB	*using
23:18:54	*	gokr quit (Quit: Leaving.)
23:19:05	EXetoC	the source has no type information so that's why it's not possible to perform a conversion
23:20:24	BlaXpirit	cast is indeed "hilariously unsafe"
23:20:52	MagusOTB	is it literally like a C cast?
23:21:06	MagusOTB	i.e. "I know what's at this memory, pretend I'm correct"
23:21:50	BlaXpirit	pretty sure
23:22:11	BlaXpirit	i am sure that it's at least somewhat like it
23:22:45	cazov	so a nim cast clobbers type information while a nim conversion is more like a function that takes something of type A and return something of type B?
23:23:13	flaviu	Almost, but it uses memcpy so that it doesn't trigger UB. (IIRC, it's been a while since I've looked into the c output)
23:23:26	flaviu	cazov: Yep, exactly.
23:24:41	cazov	so not quite like C casts then, since those might perform conversions (like from float to int or vice versa) where cast[float](someInt) in nim would do like a *(float*)(&someInt) (if that was actually legal)?
23:25:14	BlaXpirit	ye
23:25:27	cazov	nifty, thanks :]
23:25:28	EXetoC	that's true. the range is not validated in C though
23:26:05	EXetoC	in Nim it is unless that behavior is disabled
23:26:06	MagusOTB	cazov: FWIW, that's totally legal in C.
23:27:04	MagusOTB	can you use cast[] to get the pointer value of a ref?
23:27:26	EXetoC	he might have meant in nim, but all these things have equivalents
23:27:55	flaviu	MagusOTB: You don't need cast, just do addr(foo)
23:28:13	BlaXpirit	maybe addr(foo[]) ?
23:28:25	MagusOTB	flaviu: oh neat, didn't know about that.
23:28:31	BlaXpirit	definitely need []
23:28:37	MagusOTB	what does the [] do?
23:28:40	flaviu	BlaXpirit: yes, I forgot about that.
23:28:54	BlaXpirit	addr(foo[]) would be &(*foo)
23:29:05	MagusOTB	oh, so [] in this case means exactly the same thing as C
23:29:08	MagusOTB	neat
23:29:31	EXetoC	unless the data happens to be layed out in a certain way?
23:29:48	MagusOTB	well, same as foo[0] in C
23:30:20	EXetoC	no, [] dereferences
23:31:03	EXetoC	soo foo[] is like *foo. foo[i] is indexing like in C
23:31:17	MagusOTB	indexing in C is semantic sugar for dereferencing :)
23:31:22	MagusOTB	*syntactic?
23:32:11	MagusOTB	so if I wanted to assign to the value pointed to by a ref type I'd need to do that too
23:32:14	MagusOTB	like
23:32:19	MagusOTB	?
23:32:22	BlaXpirit	foo[i] is definitely not like indexing in C :|
23:33:00	BlaXpirit	yes, you need dereference when working with ref, just like with pointers. but often it can be implied so u dont need to write it
23:33:05	MagusOTB	I know. That's what I'm saying, what I'm confused about is what the empty [] means
23:33:13	BlaXpirit	dereference
23:33:26	MagusOTB	"it can be implied" ?
23:33:42	BlaXpirit	myref.x works like myref[] . x
23:33:48	BlaXpirit	etc
23:34:05	MagusOTB	so like if I have a: int32 and b: ref int32, if I said a = b it would work?
23:34:18	BlaXpirit	probably not, dunno
23:34:37	EXetoC	a = b[]
23:35:20	flaviu	.eval var a: int32 = 1; var b: ref int32 = new(int32); b = 2; a = b; echo a, b
23:35:21	Mimbus	flaviu: eval.nim(5, 4) Error: type mismatch: got (int literal(2)) but expected 'ref int32'
23:35:38	flaviu	.eval var a: int32 = 1; var b: ref int32 = new(int32); b[] = 2; a = b; echo a, b
23:35:39	Mimbus	flaviu: eval.nim(6, 4) Error: type mismatch: got (ref int32) but expected 'int32'
23:37:12	MagusOTB	but that does work for overloading purposes?
23:37:24	EXetoC	.eval var a: int32 = 1; var b: ref int32 = new(int32); b[] = 2; a = b[]; echo a, b
23:37:25	Mimbus	EXetoC: eval.nim(7, 8) Error: type mismatch: got (ref int32)
23:37:50	EXetoC	crap
23:37:56	*	MagusOTB is confused
23:37:57	BlaXpirit	MagusOTB, Nim doesn't like when you use a type as both normal and ref
23:38:11	EXetoC	.eval var a: int32 = 1; var b: ref int32 = new(int32); b[] = 2; a = b[]; echo a, b[]
23:38:12	Mimbus	EXetoC: 22
23:38:18	BlaXpirit	when you create a type, you decide if you want a normal type or a ref type
23:38:29	BlaXpirit	and then you just use it, without needing to write ref
23:38:36	MagusOTB	ah, so ref is more a property of the type than of an instance
23:38:45	MagusOTB	so really you would never use 'ref int32' in practice
23:38:54	MagusOTB	?
23:38:57	BlaXpirit	ref int32 is pointless
23:39:06	MagusOTB	ok cool, that makes more sense.
23:39:30	BlaXpirit	but if it was some other complex type, then it's not so pointless
23:39:54	MagusOTB	yeah. But you'd put ref on that types type declaration, not every time you used one, right?
23:39:55	flaviu	I wouldn't say it's pointless, although its use is limited.
23:40:12	BlaXpirit	MagusOTB, yes, that's what people usually do, but it's not enforced
23:40:22	Jehan_	ref int32 has its uses when you explicitly want to share the value between multiple places.
23:40:50	Jehan_	It's relatively rare to not have it as part of a bigger object, of course.
23:40:55	flaviu	MagusOTB: Sort of like typedef *foo_s foo; vs using *foo_s directly.
23:46:08	*	wb quit (Ping timeout: 265 seconds)
23:46:56	*	bjz joined #nim
23:47:52	BlaXpirit	def-, flaviu, so what should I do about https://github.com/BlaXpirit/nim-random/blob/d38e06e81e9e64910ff8bf77344d96e8ae66249f/src/random/common.nim#L158 ?
23:48:20	BlaXpirit	intset is exactly what I need, and even without that optimization it would be advantageous to return the results in order
23:52:14	flaviu	Well, you could require that the range == (0..(2^64 - 1)), in which case pcg32_c64 would work.
23:52:29	flaviu	You wouldn't even need temporary space!
23:53:53	Jehan_	BlaXpirit: What's the problem with the code?
23:54:25	BlaXpirit	Jehan_, it's relying on intset.items being sorted, but it's not
23:54:42	BlaXpirit	[:23:09] <BlaXpirit> IntSet with numbers 1..2000 .items gives 1536..2000 & 1024..1535 & 512..1023 & 1..511
23:55:59	Jehan_	In order for the return values to be in the same order as before?
23:56:19	BlaXpirit	that, and the optimization "not direct", "missingItems"
23:57:56	Jehan_	Hmm, I seem to remember something about a better algorithm for choosing M elements from a set of N (in order). Let me check, I think I may have some old code lying around.