<< 02-10-2015 >>

---

00:00:06	Araq	the status is: not gonna happen anytime soon.
00:00:16	tmm1	staticExec would require i call out to the compiler myself right? maybe another proc like that but with the copmiler boilerplate would be simpler for now
00:00:49	Araq	well the question is how can we avoid calling out to the C compiler? we cannot.
00:01:00	Araq	we need to ask it for its version number
00:01:10	tmm1	right definitely
00:01:40	tmm1	i just mean doing so manually with staticExec is a lot of work, staticExec("echo code > file.c; gcc -o file file.c....")
00:01:43	Araq	staticExec is actually really sweet for this since it allows to pass a caching key
00:01:50	*	saml_ joined #nim
00:02:30	Araq	er ... you don't have to generate C code at all
00:02:43	Araq	just staticExec("gcc -v")
00:10:57	*	bjz joined #nim
00:13:32	*	shodan45 joined #nim
00:15:45	*	rollo joined #nim
00:17:23	*	Kingsquee joined #nim
00:37:04	*	strcmp1 joined #nim
00:39:07	tmm1	that's true
00:40:11	tmm1	although that assumes gcc is the binary and not cc or mingw32-gcc etc
00:41:01	tmm1	where is staticExec defined, i see it says magic:"StaticExec" but i can't figure out where thats implemented
00:41:16	*	shodan45 quit (Quit: Konversation terminated!)
00:41:51	Araq	search for mStaticExec in the compiler if it says magic: "StaticExec"
00:42:23	Araq	it's built into the VM
00:47:30	tmm1	gotcha, found opGorge now
00:48:09	*	jaco60 quit (Remote host closed the connection)
00:48:34	*	jakesyl quit (Ping timeout: 246 seconds)
00:53:53	tmm1	i'm imagining a new compileTime proc like tryCompile(s: string): bool, it will use the underlying compiler and replace most common functions of autoconf/configure
00:54:24	tmm1	when tryCompile("#if GCC_MAJOR < 4\n#error to old gcc\n#endif"):
00:56:11	Araq	tmm1: this idea has come up a couple of times already
00:57:26	*	Sahnvour quit (Read error: Connection reset by peer)
00:57:36	tmm1	do you think it's a good idea?
00:58:58	Araq	I don't really like it.
00:59:57	Araq	but I don't see much better solutions either
01:00:48	tmm1	maybe i'll try to implement it
01:01:02	Araq	don't name it tryCompile please. name it probeC
01:01:12	Araq	and don't make it part of system.
01:01:40	tmm1	okay
01:01:45	*	jakesyl joined #nim
01:03:27	Araq	btw you don't have to patch the compiler to implement probeC
01:03:36	Araq	static:
01:03:49	Araq	writeFile(...)
01:04:00	Araq	gorge("gcc c file.c")
01:04:23	Araq	you only need to expose the C compiler's binary somehow
01:04:52	Araq	well, that requires patching the compiler I guess
01:07:14	Araq	good night
01:11:42	*	aw1231 quit (Ping timeout: 246 seconds)
01:12:33	tmm1	good point, it doesn't need to be an opcode
01:20:40	*	chemist69 joined #nim
01:21:48	*	FedeOmoto quit (Quit: Leaving)
01:23:48	*	chemist69_ quit (Ping timeout: 264 seconds)
01:31:34	*	pregressive joined #nim
01:42:51	*	CryptoToad joined #nim
02:01:45	*	rollo quit (Quit: Leaving)
02:21:38	*	X67r quit (Quit: leaving)
02:25:05	*	strcmp2 joined #nim
02:25:39	*	CryptoToad quit (Quit: Leaving)
02:27:06	*	strcmp1 quit (Ping timeout: 240 seconds)
02:38:21	*	pregressive quit (Read error: Connection reset by peer)
02:38:37	*	pregressive joined #nim
02:46:06	*	pregressive quit ()
02:51:59	*	jakesyl quit (Ping timeout: 240 seconds)
03:05:42	*	saml_ quit (Ping timeout: 265 seconds)
03:11:10	*	BitPuffin|osx quit (Ping timeout: 244 seconds)
03:19:56	*	darkf joined #nim
03:49:10	*	vendethiel joined #nim
04:01:06	*	silven quit (Ping timeout: 272 seconds)
04:32:13	*	silven joined #nim
04:34:19	*	strcmp2 quit (Ping timeout: 256 seconds)
04:38:48	*	Jesin quit (Quit: Leaving)
04:51:45	*	vendethiel quit (Ping timeout: 240 seconds)
05:32:42	ryu0	who
05:32:44	ryu0	err
05:58:29	*	bjz quit (Ping timeout: 252 seconds)
06:02:32	*	bjz joined #nim
06:45:06	*	bjz quit (Ping timeout: 260 seconds)
06:50:10	*	bjz joined #nim
07:23:33	*	Ven joined #nim
07:27:25	*	Trustable joined #nim
07:44:58	*	Varriount quit (Read error: Connection reset by peer)
07:49:01	*	OmIkRoNiXz joined #nim
08:07:01	*	coffeepot joined #nim
08:17:37	*	tmm1 quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
09:01:41	*	enamex joined #nim
09:03:02	*	AvinashSnl joined #nim
09:05:26	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
09:05:29	r-ku	why is some kind of artificial order enforced in enums? it complicates wrapping c++ libs a lot.
09:07:26	reactormonk	r-ku, they don't have an order?
09:07:49	r-ku	type # This is the FILE * or void * the regular output should be written to.
09:07:49	r-ku	CURLoption* = enum
09:07:49	r-ku	CURLOPT_WRITEDATA = 10000 + 1, # The full URL to get/put
09:07:49	r-ku	CURLOPT_URL = 10000 + 2, # Port number to connect to,
09:08:09	r-ku	CURLOPT_PORT = 0 + 3, # Name of proxy to use.
09:08:14	r-ku	last one is a problem
09:09:53	Araq	r-ku: c2nim reorders the enum fields for you
09:10:17	r-ku	well this is c2nim output
09:10:37	Araq	well it cannot do the additions then -.-
09:10:55	r-ku	guess thats the problem..
09:10:59	Araq	also we have a curl wrapper in nimble or the stdlib
09:11:13	r-ku	yeah, its old
09:11:38	Araq	so? you think they rewrote their API?
09:12:46	r-ku	probably not, but still
09:13:07	r-ku	would it not be better to have a solution that generates wrapper from new lib headers?
09:14:03	*	enamex quit (Ping timeout: 255 seconds)
09:14:19	*	jakesyl__ joined #nim
09:14:38	Araq	depends
09:15:21	Araq	if the new curl just adds procs that nobody uses anyway, it's good that the "old" wrapper stays slim.
09:15:26	*	M-max1 joined #nim
09:15:57	Araq	I claim 70% of a typical wrapper is never used by anybody.
09:16:06	r-ku	i dont think they add procs, more like constants used to control lib
09:16:27	*	PyHedgehog-work quit (Ping timeout: 246 seconds)
09:16:46	*	profan_ joined #nim
09:16:54	*	reactormonk_ joined #nim
09:16:55	*	rektide_ joined #nim
09:17:49	*	M-max quit (Ping timeout: 250 seconds)
09:17:57	*	rektide quit (Ping timeout: 250 seconds)
09:17:57	*	reactormonk quit (Ping timeout: 250 seconds)
09:17:57	*	jakesyl_ quit (Ping timeout: 250 seconds)
09:17:57	*	profan quit (Ping timeout: 250 seconds)
09:17:58	*	M-max1 quit (Changing host)
09:17:58	*	M-max1 joined #nim
09:18:20	Araq	well I usually just add the new definitions I need and patch the wrapper
09:18:25	*	enamex joined #nim
09:18:52	*	Ven joined #nim
09:54:47	*	coffeepot quit (Ping timeout: 264 seconds)
10:20:59	*	AvinashSnl is now known as Avinash|away
10:24:19	*	Avinash|away quit (Quit: Leaving.)
10:29:33	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
10:34:59	*	enamex quit (Ping timeout: 244 seconds)
10:39:37	*	Ven joined #nim
10:41:34	*	elrood joined #nim
10:46:20	*	bjz quit (Quit: Textual IRC Client: www.textualapp.com)
10:46:58	*	bjz joined #nim
10:56:30	*	wuehlmaus joined #nim
10:59:56	*	razaaa joined #nim
11:00:44	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
11:06:56	*	enamex joined #nim
11:19:11	*	X67r joined #nim
11:21:50	NimBot	nim-lang/Nim devel 9f33baf Aman Gupta [+0 ±1 -0]: add afterRunEvent callback to execProcesses
11:21:50	NimBot	nim-lang/Nim devel ded5753 Aman Gupta [+0 ±1 -0]: remove "rerun with --parallelBuild:1" message and show error inline
11:21:50	NimBot	nim-lang/Nim devel b093c0a Aman Gupta [+0 ±1 -0]: show full compiler output with reNimcCrash
11:21:50	NimBot	nim-lang/Nim devel 08843c6 Aman Gupta [+0 ±1 -0]: remove echo() from compiler
11:21:50	NimBot	1 more commits.
11:22:31	*	razaaa quit (Ping timeout: 244 seconds)
11:22:45	*	AvinashSnl joined #nim
11:23:43	flyx	that would be not much different from usual application software. @ 70% unused
11:38:06	*	reactormonk_ is now known as reactormonk
11:44:23	*	Kingsquee quit (Quit: http://i.imgur.com/EsXzoum.png)
11:46:26	*	Sahnvour joined #nim
12:08:37	*	drewsrem joined #nim
12:13:33	*	Ven joined #nim
12:14:42	*	AvinashSnl quit (Read error: Connection reset by peer)
12:15:17	*	AvinashSnl joined #nim
12:15:29	*	Demon_Fox quit (Quit: Leaving)
12:16:06	*	silven quit (Remote host closed the connection)
12:16:10	*	M-max1 quit (Ping timeout: 246 seconds)
12:16:43	drewsrem	Is something planned to convert an AST (/NimNode) back to its text-representation? - Would this even be possible?
12:17:50	*	Pisuke joined #nim
12:18:21	*	silven joined #nim
12:19:16	reactormonk	drewsrem, nim-pretty?
12:20:03	drewsrem	reactormonk, what's that?
12:20:23	*	MyMind quit (Ping timeout: 244 seconds)
12:20:39	reactormonk	drewsrem, compiler/nimfix/pretty.nim
12:21:56	drewsrem	reactormonk, I don't think this converts an AST back to some text-representation?
12:22:16	*	M-max joined #nim
12:22:36	flyx	drawsrem: tried repr(node)?
12:23:23	drewsrem	flyx, that just builds a pretty-printed tree of the NimNodes?
12:23:23	reactormonk	drewsrem, there's also render(node) IIRC
12:24:06	flyx	well there *was* a possibility to get Nim-like code. I used it once
12:24:14	drewsrem	Maybe I'm phrasing my question wrong, I'm talking about e.g. a function that takes a nimnode and outputs nim-code that when parsed produces the same AST
12:24:14	flyx	but if it wasn't repr, I can't remember
12:24:37	Araq	drewsrem: repr(n) does work
12:24:54	drewsrem	Oh, then excuse me
12:25:00	flyx	if it was repr, well, it has some issues
12:25:16	flyx	eg. it doesn't `` identifier names that need to
12:25:51	*	PyHedgehog-work joined #nim
12:26:02	Araq	pretty sure it does `` identifier names
12:26:35	flyx	I used it with some genSym'd names like ":whatever", they were just outputted as-is
12:26:44	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
12:27:04	Araq	well `:whatever` is not valid either
12:27:09	drewsrem	On that note, is there something to let me see the "unfolded" version of a module where all the immediate-macros are expanded?
12:27:12	Araq	that's the point of genSym.
12:27:23	Araq	to produce something that cannot be written down.
12:27:30	Araq	so that no conflict can happen.
12:27:32	drewsrem	s/immediate-macros/immediate-templates
12:27:44	flyx	well okay. then you just can't use genSym with repr
12:30:47	*	Ven joined #nim
12:36:00	softinio	@Araq @dom96 and everyone: I am thinking of putting together a talk for a python meetup on how to use nim and ffi to create python modules. Thought this would be a good indirect way of introducing python dev to nim. What do you think? A good topic ?
12:36:41	Araq	softinio: of course! will you also show nimborg?
12:38:20	softinio	Need to look into nimborg. Sure! @Araq
12:41:52	dom96	softinio: Sounds brilliant!
12:43:37	softinio	@Araq @dom96 Great I will work on doing one then. May trouble you to have a look at what I prepare so it can be perfect. Thx. So glad to have discovered nim and be part of this community.
12:44:28	dom96	softinio: no problem, will love to take a look :)
12:44:33	Araq	sure, it's always a good idea to let us look at your slides :-)
13:10:00	softinio	@Araq @dom96 great! Thanks.
13:32:13	*	FedeOmoto joined #nim
13:51:00	*	jck quit (Remote host closed the connection)
13:57:31	*	jck joined #nim
14:01:26	*	razaaa joined #nim
14:04:35	NimBot	nim-lang/Nim devel 437603a Hans Raaf [+0 ±1 -0]: Added ReadIOEffect because thats what happens on OSX.
14:04:35	NimBot	nim-lang/Nim devel 27aaa39 Andreas Rumpf [+0 ±1 -0]: Merge pull request #3411 from oderwat/fix-osx-ospaths... 2 more lines
14:26:46	*	jaco60 joined #nim
14:29:35	*	AvinashSnl quit (Quit: Leaving.)
14:41:48	*	AvinashSnl joined #nim
14:44:52	*	Varriount|Busy joined #nim
14:59:25	*	darkf quit (Quit: Leaving)
15:03:06	*	smodo joined #nim
15:03:34	Varriount|Busy	Hello smodo
15:18:34	smodo	gr頷ohl ;)
15:18:45	*	AvinashSnl left #nim (#nim)
15:30:02	*	ray- quit (Changing host)
15:30:02	*	ray- joined #nim
15:30:04	*	ray- is now known as \u
15:35:04	Varriount|Busy	\u: Interesting nickname.
15:40:56	*	UberLambda joined #nim
16:04:48	*	nim-buildbot quit (Quit: buildmaster reconfigured: bot disconnecting)
16:14:40	*	Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
16:37:20	*	Jesin joined #nim
16:37:59	*	rleblic quit (Ping timeout: 264 seconds)
16:44:56	*	rleblic joined #nim
17:02:45	*	brson joined #nim
17:03:53	*	UberLambda quit (Quit: GTG)
17:18:03	*	Varriount|Busy quit (Ping timeout: 246 seconds)
17:18:17	*	brson quit (Quit: leaving)
17:18:37	*	brson joined #nim
17:19:13	*	bjz quit (Quit: Textual IRC Client: www.textualapp.com)
17:29:42	*	r-ku is now known as bit
17:30:08	*	bit is now known as r-ku
17:42:09	*	tmm1 joined #nim
17:42:16	tmm1	morning
17:44:20	tmm1	araq could you merge https://github.com/nim-lang/c2nim/pull/54
17:45:32	Araq	done.
17:45:51	tmm1	thanks
17:46:20	*	brson quit (Quit: Changing server)
17:47:12	*	brson joined #nim
17:48:27	tmm1	any comment on https://github.com/nim-lang/Nim/pull/3403#issuecomment-144825148
17:49:28	Araq	there is msg: "Instantiation from here: file(line, col)" you can use instead of 'errormsg'
17:49:33	Araq	iirc
17:55:17	tmm1	but i want to verify the error message "type mismatch", not the instantiation message
17:59:07	Araq	you want both.
17:59:28	Araq	you can use both errormsg and msg. I hope. if not, well you know the tester ;-)
18:01:24	tmm1	yea it will require some patching
18:01:25	tmm1	of "errormsg": result.msg = e.value
18:05:22	tmm1	i'm still having a hard time figuring out how this works: when declared(gcc):
18:05:48	tmm1	where is gcc is declared? i only see "compiler gcc" as a local symbol in extccomp
18:06:06	tmm1	how is gcc undeclared when using clang for instance
18:07:00	tmm1	aha, setCC() will undefSymbol and defineSymbol
18:07:07	tmm1	ok the pieces are starting to fit together now
18:07:48	*	boopsiesisaway is now known as boopsies
18:13:48	*	Matthias247 joined #nim
18:16:43	*	bjz joined #nim
18:24:21	*	joelmo joined #nim
18:34:16	*	boopsies is now known as boopsiesisaway
18:35:28	*	boopsiesisaway is now known as boopsies
18:36:51	*	CryptoToad joined #nim
19:08:05	*	xificurC joined #nim
19:10:50	*	strcmp1 joined #nim
19:14:52	*	drewsrem quit (Quit: Leaving)
19:34:36	*	brson quit (Quit: leaving)
19:34:45	*	brson joined #nim
19:40:06	*	strcmp1 quit ()
20:22:26	*	jakesyl__ quit (Ping timeout: 240 seconds)
20:23:14	*	jakesyl_mac joined #nim
20:27:17	*	jakesyl_mac quit (Max SendQ exceeded)
20:27:50	*	jakesyl_mac joined #nim
20:31:34	tmm1	is there any way to fix these errors: https://travis-ci.org/nim-lang/Nim/builds/83372518#L2588-L2605
20:31:44	tmm1	assigning to 'NCSTRING' (aka 'char *') from incompatible type 'const char *'
20:42:49	Araq	tmm1: we currently fix these by importing gai_strerror differently when we compile to C++
20:42:53	Araq	like so:
20:43:41	Araq	proc gai_strerror() {.importc: "(char*)gai_strerror", header: "<foo.h>".}
20:43:56	Araq	which is a hack of course
20:50:33	tmm1	i see, yea definitely a hack
20:51:01	tmm1	i tried to use type cstring_const {.importc:"const char *".} = distinct cstring
20:51:26	tmm1	but the generated code still using plain NCSTRING so it didn't help
20:54:06	*	FedeOmoto quit (Ping timeout: 250 seconds)
20:59:29	Araq	IMO c++'s 'const' makes it impossible to use as a portable assembler.
21:00:11	*	boopsies is now known as boopsiesisaway
21:00:30	Araq	note that for GCC we pass -fpermissive to make it compile
21:00:50	Araq	but clang doesn't interpret this flag the same way as GCC
21:06:07	*	FedeOmoto joined #nim
21:14:03	*	Mat4 joined #nim
21:14:05	Mat4	hello
21:16:51	Araq	hi Mat4
21:17:02	Mat4	hi Araq
21:22:39	*	smodo quit (Ping timeout: 240 seconds)
21:30:42	*	yglukhov joined #nim
21:39:56	*	Kingsquee joined #nim
21:43:06	*	Guest15556 joined #nim
21:48:35	*	enquora joined #nim
21:56:21	*	FedeOmoto quit (Quit: Leaving)
22:00:37	*	Trustable quit (Remote host closed the connection)
22:09:21	*	Guest15556 quit (Quit: Leaving)
22:10:39	*	saml quit (Quit: Leaving)
22:12:31	*	enquora quit (Quit: enquora)
22:19:36	*	yglukhov quit (Remote host closed the connection)
22:34:46	*	CryptoToad quit (Quit: Leaving)
22:42:02	*	tmm1 quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
22:43:21	*	elrood quit (Quit: Leaving)
22:48:58	*	tmm1 joined #nim
22:55:06	Mat4	ciao
22:55:30	*	Mat4 quit (Quit: Leaving)
22:57:03	*	razaaa quit (Quit: Leaving)
23:28:03	*	Matthias247 quit (Read error: Connection reset by peer)
23:49:49	*	solidsnack joined #nim
23:50:30	*	solidsnack quit (Max SendQ exceeded)
23:56:38	*	Sahnvour quit (Read error: Connection reset by peer)