<<19-05-2013>>

---

00:49:31	NimBot	Araq/Nimrod 3615102 Araq [+1 ±3 -0]: fixes #433
00:49:31	NimBot	Araq/Nimrod 03127d3 Araq [+12 ±7 -1]: Merge branch 'master' of github.com:Araq/Nimrod
00:49:31	NimBot	Araq/Nimrod 5219600 Araq [+0 ±2 -0]: fixes #420
00:49:31	NimBot	Araq/Nimrod 77bcf7b Araq [+0 ±1 -0]: fixes #439
00:49:31	NimBot	Araq/Nimrod a25e167 Araq [+0 ±3 -0]: fixes #432, fixes #427
01:32:53	NimBot	Araq/Nimrod 466833e Araq [+0 ±4 -0]: updated docs
01:36:10	fowl	Araq: you wnt me to add to macros.nim
01:51:28	*	ARCADIVS quit (Quit: WeeChat 0.3.8)
02:32:15	*	q66 quit (Remote host closed the connection)
04:36:47	*	fowl quit (Ping timeout: 256 seconds)
04:41:19	*	Trixar_za is now known as Trix[a]r_za
04:48:38	*	fowl joined #nimrod
05:26:27	fowl	i get this error now entitty.nim(237, 5) Error: expression 'result' has no type (or is ambiguous)
05:26:42	fowl	macro unicast*(func): stmt =
05:50:12	fowl	test: https://gist.github.com/fowlmouth/5606825
06:21:59	*	OrionPK quit (Read error: Connection reset by peer)
07:03:09	*	Endy joined #nimrod
08:18:53	reactormonk	embedded.nim, line 45, setControlCHook uses c_signal where there are no signals - idea?
08:19:55	reactormonk	what's this hook anyway? Nimrod doesn't seem to used it
08:21:59	reactormonk	oh, I see why I don't really like the nimrod source code - too many lines per method :-)
08:23:58	Araq	reactormonk: setControlCHook is exported and used by some programs
08:24:23	Araq	you can simply wrap it in an 'when not defined(noSignalHandler)' though
08:25:34	Araq	and just what method has too many lines? and how would splitting it up help anybody?
08:25:49	reactormonk	that's the point :-/
08:26:23	Araq	this is called "system software" for the reason that the code actually DOES something
08:26:44	Araq	as opposed to "business logic" code that endless delegates stuff until you can't debug it anymore
08:27:24	Araq	"Smalltalk: Everything happens elsewhere"
08:28:09	reactormonk	if p == nil: c_raise(SIGSEGV)
08:28:13	reactormonk	how would you implement this?
08:30:17	reactormonk	proc chckNil(p: pointer) {.error: "can't assert without signals".} gives me lib/system/excpt.nim(365, 29) Error: pragmas are only allowed in the header of a proc
08:30:44	Araq	well there is a forward declaration of 'chckNil' somewhere
08:31:06	reactormonk	meh
08:31:07	Araq	you should be able to replace the c_raise by a Nimrod 'raise' statement though
08:31:18	Araq	and it's a compilerproc anyway
08:31:27	reactormonk	lib/system/excpt.nim(30, 26) Error: implementation of 'system.registerSignalHandler()' expected
08:31:28	Araq	so the 'error' can't work
08:31:36	reactormonk	ok
08:32:02	reactormonk	lib/system/excpt.nim(36, 12) Error: implementation of 'system.chckNil(p: pointer)' expected
08:32:06	reactormonk	so how do I deal with this?
08:32:20	Araq	proc chckNil ... =
08:32:27	Araq	when defined(noSignalHandler):
08:32:41	Araq	raise newException(EBase, "segfault")
08:32:45	Araq	else:
08:32:50	Araq	c_raise(...)
08:33:29	reactormonk	c_raise is defined with signal handlers
08:33:36	reactormonk	proc c_raise(sig: cint) {.importc: "raise", header: "<signal.h>".}
08:34:02	Araq	yeah but if you don't use it, it's not emitted in the C code I think
08:34:15	Araq	if it is, you need another 'when'
08:34:17	reactormonk	still waiting for fail
08:34:43	reactormonk	you can't use it without signals, that's it
08:34:48	reactormonk	that's why I wanted to add the error
08:35:03	Araq	sure you can
08:35:15	reactormonk	what do you want to do without signal.h?
08:35:32	Araq	raising an exception needs setjmp, not signals
08:35:54	Araq	setjmp is a CPU/OS specific function
08:35:57	reactormonk	ok
08:36:03	Araq	but it can be easily implemented without an OS
08:36:09	Araq	so I hope your headers support it :P
08:36:21	reactormonk	sorry, I didn't see the missing `not`
08:36:50	reactormonk	lib/system.nim(2072, 24) Error: implementation of 'system.setControlCHook(hook: proc (){.noconv.})' expected
08:36:54	reactormonk	what does this one do?
08:37:03	Araq	as I said
08:37:12	Araq	it is exported and some programs use it
08:37:25	Araq	it's not used by system.nim
08:37:25	reactormonk	yes, but what the method do?
08:37:38	Araq	sets a signal handler
08:37:50	Araq	that's called when somebody presses CTRL-C
08:38:03	reactormonk	... there are hardware interrupts, but that's about it
08:38:04	Araq	your board likely doesn't support it :P
08:38:11	reactormonk	yep.
08:38:16	reactormonk	but system wants it.
08:38:22	reactormonk	wait, macro.
08:38:26	Araq	'when' ftw
08:38:29	reactormonk	yep
08:38:48	Araq	you should really use when hostOS == "standalone"
08:38:50	Araq	instead of
08:38:55	Araq	when defined(noSignalHandler)
08:39:08	Araq	noSignalHandler doesn't really mean your OS has no signals ...
08:39:21	Araq	it means Nimrod shouldn't overwrite the default handler
08:39:52	reactormonk	kk
08:42:16	reactormonk	you sure it's when hostOS != "standalone":
08:42:36	reactormonk	because it still complains about it not being defined even if I put it inside there
08:44:07	reactormonk	can you have the compiler echo the value of hostOS?
08:44:56	Araq	well you need to compile with --os:standalone
08:45:23	Araq	the compiler outputs it with "nimrod dump" I think
08:46:04	Araq	oh and before you push I demand a code review :P
08:46:31	reactormonk	don't worry
08:46:50	Araq	today is release
08:46:52	reactormonk	os = standalone
08:47:01	reactormonk	sweet
08:47:13	reactormonk	that's inside nimrod.cfg
08:47:58	Araq	that should do
08:48:04	reactormonk	it doesn't
08:48:19	reactormonk	http://sprunge.us/bXcB
08:49:03	Araq	well "hostOS" is no defined symbol
08:49:21	Araq	ok, nimrod dump doesn't list it
08:49:27	reactormonk	does work when I specify it
08:49:36	reactormonk	http://sprunge.us/IRFW
08:49:54	Araq	--define is not the same as --os
08:50:05	reactormonk	➜ arduino-sample nimrod dump --os:standalone
08:50:12	Araq	interesting
08:50:27	Araq	ah yeah and gone is "Linux" then
08:50:39	Araq	maybe you edited the wrong configuration file :P
08:51:18	Araq	or maybe 'os' doesn't work in a .cfg ... :-/
08:51:29	reactormonk	http://sprunge.us/IJeP
08:51:50	Araq	yeah doesn't work
08:52:07	Araq	it should be 'arm' instead of 'amd64'
08:52:16	reactormonk	it is?
08:52:44	*	zahary joined #nimrod
08:52:48	Araq	no it's not: http://sprunge.us/IRFW
08:53:04	Araq	it processes the .cfg
08:53:10	reactormonk	wtf
08:53:11	Araq	and yet doesn't set OS and CPU properly
08:55:11	reactormonk	bug?
08:55:19	Araq	I dunno
08:55:35	Araq	the compiler explicitly only supports in the command line
08:55:51	Araq	I'm sure I had a reason for it
08:56:37	reactormonk	hmm
08:56:51	reactormonk	I'd prefer it in the command line here
08:56:55	reactormonk	ehh config file
09:01:46	reactormonk	in my current project, I'd have to specify each time, so I'm for changing that.
09:02:12	Araq	well fix it
09:02:35	Araq	commands.nim:427 and 435
09:03:02	Araq	--> if pass in {passCmd1, passPP}
09:03:26	Araq	I had my reasons for not supporting it via the config system but hey, if you test it again ...
09:04:03	Araq	in the worst case it should be an error to use it in a config file
09:04:09	Araq	not silently ignored
09:09:48	reactormonk	nope, works as expected
09:09:51	reactormonk	Error: system module needs 'addChar'
09:09:53	reactormonk	O.o
09:10:46	Araq	oh not that again ...
09:12:24	Araq	system.nim:934
09:12:40	Araq	perhaps get rid of the test 'hostOS != "standalone"'
09:13:19	Araq	or rather move the addChar and setStackBottom out of system/cgprocs.nim
09:13:28	Araq	these are not OS dependent
09:13:33	reactormonk	lib/system/embedded.nim(85, 22) Error: undeclared identifier: 'c_raise'
09:13:36	reactormonk	hmmm
09:15:07	Araq	the standalone target is a mess :p
09:15:35	reactormonk	lib/system/alloc.nim(85, 10) Error: Port memory manager to your platform
09:15:38	reactormonk	aww
09:15:53	Araq	yeah that's actually do-able without an OS :P
09:16:01	Araq	I should patch it
09:16:36	Araq	how much RAM do you have?
09:17:25	reactormonk	depends
09:17:44	reactormonk	Flash Memory 32 KB (ATmega328) of which 0.5 KB used by bootloader
09:17:47	reactormonk	SRAM 2 KB (ATmega328)
09:17:49	reactormonk	EEPROM 1 KB (ATmega328)
09:18:03	reactormonk	when you use the progmem macro, it goes to the flash. But iirc the flash is RO
09:19:43	*	Endy quit (Ping timeout: 264 seconds)
09:21:44	Araq	that's too little memory for a heap really
09:21:51	reactormonk	yeah.
09:22:03	reactormonk	it should store most literals in the flash memory
09:28:28	reactormonk	Araq, pushed the stuff I did onto my repo, need sleep.
09:28:46	reactormonk	cya in about... 10h
09:29:45	Araq	alright
09:58:44	*	zahary quit (Quit: Leaving.)
10:06:54	*	q66 joined #nimrod
10:09:16	*	xcombelle joined #nimrod
10:21:56	*	zahary joined #nimrod
11:09:00	*	zahary quit (Quit: Leaving.)
11:38:30	*	zahary joined #nimrod
11:48:13	*	Endy joined #nimrod
11:49:18	*	zahary quit (Quit: Leaving.)
11:54:37	*	zahary joined #nimrod
12:03:30	Araq	fowl: I know, working on it
14:05:12	*	OrionPK joined #nimrod
15:04:11	*	[1]Endy joined #nimrod
15:05:08	*	xcombelle quit (Remote host closed the connection)
15:06:34	NimBot	Araq/Nimrod 8ee185c Araq [+0 ±1 -0]: --os and --cpu work in configuration files
15:06:34	NimBot	Araq/Nimrod 4b57e2e Araq [+0 ±6 -0]: made some tests green
15:06:43	*	Endy quit (Ping timeout: 260 seconds)
15:06:44	*	[1]Endy is now known as Endy
15:21:55	*	zahary quit (Quit: Leaving.)
16:04:14	*	Endy quit (Ping timeout: 245 seconds)
16:07:34	dom96	Araq: it seems the mem profiler is broken.
16:08:04	Araq	dom96: we have a mem profiler?
16:08:24	dom96	Araq: Yeah?
16:08:41	dom96	http://build.nimrod-code.org/docs/estp.html#memory-profiler
16:10:14	Araq	well? how is it broken?
16:11:11	dom96	lib/system.nim(891, 17) Error: implementation of 'system.nimProfile()' expected
16:31:38	*	Endy joined #nimrod
16:44:09	Araq	dom96: about issue #443
16:44:36	Araq	I guess the 'fields' iterator shouldn't iterate over private fields
16:44:51	Araq	or maybe it should be granted special access
16:45:42	dom96	perhaps
16:46:38	Araq	well I dunno
16:46:55	Araq	it seems worse to ignore private fields
16:47:13	Araq	'fields' is for efficient serialization etc.
16:47:22	Araq	it can't ignore fields
16:48:37	Araq	on the other hand the current error message is not bad
17:05:40	*	[1]Endy joined #nimrod
17:07:40	*	Endy quit (Ping timeout: 246 seconds)
17:07:41	*	[1]Endy is now known as Endy
17:21:35	Araq	ping zahary_
17:34:30	NimBot	Araq/Nimrod fa174d0 Dominik Picheta [+0 ±4 -0]: Improved the performance of the SCGI module when dealing with multiple... 2 more lines
17:52:22	reactormonk	Araq, re.
17:53:02	dom96	Araq: So what about the memory profiler?
17:54:43	Araq	dom96: working on it
18:06:42	reactormonk	Araq, sooo what about the Port memory manager to your platform ? Simply don't implement one? There's not enough RAM anyway
18:09:13	Araq	yeah
18:11:01	reactormonk	and how do IU write that in nimrod?=
18:29:16	Araq	one possibility is to implement mmap et al as 'assert false'
18:50:30	*	Endy quit (Ping timeout: 264 seconds)
18:50:37	Araq	http://build.nimrod-code.org/docs/documentation.html
18:50:51	Araq	already is in the new design ... ;-)
18:51:52	dom96	:D
18:52:28	dom96	Moving the style and image files in with the output during generation proved to be a wise choice ;)
18:52:53	fowl	oo looks nice
18:53:21	Araq	fowl: make a pull request for your macro stuff please, add it to macros.nim
18:53:30	Araq	nobody wants a 2nd import for that anyway
18:53:54	fowl	ok
18:55:31	*	Trix[a]r_za is now known as Trixar_za
19:10:18	*	fowl quit (Ping timeout: 256 seconds)
19:23:27	*	fowl joined #nimrod
19:58:54	*	Trixar_za is now known as Trix[a]r_za
20:20:39	*	Trix[a]r_za is now known as Trixar_za
20:45:10	NimBot	Araq/Nimrod 7faaf67 Dominik Picheta [+0 ±1 -0]: Added news about the new website.
20:48:41	reactormonk	lib/system/alloc.nim(87, 10) Error: undeclared identifier: 'assert'
20:48:43	reactormonk	hmmm
20:49:11	fowl	what is "libmongoc.so" i installed mongo but dont have this ._.
20:49:24	reactormonk	looks like a C api to me
20:50:12	dom96	^
20:50:37	Araq	fowl: maybe a typo? is it libmongo.so ?
20:51:01	Araq	but hrm no the 'c' stands for client I think
20:51:07	fowl	i have libmongo-client.so
20:51:13	Araq	I compiled mongo from source I think
20:51:18	dom96	yeah, libmongoc is the C mongo Driver.
20:51:41	fowl	ah thats not in my repos
20:51:56	Araq	well I did test the wrapper for some project I can't remember
20:52:12	Araq	so it should work
20:52:37	*	q66_ joined #nimrod
20:53:27	*	zahary__ joined #nimrod
20:55:05	*	q66 quit (Ping timeout: 240 seconds)
20:56:15	*	zahary_ quit (Ping timeout: 260 seconds)
20:57:20	fowl	maybe it should be noted that it requires the c driver https://github.com/mongodb/mongo-c-driver
20:58:54	reactormonk	Error: system module needs 'copyStringRC1'
20:58:56	reactormonk	argh
20:59:37	Araq	reactormonk: I checked if I can improve that error message
20:59:45	Araq	unfortunately I can't easily
21:01:54	fowl	*** stack smashing detected ***: /tmp/aporia/a15 terminated
21:02:11	fowl	on the db_mongo example
21:03:26	reactormonk	hmm
21:06:11	reactormonk	Araq, well, copying strings around is not that useful
21:06:27	Araq	yeah indeed
21:06:35	Araq	so you need to disable the code that does that
21:10:16	fowl	WARNING: mongo_connect() is deprecated, please use mongo_client()
21:16:35	reactormonk	Araq, where do I define my own symbols, like noDynmaicAlloc?
21:18:02	Araq	you don't
21:18:15	Araq	you check with 'defined' and pass it on the command line
21:18:16	reactormonk	hm
21:18:29	Araq	that means it doesn't notice your typos :P
21:18:29	reactormonk	lib/system/alloc.nim(84, 19) Error: undeclared identifier: 'noDynamicAlloc'
21:18:50	Araq	when defined(noDynamicAlloc)
21:19:39	reactormonk	oh.
21:20:15	reactormonk	gotta say, I don't know where to disable the copyStringRC1, I'd have to mess with the compiler and add something in there.
21:20:19	reactormonk	compiler/ccgexprs.nim: linefmt(p, cpsStmts, "$3 = $1; $1 = #copyStringRC1($2);$n",
21:21:32	Araq	just add some Warning(e.info, warnUser, "copystring used from here")
21:21:40	Araq	and bootstrap
21:38:46	Araq	fowl: sorry, I made some comments for your macro changes
21:42:11	NimBot	Araq/Nimrod cd96412 Araq [+1 ±13 -2]: made some tests green; implemented 'from module import nil'
21:42:11	NimBot	Araq/Nimrod 3366925 Araq [+0 ±4 -0]: Merge branch 'master' of github.com:Araq/Nimrod
21:42:11	NimBot	Araq/Nimrod 3f2486f Araq [+0 ±4 -0]: 'from' statement documented
21:42:47	*	q66_ is now known as q66
21:47:20	NimBot	nimrod-code/nimbuild c3715b1 Dominik Picheta [+0 ±1 -0]: Some fixes for the builder.
21:47:20	NimBot	nimrod-code/nimbuild 969cbd6 Dominik Picheta [+0 ±1 -0]: Merge branch 'master' of github.com:nimrod-code/nimbuild
21:51:18	reactormonk	Araq, well, I assume the compiler needs that
21:53:33	reactormonk	and warning() doesn't work inside the compiler, oh well
21:53:39	Araq	reactormonk: yeah you get meaningless warning during bootstrapping
21:53:47	fowl	there should be a suggestion to use --depth 1 when cloning nimrod, it make git only pull the minimum needed and not the whole (huge) repos and history
21:54:10	Araq	there is a --depth 1 command? o.O
21:54:17	Araq	lol I didn't know that
21:54:33	fowl	yea you can use it with clone or pull and it speeds it up quite a bit
21:56:33	reactormonk	Araq, compiler/ccgexprs.nim(280, 15) Error: undeclared identifier: 'Warning'
21:56:36	reactormonk	FAILURE
21:57:35	Araq	s/Warning/Message
21:57:35	fowl	Araq: i didnt like !! either, just used it since ! was already for string->ident
21:57:43	fowl	string -> tident
21:58:20	Araq	yeah we should deprecate '!' instead I think
22:01:35	reactormonk	compiler/ccgexprs.nim(280, 18) Error: undeclared field: 'info'
22:02:00	Araq	that's where you need to use your brain
22:02:14	Araq	the node is likely called 'n' or 'e'
22:02:19	Araq	so it's n.info or e.info
22:07:12	reactormonk	proc genAssignment(p: BProc, dest, src: TLoc, flags: TAssignmentFlags) =
22:07:19	reactormonk	doesn't work on `p` either
22:07:34	fowl	Araq: changing the sets into procs isRoutine/isAtomic is annoying, i use them in `of` branches
22:07:50	NimBot	nimrod-code/nimbuild 4ba6618 Dominik Picheta [+0 ±1 -0]: Another fix for the build process.
22:07:53	Araq	fowl: aye make them sets then but not templates
22:08:04	fowl	ok
22:08:49	fowl	"RoutineNodes"/"AtomicNodes" good?
22:10:02	Araq	yeah
22:10:10	Araq	we also desparately need CallNodes
22:10:41	Araq	= {nnkCall, nnkInfix, nnkPrefix, nnkPostfix,
22:10:43	Araq	nnkCommand, nnkCallStrLit, nnkHiddenCallConv}
22:11:34	*	Trixar_za is now known as Trix[a]r_za
22:11:44	Araq	btw AtomicNodes = {nnkNone..nnkNilLit}
22:11:47	fowl	isnt there one for macro calls too
22:13:25	Araq	I think zahary__ removed it as it's now translated into a 'do'
22:13:56	fowl	what's a postfix call
22:14:14	Araq	a*
22:14:40	fowl	thats allowed?
22:14:46	Araq	it's not part of the concrete syntax for now ;-)
22:14:59	Araq	but it may be one day
22:15:20	fowl	o
22:17:02	reactormonk	Araq, there is no node in that scope
22:17:25	Araq	reactormonk: use p.prc.info then
22:18:33	reactormonk	http://sprunge.us/IPjO
22:19:46	Araq	jezz only do it if p.prc != nil then
22:19:59	fowl	proc ident*(n: PNimrodNode): TNimrodIdent {.magic: "NIdent".}
22:20:36	fowl	wait no thats not causing problems
22:20:53	reactormonk	what's `prc`?
22:21:29	reactormonk	raise newException(EBase, "dynamic memory allocation not supported.") <- uses copyStringRC1
22:21:46	Araq	quite ironic, hmm?
22:21:51	Araq	XD
22:22:00	reactormonk	.... :-(
22:22:05	Araq	but yeah you shouldn't use exceptions if you have no heap ...
22:22:15	reactormonk	figured so
22:22:25	reactormonk	any idea how to do exception handling? Go-style with return codes?
22:22:41	reactormonk	or an either type?
22:22:43	Araq	just make sure you have some 'quit'
22:22:54	Araq	and some display to display an error code
22:26:47	reactormonk	there is no display.
22:26:55	reactormonk	well, serial port
22:27:44	Araq	well you better teach system.nim how to output *something* and quit
22:28:01	Araq	otherwise you may as well start proving all your array bounds correct etc.
22:28:09	reactormonk	:-)
22:31:52	reactormonk	let's see
22:32:06	reactormonk	Araq, any way I can do that without hacking system.nim?
22:32:20	Araq	no, sorry
22:32:30	Araq	I figured system.nim ain't change friendly
22:32:43	Araq	for embedded stuff
22:32:58	Araq	but it's hard to improve the situation
22:34:17	Araq	on the other hand you almost always get nice compile time errors
22:36:15	Araq	fowl: do you use widestrs.nim?
22:36:54	reactormonk	can I at least fuck with in in embedded.nim?
22:37:11	Araq	reactormonk: I think so
22:38:37	*	q66 quit (Remote host closed the connection)
22:41:32	reactormonk	oh god...
22:42:19	reactormonk	Araq, so no exceptions without dynamic allocation?
22:42:57	Araq	what would you do with them anyway?
22:44:13	reactormonk	hmm, either a quit or link them statically somewhere... I assume an Either type would be useful here
22:44:59	reactormonk	can you check at compile time if you can have exceptions? Wait, OOM n stuff can occur basically everywhere
22:45:23	reactormonk	except there's no OOM without dynamic allocation, is there?
22:45:54	Araq	of course there is
22:46:01	Araq	you can always run out of stack space
22:46:08	reactormonk	right
22:46:34	reactormonk	so what to do with exceptions? Don't use them and instead just quit?
22:46:45	reactormonk	oh, and according to the people in #arduino, there is no quit
22:47:02	Araq	how do you tell the CPU to stop then?
22:47:15	reactormonk	you can sleep it
22:47:19	Araq	while true: then?
22:48:03	reactormonk	reactormonk | how do you stop the processor? aka an assertion failed or similar.
22:48:11	reactormonk	Yotson | reactormonk: ehm, you don't.
22:48:29	reactormonk	Yotson | reactormonk: yup, as last resort you could reset the controller.
22:48:34	reactormonk	Yotson | reactormonk: you have programming experience on/under operating systems? forget most of that. lol
22:49:32	Araq	well I suppose getting some error message out of the device is all you really need
22:49:48	Araq	then you can reset it manually
22:51:14	Araq	my mindstorm can output things like E42 on its 4 char display :P
22:51:24	Araq	and make some beep
22:52:18	reactormonk	http://sprunge.us/jSca
22:53:03	Araq	great then use Serial.println()
22:54:26	Araq	you should get a mindstorm instead; it has a display :P
22:55:01	reactormonk	^^
22:55:16	reactormonk	does it have 10 pins each as well?
22:55:39	Araq	even better: It has Lego pins
22:56:32	Araq	you can use it to control your Lego motors
22:56:54	reactormonk	yeah yeah ^^
22:57:25	reactormonk	so to integrate it into system.nim, I have to put it somewhere into arduino.nim in lib/system/ ?
22:58:03	Araq	yeah why not
23:03:17	reactormonk	don't like that, but ok
23:03:36	Araq	you can also rewrite system.nim and put it to some different dir
23:04:33	Araq	and use the --lib:DIR option to override the default dir
23:10:26	reactormonk	and reimplement the whole system.nim?
23:10:38	Araq	yeah
23:10:45	Araq	maybe you like that better :P
23:11:02	reactormonk	no way to hook into system.nim procs and overwrite them? ;-)
23:11:12	Araq	the headers should be categorized and put into separate includes
23:11:29	Araq	then we can have multple system.nim that cherrypick what's available
23:11:31	reactormonk	should. Oh well, fuck it, let's write that arduino.nim
23:11:58	Araq	but not now, release is overdue
23:14:02	reactormonk	sure, I'll not merge it
23:16:14	fowl	Araq: no
23:16:20	fowl	irt widestrs
23:17:37	fowl	i didnt know it existed
23:25:44	Araq	good cause I'm changing the interface
23:27:59	fowl	i updated the pr
23:28:45	fowl	oh i didnt mean to deprecate `!`
23:28:59	fowl	i dont know what to replace it with
23:29:30	Araq	proc ident*(...)
23:29:42	Araq	but this needs your judgement
23:29:55	Araq	is !"abc" common?
23:29:59	fowl	thought you wanted ident*(s: string): PNimrodNode to replace !!"foo"
23:30:20	Araq	well I dunno
23:30:30	Araq	that would work too ...
23:30:49	Araq	you really should use 'bindSym' in macros anyway
23:31:03	Araq	this identifier stuff is fragile
23:31:26	fowl	i think it should be ~"foo" for newidentnode("foo")
23:31:49	fowl	i meant to change it to ~ hellas ago but never got around to it
23:34:39	NimBot	Araq/Nimrod eda8a96 Dominik Picheta [+0 ±1 -0]: Updated news.