IRC log for #perl6, 2010-04-29

Time	Nick	Message
00:00	sorear	hmm
00:01	sorear	$perl5string ~ $perl5string doesn't work
00:02	sorear	because it's defined as our multi sub infix:<~>($a, $b) {
00:02	sorear	and the Rakudo multisub dispatcher doesn't conster foreign objects suitable for 'any'
00:03	sorear	rakudo: multi sub foo($x) { say($x); }; foo(pir::new__PS('String'))
00:03	p6eval	rakudo 72f914: OUTPUT«␤»
00:03	sorear	rakudo: multi sub foo($x) { say($x); }; foo(pir::new__PS('Hash'))
00:03	p6eval	rakudo 72f914: OUTPUT«No applicable candidates found to dispatch to for 'foo'. Available candidates are:␤:(Any $x)␤␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
00:03	sorear	rakudo: multi sub foo($x) { say($x); }; say(pir::new__PS('Hash'))
00:03	p6eval	rakudo 72f914: OUTPUT«Hash[0x53815a8]␤»
00:03	sorear	rakudo: sub foo($x) { say($x); }; foo(pir::new__PS('Hash'))
00:03	p6eval	rakudo 72f914: OUTPUT«Null PMC access in invoke()␤current instr.: 'foo' pc 191 (EVAL_1:81)␤... call repeated 1 times␤»
00:04	sorear	Rakudobug?
00:05	pugssvn	r30495 \| lwall++ \| [Cursor] lexer generator mislaid a bare . pattern so cursor_fate never called it
00:05	pugssvn	r30495 \| [Cursor] added .looks_like_cclass method to detect accidental use of P5 ranges
00:05	pugssvn	r30495 \| [STD] various character classes needed to backslash #
00:05	pugssvn	r30495 \| [STD] change :tr language to :cc langauge since character classes share it
00:05	pugssvn	r30495 \| [STD] remove old-school-ish character classes
00:05	pugssvn	r30495 \| [STD] allow \# in character classes instead of misparsing as unspace
00:06	pugssvn	r30495 \| [STD] remove unused %*LANG<Trans>
00:06	pugssvn	r30495 \| [STD] on invalid - in regex, presume we're in an old-school character class
00:06	pugssvn	r30495 \| [STD] check a normal regex bracket's innards for old-school character class, and warn if found
00:06	pugssvn	r30495 \| [STD] use ~ for regex brackets to set $*GOAL correctly
00:06	pugssvn	r30495 \| [STD] clean up recursive panic detection to avoid both false positives and negatives
00:06	pugssvn	r30495 \| [STD] don't use 'note' to emit a panic inside a suppose
00:06	pugssvn	r30495 \| [STD] suppress duplicate sorry messages
00:06	pugssvn	r30495 \| [STD] sorry no longer uses panic in supposition, but dies directly
00:14	sjohnson	i had some fun programming in Perl today
00:15	sorear	6?
00:18		[Coke] joined #perl6
00:18		cosimo joined #perl6
00:18		slavik joined #perl6
00:22		stepnem joined #perl6
00:23		yinyin joined #perl6
00:29	sjohnson	sorear: p5 ... writing some interfaces for git status outputs to make my life easier...
00:29	sjohnson	i'll try out some more p6 soon once i get through all the apocolypses
00:38		rv2733 joined #perl6
00:44	sjohnson	my knowledge of perl is limited to only doing c00l things with p5
00:45	colomon	you don't have to get through all the apocalypses, you know.
00:46	sorear	the apocalypses are mostly irrelevant now for practical purposes
00:46	sorear	they're good if you need a more detailed rationale
00:47	sorear	although I prefer the flesh version if it's available
00:48		meppl joined #perl6
00:49	sjohnson	not sure what the most complete p5->p6 transition guide is
00:49	sjohnson	maybe of your helpful ppl can recommend something for me
00:49	sorear	moritz' 5->6 maybe?
00:50	sorear	hey, it's even linked from perl6.org
00:50	PerlJam	http://perlgeek.de/blog-en/perl-5-to-6/
00:50	PerlJam	mortiz++
00:51	PerlJam	oops, my link has an extra slash at the end.
00:51	PerlJam	sjohnson: for specific items, you can ask here when people are active.
00:51	* sorear	, having finished the core functionality in blizkost, goes back to hacking his irc client (which is what needed :from<perl5> in the first place, for DBI, Curses, and POSIX)
00:51	sorear	which is all the time.
00:52	snarkyboojum	sorear++
00:53		orafu joined #perl6
00:56	sjohnson	thanks
00:57	sjohnson	moritz_: typo on word "Aprili" on latest rakudo moscow news page
01:17	dalek	rakudo: 3966c62 \| (Solomon Foster)++ \| src/core/ (2 files):
01:17	dalek	rakudo: Straighten out type issues with Real.exp. In the process, add Real * Real, Complex Real, and Real Complex.
01:17	dalek	rakudo: review: http://github.com/rakudo/rakud[…]0301dba2546134af5
01:17	dalek	rakudo: 0857098 \| (Solomon Foster)++ \| src/core/ (2 files):
01:17	dalek	rakudo: Remove Complex Any and Any Complex, as they should no longer be needed. Change Any * Any to convert its arguments to Numeric and call multiply on them then.
01:17	dalek	rakudo: review: http://github.com/rakudo/rakud[…]f3e1852263f173c10
01:18	pugssvn	r30496 \| colomon++ \| [t/spec] Additional exp tests using mixed argument types.
01:24	sorear	hmm
01:24	diakopter	hmm
01:25	sorear	rakudo: sub foo(:$bar) {}; foo("bar" => 1);
01:25	p6eval	rakudo 72f914: OUTPUT«Too many positional parameters passed; got 1 but expected 0␤current instr.: 'foo' pc 197 (EVAL_1:80)␤»
01:25	sorear	Is it possible to pass a named argument with weird characters in p6?
01:25	sorear	like initial "-" (quite common in perl5 APIs)
01:25	diakopter	rakudo: my method foo(:$bar) {}; foo("bar" => 1);
01:25	p6eval	rakudo 72f914: ( no output )
01:26	diakopter	rakudo: my method foo(:$bar) {}; foo("bar" => 1, "foo" => 2);
01:26	p6eval	rakudo 72f914: OUTPUT«Too many positional parameters passed; got 2 but expected 1␤current instr.: 'foo' pc 338 (EVAL_1:133)␤»
01:26	diakopter	rakudo: my method foo(:$bar) { say $_; }; foo("bar" => 1);
01:26	p6eval	rakudo 72f914: OUTPUT«Any()␤»
01:27	* PerlJam	wonders what class or object that method belongs to
01:27	sorear	diakopter: the pair is being bound to 'self'
01:27	sorear	PerlJam: none
01:27	sorear	my method foo( is sugar for my sub foo(self,
01:27	sorear	except that it's valid syntax
01:28	sorear	methods in perl 6 are simply subs with a hidden first positional argument
01:29	snarkyboojum	sorear: are you working on a Perl 6 port of soric?
01:29	sorear	ok now that's just freaky.
01:30	sorear	soric is the collective code name for three broken prototypes I've played with
01:30	snarkyboojum	oh ok
01:30	snarkyboojum	good-o
01:30	sorear	#4 is being done in perl6 based on several insights from #3
01:30	sorear	how did you know about soric?
01:30	sorear	the name, I mean
01:30	snarkyboojum	github
01:30	snarkyboojum	not terribly hard to find out :)
01:30	sorear	how did I forget :/
01:31	snarkyboojum	github lurkers unite!
01:31	sorear	anyway, yes
01:32	snarkyboojum	cool, looks interesting
01:32	sorear	#3 was going along nicely until I realized that 90% of my code consisted of emulations of the new .* multicall syntax
01:32	sorear	at which point I decided to learn perl 6
01:34	snarkyboojum	at which point you decided to help get blizkost going :)
01:34	snarkyboojum	nice progression :)
01:37		bradb joined #perl6
01:38	colomon	"will review Solomon Foster's Mandlebrot example, especially with regard to performance" -- woot!
01:38	diakopter	phenny: tell pmurias I fixed ur buggo
01:38	phenny	diakopter: I'll pass that on when pmurias is around.
01:38	diakopter	phenny: tell pmurias by rejiggering the grammar
01:38	phenny	diakopter: I'll pass that on when pmurias is around.
01:38		orafu joined #perl6
01:38	diakopter	phenny: tell pmurias good luck :)
01:38	phenny	diakopter: I'll pass that on when pmurias is around.
01:44		dalek left #perl6
01:44		dalek joined #perl6
01:45	diakopter	phenny: tell masak k, yapsi notifications should work now. it would help if I didn't try to load "/yapsilog.com" from botnix.conf and tried "/yapsi.pm" instead. :/
01:45	phenny	diakopter: I'll pass that on when masak is around.
01:53		sorear joined #perl6
01:55	sorear	please don't op me without also giving out an ACL bit
01:55	sorear	it... won't last
01:55	diakopter	heh
01:56	diakopter	we actually don't have access to that :( I need to get TimToady to fill out the freenode form and mail it in
01:57	diakopter	so that the channel can get registered again and get an actual "founder"
01:58	sorear	I'm really looking forward to IPv6 in the US
01:58	sorear	and by extension the end of unstable NATs
01:59		Psyche^ joined #perl6
02:00	diakopter	urgh. silly dalek.
02:01	diakopter	it just missed my http://code.google.com/p/csmet[…]urce/detail?r=220
02:01	sorear	alternatively, I could simply move to a country with real internet
02:02	diakopter	yapsi: say 5
02:02	diakopter	perlesque: say(4)
02:02	p6eval	perlesque: OUTPUT«4␤»
02:02	diakopter	perlesquel: .
02:02	p6eval	perlesquel: OUTPUT«Cannot open assembly 'asmbly_1.exe': No such file or directory.␤Command exited with non-zero status 2␤real 0.23␤user 0.01␤sys 0.01␤»
02:03	diakopter	heh
02:04	TimToady	std: /[a-z]/
02:04	p6eval	std 30496: OUTPUT«===[0mSORRY!===[0m␤Invalid regex metacharacter; for a character class, use <[...]> instead of [...]␤ (and use .. instead of - to indicate a range) at /tmp/0HSYFtJNOt line 1:␤------> /[a-⏏z]/␤Unable to parse bracketed regex at /tmp/0HSYFtJNOt line
02:04	p6eval	..1:␤…
02:04	TimToady	std: /[aeiou]/
02:04	p6eval	std 30496: OUTPUT«Potential difficulties:␤ This appears to be an old-school character class; please use <[aeiou]> if you␤ mean a character class, or put whitespace inside like [ aeiou ] to disable␤ this warning at /tmp/r4tAr7xqIJ line 1:␤------> /⏏[aeiou]/␤ok 00:01 108m␤»
02:05	PerlJam	TimToady++
02:05	diakopter	std: /[ aeiou]/
02:05	p6eval	std 30496: OUTPUT«Potential difficulties:␤ This appears to be an old-school character class; please use <[ aeiou]> if you␤ mean a character class, or put whitespace inside like [ aeiou ] to disable␤ this warning at /tmp/rF4lulyzxZ line 1:␤------> /⏏[ aeiou]/␤ok 00:01
02:05	p6eval	..108m␤»
02:05	diakopter	std: /[ aeio u]/
02:05	p6eval	std 30496: OUTPUT«ok 00:01 108m␤»
02:06	diakopter	std: /[aeio u]/
02:06	p6eval	std 30496: OUTPUT«Potential difficulties:␤ This appears to be an old-school character class; please use <[aeio u]> if you␤ mean a character class, or put whitespace inside like [ aeio u ] to disable␤ this warning at /tmp/PTnJmJVgN7 line 1:␤------> /⏏[aeio u]/␤ok 00:01
02:06	p6eval	..108m␤»
02:06	diakopter	std: /[aeio u ]/
02:06	p6eval	std 30496: OUTPUT«ok 00:01 108m␤»
02:06	diakopter	heh
02:06	TimToady	std: /[ \t]/
02:06	p6eval	std 30496: OUTPUT«ok 00:01 109m␤»
02:06	TimToady	hmm, that one didn't work
02:06	diakopter	std: /[ \\t]/
02:06	TimToady	supposed to recommend \h
02:06	p6eval	std 30496: OUTPUT«ok 00:01 108m␤»
02:06	diakopter	std: /[ \\\t]/
02:06	p6eval	std 30496: OUTPUT«ok 00:01 108m␤»
02:06	TimToady	std: /[ ]/
02:06	diakopter	std: /[\t ]/
02:06	p6eval	std 30496: OUTPUT«===[0mSORRY!===[0m␤Null pattern not allowed at /tmp/qXb0RC3H5V line 1:␤------> /[ ⏏]/␤Unrecognized regex metacharacter (must be quoted to match literally) at /tmp/qXb0RC3H5V line 1:␤------> /[⏏ ]/␤Can't call method "_REDUCE" on
02:06	p6eval	..unbl…
02:07	p6eval	std 30496: OUTPUT«ok 00:01 108m␤»
02:07	PerlJam	std: /[\t ]/
02:07	p6eval	std 30496: OUTPUT«ok 00:01 108m␤»
02:07	diakopter	std: /[\t]/
02:07	p6eval	std 30496: OUTPUT«ok 00:01 108m␤»
02:07	diakopter	std: /[ t]/
02:07	p6eval	std 30496: OUTPUT«Potential difficulties:␤ This appears to be an old-school character class; please use <[ t]> if you␤ mean a character class, or put whitespace inside like [ t ] to disable␤ this warning at /tmp/29NyFiT1G9 line 1:␤------> /⏏[ t]/␤ok 00:01 108m␤»
02:07	diakopter	std: /[ t ]/
02:07	p6eval	std 30496: OUTPUT«ok 00:01 108m␤»
02:08	TimToady	std: /[\n]/
02:08	p6eval	std 30496: OUTPUT«ok 00:01 108m␤»
02:08	TimToady	hmm
02:08	PerlJam	TimToady: is { $a = 5; say $a; my $a } correct and analogous to { foo; sub foo { say "hi" } } ?
02:08	PerlJam	rakudo: { $a = 5; say $a; my $a }
02:08	p6eval	rakudo 085709: OUTPUT«5␤»
02:08	PerlJam	rakudo: { foo; sub foo { say "hi" } }
02:08	p6eval	rakudo 085709: OUTPUT«hi␤»
02:10		ash__ joined #perl6
02:11		ash__ left #perl6
02:11	diakopter	rakudo: { our sub foo { say "hi" } }(foo)
02:11	p6eval	rakudo 085709: OUTPUT«hi␤»
02:12	* diakopter	whimpers
02:12	diakopter	er, simpers
02:12	TimToady	PerlJam: no
02:13	diakopter	std: { $a = 5; say $a; my $a }
02:13	p6eval	std 30496: OUTPUT«===[0mSORRY!===[0m␤Variable $a is not predeclared at /tmp/2WwonoU5ya line 1:␤------> { $a⏏ = 5; say $a; my $a }␤Variable $a is not predeclared at /tmp/2WwonoU5ya line 1:␤------> { $a = 5; say $a⏏; my $a }␤Check failed␤FAILED 00:01
02:13	p6eval	..1…
02:13	pugssvn	r30497 \| colomon++ \| [t/spec] Unfudge a bunch of comb tests which now work.
02:13	colomon	31740 passing tests now.
02:15	dalek	rakudo: ed321b4 \| (Solomon Foster)++ \| src/core/Cool-str.pm:
02:15	dalek	rakudo: Fix comb version of the match with take bug fixed earlier today in split.
02:15	dalek	rakudo: review: http://github.com/rakudo/rakud[…]73d680a152e5de0e4
02:15	PerlJam	TimToady: this is what I expected but, then I also tended to expect { XXX; my XXX } in general to carp. Would it be safe to say that it only carps if XXX is a variable?
02:16	TimToady	only functions may be post-declared
02:16	sorear	it's kinda like ANSI C
02:16	sorear	when an undeclared function is seen, a declaration is automatically generated assuming that it has listop precedence
02:17	sorear	so in the common case, postdeclaring works
02:17	TimToady	well, all C functions are global, to the first approximation, and the call syntax is relatively unambiguous
02:17	sorear	Haskell allows postdeclaration of functions with /any/ precedence
02:18	sorear	which makes the parsing algorithm completely insane
02:18	TimToady	probably forces multiple passes
02:18	* diakopter	murmurs about the insanity defense of accusations of insanity
02:18	PerlJam	I understand it, but my brain isn't accepting it for some reason. I used to think that "my XXX" meant than XXX existed from that point forward to the end of the inner-most enclosing scope. Having functions not do that and be lexical is bothersome
02:19	PerlJam	s/than/that/
02:19	sorear	for a long time I thought it was undecidable, but eventually I came up with an EXPTIME algorithm which tentatively assigns each possible combination of precedences before trying to parse
02:19	TimToady	meeting &
02:19	sorear	TimToady: In practice, yes
02:19	diakopter	std: foo(); { our sub foo { say "hi" } };
02:19	p6eval	std 30496: OUTPUT«ok 00:01 108m␤»
02:19	sorear	TimToady: In theory, the language is not sufficiently self-synchronizing for a multi-pass parser to work
02:20	sorear	but the heuristics work ok on real code
02:20	diakopter	nobody's created types whose dependency knots can't be untangled?
02:22	sorear	diakopter: there are only a finite number of possible untanglings, and you can just check each one for consistency
02:22	diakopter	ok.
02:38		bradb joined #perl6
02:55	dalek	rakudo: e16cf45 \| duff++ \| tools/contributors.pl:
02:55	dalek	rakudo: [tools] fix typo and output sorted keys
02:55	dalek	rakudo: review: http://github.com/rakudo/rakud[…]5d4c4a7592334f89b
02:58		JimmyZ joined #perl6
03:10		ingy joined #perl6
03:20		_jaldhar joined #perl6
03:20		ingy joined #perl6
03:37		crythias joined #perl6
03:47		ingy joined #perl6
03:49		molaf joined #perl6
03:52		jaldhar_ joined #perl6
04:04	dalek	csmeta: r221 \| diakopter++ \| trunk/Sprixel/ (5 files):
04:04	dalek	csmeta: [perlesque] lots o' fixes :)
04:04	dalek	csmeta: review: http://code.google.com/p/csmet[…]urce/detail?r=221
04:14		bradb joined #perl6
04:26		mikehh_ joined #perl6
04:38	dalek	csmeta: r222 \| diakopter++ \| trunk/Sprixel/ (4 files):
04:38	dalek	csmeta: [perlesque] more fixes. stopped using the trampoline for now.
04:38	dalek	csmeta: review: http://code.google.com/p/csmet[…]urce/detail?r=222
04:40		snarkyboojum joined #perl6
04:40		ingy joined #perl6
04:40	diakopter	phenny: tell pmurias I can re-enable the trampoline when a possibly-recursive invocation is detected, but the addl cost will be another slot for the return value per callsite in each recursive frame.
04:41	phenny	diakopter: I'll pass that on when pmurias is around.
04:41	diakopter	that, and I didn't want to have to implement that additional callsite-slot generation tonight :D
04:41	diakopter	phenny: tell pmurias that, and I didn't want to have to implement that additional callsite-slot generation tonight :D
04:41	phenny	diakopter: I'll pass that on when pmurias is around. I'll have to use a pastebin, though, so your message may get lost.
04:42	diakopter	phenny: bleh
04:43	sorear	What
04:43	sorear	phenny: help
04:43	phenny	sorear: Hi, I'm a bot. Say ".commands" to me in private for a list of my commands, or see http://inamidst.com/phenny/ for more general details. My owner is sbp.
04:44		alester joined #perl6
04:45	sorear	ah
04:46	sorear	phenny keeps all messages, but pastes any after 4
04:46	sorear	using paste.lisp.org and a hardcoded captcha
04:46	* sorear	plugs App::Nopaste
04:53		ash__ joined #perl6
04:56		jaldhar_ joined #perl6
05:01		_jaldhar_ joined #perl6
05:03		ReiniUrban joined #perl6
05:04	diakopter	sorear: "What"?
05:05	sorear	diakopter: phenny's comment about a pastebin
05:08	diakopter	oh
05:11	diakopter	actually...
05:21		meppl joined #perl6
05:29		[particle]1 joined #perl6
05:37	dalek	csmeta: r223 \| diakopter++ \| trunk/Sprixel/ (3 files):
05:37	dalek	csmeta: more parsing fixes.
05:37	dalek	csmeta: review: http://code.google.com/p/csmet[…]urce/detail?r=223
05:38		redicaps joined #perl6
05:41	diakopter	sorear: around?
05:41		agentzh joined #perl6
05:42	diakopter	I added a "frame" keyword that snags a ref to the current frame :)
05:44	sorear	cute.
05:45	* sorear	remembers implementing closures in Pascal using frame pointers and horrible function reference casting
05:47	* snarkyboojum	just discovered cd ~- in bash
05:50	moritz_	good morning
05:50	snarkyboojum	oh it's just cd -
05:50	diakopter	zugh
05:50	snarkyboojum	:\|
05:50	snarkyboojum	mortiz_ o/
05:50	snarkyboojum	moritz_ o/
05:50	snarkyboojum	:)
05:50	moritz_	diakopter: like &?BLOCK?
05:50	moritz_	\o
05:51	diakopter	maybe
05:51	diakopter	what does &?BLOCK? mean
05:51		bradb joined #perl6
05:52	moritz_	&?BLOCK is the reference to the current block
05:53	diakopter	what does "block" mean.. (what can you do with it?)
05:53		kaare joined #perl6
05:53	moritz_	block = { ... }
05:54	moritz_	you can store it, invoke it, introspect it
05:55	sorear	kinda like RECURSE in FORTH?
05:55	diakopter	the keyword I added is "the instance of the current block that represents thisexecution* of the current block, including all locals"
05:55	sorear	diakopter: in parrotland we call those CallContexts
05:56	moritz_	and in Perl 6 callframe
05:56	moritz_	(though it's a bit more general)
05:56	diakopter	ok
05:56	diakopter	what can you do with a Perl 6 callframe?
05:58		viklund joined #perl6
05:58	moritz_	search outwards, obtain file and line number etc
05:58	moritz_	see S06/The callframe and caller functions/
06:01	* diakopter	is reminded of "what do you get when you guzzle down sweets?"
06:01	viklund	"...Eating as much as an elephant eats"
06:03	snarkyboojum	would be cool to see Perl 6 on this list http://emerginglangs.com/speakers/ :)
06:03	snarkyboojum	and/or Parrot :)
06:04	diakopter	parrot ain't a language
06:04	snarkyboojum	true, but as an emerging VM :P
06:05	snarkyboojum	they're keen to see VM people there too apparently
06:08	diakopter	Yeah, I guess since PyPy & Cola are on there.
06:08	sorear	Do we want to be associated with that crowd?
06:09	diakopter	TimToady said that a whole conference-long track dedicated to Perl is enough... and that Perl isn't an emerging language
06:09	snarkyboojum	it's not?
06:09	snarkyboojum	ok then
06:10	snarkyboojum	what crowd is that sorear?
06:11	diakopter	the in-your-face lets-debate-whose-favorite-language-is-better crowd?
06:15	sorear	yes
06:15	snarkyboojum	I didn't think that's what this was about - perhaps I'm just naive :)
06:15	avar	the speakers don't sound like the sort that'll be "lets-debate-whose-favorite-language-is-better"
06:16	avar	e.g. Rich Hickey and Rob Pike
06:16	diakopter	not the speakers; the attendees :P
06:17	snarkyboojum	diakopter: programmers will do that regardless won't they? :)
06:17	avar	I think you're being too pessamistic
06:17	avar	Anyway, would be nice to have Perl 6 there too
06:17		uniejo joined #perl6
06:22	dalek	book: 971d45f \| moritz++ \| src/subs-n-sigs.pod:
06:22	dalek	book: [subs] make that parameter named
06:22	dalek	book: review: http://github.com/perl6/book/c[…]a2ade1ff30a91bd80
06:24		szabgab joined #perl6
06:32		iblechbot joined #perl6
06:33		kst joined #perl6
06:57	dalek	yapsi: 133495c \| snarkyboojum++ \| t/runtime.t:
06:57	dalek	yapsi: Test to ensure blocks at the same level as variable definitons don't
06:57	dalek	yapsi: trounce variable values.
06:57	dalek	yapsi: review: http://github.com/masak/yapsi/[…]810e00b39d29c16f2
07:31		eternaleye joined #perl6
07:33	sorear	It might be an interesting project to look through MooseX::* on CPAN and see how many apply to Perl6
07:38	JimmyZ	Heh
07:41		snarkyboojum joined #perl6
07:44	sorear	moose has actually diverged quite a bit from perl6
07:45	sorear	we might need to start stealing stuff from them - they've got some cool bits like variable attribute representations
07:46	viklund	how much of the divergence is bacause of adjustments to how perl5 works and how much is divergence from discovering "best practices", would be interesting to know...
07:47	viklund	what is variable attribute representations?
07:47	moritz_	sorear: re named arguments with weird names: you can always do foo(\|%h)
07:48	moritz_	sorear: where %h can has arbitrary keys
07:48	* moritz_	backlogs
07:55	moritz_	Herve++ posted an interesting comment on my blog
07:55	moritz_	he suggested to change the triangle form of reduce from [\+] to [@+]
07:56	moritz_	with the rationale that @ indicates a list-ish return type
07:56	moritz_	I wonder if we could even change to that @[+] to disambiguate infix operators starting with @
07:57	moritz_	TimToady: what do you think about that? (when you backlog :-)
08:01	sorear	viklund: in perl6, attributes are just variables. in Moose, attributes have a meta object with overridable accessor code
08:01	sorear	so you can have attributes that do fancy things when written to, etc
08:02	moritz_	you can have that too in Perl 6
08:02	sorear	moritz_: You could hear that behind the noise of "OMGZ METAOPS ARE WRITE ONLY"?
08:02	viklund	ahh, yes I knew that, just hadn't heard the phrase before...
08:02	moritz_	if you use a modified storage backend
08:02	viklund	sorear: thx
08:02	moritz_	sorear: I try hard to focus on the constructive part of the critic
08:02	sorear	I think the divergence is because the Moose cabal has given up on Perl 6 and decided to drop all connection with the Synopses
08:02		bbkr_ joined #perl6
08:02	sorear	but then I'm just a cynic
08:03	moritz_	sorear: for sure the Moose folks chose whatever fits best for Perl 5
08:04	sorear	moritz_: can you give me a short code sample for that?
08:04	moritz_	sorear: and introducing $!foo variables isn't easy in Perl 5 at all :-)
08:04	moritz_	sorear: no
08:05	moritz_	it would have to be some sort of object being passed as the REPR to YourClass.bless
08:05	moritz_	which has FETCH and STORE methods
08:05	sorear	$!foo is hard-wired to access the object representation
08:05	moritz_	that do the fancy stuff
08:05	moritz_	yes, but the object representation isn't hard wired
08:05	sorear	with pervasive accessors, you get an extra layer of indirection
08:05	sorear	which makes this absolutely trivial
08:05	sorear	has 'foo' => ( trigger => $SUB )
08:05	sorear	just sets an attribute on the attribute
08:06	sorear	which changes code generation
08:06	sorear	I'm not following the Perl 6 idea of "our approach must be better because it uses $! variables"
08:06	moritz_	well, you can still use lvalue accessors if that's what you want
08:07	moritz_	and don't use the $!foo variables at all
08:07	moritz_	just like you can also use $self->{foo} in Moose, if you want (iirc)
08:09	sorear	how does an lvalue accessor do anything interesting?
08:09	sorear	the only way to make lvalues 'interesting' is to tie them
08:09	sorear	does perl 6 discard the taboo against ties?
08:09	moritz_	I'm not sure; it's something I forget everytime I read it up because nobody implements it yet
08:10	moritz_	should be in S06 somewhere
08:10	sorear	Why isn't it trivial?
08:11	moritz_	ask @Larry :-)
08:11	moritz_	I have no idea
08:12	moritz_	I can't remember how lvalue subroutines work in perl 5 either
08:13	JimmyZ	rakudo: say [\+] 1..*
08:13	p6eval	rakudo e16cf4: ( no output )
08:13	JimmyZ	rakudo: say [\+] 1..5
08:13	p6eval	rakudo e16cf4: OUTPUT«1361015␤»
08:14	JimmyZ	rakudo: say [\,] 1..5
08:14	p6eval	rakudo e16cf4: OUTPUT«11 21 2 31 2 3 41 2 3 4 5␤»
08:14	JimmyZ	rakudo: say [\, ] 1..5
08:14	p6eval	rakudo e16cf4: OUTPUT«You can't backslash that at line 11, near ", ] 1..5"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
08:15	JimmyZ	rakudo: say ([\,] 1..5).perl
08:15	p6eval	rakudo e16cf4: OUTPUT«(1, (1, 2), ((1, 2), 3), (((1, 2), 3), 4), ((((1, 2), 3), 4), 5))␤»
08:15	JimmyZ	rakudo: say [@,] 1..5
08:15	p6eval	rakudo e16cf4: OUTPUT«Unable to parse postcircumfix:sym<[ ]>, couldn't find final ']' at line 11␤current instr.: 'perl6;Regex;Cursor;FAILGOAL' pc 1664 (ext/nqp-rx/src/stage0/Regex-s0.pir:907)␤»
08:15	JimmyZ	rakudo: say @[,] 1..5
08:15	p6eval	rakudo e16cf4: OUTPUT«Unable to parse postcircumfix:sym<[ ]>, couldn't find final ']' at line 11␤current instr.: 'perl6;Regex;Cursor;FAILGOAL' pc 1664 (ext/nqp-rx/src/stage0/Regex-s0.pir:907)␤»
08:15	moritz_	std: @[+] 1, 2, 3
08:15	p6eval	std 30497: OUTPUT«===[0mSORRY!===[0m␤Non-declarative sigil is missing its name at /tmp/GLM74EhuSg line 1:␤------> <BOL>⏏@[+] 1, 2, 3␤Bogus statement at /tmp/GLM74EhuSg line 1:␤------> @⏏[+] 1, 2, 3␤ expecting twigil␤Parse failed␤FAILED 00:01
08:15	p6eval	..109m␤»
08:16	moritz_	hm
08:16	moritz_	might be sufficienlty distinct from existing syntax
08:18	JimmyZ	-1
08:18	moritz_	JimmyZ: why?
08:18	JimmyZ	[] means array, I don't think need @ again.
08:19	moritz_	but [+] doesn't return an Array either
08:19	moritz_	here the [...] means more "operate on a list", not "return a list"
08:20	JimmyZ	it mean [] deals with array, not return array
08:21	JimmyZ	that is , it's args should be array.
08:22	JimmyZ	std: {+} 'f' => 1, 'd' => 2;
08:22	p6eval	std 30497: OUTPUT«===[0mSORRY!===[0m␤Prefix requires an argument at /tmp/484Obkwpot line 1:␤------> {+⏏} 'f' => 1, 'd' => 2;␤Parse failed␤FAILED 00:01 108m␤»
08:23	JimmyZ	rakudo: [+] @('3' => 1, '4' => 2);
08:23	p6eval	rakudo e16cf4: OUTPUT«Method 'Num' not found for invocant of class 'Pair'␤current instr.: 'perl6;Mu;' pc -1 ((unknown file):-1)␤»
08:23	JimmyZ	std: [+] @('3' => 1, '4' => 2);
08:23	p6eval	std 30497: OUTPUT«ok 00:01 109m␤»
08:25		kst joined #perl6
08:28		ive joined #perl6
08:29	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Num() { ~$.value;} }; say [+] @('3' => 1, '4' => 2);
08:29	p6eval	rakudo e16cf4: OUTPUT«3␤»
08:29	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Num() { ~$.value;} }; say [\+] @('3' => 1, '4' => 5);
08:29	p6eval	rakudo e16cf4: OUTPUT«3 16␤»
08:29	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Num() { ~$.value;} }; say ([\,] @('3' => 1, '4' => 5)).perl;
08:29	p6eval	rakudo e16cf4: OUTPUT«("3" => 1, ("3" => 1, "4" => 5))␤»
08:30	moritz_	JimmyZ: try method Numeric, not Num
08:30	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Numeric() { ~$.value;} }; say ([\,] @('3' => 1, '4' => 5)).perl;
08:30	p6eval	rakudo e16cf4: OUTPUT«("3" => 1, ("3" => 1, "4" => 5))␤»
08:30	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Numeric() { ~$.value;} }; say [+] @('3' => 1, '4' => 2);
08:30	p6eval	rakudo e16cf4: OUTPUT«Method 'Num' not found for invocant of class 'Pair'␤current instr.: 'perl6;Mu;' pc -1 ((unknown file):-1)␤»
08:30	moritz_	hum.
08:31	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Num() { ~self;} }; say ([\,] @('3' => 1, '4' => 5)).perl;
08:31	p6eval	rakudo e16cf4: OUTPUT«("3" => 1, ("3" => 1, "4" => 5))␤»
08:31	moritz_	colomon: it seems that infix:<+> still uses .Num instead of .Numeric
08:31	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Num() { ~self;} }; say ([\+] @('3' => 1, '4' => 5)).perl;
08:31	p6eval	rakudo e16cf4: OUTPUT«("3" => 1, 7)␤»
08:31	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Num() { ~$.value;} }; say ([\+] @('3' => 1, '4' => 5)).perl;
08:31	p6eval	rakudo e16cf4: OUTPUT«("3" => 1, 6)␤»
08:31		vorner left #perl6
08:31	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Num() { ~$.value;} }; say ([\,] @('3' => 1, '4' => 5)).perl;
08:31	p6eval	rakudo e16cf4: OUTPUT«("3" => 1, ("3" => 1, "4" => 5))␤»
08:32	JimmyZ	what's the different between self and $.value?
08:32	moritz_	self is the whole pair
08:33	moritz_	but in the example above, you're not doing anything that would caause a call to Num
08:34	JimmyZ	using [\,] are the same result, but [\+] not the same result
08:35	moritz_	sure; [\,] has no reason to treat its arguments as numbers
08:35	JimmyZ	Oh, so it is.
08:36	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~$.value;} }; say ([\,] @('3' => 1, '4' => 5)).perl;
08:36	p6eval	rakudo e16cf4: OUTPUT«("3" => 1, ("3" => 1, "4" => 5))␤»
08:36	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~self;} }; say ([\,] @('3' => 1, '4' => 5)).perl;
08:36	p6eval	rakudo e16cf4: OUTPUT«("3" => 1, ("3" => 1, "4" => 5))␤»
08:36	moritz_	it doesn't really call any method on object
08:38	JimmyZ	so I don't like [@,] @('3' => 1, '4' => 5) or @[,] @()
08:38	JimmyZ	that's duplicate semanteme
08:39	moritz_	there's no reason for the second @(...) at all
08:39	JimmyZ	I mean @[] or [@] is duplicate semanteme ;)
08:40	JimmyZ	just in my opinion
08:40	moritz_	and I disagree, becaue the @ is about the return value and [] is about the arguments
08:41	moritz_	this is just like saying "I don't like our Num sub sqrt(Num $x) { ... } because Num appears twice"
08:41	moritz_	but they serve different purposes
08:41	JimmyZ	I don't think operaters need return type .
08:41	moritz_	well, I guess liking and disliking can't always be rationalized
08:42	moritz_	JimmyZ: they don't, in general, but in the case of triangle reduction we need something to distinguish it from the non-triangle form
08:43	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~self;} }; say [\,] '3' => 1, '4' => 5;
08:43	p6eval	rakudo e16cf4: ( no output )
08:43	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~self;} }; say ([\,] '3' => 1, '4' => 5);
08:43	p6eval	rakudo e16cf4: ( no output )
08:43	moritz_	oh
08:43	moritz_	I think that's a bug
08:43	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~self;} }; say [\,] ('3' => 1, '4' => 5);
08:43	p6eval	rakudo e16cf4: ( no output )
08:43	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~self;} }; say [\,] @('3' => 1, '4' => 5);
08:43	p6eval	rakudo e16cf4: ( no output )
08:44	moritz_	oh
08:44	moritz_	I know why
08:44	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Num() { ~self;} }; say [\,] '3' => 1, '4' => 5;
08:44	p6eval	rakudo e16cf4: OUTPUT«3 13 1 4 5␤»
08:44	moritz_	you're causing an infinite recursion
08:44	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Num() { ~self;} }; say [\+] '3' => 1, '4' => 5;
08:44	p6eval	rakudo e16cf4: OUTPUT«3 17␤»
08:44	moritz_	because prefix:<~> calls .Str under teh hood
08:45	JimmyZ	>rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~self;} }; say ([\,] @('3' => 1, '4' => 5)).perl;
08:45	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~self;} }; say ([\,] @('3' => 1, '4' => 5)).perl;
08:45	p6eval	rakudo e16cf4: OUTPUT«("3" => 1, ("3" => 1, "4" => 5))␤»
08:45		dual joined #perl6
08:46	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~self;} }; say [\,] '3' => 1, '4' => 5;
08:46	p6eval	rakudo e16cf4: ( no output )
08:46	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~self;} }; say ([\,] '3' => 1, '4' => 5).perl;
08:46	p6eval	rakudo e16cf4: OUTPUT«("3" => 1, ("3" => 1, "4" => 5))␤»
08:47	JimmyZ	it's strange
08:47	moritz_	I've explained it above
08:47	JimmyZ	moritz_: yes, thanks
08:48	JimmyZ	rakudo: use MONKEY_TYPING; augment class Pair { method Str() { ~self;} }; say ([\,] '3' => 1, '4' => 5).Str; # wrong thing
08:48	p6eval	rakudo e16cf4: OUTPUT«maximum recursion depth exceeded␤current instr.: 'parrot;P6object;HOW' pc 54 (runtime/parrot/library/P6object.pir:98)␤»
08:53		dakkar joined #perl6
08:59		mikehh joined #perl6
09:05	sorear	Who gets the blame for PseudoPod and POD6 being nearly identical?
09:06	moritz_	Damian and p6l
09:07		slavik1 joined #perl6
09:07	slavik1	rakudo: class MyClass { has $.val is rw; }; my @list = MyClass.new xx 5; say @list.perl;
09:07	p6eval	rakudo e16cf4: OUTPUT«[MyClass.new(), MyClass.new(), MyClass.new(), MyClass.new(), MyClass.new()]␤»
09:08	slavik1	rakudo: class MyClass { has $.val is rw; }; my @list = MyClass.new xx 5; say @list.$val;
09:08	p6eval	rakudo e16cf4: OUTPUT«Symbol '$val' not predeclared in <anonymous>␤current instr.: 'perl6;PCT;HLLCompiler;panic' pc 152 (compilers/pct/src/PCT/HLLCompiler.pir:108)␤»
09:08	slavik1	rakudo: class MyClass { has $.val is rw; }; my @list = MyClass.new xx 5; say @list.val;
09:08	p6eval	rakudo e16cf4: OUTPUT«Method 'val' not found for invocant of class 'Array'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
09:08	sorear	rakudo: role A { multi method handler() { say "Foo!" }; }; role B { multi method handler() { say "Bar!" }; }; class C does A does B { }; C.new.*handler;
09:08	p6eval	rakudo e16cf4: OUTPUT«Method 'handler' collides and a resolution must be provided by the class␤current instr.: 'perl6;Perl6;Metamodel;RoleToClassApplier;_block132' pc 9253 (src/gen/RoleToClassApplier.pir:306)␤»
09:09	sorear	Opinion: this should be legal
09:09	slavik1	rakudo: class MyClass { has $.val is rw; does p { say 1; } }; my $list = MyClass.new; $list.p;
09:09	p6eval	rakudo e16cf4: OUTPUT«Could not find sub &p␤current instr.: 'perl6;MyClass;_block46' pc 370 (EVAL_1:150)␤»
09:10	slavik1	rakudo: class MyClass { has $.val is rw; method p { say 1; } }; my $list = MyClass.new; $list.p;
09:10	p6eval	rakudo e16cf4: OUTPUT«1␤»
09:10	slavik1	how do I access a class variable ?
09:10	slavik1	$class.$.var ?
09:10	sorear	classes don't have variables
09:10	slavik1	attributes?
09:11	sorear	Still not sure what you mean
09:11	sorear	Unless
09:11	sorear	rakudo: class MyClass { has $.val is rw }; my $list = MyClass.new; $list.val = 5; say $list.val;
09:11	slavik1	class a { has $.blah } ... how do I access that $.blah?
09:11	p6eval	rakudo e16cf4: OUTPUT«5␤»
09:12	jnthn	slavik1: The generated accessor method is just called blah
09:12	slavik1	wtf ...
09:12	sorear	$.val is sugar for self.val
09:12	slavik1	ok, I think I see it
09:12	moritz_	rakudo: class A { has $.blah; method buh { say $.blah } }; A.new(blah => "fasel").buh
09:12	p6eval	rakudo e16cf4: OUTPUT«fasel␤»
09:13	sorear	does anyone here have a position on that roles thing?
09:13	slavik1	rakudo: class MyClass { has $.val is rw; method inc { $.val++; } }; my @list = MyClass.new(val => 1) xx 5; say @list.per;;
09:13	p6eval	rakudo e16cf4: OUTPUT«Method 'per' not found for invocant of class 'Array'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
09:13	* sorear	is trying to implement an event framework in perl 6
09:13	slavik1	rakudo: class MyClass { has $.val is rw; method inc { $.val++; } }; my @list = MyClass.new(val => 1) xx 5; say @list.perl;
09:13	p6eval	rakudo e16cf4: OUTPUT«[MyClass.new(), MyClass.new(), MyClass.new(), MyClass.new(), MyClass.new()]␤»
09:14	slavik1	rakudo: class MyClass { has $.val is rw; method inc { $.val++; } }; my @list = MyClass.new(val => 1) xx 5; say @list.val;
09:14	p6eval	rakudo e16cf4: OUTPUT«Method 'val' not found for invocant of class 'Array'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
09:14	slavik1	rakudo: class MyClass { has $.val is rw; method inc { $.val++; } }; my @list = MyClass.new(val => 1) xx 5; @list.val;
09:14	p6eval	rakudo e16cf4: OUTPUT«Method 'val' not found for invocant of class 'Array'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
09:14	slavik1	wait ...
09:14	slavik1	how do I get the same method to get invoked for a list of classes?
09:15	slavik1	this was in spec I believe
09:15	slavik1	sorear: that looks like the diamond of death
09:15	slavik1	sorear: which one should be invoked in that sutation?
09:17	sorear	slavik1: Both of them. .* means "ignore the usual type dispatch rules and just invoke all applicable methods in some order"
09:17	JimmyZ	rakudo: augment class Pair { method Str() { ~self;} };
09:17	p6eval	rakudo e16cf4: OUTPUT«Can't augment class Pair without 'use MONKEY_TYPING'␤current instr.: 'perl6;Perl6;Grammar;add_name' pc 42212 (src/gen/perl6-grammar.pir:374)␤»
09:17	sorear	The kicker is that it is the diamond of death in the absense of .*
09:17	JimmyZ	rakudo: augment class Paiir { method Str() { ~self;} };
09:17	p6eval	rakudo e16cf4: OUTPUT«Can't augment class that doesn't exist␤current instr.: 'perl6;Perl6;Grammar;add_name' pc 42212 (src/gen/perl6-grammar.pir:374)␤»
09:18	jnthn	sorear: On the role/multi thing, maybe it should work. I suspect if you write a proto in the class it will work.
09:18	slavik1	sorear: oh, nice
09:18	slavik1	perl6++
09:19	sorear	rakudo: role A { proto method handler(); multi method handler() { say "Foo!" }; }; role B { multi method handler() { say "Bar!" }; }; class C does A does B { }; C.new.*handler;
09:19	slavik1	can a class know it's own runtime type/name?
09:19	p6eval	rakudo e16cf4: OUTPUT«Malformed method at line 11, near "handler();"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
09:19	sorear	rakudo: role A { proto method handler() { ... }; multi method handler() { say "Foo!" }; }; role B { multi method handler() { say "Bar!" }; }; class C does A does B { }; C.new.*handler;
09:19	p6eval	rakudo e16cf4: OUTPUT«Method 'handler' collides and a resolution must be provided by the class␤current instr.: 'perl6;Perl6;Metamodel;RoleToClassApplier;_block132' pc 9253 (src/gen/RoleToClassApplier.pir:306)␤»
09:19	slavik1	so that it would be possible to do what sorear is doing but also output from what class the method is inherited?
09:19	sorear	jnthn: Well, I don't want the class to know the names of all legal event names.
09:22		pmurias joined #perl6
09:22	pmurias	hi
09:22	phenny	pmurias: 01:38Z <diakopter> tell pmurias I fixed ur buggo
09:22	phenny	pmurias: 01:38Z <diakopter> tell pmurias by rejiggering the grammar
09:22	jnthn	sorear: Yeah, good point.
09:22	phenny	pmurias: 01:38Z <diakopter> tell pmurias good luck :)
09:22	phenny	pmurias: 04:40Z <diakopter> tell pmurias I can re-enable the trampoline when a possibly-recursive invocation is detected, but the addl cost will be another slot for the return value per callsite in each recursive frame.
09:22	phenny	[Sorry, some messages were elided and lost...]
09:23	pmurias	why are messages elided?
09:23	jnthn	sorear: I think it's maybe a grey area in the spec, though I'd like it to work.
09:23	sorear	because phenny tried to paste the remainder to paste.lisp.org with an outdated captcha key
09:23	sorear	it should just be using App::Nopaste
09:24	sorear	jnthn: I don't like it all that much because if I misspell an event, nothing will warn me.
09:24	sorear	However, my options are rather limited by Rakudo's limited support for the MOP
09:27	sorear	I think for now I'll just give my methods names like _handler_FOO_uniquey() and rely on introspection for event dispatch
09:27	sorear	With an eye towards moving to a macro & trait based system in the future
09:28	slavik1	if anyone cares, I had my phone HR phone screen with google ... if luck is on our side, there will be a Perl6 binding for MapReduce :)
09:28	sorear	Is that better than » ?
09:30	pmurias	>> is just map
09:30	pmurias	and MapReduce is intended for distributed things while >> is for concurrency on a single box
09:31	slavik1	pmurias: err?
09:32	pmurias	sorear: you asked if MapReduce was better than »?
09:34	sorear	yes
09:34	sorear	[] too
09:35		masak joined #perl6
09:35	masak	oh hai, #Perl6
09:35	phenny	masak: 01:45Z <diakopter> tell masak k, yapsi notifications should work now. it would help if I didn't try to load "/yapsilog.com" from botnix.conf and tried "/yapsi.pm" instead. :/
09:35	masak	wtf ;)
09:36	moritz_	good morning masak
09:36	moritz_	I've already seen a notification for a yapsi commit
09:37	masak	ah. snarkyboojum++
09:38	jnthn	oh hai masak
09:42	pmurias	std: CALLER
09:42	p6eval	std 30497: OUTPUT«===[0mSORRY!===[0m␤Undeclared name:␤ 'CALLER' used at line 1␤Check failed␤FAILED 00:01 109m␤»
09:42	pmurias	std: CALLER::<$foo>
09:42	p6eval	std 30497: OUTPUT«ok 00:01 107m␤»
09:42	pmurias	std: CALLER::
09:42	p6eval	std 30497: OUTPUT«ok 00:01 107m␤»
09:43		riffraff joined #perl6
09:44		ive joined #perl6
09:46	sorear	#70469 #73034 are duplicates
09:46	sorear	also, why do we do this?
09:46	sorear	it seems like a premature optimization
09:46	sorear	it's been bugrepped three times.
09:47	sorear	(this = omitting the pir::newclosure before pir::capturelex in nested sub definitions)
09:47		ive joined #perl6
09:48	moritz_	sorear: last time somebody (=pmichaud) tried to fix it, spectest broke horribly. I have no idea what his fix was, though
09:49	moritz_	sorear: if you have a patch that fixed it and doesn't break spectest, please share :-)
09:56		gurjeet_ joined #perl6
09:57	* masak	merges #73034 into #70469
09:57	sorear	spectest is still /checking out/
09:58	sorear	rakudo: my @x; my $n; for @x.grep(rx/^ _handler_ $n/) -> $handler { $handler() }
09:58	p6eval	rakudo e16cf4: ( no output )
09:58	sorear	rakudo: my @x; my $n; for @x.grep(rx/^ _handler_ $n/) -> $handler { $handler(2, 3) }
09:58	p6eval	rakudo e16cf4: ( no output )
09:58	sorear	oh
09:58	moritz_	@x being empty, and all that
09:59	sorear	moritz_: rakudo was locally generating an error at compile time
09:59	sorear	however, %_ was bound in lexical scope
09:59		_jaldhar_ joined #perl6
09:59	sorear	and I just remembered that was a known bug
10:00		stepnem joined #perl6
10:00	colomon	moritz_: re [+] using Num ... yes, the grand Numeric conversion definitely isn't done yet. :)
10:00	sorear	Method 'perl' not found for non-object
10:00	sorear	current instr.: 'trait_mod:<does>' pc 291390 (src/Perl6/Compiler.pir:112)
10:00	sorear	yikes
10:00	moritz_	colomon: ok
10:00	moritz_	sorear: some error messages contain .perl output...
10:00		simcop2387 joined #perl6
10:01	moritz_	sorear: maybe it's throwing an error while trying to build an error message
10:01	jnthn	Sounds like it.
10:02	colomon	moritz_: BTW, I did fix the split bug yesterday, so the subst call from SVG.pm should work now.
10:02	pmurias	what's the state of APL on parrot? (i'm currently reading "A Programming Language")
10:02	jnthn	(I know there's a fallback multi for that trait mod that gives an error.)
10:03	sorear	s/module/role/ in the role's definition fixed it.
10:03	moritz_	colomon: thanks, I'll try once rakudo compilation is finished
10:04	pugssvn	r30498 \| pmurias++ \| [mildew] fix the building of CORE (one test fails)
10:08	moritz_	colomon++
10:08	moritz_	and indeed, SVG.pm seems to work fine with master now
10:09	masak	\o/
10:09	moritz_	now on to SVG::Plot
10:09	jnthn	\o/
10:09	moritz_	Method 'Num' not found for invocant of class 'Block'
10:10	jnthn	Whatever.
10:10	moritz_	that sound suspiciously like Whatever-currying
10:10	colomon	indeed.
10:10	moritz_	:style("fill:{ @.colors[$d % *] }"),
10:11	moritz_	rakudo: <a b c>[-> $x { $x - 1 }]
10:11	p6eval	rakudo e16cf4: OUTPUT«Method 'Int' not found for invocant of class 'Block'␤current instr.: 'perl6;Mu;' pc -1 ((unknown file):-1)␤»
10:12	moritz_	rakudo: say (24 % *).(5)
10:12	p6eval	rakudo e16cf4: OUTPUT«Method 'Num' not found for invocant of class 'Block'␤current instr.: 'perl6;Mu;' pc -1 ((unknown file):-1)␤»
10:12	sorear	rakudo: class Foo { }; my $name = 'xyzzy'; Foo.^methods(:local);
10:12	p6eval	rakudo e16cf4: ( no output )
10:13	sorear	rakudo: class Foo { }; my $name = 'xyzzy'; Foo.^methods(:local);
10:13	p6eval	rakudo e16cf4: ( no output )
10:13	sorear	is p6eval mangling lexicals?
10:14	moritz_	it should not... why do you think it does?
10:14	moritz_	rakudo: say log10(100)
10:14	p6eval	rakudo e16cf4: OUTPUT«2␤»
10:15	sorear	moritz_: rakudo 082caee local blows up on that code.
10:15	moritz_	why do I get an error that &log10 wasn't found?
10:15	sorear	if you rename or remove $name, it works.
10:16	moritz_	oh, I remember diakopter hacking something in for safe mode... don't know if that's still active
10:16	sorear	can someone with an unpatched rakudo, ideally 082caee, tell me if I'm insane/
10:17	sorear	make spectest is running 3 simultaneous rakudo processes
10:17	sorear	how do I stop it?
10:17	moritz_	Ctrl+C
10:17	sorear	I have 1 core are < 3*RakudoHeap RAM
10:17	sorear	*and
10:18	moritz_	exprot TEST_JOBS=1
10:19	moritz_	sorear: my build of rakudo HEAD doesn't blow up on class Foo { }; my $name = 'xyzzy'; Foo.^methods(:local);
10:19	moritz_	if that's what you wanted to know
10:19	moritz_	\o/ SVG::Plot now working on master
10:20	colomon	moritz_++
10:21	moritz_	just three changed lines
10:23	colomon	hey, you ++'d me for fixing split, and that was just changing one line and adding another.
10:24	moritz_	I wasn't trying to dminish my fixes, just wanted to tell you that master was working pretty well :-)
10:24	colomon	:)
10:24	avar	does rakudo have an sqlite interface?
10:25	moritz_	avar: masak wrote "Squerl" which uses sqlite... no idea how well it works
10:25	moritz_	std: s[foo];
10:25	p6eval	std 30497: OUTPUT«===[0mSORRY!===[0m␤Missing assignment operator at /tmp/SfWppLSxKr line 1:␤------> s[foo]⏏;␤ expecting quantifier␤Parse failed␤FAILED 00:01 111m␤»
10:25	sorear	just use DBD::SQLite
10:26	avar	sorear: maybe if you make your patches public :)
10:26	sorear	I did.
10:26	avar	oh, where is it?
10:26	sorear	Rakudo master.
10:27	sorear	A one-line bugfix which allows you to write 'DBI' instead of '::DBI' is still waiting on jnthn though
10:27	moritz_	avar: and blizkost is in the blizkost repo :-)
10:28	* moritz_	currently spectests a patche that implements rx{...} (as opposed to rx/.../)
10:28	avar	sorear: what are the relevant commits/tests?
10:28	avar	or example program :)
10:28	sorear	Blizkost can currently marshal hashes, strings, numbers, coderefs, and arrays(arguments only)
10:29	jnthn	rakudo: s[foo];
10:29	p6eval	rakudo e16cf4: OUTPUT«Missing assignment operator at line 11, near ";"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
10:29	sorear	avar: examples/ in github.com/jnthn/blizkost
10:29	moritz_	jnthn: I've just seen the error message in rakudo, and wondered if we should add "after s[...]", but decided that we don't need to more awesome than STD.pm
10:30	avar	sorear: You you patched blizkost but didn't create some new project?
10:30	avar	ah, right, thought it was some new glue
10:30	sorear	avar: I more or less took over blizkost. It is my new project
10:30	moritz_	and I think jnthn is glad somebody else takes care :-)
10:31	avar	cool:)
10:31	avar	I was going to test out rewriting Hailo in Perl 6, but I was always blocking on "oh that'll be a total PITA without modules" :)
10:32	* jnthn	is very glad somebody else is taking care of Blizkost. sorear++
10:33	sorear	There's a very good chance I'll need to implement arrayref and undef returns before DBI works fully
10:34	sorear	however, you can do stuff like my &p5defined = eval ("sub { defined $_[0] }", :lang<perl5>);
10:34	avar	Right, or implement a glue module on the p5 side
10:36	avar	w 8
10:39	avar	sorear: patch: http://github.com/avar/blizkos[…]d59412d031f049b25 :)
10:42		am0c joined #perl6
10:45	masak	Squerl should still work under alpha. probably not under master without some modifications.
10:59		rv2733 joined #perl6
10:59	m6locks	crap. i got only master
11:09		Su-Shee joined #perl6
11:23		takadonet left #perl6
11:27	masak	huh. Python has combinations() and permutations() built in: http://docs.python.org/library/itertools.html
11:28	Su-Shee	the doc say it's a module.
11:28	Su-Shee	hi all.
11:28	avar	Will Perl 6 have batteries ?:)
11:28	avar	sorear: It's part of the stdlib
11:29	pugssvn	r30499 \| moritz++ \| [t/spec] unfudge passing split :all tests, colomon++
11:30	masak	Su-Shee: I think it's one of them 'core modules'.
11:30		am0c joined #perl6
11:30	masak	ah. what avar said, to the wrong tab-person.
11:31	Su-Shee	which means one can easily define whatover one wants to be a core module. :)
11:31	Su-Shee	ever
11:31		kst joined #perl6
11:34	masak	well, it seems that that kind of 'core' will be up to distributions of Perl 6, not the spec.
11:36	Su-Shee	masak: I like tiny cores. I like adding modules depending on the need.
11:36	Su-Shee	and why not make a "security related math libs" distribution or a "gui-all-you-ever-need" distribution.. :)
11:36	moritz_	tiny core... then Perl 6 is not for you :-)
11:36	masak	Su-Shee: me too. but I also know I draw the line somewhere.
11:37	Su-Shee	I realized that already. ;)
11:37	masak	Su-Shee: for example, impoting 'stdio' for IO in C feels silly.
11:37	Su-Shee	masak: excellent example I would totally agree.
11:38	moritz_	anyway, I'd welcome some attempts to make DateTime, Date and various container types non-core, but still with a somewhat built-in feeling
11:38	Su-Shee	moritz_: from a user's point of view I essentially don't want to "feel" the difference between "is core, is builtin, is cpan"
11:40	masak	moritz_: not sure what you mean. what more would it take than putting `use DateTime;` at the top of the program?
11:41	colomon	rakudo: say 1i * "3"
11:41	p6eval	rakudo e16cf4: OUTPUT«0 + 3i␤»
11:41	colomon	rakudo: say 1i + "3"
11:41	p6eval	rakudo e16cf4: OUTPUT«3 + 1i␤»
11:42	moritz_	masak: integration with other modules or core things...
11:42	masak	moritz_: it's the latter thing that concerns me.
11:42	moritz_	for example now that 'now' returns an Instant, I think time() could return a DateTime
11:42	moritz_	(if we need time() at all)
11:43	masak	seems that we can't have both 'time' as a builtin and DateTime as non-core.
11:44	moritz_	so... move DateTime and time() both out of core?
11:46	masak	well, I'm still a bit doubtful.
11:46	masak	wouldn't filetests need DateTime, for example?
11:46	masak	those are definitely core.
11:46	moritz_	no, they need Instant only
11:46	masak	ok.
11:47	moritz_	maybe I should compile a list of types that could be non-core
11:47	masak	I'd like for DateTime to end up as 'almost core', i.e. something that most distributions will want to pull in.
11:47	moritz_	aye
11:47	masak	things like timezones and localizations may well be CPAN-style modules.
11:48	masak	i.e. not necessarily 'almost core'.
11:50		snarkyboojum joined #perl6
11:51	moritz_	Set, KeyBag, KeySet, KeyHash # feel like they could be "almost core"
11:54		plobsing joined #perl6
11:59		takadonet joined #perl6
11:59	takadonet	morning all
11:59	masak	takadonet: \o
11:59	masak	moritz_: aye.
12:00	* masak	would still like to 'flesh out' the specification of those classes to contain a richer set of common operations
12:07		envi^home joined #perl6
12:08	pugssvn	r30500 \| colomon++ \| [t/spec] All those 0.0s and 2.5s were meant to be Nums -- and now they are again.
12:34		ruoso joined #perl6
12:42	colomon	jnthn: ping?
12:44		SmokeMachine joined #perl6
12:44		tylerni7 joined #perl6
12:45		kst joined #perl6
12:49	moritz_	http://www.perlmonks.org/?node_id=837535 # rir++ for asking the right questions
12:51	moritz_	if you have things to add to my answer, please do so now
12:52		iblechbot joined #perl6
12:53	moritz_	while I was writing my reply, I realized that proto and cpan differ quite significantly
12:53	moritz_	cpan is push-style: each version is uploaded to pause.cpan.org
12:54	moritz_	while our current approach is pull-style: the meta data is kept central, but the source code isn't kept in one place
12:54	masak	nod
12:54	moritz_	not sure which approach is preferable in the long run
12:54	masak	proto was never meant to be like CPAN. it was created to solve a piece of immediate pain.
12:54	moritz_	I know, you keep repeating that :-)
12:54	pmurias	keeping the source code in one place seems very sensible
12:55	masak	moritz_: I keep repeating it because I'm constantly surprised about the new things that proto starts to mean to people. :)
12:55	masak	I consider that project to be my most successful failure so far.
12:56	moritz_	masak: :-) linux started out as an exploration of i386 architecture
12:56	moritz_	masak: pugs started as an exercise in compiler writing
12:56	moritz_	etc.
12:56	masak	aye. start small.
12:57	pugssvn	r30501 \| pmurias++ \| [mildew] avoid duplication of the output method
12:57	pugssvn	r30502 \| pmurias++ \| [mildew] only the C backends load the setting through MildewSOLoader
12:59	pugssvn	r30503 \| moritz++ \| [helpnow] new idea: overview page for proto projects
13:02		tylerni7 joined #perl6
13:02	[Coke]	(most successful failure)++
13:03		ReiniUrban joined #perl6
13:04	masak	it's still in the README. 'It's only purpose is to help you set up a running environment where you can play with Perl 6 modules with minimal hassle.'
13:04	masak	nowadays proto compiles and installs modules, and it can help people create new modules.
13:05	masak	I'm not saying that's bad. just not what I had in mind. :)
13:11	colomon	has anyone figured out how to output data from inside the core when &say cannot be found?
13:11	masak	moritz_: 'blizkost' is a noun; it means 'closeness' or 'proximity' rather than 'close'. just a teeny tiny nit.
13:12	masak	colomon: assign a global variable &s and use that.
13:12	moritz_	masak: thanks, fixed
13:16	pmurias	diakopter: hi
13:17		[particle] joined #perl6
13:19		sahadev joined #perl6
13:20	colomon	masak++
13:20	masak	colomon: :) discovered it just the other day
13:20	colomon	of course, now that I know more, I've gone from confused to utterly mystified
13:21	masak	is it possible to isolate and golf that mystification?
13:21	[Coke]	http://merlyn.posterous.com/pr[…]-perl-6-cover-art
13:22	colomon	I dunno.
13:22		meppl joined #perl6
13:24	colomon	here's the deal.
13:25	colomon	moritz_++ pointed out that arithmetic was still using .Num instead of .Numeric internally.
13:25	colomon	as a result of thinking about that a bit, I started adding tests like "1" + 3i to complex.t
13:26	colomon	in theory, the way they should work properly now is:
13:26	colomon	"1" + 3i dispatches to infix:<+>(Any, Any)
13:26	moritz_	rakudo: say '1' + 3i
13:26	p6eval	rakudo e16cf4: OUTPUT«1 + 3i␤»
13:26	colomon	that calls +"1" + +3i
13:27	colomon	moritz_: that works now because there is an infix:<+>(Any, Complex)
13:27	colomon	I'm trying to get rid of that.
13:27	colomon	and (concluding) +"1" + +3i dispatches to infix:<+>(Real, Complex)
13:28	moritz_	but prefix:<+> should already call .Numeric, no?
13:29	colomon	moritz_: yes, it does. but the existing versions of infix:<+>(Any, Any) calls .Num on each argument instead.
13:29	moritz_	ok
13:29	moritz_	so that needs fixing too
13:29	colomon	the change to use +$a + +$b is only local to my system at the moment.
13:29	colomon	thing is, that works great
13:29	colomon	afk for sec
13:32	colomon	but the exact same approach fails miserably for infix:<**>
13:33	moritz_	rakudo: infix:<**>(Mu, Mu)
13:33	p6eval	rakudo e16cf4: OUTPUT«No applicable candidates found to dispatch to for 'infix:<**>'. Available candidates are:␤:(Int $a, Int $b)␤:(Complex $a, Complex $b)␤:()␤:(Num $a, Num $b)␤:(Real $a, Real $b)␤:(Complex $a, Real $b)␤:(Real $a, Complex $b)␤:(Any $a, Any $b)␤␤current instr.: '_block14' pc 29
13:33	p6eval	..(EVAL_1:…
13:33	colomon	infix:<**>(Any, Any) calls itself recursively 1000 times
13:33	moritz_	so what does $a.Numeric and $b.Numeric return?
13:33	colomon	and when I use masak++'s trick to dump the Parrot version of the types involved, it is Perl6Str and Complex all the way down.
13:34	colomon	so +Perl6Str seems to return a Perl6Str.
13:34	moritz_	that's wrong.
13:34	colomon	well yes.
13:35	colomon	what I don't get is why +Perl6Str doesn't work for **, but seems to work fine for +
13:35	colomon	(last + is infix:<+>)
13:36	moritz_	and that's for '1' ** 3i, or what?
13:36	colomon	moritz_: "1" ** 3i
13:36	moritz_	rakudo: say "1".PARROT
13:36	p6eval	rakudo e16cf4: OUTPUT«Perl6Str␤»
13:37	moritz_	rakudo: say (+"1").PARROT
13:37	p6eval	rakudo e16cf4: OUTPUT«Num␤»
13:37	moritz_	colomon: maybe.... a very stupid precedence issue?
13:37	colomon	moritz_: I do believe that must be it!
13:37	moritz_	(+$a) + (+$b) or so?
13:37	colomon	moritz_++
13:38	colomon	+$a +$b is probably +($a +$b)
13:38	colomon	that would explain everything.
13:38	masak	colomon: can you demonstrate it on p6eval?
13:39	colomon	masak: let me think
13:40	colomon	masak: behaving that way is spec
13:40	colomon	exponentiation has higher precedence than symbolic_unary
13:40	masak	o.O
13:41	* masak	looks
13:41	colomon	or at least, that's what rakudo's grammar has for it.
13:41	masak	by blog, S03 agrees with you.
13:41	moritz_	aye
13:42	masak	oh right, that's so that -2**4 will mean the Right Thing.
13:42	masak	TimToady mentioned that the other day.
13:42	moritz_	for the right value of "right" :-)
13:42	masak	well, it means what mathematicians mean by it.
13:43	masak	TimToady: should p5=> really remain in the S03 precedence table?
13:46	colomon	moritz_++ # bug appears to be fixed!
13:49		ruoso joined #perl6
13:51	pmurias	ruoso: using Regexp::Grammars to replace the haskell m0ld parser seems sane?
13:52	ruoso	I never used it, but it being in perl5 is already good enough when compared to the haskell dependency ;)
13:54	ruoso	it doesn't support rec descent well... but m0ld doesn't require it... so it's probably a good choice, since it's going to be faster
13:56	pmurias	i would be so sure about it being faster but it certainly won't require setting up cabal
13:59	pmurias	do you think carefull placing of #line would allow using gdb to debug code spit out by mildew's C backends?
14:02	pugssvn	r30504 \| moritz++ \| [helpnow] draft announcement
14:08		alester joined #perl6
14:12	jnthn	colomon: belated pong (meeting... :-))
14:12	colomon	jnthn: no worries, we actually solved the problem without you. :)
14:12	colomon	masak++ and moritz_++
14:13	jnthn	Yay!
14:15	jnthn	btw, I'm heading off the Iceland tomorrow, all being well. I'll probably have good connectivity until Monday while in the capital, but may be sketchy at times next week when I head out to the countryside (hopefully). Just a heads up. :)
14:15	mathw	Iceland? Hmmm
14:15	mathw	Watch out for volcanoes
14:15	colomon	:)
14:15	colomon	afk # exercising
14:16	moritz_	jnthn: have fun in Iceland :-)
14:16	* moritz_	wonders if NPW is actually cancelled
14:18	jnthn	moritz_: Yes
14:18	jnthn	moritz_: But since I'd planned to take the week after it having a vacation anyway, I decided to go ahead with the trip.
14:18	moritz_	jnthn: then I hope your weather will be a nice as it is here a the moment
14:19	moritz_	(sunny, 26C)
14:19		JimmyZ joined #perl6
14:22	jnthn	moritz_: ooh, that is nice :-)
14:22	jnthn	It'll likely be a bit cooler.
14:22	masak	jnthn: enjoy Iceland for me! :)
14:22	moritz_	I'll leave $work in a few minutes, buy some foodstuff and then exercise my inline skates
14:24	moritz_	at times I'm very happy for not having fixed (or even defined) work hours
14:25	[Coke]	moritz_: same here. though some times I can use the structure that 9-5 woudl provide. =-)
14:25	moritz_	aye. External constraints help not let things slip
14:25	moritz_	decommute&
14:26		KeithWolters joined #perl6
14:27	KeithWolters	should I be able to return from a sub with something like "retturn 0 but true"?
14:28	masak	rakudo: sub foo { return 0 but True }; say foo
14:28	p6eval	rakudo e16cf4: OUTPUT«No applicable candidates found to dispatch to for 'infix:<does>'. Available candidates are:␤:(Mu $do-it-to-me, Role $r)␤:(Mu $do-it-to-me, ConcreteRole $r)␤␤current instr.: 'infix:<but>' pc 311315 (src/gen/core.pir:0)␤»
14:29	masak	rakudo: role Troo {}; sub foo { return 0 but Troo }; say foo
14:29	p6eval	rakudo e16cf4: OUTPUT«0␤»
14:29	masak	KeithWolters: yes. it should work in Rakudo too, but enums (Bool::True in particular) are not sufficiently role-y yet.
14:33	KeithWolters	thanks
14:34		rgrau_ joined #perl6
14:34		[particle]1 joined #perl6
14:40	jnthn	enums are not sufficiently existential in Rakudo yet either...
14:41		envi^home joined #perl6
14:41		REPLeffect joined #perl6
14:43		[particle]1 joined #perl6
14:43	mathw	They're not sufficiently named
14:44	mathw	rakudo: enum Mode <Binary ASCII>; my $m = Mode::Binary; say $m;
14:44	p6eval	rakudo e16cf4: OUTPUT«Can not find sub Mode::Binary␤current instr.: 'perl6;Perl6Exception;throw' pc 14877 (src/builtins/Seq.pir:28)␤»
14:52		pmurias joined #perl6
14:56	diakopter	pmurias: hi
14:57	pmurias	diakopter: hi
14:58		kst joined #perl6
14:58		proller joined #perl6
14:58		justatheory joined #perl6
14:59	sorear	I figured out my problem yesterday
14:59	pmurias	diakopter: the catch/try in the trampoline is used for catching the case when next is null?
14:59	sorear	It involved perl6 -e ' ... ... ... my $name = 'foo' ... ...'
14:59	sorear	sh, I hate you
15:00	masak	sorear: I do that one far too often too :)
15:00	jnthn	Oh, sh..
15:00	jnthn	:-)
15:00	colomon	not ok 6 - increment of undefined variable does not warn
15:00	colomon	# got err: "Use of uninitalized value in numeric context\n"
15:00	colomon	# expected err: ""
15:01	colomon	my $a; $a += 1
15:01	colomon	rakudo: my $a; $a += 1
15:01	pmurias	diakopter: could that be done by an if as it hides exceptions from methods invoked by sprixel code
15:01	p6eval	rakudo e16cf4: ( no output )
15:02	colomon	> my $a; $a += 1; say $a;
15:02	colomon	Use of uninitalized value in numeric context
15:02	colomon	1
15:02	colomon	I thought we'd agreed it should warn in cases like this? Opinions?
15:02	masak	omgiblogged: http://use.perl.org/~masak/journal/40333
15:03	masak	colomon: I thought we'd agreed it shouldn't warn in that case.
15:03	masak	colomon: it should warn in the `$a = $a + 1` case.
15:03	masak	but not in the assignment metaop cases.
15:03	diakopter	pmurias: but I stopped using the trampoline anyway
15:05	colomon	masak: In that case, how can I check definedness in PIR? :)
15:06	masak	dunno.
15:07	diakopter	pmurias: but yes, it does hide those exceptions..... lemme fix that.
15:08	jnthn	omgthreethirds!
15:08	pmurias	diakopter: the trampoline is still used in commited code?
15:08	diakopter	yes, but only once
15:08	diakopter	it's easy to stop using entirely
15:08	colomon	masak: on your blog post, part 2 -- contextuals exist only for the scope they're defined in and any scopes that it calls? I've never quite understood them...
15:09	masak	colomon: that's right.
15:09	diakopter	pmurias: ok.
15:10	masak	colomon: which makes them perfect for when you have a routine at the top, sharing data with all the routines transitively called.
15:10	colomon	masak: sweet, yes, masak++ for blogging that. (and thinking to do it.)
15:11	sorear	Is there any short way to say git log avar/master..HEAD; git log HEAD..avar/master ?
15:11	sorear	I want the symmetric difference
15:11	colomon	jnthn: any hints on testing for definedness in pir? I'd like to hack assign_metaop so that it checks to see if the first arg is defined and calls $P0() if not....
15:11	masak	sorear: that might be what '...' does.
15:11	* masak	doesn't quite recall, though
15:12	jnthn	colomon: Can't you just call .defined?
15:12	jnthn	$P1 = $P0.'defined'()
15:12	jnthn	if $P1 goto omgyayitsdefined
15:12	colomon	jnthn: I dunno, I'm not very good at PIR. I'll give that a try.
15:13	sorear	that'll work, except that if $P0 is null it may segfault
15:13	jnthn	no it own't
15:13	jnthn	it'll give a null PMC access excpetion.
15:13	diakopter	afk&
15:13	jnthn	colomon: Should work
15:13	sorear	and of course it won't work well if $P0 is a non-parrot object
15:13	colomon	will it work if I use .param pmc a a.'defined'() instead?
15:13	jnthn	sure
15:13	sorear	colomon: That changes nothing, so yes
15:14	jnthn	The other way is $I0 = defined a
15:14	jnthn	if $I0 goto yayitsdeinfed
15:14	sorear	jnthn: PARROT_CATCH_NULL is an option
15:14	sorear	avar: Still here?
15:14	dalek	csmeta: r224 \| diakopter++ \| trunk/Sprixel/ (3 files):
15:14	dalek	csmeta: [perlesque] see if this works.
15:14	dalek	csmeta: review: http://code.google.com/p/csmet[…]urce/detail?r=224
15:15	[Coke]	(parrot nulliness) ISTR there is a ticket that is eligible to go in shortly that will impact that.
15:15	avar	sorear: yah
15:15	avar	sorear: git remote add avar my-git-url
15:15	avar	git diff master..avar/master
15:15	[Coke]	https://trac.parrot.org/parrot/ticket/1207 ?
15:16	pmurias	diakopter: would i be possible to add an option to sprixel to spit out the disassembled version of the code instead of running it?
15:18	sorear	avar: that only shows me patches you've added.
15:18	sorear	I want to see all differences, including ones you may have removed
15:18	sorear	avar: btw I've already committed your patch
15:20	avar	sorear: I don't get what you mean. I just added this: http://github.com/avar/blizkos[…]d59412d031f049b25
15:20	avar	so that's all git dif master..avar/master would show you. I didn't remove anything
15:20	sorear	Yes. And. http://github.com/jnthn/blizko[…]d59412d031f049b25
15:20	sorear	avar: git diff master..avar/master will show me ABSOLUTELY NOTHING if you have removed a patch.
15:21	sorear	As a rule I don't trust people when they tell me about why their code is safe. It's the same with branches.
15:21	sorear	git diff avar/master..master shows me the other side of the story.
15:21	sorear	I'd like to only run one command to see both.
15:21	avar	well, then alias it in your ~/.gitconfig :)
15:22	avar	Anyway, git-cherry-pick is probably what you want for single patches like that
15:22		proller joined #perl6
15:22	avar	Anyway, 'git diff master..avar/master' will show you the complete story, it's all the patches on that branch since it diverged.
15:23	avar	so a subsequent git merge avar/master is safe, the worst that can happen is a merge conflict because the patches changed something you changed too.
15:23	sorear	If you did git reset --hard HEAD^
15:23	sorear	that wouldn't register on git diff master..avar/master
15:24	sorear	are you disputing this?
15:24	colomon	rakudo: my $a; $a *= 5; say $a
15:24	p6eval	rakudo e16cf4: OUTPUT«Use of uninitalized value in numeric context␤0␤»
15:24	colomon	> my $a; $a *= 5; say $a
15:24	colomon	5
15:24	sorear	incidentally, ... works
15:24	[Coke]	rakudo: say ...
15:24	p6eval	rakudo e16cf4: OUTPUT«No exception handler and no message␤current instr.: '&fail' pc 17511 (src/builtins/Junction.pir:489)␤»
15:24	avar	No. I didn't account for reset, anyway if someone was evil like that you'd get a push conflict about a non-fast forward when pushing to github unless you did git push -f
15:25	[Coke]	std: say ...
15:25	p6eval	std 30504: OUTPUT«ok 00:01 109m␤»
15:25	* [Coke]	rings up masak.
15:25	masak	hello?
15:25	sorear	masak++
15:25	[Coke]	see last two evalbots?
15:25	masak	yeah.
15:25	masak	but I see no bug.
15:26	moritz_	well, ... in rakudo is less than awesome
15:26	masak	LTA might be the bug, sure.
15:26	masak	what would you want the error message to be?
15:26	[Coke]	I see a parrot error message. shouldn't rakudo never emit one?
15:26	* masak	submits LTA rakudobug
15:26	[particle]1	it's masakworthy!
15:27	[Coke]	woot.
15:31	jnthn	We already have a ticket on ... saying it probably wants to use leave, not return, so we don't run into issues like the above, fwiw.
15:31	[Coke]	single tear
15:33	jnthn	That error still sucks though, and
15:33	jnthn	rakudo: return
15:33	p6eval	rakudo e16cf4: OUTPUT«No exception handler and no message␤current instr.: '&return' pc 17409 (src/builtins/Junction.pir:416)␤»
15:33	jnthn	Needs fixing for that reason. :-)
15:35	* [Coke]	sees that theDamian is giving a rakudo class at yapc::na this year.
15:35		gurjeet joined #perl6
15:35	masak	rakudo: return
15:35	p6eval	rakudo e16cf4: OUTPUT«No exception handler and no message␤current instr.: '&return' pc 17409 (src/builtins/Junction.pir:416)␤»
15:35	masak	ah, jnthn already did that :)
15:35	diakopter	pmurias: it does.. to asmbly_1.exe
15:36	jnthn	masak: Well, if you must return to the issue... :-P
15:36	masak	:P
15:36	diakopter	pmurias: it's in CIL
15:37	diakopter	pmurias: into what language would you have it disassembled?
15:37	masak	[Coke]: ah, and that bug is already reported: http://rt.perl.org/rt3/Ticket/[…]lay.html?id=57268
15:37	diakopter	pmurias: it's in CIL bytecode I should say.
15:38	* masak	reopens that ticket instead of submitting a new one
15:38	diakopter	pmurias: it could be disassembled to human-readable CIL with `monodis asmbly_1.exe`
15:38	diakopter	that may help
15:39	diakopter	pmurias: but otherwise, you'd need something like Reflector, on Windows
15:39	diakopter	though I think a very old version of Reflector ran on mono/linux...
15:42	pmurias	usr/local/bin/monodis: error while loading shared libraries: libmono.so.0: cannot open shared object file: No such file or directory
15:43	moritz_	looks like it's not properl installed
15:43	diakopter	pmurias: you just have to run libtool
15:43	diakopter	or whatever it's calls
15:43	diakopter	called
15:44	sorear	ldconfig ITYM
15:44	diakopter	yeah that :)
15:44	sorear	libtool is something entirely different
15:44	dalek	rakudo: fe59fa8 \| moritz++ \| src/Perl6/Grammar.pm:
15:44	dalek	rakudo: enable rx{...} quoting form. At some point we need a more general solution for these quoting mechanismsn...
15:44	dalek	rakudo: review: http://github.com/rakudo/rakud[…]bfa76db8ff916beea
15:44	sorear	also, add /usr/local/lib to your /etc/ld.so.conf
15:44	diakopter	pmurias: sorry :D what sorear said
15:45	sorear	note: these two steps must be done in a specific order. Namely, the opposite of the one I listed them in.
15:45		wasy_ joined #perl6
15:45		wasy_ left #perl6
15:45	diakopter	usually /usr/local/lib is already added in debian/ubuntu, I think.
15:46	pmurias	ldconfig worked
15:46	pmurias	but monodis segfaults
15:46	diakopter	heh; hm
15:46		nihiliad joined #perl6
15:47	moritz_	disassembling is an evil hacker technique and forbidden in your country!
15:47	moritz_	or so :-)
15:47	sorear	diakopter: maybe it is now, I've been running this debian install for about seven years
15:48	diakopter	pmurias: does it output anything before it segfaults?
15:49	pmurias	yes
15:49	pmurias	but CIL won't be helpfull
15:49	diakopter	oh :P
15:50	diakopter	disassembling to C# is very nontrivial, hence the lack of availability of tools
15:50	diakopter	open-source tools, I mean
15:51	diakopter	http://blog.coryfoy.com/2006/0[…]oes-work-on-mono/
15:52	diakopter	a very old version ran in linux
15:58	diakopter	pmurias: I'd guess you'd have to somehow find one of those old versions of reflector to run it on linux
15:58	diakopter	b/c I doubt it runs on it anymore.
15:58	diakopter	I haven't tried it lately though.
16:02		chitragupt joined #perl6
16:04		nihiliad joined #perl6
16:05		bradb joined #perl6
16:06		isBEKaml joined #perl6
16:13		cdarroch joined #perl6
16:13		cdarroch joined #perl6
16:15		am0c joined #perl6
16:17		ashizawa joined #perl6
16:17	isBEKaml	hi #perl6!
16:18	masak	hi isBEKaml!
16:18	isBEKaml	I feel like looking at a module today... what's already in? :)
16:19	masak	isBEKaml: you mean writing one?
16:19	sorear	try src/core/*.pm
16:19	isBEKaml	masak: no, reading one! ;)
16:19	masak	oh! well, what sorear said.
16:20	masak	isBEKaml: I can recommend Druid, too. it's a documented, OO board game in Perl 6. http://github.com/masak/druid
16:20	sorear	src/core/*.pm isn't exactly modules, but it's the largest body of reusable Perl 6 code we have
16:20	masak	sorear: largest? on what do you base that?
16:20	masak	the spectest suite is pretty large.
16:20	masak	November and GGE are large projects.
16:21	[particle]1	STD.pm is pretty large
16:21	[particle]1	...and that's just one module...
16:21	diakopter	_reusable_, sorear said
16:21	masak	STD.pm might be the largest reusable piece of Perl 6 right now.
16:21	sorear	masak: A weighted function of size and reusability
16:21	isBEKaml	YOU_ARE_HERE.pm; a single line module, yay! :D
16:21	sorear	STD.pm is huge but it's only good for one thing
16:21	* diakopter	agrees
16:22	masak	hm. I need to think about the difference there.
16:22	masak	STD.pm5 isn't reusable because it's only good for one thing?
16:22	[particle]1	where on earth are the p6 acme modules??!?!!!?
16:22	masak	even though that one thing is parsing Perl 6?
16:22	sorear	masak: it is reusable. just, less so than stuff like Hash.pm
16:22	jnthn	isBEKaml: That's just there to make people think, "wtf?" ;-)
16:23	colomon	pir folk: any way to do the equivalent of .?call in pir?
16:23	jnthn	Ew!
16:23	jnthn	Yes
16:23	jnthn	But I'll have to look it up
16:23	colomon	maybe there's a better way?
16:23	sorear	colomon: src/glue/dispatch.pir
16:23	jnthn	See src/glue/dispatch.pir for the thingy to call.
16:23	sorear	!dispatch_.?
16:23	jnthn	It's got .* in its name.
16:23	jnthn	oh, .?
16:24		ash_ joined #perl6
16:24	jnthn	What sorear++ said.
16:24	sorear	.? actually does two things
16:24	colomon	I've patched assign_metaop to take the value of the op called with no parameters.
16:24	sorear	it guards against the method not existing
16:24	sorear	and it guards against MMD failing
16:24	isBEKaml	jnthn: :). Anyway, I was thinking of PIR code generated when building Rakudo after checking out Parrot... Was that all auto generated?
16:24	sorear	if you only want to do one of these things, you can get away with less code
16:24	isBEKaml	jnthn: I tried reading them, looked close to heavy metal bytecode..
16:24	colomon	which works great unless there is no version that takes no parameters.
16:25	sorear	isBEKaml: Everything in src/gen is autogenerated compiler output.
16:25	jnthn	Yup
16:25	diakopter	\_o_/
16:25	sorear	There's a fair amount of handwritten (i.e. commented and with sane register names) pir, in src/builtins, src/cheats, and src/glue
16:25	jnthn	Pretty much the only hand-written PIR lives in src/builtins/ and src/glue/ iirc.
16:25	sorear	oh and src/metamodel
16:26	isBEKaml	eww, does anybody ever handcode PIRs?
16:26	sorear	Yes
16:26	sorear	Look inside your libc sometime, you'll see plenty of assembly there too
16:26	sorear	same reason, really
16:27	sorear	implementing the entire Perl 6 standard library in Perl 6 has bootstrap issues
16:27	colomon	errr, ick. actually I need the sub equivalent of that, if there is such a thing, now that I think about it.
16:27	sorear	$P0 = get_hll_global "&my_sub"
16:27	isBEKaml	I thought handcoding assembly was so long ago, unless you really need to get close to hardware.. That isn't the case with perl6, right?
16:27	sorear	if null $P0 goto not_found
16:27	sorear	$P0()
16:28	colomon	sorear: in this case &my_sub should exist, just with a different signature than I need to use.
16:28	sorear	isBEKaml: Libraries can't just float in mid-air, they need to be anchored to the bare metal somewhere
16:28	ash_	isBEKaml: anytime you port say the GCC to new hardware there will always be handcoded assembly, in embedded systems its happens all the time
16:28	masak	swimming &
16:29	jnthn	isBEKaml: Calling it assembly and thinking the same was as hardware-level assembly is making it out a bit worse than it is though.
16:29	isBEKaml	ash_: I can understand embedded systems, yes...
16:29	jnthn	isBEKaml: All your objects are garbage collected, don't need to manage calling conventions by hand, etc.
16:29	ash_	pir has a lot of nice things, its a bit above assembly
16:30	isBEKaml	jnthn: I didn't mean that way.. I didn't even call them assembly, I said they looked like auto generated bytecode, akin to java bytecodes..
16:30	[particle]1	isBEKaml: how would you implement infix:<+> in pure perl 6?
16:30	isBEKaml	jnthn: sorry if I came out wrong..
16:31	sorear	isBEKaml: Yes, they're evil. We know.
16:31	sorear	Are you complaining?
16:31	sorear	There's no alternative.
16:31	ash_	[particle]1: I would use a NCI system to call libc to add them together, :P
16:32	isBEKaml	sorear: No, not complaining. Just curious why people would still hand-code those bytecodes...
16:32	sorear	ash_: How would you implement the NCI in perl 6?
16:32	[particle]1	my platform doesn't have libc :P
16:32	sorear	isBEKaml: see [particle]1's responce before last.
16:32	ash_	sorear: thats a different question all together
16:33	jnthn	isBEKaml: The hand-written stuff looks rather nicer than the generated stuff. :-)
16:33	jnthn	isBEKaml: The overall direction is to write less in PIR rather than more, though, that's for sure.
16:33	[particle]1	the generator isn't as smart as the humans yet
16:33	isBEKaml	sorear: I'd have thought C, but not assembly.. :)
16:33	colomon	isBEKaml: which is why I flounder when I have to break down and write PIR.
16:34	sahadev	http://rt.perl.org/rt3/Public/[…]lay.html?id=60672 says: "According to S29, there should be an sprintf directive %C, ...". I looked at S29 (http://perlcabal.org/syn/S29.html), which points to S32-setting-library/IO.pod for printf. But the latter (http://svn.pugscode.org/pugs/d[…]g-library/IO.pod) also doesn't contain any reference to sprintf.
16:34	sahadev	what exactly does %C directive do?
16:34	ash_	i'd probably bootstrap a subset of perl6 that can self compile, a subset that can be directly translated to native code, then go on from there
16:34	TimToady	sahadev: beats me..complex?
16:34	moritz_	sahadev: it's actually in S32-setting-library/Str.pod
16:35	sorear	isBEKaml: Well... we do that. Everything in src/Perl6/ is written in our "C"
16:35	moritz_	The special format directive, C<%C> invokes the target argument as
16:35	moritz_	code, passing it the result string that has been generated thus
16:35	moritz_	far and the argument array.
16:35	sorear	As well as the .pm modules in builtins/ and cheats/
16:35	sahadev	moritz_: thanks. found it.
16:35	TimToady	moritz_: I don't like @[] because I don't like noun markers attaching themselves to verbs.
16:35	isBEKaml	colomon: gee, looks like I'll be learning the way around things here too... :D
16:35	TimToady	how do we format a complex?
16:36	colomon	rakudo: say -1 + 3i
16:36	p6eval	rakudo fe59fa: OUTPUT«-1 + 3i␤»
16:36	moritz_	is there any free format string? :-)
16:36	TimToady	rakudo: say -1+3i
16:36	p6eval	rakudo fe59fa: OUTPUT«-1 + 3i␤»
16:37	TimToady	what if I don't want to output the spaces?
16:37	colomon	rakudo: my $a = -1 + 3i; $a.comb(/\S+/).join.say
16:37	p6eval	rakudo fe59fa: OUTPUT«-1+3i␤»
16:37	moritz_	colomon++ :-)
16:37	colomon	;)
16:37	TimToady	rakudo: my $a = -1 + 3i; $a.words.join.say
16:38	p6eval	rakudo fe59fa: OUTPUT«-1+3i␤»
16:38	moritz_	or you sprintf $c.re and $c.im
16:38	TimToady	:P
16:38	sorear	What's the canonical Perl6 way to lift a smartmatcher?
16:39	jnthn	rakudo: my $a = -1 + 3i; say "$a.re()+$a.im()i";
16:39	p6eval	rakudo fe59fa: OUTPUT«-1+3i␤»
16:39	sorear	I want to find everything for which { $^a.Str ~~ m/.../ }
16:39	isBEKaml	What's a GatherIterator() ?
16:39	ash_	rakudo: my $a = -1 + 3i; say $a.perl;
16:39	p6eval	rakudo fe59fa: OUTPUT«Complex.new(-1, 3)␤»
16:39	sorear	Rakudoese for a lazy list ATM
16:39	sorear	(subject to change)
16:39	isBEKaml	rakudo: my $str="helpoworld".split(/o/,:all); say $str; say $str.WHAT;
16:39	p6eval	rakudo fe59fa: OUTPUT«help o w o rld␤GatherIterator()␤»
16:40	sorear	a GatherIterator looks kinda like an array but can be infinite and is not calculated all at once
16:40	moritz_	and should be called List or Seq or Array or so
16:40	TimToady	"everything" seems underspecified
16:40	sorear	tokens are only split from that string when you need them
16:41	sorear	TimToady: everything is a list of methods.
16:41	sorear	currently .grep({ $^a.Str ~~ m/.../ }) but that feels wrong
16:41	isBEKaml	sorear: iterator, hmmm... why can I just display it in one fell swoop then?
16:41	TimToady	why does it feel wrong?
16:41	moritz_	isBEKaml: because the consumer (here say) iterates it
16:42	sorear	TimToady: the point, mostly
16:42	isBEKaml	sorear: if that's an iterator, I should only be able to do something like next() until it exhausts itself, right?
16:42	moritz_	sorear: .grep(/.../)
16:42	moritz_	sorear: it will stringify internally anyway
16:42	moritz_	sorear: and .grep does a smart match against its argument anyway
16:42		molaf joined #perl6
16:43	isBEKaml	moritz_: I don't understand the behaviour of "say", maybe I should go look at it again... You're saying it just doesn't do printing stuff, but more than that, like eat a _potentially_ infinite array/seq... :\|
16:44	moritz_	isBEKaml: well, how else would it print their contents?
16:44	moritz_	isBEKaml: actually it just calls .Str on its argument
16:45	TimToady	each argument, surely
16:45	moritz_	isBEKaml: the the iterator consumes itself to generate the string representation of its argument
16:45	isBEKaml	moritz_: so it stringifies whatever is passed to it?
16:45	sorear	rakudo: (1, 1, + ...).batch(20).say
16:45	moritz_	right
16:45	p6eval	rakudo fe59fa: OUTPUT«Confused at line 11, near "(1, 1, +"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
16:45	sorear	rakudo: (1, 1, + ... *).batch(20).say
16:45	p6eval	rakudo fe59fa: OUTPUT«11235813213455891442333776109871597258441816765␤»
16:45	TimToady	.Str would insert spaces
16:45	sorear	rakudo: (1, 1, + ... *).batch(20).Str.say
16:45	p6eval	rakudo fe59fa: OUTPUT«1 1 2 3 5 8 13 21 34 55 89 144 233 377 610 987 1597 2584 4181 6765␤»
16:45	moritz_	isBEKaml: yes
16:46	TimToady	rakudo: (1\|2\|3).Str.say
16:46	hejki	rakudo: (1, 1, -1+-2 ... *).batch(10).Str.say
16:46	p6eval	rakudo fe59fa: OUTPUT«any(1, 2, 3)␤»
16:46	p6eval	rakudo fe59fa: OUTPUT«Method 'Num' not found for invocant of class 'Block'␤current instr.: 'perl6;Mu;' pc -1 ((unknown file):-1)␤»
16:46	moritz_	Whatever currying not yet fully implemented
16:46	hejki	rakudo: (1, 1, (-1)+(-2) ... *).batch(10).Str.say
16:46	isBEKaml	Actually, I got a slightly different output on my local perl6 interpreter. It didn't display the "o"s at all..
16:46	p6eval	rakudo fe59fa: OUTPUT«Method 'Num' not found for invocant of class 'Block'␤current instr.: 'perl6;Mu;' pc -1 ((unknown file):-1)␤»
16:47	TimToady	rakudo: (1\|2\|3).Str.WHAT.say
16:47	p6eval	rakudo fe59fa: OUTPUT«Str()␤»
16:47	TimToady	rakudo: (1\|2\|3).Stringy.WHAT.say
16:47	p6eval	rakudo fe59fa: OUTPUT«Method 'Stringy' not found for invocant of class 'Integer'␤current instr.: '!DISPATCH_JUNCTION_METHOD' pc 16732 (src/builtins/Junction.pir:76)␤»
16:47	isBEKaml	rakudo: my $str="helpoworld".strip(/o/,:all); say $str; say $str.WHAT;
16:47	p6eval	rakudo fe59fa: OUTPUT«Method 'strip' not found for invocant of class 'Perl6Str'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
16:47	TimToady	rakudo: (~(1\|2\|3))WHAT.say
16:47	p6eval	rakudo fe59fa: OUTPUT«Confused at line 11, near "(~(1\|2\|3))"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
16:47	isBEKaml	rakudo: my $str="helpoworld".split(/o/,:all); say $str; say $str.WHAT;
16:47	TimToady	rakudo: (~(1\|2\|3)).WHAT.say
16:47	p6eval	rakudo fe59fa: OUTPUT«help o w o rld␤GatherIterator()␤»
16:47	p6eval	rakudo fe59fa: OUTPUT«Str()␤»
16:48	isBEKaml	gack..
16:48	TimToady	rakudo: (1\|2\|3).WHAT.say
16:48	isBEKaml	why are the "o"s there?
16:48	p6eval	rakudo fe59fa: OUTPUT«Junction()␤»
16:49	TimToady	~ doesn't seem to autothread correctly yet
16:49	moritz_	rakudo: say "helpoworld".subst(/o/, '', :g)
16:49	p6eval	rakudo fe59fa: OUTPUT«helpwrld␤»
16:49	moritz_	isBEKaml: because your build of rakudo is too old
16:49	moritz_	isBEKaml: colomon++ fixed that about a day ago
16:49	jnthn	TimToady: It's the Stringy / Str discintion, iirc.
16:49	isBEKaml	moritz_: A few days is too old.. :)
16:50	TimToady	yes
16:50	sorear	Unable to parse blockoid, my favorite error
16:50	moritz_	TimToady: I hope masak is going to correct that as part of his gsoc project
16:50	TimToady	rakudo: (+(1\|2\|3)).WHAT.say
16:50	p6eval	rakudo fe59fa: OUTPUT«Junction()␤»
16:50	TimToady	looks like Numeric is right tho
16:50	jnthn	Yeah
16:50	TimToady	rakudo: (1\|2\|3).Num.WHAT.say
16:51	p6eval	rakudo fe59fa: OUTPUT«Junction()␤»
16:51	TimToady	well, maybe Num is different from Str that way
16:52	sorear	it seems that smartmatching against rx// is doing something extremely wrong
16:52	TimToady	but perhaps .Num ought to blow up on a Junction, so we can make sure + is calling Numeric
16:54	jnthn	Auto-threading methods on junctions plays interestingly with coercions. :-)
16:54	jnthn	SomeType(1\|2) => Junction of SomeType instances, not SomeType
16:56	jnthn	Thankfully though not a problem for
16:57	jnthn	sub foo($x as SomeType) { ... }
16:57	jnthn	Which will still auto-thread just fine.
16:57	jnthn	(Because a junction is not Any)
16:58	TimToady	well, maybe Str is just special in having a multi that takes Junction
16:58		spinclad joined #perl6
16:58	TimToady	we have to give the low level stringifier used by say and print a name of some sort
16:58	TimToady	and .Str doesn't seem so bad as long as .Stringy provides ~ semantics, which do autothread
16:59	jnthn	nod
16:59	jnthn	I'm not complaining as such, just pointing out the various interesting semantics. :-)
16:59	lisppaste3	sorear pasted "Insane bug w/ rx and for?" at http://paste.lisp.org/display/98544
17:01	sorear	does anyone have a clue what's going on there?
17:03	TimToady	it's interpreting it as "go there" instead of "got here"? :)
17:04	jnthn	I wish I had a better guess than TimToady...
17:04	jnthn	epic wtf.
17:05		kst joined #perl6
17:05	ash_	if you take out the ~~ rx/./ it says got here twice
17:06	sorear	yes, I mentioned that in a comment
17:06	* [Coke]	ponders how to setup his dev machine so he can have master and parrot-latest installed simultaneously.
17:06	moritz_	sorear: it's $^a brokenness
17:07	moritz_	sorear: if you use -> $a { ... } instead of { $^a; $^a; } it's all good
17:07	sorear	Whaaat
17:07	sorear	Hoowwww
17:07	sorear	Known?
17:07	moritz_	sorear: there are various bug reports wrt $^a in RT - non of them affects control flow, to the best of my knowledge
17:08	* sorear	wants a streaming PIR pretty-printer
17:08	* moritz_	want's production ready Perl 6 now!
17:08	moritz_	and a ponie
17:09	* ash_	wants more time to get nq-nqp working
17:09	sorear	Production is ready for Perl 6.
17:09	* moritz_	hands ash_ a round tuit
17:10	jnthn	oh ouch, is it somehow added that placeholder to the block twice...
17:10	moritz_	seems like
17:10	jnthn	rakudo: for 1,2 { say $^a, $^a }
17:10	p6eval	rakudo fe59fa: OUTPUT«22␤»
17:10	jnthn	arse.
17:10	jnthn	Well, that explains it.
17:10	moritz_	rakudo: say ({$^a, $^a}).signature.perl
17:10	p6eval	rakudo fe59fa: OUTPUT«:(;; Mu $a, Mu $a)␤»
17:11	jnthn	Who implemented this crap...
17:11	[Coke]	jnthn: you?
17:11	* jnthn	would fix, but his Rakudo build is busted due to patches to try and get it to build on latest Rakudo.
17:11	jnthn	[Coke]: :-P
17:12	[particle]	s:2nd/Rakudo/Parrot/
17:12	jnthn	Er, yes.
17:12	jnthn	:-)
17:12	* jnthn	svn ups to get latest Parrot
17:12	jnthn	with fixes
17:12	sorear	Do I file a bug or...
17:12	moritz_	sorear: please do
17:13	sorear	Also, the original manifestation had only one reference to $^a
17:13	sorear	this double $^a weirdness came about while trying to debug the first bug
17:13	sorear	I don't know if these are the same
17:13	* isBEKaml	is svn upping Parrot and git pulling Rakudo to the latest build...
17:14	sorear	rakudobugperl ?
17:14	moritz_	@perl.org, yes
17:16	* moritz_	has a working desktop PC again, with enough RAM to build a few copies of rakudo at the same time :-)
17:16	sorear	sent
17:16	moritz_	sorear++
17:17	ash_	moritz_: ooo its always fun getting new hardware
17:18	moritz_	ash_: well, I had the RAM and the machine before, but the power supply unit made "plopp" on the last weekend, and smelled like busted condensators
17:18	moritz_	on that occasion I also ordered a new hard disc :-)
17:19		pacificpelican joined #perl6
17:20	[particle]	...unless your hardware refuses to work, like the case of my new cpu's...
17:20	sorear	Condensators!
17:20	sorear	rakudo: my $name = "moo"; say "_handler_moo" ~~ m/_handler_ moo/;
17:20	p6eval	rakudo fe59fa: OUTPUT«_handler_moo␤»
17:20	sorear	rakudo: my $name = "moo"; say "_handler_moo" ~~ m/_handler_ $name/;
17:20	p6eval	rakudo fe59fa: OUTPUT«␤»
17:20	sorear	?
17:20	sorear	Is regex interpolation just not implemented yet?
17:20	jnthn	NYI
17:21	jnthn	bkeeler++ is working on it.
17:21	slavik	moritz_: what's your native language?
17:21	jnthn	lolspeak
17:22	jnthn	;-)
17:22	sorear	What's the neatest way to say `startsWith` in rakudo?
17:22	sorear	let's see... Lenz... perl6geek.de...
17:22	sorear	I've got some guesses
17:22	jnthn	Must be .DEnmark!
17:22	sorear	er, just perlgeek
17:23	moritz_	~~ /^startsWith/ ?
17:23	slavik	moritz_: condenser/capacitor :P
17:23	moritz_	slavik: Perl :-)
17:24	moritz_	slavik: German, actually
17:24	sorear	moritz_: NYA.
17:24	sorear	NYI.
17:24	moritz_	huh?
17:24	slavik	moritz_: I guess condensator is from german?
17:24	moritz_	slavik: yes, actually with s/c/k/
17:24	slavik	which would prolly explain the russian word
17:24	sorear	rakudo: my $pfx = "_handler_"; say "_handler_moo" ~~ m/^$pfx/; #moritz_
17:24	p6eval	rakudo fe59fa: OUTPUT«␤»
17:25	moritz_	oh
17:25	moritz_	substr then
17:28	sorear	finally, it works!
17:29		REPLeffect joined #perl6
17:35		M_o_C joined #perl6
17:46		SmokeMachine joined #perl6
17:48	* [Coke]	finds mst talking smack in #yapc!
17:56		kst joined #perl6
18:00	[Coke]	yay, I now have an installed stable-parrot and regular parrot and finally build rakudo master again.
18:01	* [Coke]	wonders if ./perl6 will ever be renamed ./rakudo
18:01	isBEKaml	[Coke]: unstable is regular? :D
18:01	[Coke]	er, supported-parrot.
18:02	[Coke]	"2.3.0"
18:02	jnthn	current instr.: 'perl6;Callable[::T];' pc 13147 (src\builtins\Num.pir:185)
18:02	[Coke]	I tend to track parrot-trunk normally; just installed the 2.3.0 tarball separately and modified my rakudo build script.
18:02	jnthn	...I think the erros might not be fixed yet. :-\|
18:04	isBEKaml	[Coke]: oh, I thought you were tracking trunk when you said regular.. ;)
18:05	* isBEKaml	wonders how often he should build when updates yied so many changes all in a span of few days.. :\|
18:11	* jnthn	gets the stage 1 compiler to build and startup on latest Parrot
18:11	jnthn	Too bad it explodes as soon as it then tries to parse anything.
18:11		ash_ joined #perl6
18:25	jnthn	OK, Rakudo builds on latest Parrot now, but we has a startup crash.
18:25	jnthn	Dinnertime.
18:30		pyrimidine joined #perl6
18:31		[particle] joined #perl6
18:53		dakkar joined #perl6
18:57	diakopter	There's More Than One Way To Implement It
19:05	dalek	rakudo: c4857ab \| (Solomon Foster)++ \| src/core/metaops.pm:
19:05	dalek	rakudo: Turn on zero-argument versions of infix:<~> and infix:<//>.
19:05	dalek	rakudo: review: http://github.com/rakudo/rakud[…]e0903c731b13df748
19:05	dalek	rakudo: 4aeb5a2 \| (Solomon Foster)++ \| src/builtins/assign.pir:
19:05	dalek	rakudo: Have $a op= $b call op() to get a value for $a if $a is not defined.
19:05	dalek	rakudo: review: http://github.com/rakudo/rakud[…]742eaa9a26e352402
19:05	dalek	rakudo: 9d955f2 \| (Solomon Foster)++ \| src/core/ (2 files):
19:05	dalek	rakudo: Be more specific about types for Complex mixed-type math. Change Any, Any versions of the basic math functions to coerce to Numeric (instead of called PIR functions which coerce to Num).
19:05	dalek	rakudo: review: http://github.com/rakudo/rakud[…]47cc15280ed18ada7
19:05	pugssvn	r30505 \| colomon++ \| [t/spec] New tests of basic arithmetic mixing Complex and numeric Strs.
19:06	jnthn	colomon++ :-D
19:06	colomon	Took me quite a while to get all the tests passing again. :)
19:08	jnthn	> say "yes, I work with latest Parrot"
19:08	jnthn	yes, I work with latest Parrot
19:08	colomon	\o/
19:08	pugssvn	r30506 \| colomon++ \| [t/spec] Turn on two more zero-arg tests which now work.
19:08		colomon left #perl6
19:08	[particle]	say "YES, I CAN HAZ SPEEDUPS!!!"
19:08	jnthn	[particle]: It feels...erm...slow. :-\|
19:08		colomon joined #perl6
19:09	* [particle]	blames chromatic
19:09	[particle]	he's looking forward to benchmarking it
19:10	jnthn	erm.
19:10	jnthn	The benchmarks...segfault.
19:10	jnthn	(one of 'em, anyway)
19:11	* jnthn	runs the spectests to see what the fallout is.
19:11	colomon	It's probably segfaulting because it's running so incredibly fast it melts your CPU.
19:12	[particle]	i wish... parrot cut out the 'hcf' opcode last year, though :(
19:14		ShaneC joined #perl6
19:15	colomon	I think this would be the grfcf opcode...
19:16		ShaneC left #perl6
19:17	jnthn	Down to S04 and things are looking fairly OK.
19:18	jnthn	We epic fail S05 though.
19:19	* moritz_	hopes it's not his nqp-rx update
19:19	moritz_	well, other parrot based languages haven't complained yet
19:20	jnthn	One fail in S10, but probably not too concerning.
19:20	jnthn	S12 and S14 are in great shape.
19:21		pyrimidine left #perl6
19:21	jnthn	I don't actually have latest changes from @other mind, so some of my test fails may well be to do with that too.
19:23	colomon	oh, you have the latest tests but not the latest commits?
19:26	jnthn	Right.
19:26	jnthn	Which means I probably needn't worry too much over having a couple of fails here and there.
19:26	jnthn	At this rate, it looks like the only big issue left is going to be why we fail S05 so much.
19:27	jnthn	(We pass the majority of test files, but there's a streak of them that we don't.)
19:27	moritz_	and the rest might be due to regex failures in other files (/me hopes)
19:29	jnthn	Oh, I think the S05 ones are probably all for the same underlying reason.
19:29	moritz_	that's good
19:30	jnthn	(namely, crash in GrammarHOW.compose, investigating.)
19:30	[Coke]	jnthn++
19:33		dual joined #perl6
19:34		lichtkind joined #perl6
19:38	moritz_	rakudo: my $x *= 5; say $x
19:38	p6eval	rakudo fe59fa: OUTPUT«Use of uninitalized value in numeric context␤0␤»
19:39		sahadev joined #perl6
19:39	diakopter	pugs: my $x *= 5; say $x
19:39	p6eval	pugs: OUTPUT«0␤»
19:39	moritz_	that's wrong
19:39	diakopter	which?
19:39	moritz_	both
19:39	diakopter	heh
19:39	ash_	alpha: my $x *= 5;
19:40	p6eval	alpha 30e0ed: ( no output )
19:40	ash_	alpha: my $x *= 5; say $x;
19:40	p6eval	alpha 30e0ed: OUTPUT«5␤»
19:40	diakopter	moritz_: what should it be
19:40	moritz_	diakopter: 5
19:40	ash_	o.0
19:40	diakopter	why 5
19:40	[Coke]	I would expect 0. =-)
19:40	ash_	yeah, why 5?
19:41	jnthn	ESRSLYMAGIC
19:41	* [Coke]	grabs some popcorn. this is gonna be great!
19:41	moritz_	because op= treats an undefined value on the LHS specially
19:41	diakopter	ESRSLYUNIMPLEMENTABLE
19:41	moritz_	it takes the neutral element for this operation
19:41	diakopter	INTHEGENERALCASE
19:41	jnthn	diakopter: Just falls out of multi-dispatch. ;)
19:41	[Coke]	moritz_: ... my initial reaction is one of revulsion. =-)
19:42	moritz_	and the neutral element is provided by the 0-ary form of the operator, e.g. infix:<*>()
19:42	ash_	thats going to be confusing....
19:42	diakopter	jnthn: that's an easy way of saying, the general case isn't handled.
19:42	* moritz_	likes it
19:42	jnthn	diakopter: ?
19:42	* jnthn	was a bit surprised by it.
19:42	moritz_	it means you can say things like for @a.kv -> $k, $v { %hash{$k} *= $v }
19:42	diakopter	I mean, cases for which there isn't a multi
19:42	diakopter	that matches
19:42	moritz_	and DWIM
19:42	ash_	std: my $x *= 4;
19:43	p6eval	std 30506: OUTPUT«ok 00:01 110m␤»
19:43	jnthn	diakopter: Sure, well, then you just gotta supply a multi that matches. :-)
19:43	jnthn	I think it defaults to undef if there isn't one though
19:43	* jnthn	didn't check the patch yet
19:43	jnthn	BTW, spectests actually aren't horrifyingly slower, it seems.
19:43	moritz_	s/undef/Any/
19:43	jnthn	moritz_: er, yes. :-)
19:44	diakopter	s/er//, no
19:44	diakopter	oh wait; heh; nm.
19:45	moritz_	colomon++ # implementing this insanity
19:45	[Coke]	jnthn: weren't we hoping for 10% faster?
19:45	moritz_	[Coke]: that's kinda defied by code generation using many, many concatenations, no?
19:45	moritz_	I remember such a mail on parrot-dev
19:46	[Coke]	ah, yes. I am reminded to dig into CodeString to see if there was an easy fix there.
19:46	diakopter	rakudo: my $a = 'hi'; $a ~= $a; say $a;
19:46	p6eval	rakudo fe59fa: OUTPUT«hihi␤»
19:46	diakopter	rakudo: my $a ~= 'hi; say $a;
19:46	p6eval	rakudo fe59fa: OUTPUT«Confused at line 11, near "my $a ~= '"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
19:46	diakopter	rakudo: my $a ~= 'hi'; say $a;
19:46	jnthn	I have a grammars fix.
19:46	p6eval	rakudo fe59fa: OUTPUT«Any()hi␤»
19:47	diakopter	that one needz an empty string
19:47	[Coke]	moritz_: looks like it would be pretty easy to convert CodeString to do it the fast way internally.
19:47	moritz_	diakopter: works here, with newest rakudo
19:47	moritz_	[Coke]: then go for it :-)
19:47	diakopter	? p6eval should be newest
19:47	[Coke]	If someone volunteers to benchmark it, I can craft a patch for that this $evening.
19:48	diakopter	unless that was committed in the last hour
19:49	moritz_	diakopter: it was
19:49	diakopter	oh :)
19:49		jaldhar joined #perl6
19:50	jnthn	Going for a spectest run now, and then probably I push the update that gets us running on latest Parrot for wider feedback.
19:51	moritz_	jnthn++
19:52	ash_	nqp: class Foo { has $a }; my $a := Foo.new; $a.a := 3; my $b := $a.a; my $a.a = 4; say($b);
19:52	p6eval	nqp: OUTPUT«Redeclaration of symbol $a at line 1, near ".a = 4; sa"␤current instr.: 'parrot;HLL;Grammar;panic' pc 500 (src/cheats/hll-grammar.pir:197)␤»
19:53	ash_	are object variables write-once in nqp?
19:53	moritz_	ash_: has $.a
19:54	moritz_	my $a.a = 4;
19:54	moritz_	ENOSENSE
19:54	ash_	oh, okay
19:54	jnthn	$a.a = 4 # works fine
19:54	moritz_	[Coke]: would "compiling rakudo" be a sufficient benchmark?
19:56	ash_	hmmm odd...
19:56	ash_	nqp: class Foo { has $!a }; class Bar { has $a }; class Broke { has $.a };
19:56	p6eval	nqp: OUTPUT«Unable to parse blockoid, couldn't find final '}' at line 1␤current instr.: 'parrot;Regex;Cursor;FAILGOAL' pc 1664 (src/Regex/Cursor-builtins.pir:179)␤»
19:56	[Coke]	moritz_: absolutely. I guess I can do that myself, huh? =-)
19:56	ash_	it does $!a, and $a okay, but $.a gives me a parse error
19:57	jnthn	ash_: NQP doesn't implement that.
19:57	moritz_	ash_: ah, then use $!a
19:57	ash_	oh, okay
19:57	jnthn	(Intentionally.)
19:57	moritz_	and write an accessor if you want one
19:57	ash_	so in nqp is "has $a" a mistake?
19:58	moritz_	std: has $a
19:58	p6eval	std 30506: OUTPUT«ok 00:01 107m␤»
19:58		goksie joined #perl6
19:58	moritz_	ash_: I'm not sure what it is meant to mean... a class attribute maybe?
19:59	ash_	in perl6 the spec says class Foo { has $a } is the same as class Foo { has $.a } (i believe)
19:59	ash_	i was just wondering what it meant in nqp terms
20:00	moritz_	maybe look at the PIR that nqp generates for it
20:00	[Coke]	moritz_: oooh. there's a lot of concat going on in CodeString. glee
20:03		chitragupt joined #perl6
20:05	moritz_	[Coke]: is .'emit' a method in CodeString?
20:05		kensanata joined #perl6
20:05	moritz_	seems like, yes
20:06	moritz_	and that method looks like a hot path to me
20:07	ash_	interesting
20:07	ash_	$a and $!b both act the same, as private variables
20:08	ash_	in nqp
20:08	[Coke]	moritz_: yup. I can definitely clean tht up.
20:08		pyrimidine joined #perl6
20:09	ash_	since they are attributes, they both end up calling " getattribute $P34, self, "$a" " or getattribute $P34, self, "$!b"
20:09	ash_	so functionally, they are equivalent
20:10	colomon	actually, op= defaults to failing at the moment if there isn't a 0-arg form of the operator defined.
20:10	colomon	patches welcome. ;)
20:11	ash_	hmm, thats odd, one minor difference between the two, one made a find_lex lookup for "self" the other just used self
20:14	colomon	(defaults to failing if $a (of $a op= $b) isn't defined, that is.)
20:14		wolv joined #perl6
20:17	jnthn	Hold on tight... :-)
20:17		_buno_ joined #perl6
20:17	* jnthn	pushes
20:20	dalek	rakudo: cdd3430 \| jonathan++ \| src/ (6 files):
20:20	dalek	rakudo: A bunch of changes required to make us run under latest Parrot, mostly because we relied wrongly on :method's old semantics of putting things into the namespace, and the 4-argument substr op has gone.
20:20	dalek	rakudo: review: http://github.com/rakudo/rakud[…]14dad56c28dc86939
20:20	dalek	rakudo: 9d927ac \| jonathan++ \| build/PARROT_REVISION:
20:20	dalek	rakudo: Bump PARROT_REVISION to latest Parrot.
20:20	dalek	rakudo: review: http://github.com/rakudo/rakud[…]4848849b132cacacb
20:20	dalek	rakudo: 6e91b9b \| jonathan++ \| (10 files):
20:20	dalek	rakudo: Merge branch 'master' of gitgithub.com:rakudo/rakudo
20:20	dalek	rakudo: review: http://github.com/rakudo/rakud[…]55dcd187ca4b095f4
20:24		ash_ joined #perl6
20:31	lichtkind	there is no slurp in list context?
20:31	lichtkind	ah lines is the command
20:32	TimToady	rakudo: say infix:<min>()
20:32	p6eval	rakudo c4857a: OUTPUT«Inf␤»
20:32	TimToady	rakudo: say infix:<max>()
20:32	moritz_	there is no inward flowing context at all
20:32	p6eval	rakudo c4857a: OUTPUT«-Inf␤»
20:34		am0c joined #perl6
20:34		spinclad joined #perl6
20:35	[Coke]	How often do Codestrings have to return their string value more than once, I wonder.
20:36	moritz_	jnthn: rx.t fails here...
20:36	moritz_	Unrecognized backslash sequence: '\e' at line 1171, near "def\" ~~ /c"
20:36	moritz_	probably because pmichaud has removed that stuff from NQP-rx
20:37	moritz_	need to add that to rakudo
20:37	* moritz_	will take a look tomorrow
20:39		_buno_ joined #perl6
20:43	jnthn	moritz_: Here too.
20:43	jnthn	moritz_: oh argh...why remove it rather than move it. :-/
20:43	jnthn	moritz_++
20:44		adamsj joined #perl6
20:45	colomon	moritz_: was just thinking that fixing op= to work fully when op() isn't defined might be a good LHF for your blog project.
20:46	moritz_	colomon: then add it to the pugs repo, misc/helpnow/README
20:46	moritz_	that's where I collect ideas for that project
20:47	moritz_	colomon: I've now closed the masabug wrt op= defaults
20:48	* moritz_	-> bed
20:52	lichtkind	good night moritz_
20:52	* lichtkind	added term phaser to index
20:53		justatheory joined #perl6
20:54	lichtkind	jnthn: hej
20:56		chitragupt joined #perl6
20:56	jnthn	lichtkind: hej hej
20:58	pugssvn	r30507 \| lwall++ \| [prior.t] fix parsefail
21:01		[particle] joined #perl6
21:02		Intensity joined #perl6
21:03	pugssvn	r30508 \| lwall++ \| [charset.t] fix parsefail
21:03		ReiniUrban joined #perl6
21:26		k23z__ joined #perl6
21:28		[particle] joined #perl6
21:28		alexn_org joined #perl6
21:31		[particle] joined #perl6
21:34	dalek	csmeta: r225 \| diakopter++ \| trunk/Sprixel/ (4 files):
21:34	dalek	csmeta: [perlesque] try to fix mono's not-quite-as-correct-as-.NET's
21:34	dalek	csmeta: Type.IsAssignableFrom(Type t) by creating my own edition.
21:34	dalek	csmeta: stop catching exceptions in the reified frame trampoline; test for null instead.
21:34	dalek	csmeta: review: http://code.google.com/p/csmet[…]urce/detail?r=225
21:35		pmurias joined #perl6
21:35	pmurias	diakopter: sprixel HEAD is broken
21:37	diakopter	fixing
21:37	diakopter	:)
21:37	diakopter	sry
21:37	diakopter	(it worked on .NET but not on mono)
21:38	diakopter	pmurias: fixed :D
21:38	dalek	csmeta: r226 \| diakopter++ \| trunk/Sprixel/RunSharp/Conversion.cs:
21:38	dalek	csmeta: [perlesque] fix
21:38	dalek	csmeta: review: http://code.google.com/p/csmet[…]urce/detail?r=226
21:41	diakopter	pmurias: and yet still I get this:
21:41	diakopter	perlesque: my $a = sub (string $b) { say($b ~ $b); return 1 }; my $c = $a.Bind(frame, 'hi'); $c.Exec();
21:41	p6eval	perlesque: OUTPUT«␤Unhandled Exception: System.MissingMemberException: Cannot find property Bind in _IClosure_2␤ at TriAxis.RunSharp.TypeInfo.FindProperty (System.Type t, System.String name, TriAxis.RunSharp.Operand[] indexes, Boolean static) [0x00000] in <filename unknown>:0 ␤ at
21:41	p6eval	..TriAxis.RunSharp…
21:41	diakopter	but the same thing works fine in .NET
21:41	diakopter	sigh.
21:41		kensanata joined #perl6
21:42	pmurias	diakopter: you commit fix the bug here
21:42	pmurias	diakopter: see the daily failing test
21:42	diakopter	ok
21:43	wolv	are the perlesque .cs files generated from something? they don't look exactly something I'd want to code... :)
21:44	dalek	csmeta: r227 \| pawelmurias++ \| trunk/Sprixel/ (2 files):
21:44	dalek	csmeta: failing test with P6capture
21:44	dalek	csmeta: review: http://code.google.com/p/csmet[…]urce/detail?r=227
21:44	pmurias	wolv: you mean the grammar ones?
21:44	wolv	yes
21:44	diakopter	heh
21:44	wolv	(the only ones under Perlesque/ I think?)
21:44	diakopter	it's code-gen-gen
21:45	diakopter	twice-deferred code
21:45	diakopter	input language syntax directly transformed to runtime code, via parameterized codegen, where the compiler action routines themselves are also codegen-d
21:46	wolv	ah, thanks
21:46	diakopter	wolv: you're the first that understood my explanation of that
21:46	diakopter	my (first) explanation, I mean.
21:47	wolv	I took that CS course ;)
21:47	diakopter	wait, what?
21:47	diakopter	what cs course
21:47	pmurias	somebody in your uni suggested implementing stuff that way?!
21:48	diakopter	heh
21:48	wolv	mostly I was joking. but I did read up on all kinds of weird stuff during the compiler course...
21:48	* diakopter	hasn't taken a cs course
21:48	diakopter	<-- (and it shows)
21:49	pmurias	what did you study?
21:49	wolv	me?
21:49	diakopter	nothing, but I got a degree in business (finance/investments)
21:49	diakopter	actually, during college I studied theology/hermeneutics
21:49	wolv	by doing nothing? :)
21:50	diakopter	yeah; the degree isn't worth much.
21:50		bradb left #perl6
21:50	diakopter	(or anything, at this point)
21:53	diakopter	pmurias: actually it's not parsefailing
21:53	diakopter	it runs
21:54	pmurias	yes
21:54	diakopter	the positionals member is not initialized
21:54	diakopter	want me to fix it?
21:55	pmurias	ok
21:55	diakopter	also I need to move the "Catch" in TopLevel.cs up several lines
21:55	diakopter	so it doesn't catch that exception
21:55	pmurias	you shouldn't catch all exceptions
21:55	diakopter	that's what I said
21:55	pmurias	the parse ones should inherit from a parse exception subclass
21:55	diakopter	yes, someday.
21:56	diakopter	here's the error it throws now:
21:56	diakopter	Unhandled Exception: System.Reflection.TargetInvocationException: Exception has been thrown by the target of an invocati
21:56	diakopter	on. ---> System.NullReferenceException: Object reference not set to an instance of an object. at Sprixel.Runtime.P6capture.add_positional(P6object pos) in C:\Users\mwilson\Documents\src\csmeta\Sprixel\src\Types\
21:56	diakopter	P6capture.cs:line 11
21:57	diakopter	(fixing and comitting)
21:59	diakopter	sigh.
21:59	pmurias	?
21:59	diakopter	svn up :P
22:00	pmurias	i don't do svn up
22:00	diakopter	what do you do
22:00	pmurias	i use git svn rebase
22:00	diakopter	ok
22:00	diakopter	:)
22:00	pmurias	as i'm not on speaking terms with the svn CLI client
22:01	diakopter	lol.
22:03	dalek	csmeta: r228 \| diakopter++ \| trunk/Sprixel/ (3 files):
22:03	dalek	csmeta: [perlesque] program exceptions aren't caught as a "parsefail"
22:03	dalek	csmeta: fix P6capture to initialize its positionals member.
22:03	dalek	csmeta: review: http://code.google.com/p/csmet[…]urce/detail?r=228
22:03	dalek	csmeta: r229 \| diakopter++ \| trunk/Sprixel/Main.cs:
22:03	dalek	csmeta: [perlesque] disable debug mode
22:03	dalek	csmeta: review: http://code.google.com/p/csmet[…]urce/detail?r=229
22:06		[particle] joined #perl6
22:06		sahadev left #perl6
22:13	diakopter	wolv: much of that action code can feasibly be extracted out into combinators in only singly-deferred C#
22:13	diakopter	so that one could call a routine like c.DescendNewLexicalScope()
22:14	diakopter	instead of the 200-character 4 lines
22:14	wolv	heh
22:18		pyrimidine left #perl6
22:29		crythias joined #perl6
22:45		crythias left #perl6
22:57		wknight8111 joined #perl6
22:57		rgrau__ joined #perl6
23:17		cls_bsd joined #perl6
23:33	dalek	book: 1106aab \| chromatic++ \| src/basics.pod:
23:33	dalek	book: Edited chapter. It looks good.
23:33	dalek	book: review: http://github.com/perl6/book/c[…]0b48d5ea93ca080ee
23:35		crythias1 joined #perl6
23:41		cls_bsd joined #perl6
23:45		jaldhar joined #perl6
23:49	pugssvn	r30509 \| lwall++ \| [STD] allow subscripts on regex variables so $x[0] isn't taken as a character class
23:49	pugssvn	r30509 \| [STD,Cursor] revamp of character class parsing to be more helpful/correct
23:54	pugssvn	r30510 \| lwall++ \| [S03] remove p5=> description since it's not supported in core
23:54		kst joined #perl6

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs