IRC log for #perl6, 2010-09-04

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs

← Previous day | Channels | #perl6 index | Today | Next day → | Search | Google Search | Plain-Text | plain, newest first | summary

All times shown according to UTC.

Time	S	Nick	Message
00:00		masak	rakudo: multi infix:<->(Str $a, Str $b) { $a.subst($b, "") }; say "Perl 6 sucks rocks" - "sucks "
00:00		p6eval	rakudo 52f9ea: OUTPUT«Perl 6 rocks␤»
00:00		masak	ingy: ^^
00:00		masak	rakudo: multi infix:<->(Str $a, Str $b) { $a.subst($b, "") }; say 44 - 2
00:00		p6eval	rakudo 52f9ea: OUTPUT«42␤»
00:00		masak	\o/
00:01		masak	hugme: hug whoever fixed that
00:01		* hugme	hugs whoever
00:01		masonkramer	how does @urn.pick(*, :replace) work?  What is :replace?
00:02		masak	masonkramer: the combination of * and :replace will give you an infinite sequence of randomly picked elements.
00:02		masonkramer	I get what it does, but what is :replace?
00:02		masak	masonkramer: :replace means that there's the option of getting an element that was already picked.
00:02		masonkramer	I mean, what type of thing is it?
00:02		masak	oh.
00:02		masonkramer	I know what it means, but what do I call it?
00:02		masak	it's called a named argument.
00:03		masak	it's sugar for :replace(Bool::True)
00:03		masonkramer	ahhh
00:03		masonkramer	there we go
00:03		masak	sometimes, especially in the case of operators, we tend to call them "adverbs".
00:04		masonkramer	I can't figure out what the difference between an operator and a method is, is there a clear definition?
00:04		masonkramer	you're calling .pick an operator, but it looks like a method to me
00:04		masak	no, it's a method.
00:05		masonkramer	ok
00:05		masak	I call things like + and ~~ and , operators.
00:05		masonkramer	but those are functions with fancy syntax
00:05		masak	though it's worth mentioning that in Perl 6... right.
00:05		masak	either they're subs or they're methods.
00:06		masak	postfix operators tend to be methods.
00:06		masonkramer	a method is what?  A function that has $self in socpe?
00:06		masak	(though I admit I still haven't seen the real system there. I often get it wrong)
00:06		masak	masonkramer: a method is a separate thing in Perl 6.
00:07		masak	rakudo: class A { our sub foo() { say "sub foo" }; method foo { say "method foo" }; A::foo; A.new.foo
00:07		p6eval	rakudo 52f9ea: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 22␤»
00:07		masak	rakudo: class A { our sub foo() { say "sub foo" }; method foo { say "method foo" } }; A::foo; A.new.foo
00:07		p6eval	rakudo 52f9ea: OUTPUT«sub foo␤method foo␤»
00:07		masonkramer	that's not going to be confusing
00:08		masonkramer	;)
00:08		masonkramer	Just kidding, it's even more confusing in Perl 5
00:08		masak	it has been known to confuse some people, yes.
00:09		masak	I'm only starting to see the real wins with having a clean separation.
00:09		masonkramer	what does method have in there?  $ stands in for $self?
00:09		masak	every method has a 'self'
00:10		masak	rakudo: class A { has $.x; method foo { say self.x } }; A.new(:x(42)).foo
00:10		p6eval	rakudo 52f9ea: OUTPUT«42␤»
00:10		masonkramer	rakudo: class A { method foo { say self } }
00:10		p6eval	rakudo 52f9ea:  ( no output )
00:10		masonkramer	rakudo: class A { method foo { say self } }; A.new.foo
00:10		p6eval	rakudo 52f9ea: OUTPUT«A()<0x559fce0>␤»
00:11		masonkramer	rakudo: class A { has $.x; method foo { say self.x } }; A.new(:x).foo
00:11		p6eval	rakudo 52f9ea: OUTPUT«1␤»
00:11			cibs left #perl6
00:12		masak	it says '1', but it should say 'True'
00:12		masak	(according to the latest spec on enums)
00:12		masak	rakudo: class A { has $.x; method foo { say self.x.perl } }; A.new(:x).foo
00:12		p6eval	rakudo 52f9ea: OUTPUT«Bool::True␤»
00:13		masonkramer	I see .perl a lot, is that the p6 equivalent of Data::Dumper?
00:13		masak	yes.
00:14		masak	built in for your convenience.
00:14		masonkramer	beautiful, beautiful
00:15		masonkramer	my programs already define a UNIVERSAL::Dump which dumps $self
00:15		masonkramer	but I always feel bad about doing things like that, for the obvious reason of namespace pollution; it should be built in
00:15			svetlins_ left #perl6
00:16			svetlins_ joined #perl6
00:17			sftp left #perl6
00:26			justatheory left #perl6
00:26			svetlins_ left #perl6
00:42		perigrin	masonkramer: Moose::Object implements one too
00:43		perigrin	so you are at least in good company with p6 and moose.
00:43			svetlins_ joined #perl6
00:44		masak	perigrin: I didn't know that. cool.
00:44		masonkramer	very nifty
00:44		perigrin	masak: yeah it has since forever.
00:45		perigrin	confess $obj->dump; is a *very* common debugging idiom in my world.
00:46		perigrin	I'm not sure what the p6 form of confess is/will be ... but $obj.perl is a nice feature :)
00:46		masak	die
00:46			svetlins_ left #perl6
00:46		masonkramer	$obj.explode
00:47		masak	I don't know how to die in Rakudo *without* the backtrace. I want that sometimes.
00:47		gfldex	kill -9 ?
00:47		perigrin	yeah ... when the masses arrive they'll do nothing but complain
00:47			whiteknight left #perl6
00:47		perigrin	it is one of the largest complaints about moose from people learning it
00:48		masak	what is?
00:48		perigrin	300 lines of "garbage" when you trip an error
00:49		perigrin	catalyst can get kind of ... intense ... with it's debugging output ... if you trip something in say a runtime trait used in a controller that inherits from a controller ...
00:53			Psyche^ joined #perl6
00:54		tylercurtis	masak: maybe there should be a :quietly adverb for &die.
00:55		tylercurtis	Or maybe :concisely would be more fitting.
00:55		masak	oh, I wouldn't mind the error message as such.
00:55			Psyche^ is now known as Patterner
00:55		masak	but the stack trace is sometimes just not needed.
00:55		masak	applications like pls are perhaps the most obvious example.
00:55		masak	right now I do &say, then &exit.
00:56		masak	or &warn, then &exit.
00:56		masak	or &note, then &exit :)
00:56		gfldex	die :quietly
00:57		gfldex	looks about right :->
00:57		tylercurtis	:quietly seems more like something that would totally silence the error message, though.
00:57		* hugme	hugs tylercurtis, good vi(m) user!
00:58			hercynium left #perl6
00:58		masak	hugme: that was a misfire, mind you.
00:58		gfldex	:q
00:58		* hugme	hugs gfldex, good vi(m) user!
00:58		gfldex	:->
00:59		perigrin	tylercurtis: no :silently would silence the error message ...
00:59		gfldex	i learned a new trick and therefor can go to bed now
00:59		perigrin	:quietly would just muffle it
00:59		* hugme	hugs perigrin, good vi(m) user!
00:59		gfldex	good * #perl6
01:04			svetlins_ joined #perl6
01:04		masak	o/
01:05		sorear	really :concisely belongs on try, or maybe as a contextual
01:05		sorear	I hate confess; -MCarp::Always
01:06		sorear	the call site is the wrong place to put knowledge of how much verbosity is needed
01:06			svetlins_ left #perl6
01:06		perigrin	yes
01:06		perigrin	try { ... } catch :loudly { ... }
01:07		masak	in the case of a die outside of any try, there wouldn't be a try to peg the adverb on.
01:08		perigrin	this is where my knowledge of exception in Perl6 and Rakudo falls down entirely
01:09		masak	perigrin: there's some info in S04.
01:09		masak	perigrin: generally, CATCH blocks appear inside the try blocks.
01:10		* perigrin	has to wait for November before he can multitask and look up S04 without losing SSH
01:10		perigrin	apple--
01:10		perigrin	I will however check in a bit
01:11			MarcoS joined #perl6
01:11		MarcoS	Totally new here... Is it possible to install perl on Vista?
01:12		masak	there is a win32 installer for Rakudo Star somewhere.
01:13		MarcoS	OK, thanks!
01:13			justatheory joined #perl6
01:13		masak	here. http://github.com/rakudo/star/downloads
01:13		masak	the .msi file in the middle.
01:13			meppl left #perl6
01:15		masak	nom &
01:15			masak left #perl6
01:16		MarcoS	..getting it now...
01:17		sorear	perigrin: er... screen on the far side of ssh?
01:21			MarcoS left #perl6
01:23			svetlins_ joined #perl6
01:27			Italian_Plumber left #perl6
01:41		Util	Instead of writing `for 0..@array.end -> $i {...}`, is there a convenience method on Array that returns a Range object equiv to `0..@array.end`? If not, should there be?
01:41		Util	Oh, @array.keys. Doh!
01:49			risou joined #perl6
01:49			svetlins_ left #perl6
01:50			jhuni joined #perl6
01:51		tylercurtis	Util: and if you want to get at the item at index $i, too, "for @array.kv -> $i, $v {...}"
01:53		Util	tylercurtis: Thanks! I actually showed this off at Atlanta.pm this month, and had already forgotten it.
01:57		lue	hai again o/
02:04			araujo joined #perl6
02:04		ingy	goto 17:00
02:06		ingy	phenny: tell masak That's (the multi redef of <->) so cool!
02:06		phenny	ingy: I'll pass that on when masak is around.
02:06		ingy	thanks
02:06			svetlins_ joined #perl6
02:13			_jaldhar left #perl6
02:15			Schwern joined #perl6
02:23			literal joined #perl6
02:27			bjarneh joined #perl6
02:36			risou left #perl6
02:36		perigrin	sorear: yes
02:40			risou joined #perl6
02:53			risou left #perl6
03:18			Chillance left #perl6
03:25			masak joined #perl6
03:26		masak	ingy: I know! Perl 6 <del>sucks</del> rocks!
03:26		phenny	masak: 02:06Z <ingy> tell masak That's (the multi redef of <->) so cool!
03:29			tylercurtis left #perl6
03:29		masak	today on Twitter we find something as unusual as a reformed Perl6/DNF-memer: http://twitter.com/batchout/status/22925307636
03:29		masak	and other examples of people generally getting it: http://twitter.com/outZider/status/22925279162
03:34		lue	I'm not a real person!? :*(
03:35		masak	lue: I sincerely believe you are.
03:35		masak	note that e didn't write '*only* real people are playing...'
03:36		masak	or '*all* real people are playing...'
03:36		masak	he's basically just pointing out that the intersection of the two sets 'real people' and 'people who are playing a prerelease of Duke Nukem Forever' is non-empty.
03:37		masak	lue: oh by the way, I've been enjoying watching your progress with the <A::B> feature over the past few days.
03:38		masak	lue: I recognize the steps you've been going through from when I was discovering Rakudo back in 2008.
03:39		lue	.proof(real people = duke nukem. me = !(duke nukem), ∴ me = !(real people). QED at my expense)
03:40		masak	I stand in awe at your use of the Unicode THEREFORE symbol.
03:41		masak	I would advise, however, a slight refresher on the meaning of the assignment symbol, and its various uses.
03:42		lue	I think all hope of me making <A::B> work was murdered when I found out there would be PIR involved on top of the parsing business. :)
03:42		masak	lue: luckily, you got pmichaud to promise to do it for you.
03:42		lue	rakudo: say $_.WHAT
03:42		p6eval	rakudo 52f9ea: OUTPUT«Any()␤»
03:43		masak	lue: are you aware of the LHF exchange program over at RT? :)
03:43			_jaldhar joined #perl6
03:44		lue	rakudo: use MONKEY_TYPING; augment Any { method oO($text) { ... }; }; .oO(I invented a useless method that behaves like a comment!)
03:44		p6eval	rakudo 52f9ea: OUTPUT«===SORRY!===␤In "augment" declaration, typename Any must be predeclared (or marked as declarative with :: prefix) at line 22, near " { method "␤»
03:44		masak	lue: text still needs to be quoted.
03:45		masak	lue: you're not creating a new sublanguage, just doing a normal method signature.
03:46		lue	Eh, it was a quick attempt to let you use .oO() thought bubbles. .oO(P6::Acme::Thoughts ?)
03:46		masak	if you want to get rid of the quotes, you need to make .oO() a quote operator.
03:46		masak	see S02 and S03.
03:47		masak	specifically, S02:3699.
03:47		lue	I would much rather implement <A::B> myself, but I think we all learned today I wouldn't be able to do so anytime soon :)
03:48		masak	lue: right. that's why I suggested the LHF exchange program.
03:48		masak	it might be a simple way for you to show your gratitude to pmichaud++ for eventually implementing your feature.
03:48		masak	not that that need be your driving force, of course.
03:48		masak	you could do it just because it's fun!
03:51		masak	lue: http://rt.perl.org/rt3/Search/[…]!%3D%20'resolved'
03:52		masak	lue: I'll even review and apply any patch you write.
03:53		lue	I believe I already have a search ready for LHF. Let me check
04:00			lue joined #perl6
04:00		lue	don't you love random unexpected shutdowns of your 12-year-old laptop ?
04:04			JimmyZ joined #perl6
04:10		JimmyZ	rakudo: say ~(my @Fibonacci := (0, 1, { * + * } ... 900));
04:10		p6eval	rakudo 52f9ea: OUTPUT«0 1␤»
04:10		lue	rakudo: say 10000000[1]
04:10		p6eval	rakudo 52f9ea:  ( no output )
04:11		JimmyZ	rakudo: say ~(my @Fibonacci := (0, 1, -> $a, $b { $a + $b } ... 900));
04:11		p6eval	rakudo 52f9ea: OUTPUT«0 1 1 2 3 5 8 13 21 34 55 89 144 233 377 610␤»
04:12		masak	卓明亮! \o/
04:12			jimi_hendrix left #perl6
04:13		masak	我看你读过一些博客文章。 哈哈
04:14		lue	.oO(#perl6 reminds me almost every day to start learning japanese)
04:14		masak	for the record, those are hanzi, not kanji :)
04:15		lue	the 1000000[1] error message only displays in the REPL, it's weird.
04:18		lue	Could it be because the error mesage is invoked using fail ?
04:18		masak	yes.
04:19		JimmyZ	masak: 是啊
04:20		JimmyZ	masak: why { * + * } doesn't work?
04:20		masak	JimmyZ: because it's a closure in a closure.
04:20		masak	JimmyZ: * + * already means { $^a + $^b }
04:20		JimmyZ	yes
04:21		masak	with the extra curly braces, you get two levels of curlies. and it doesn't work.
04:21		JimmyZ	rakudo: say ~(my @Fibonacci := (0, 1, * + * ... 900));
04:21		p6eval	rakudo 52f9ea: OUTPUT«0 1 1 2 3 5 8 13 21 34 55 89 144 233 377 610␤»
04:22		JimmyZ	masak: my fault :0
04:22		masak	no sweat. now you know :)
04:23		JimmyZ	masak: yes, thanks ;)
04:24		\xF0	nice
04:25		* masak	really likes Haskell's </> operator for FilePath. http://www.haskell.org/ghc/doc[…]tml#v%3A%3C%2F%3E
04:25		masak	maybe we should steal it.
04:25		JimmyZ	rakudo: sub infix:<+-*/>($a, $b)  { ( { $a + $b }, { $a - $b }, { $a * $b }, { $a / $b } ).pick(1) }; say 5+-*/2;
04:26		p6eval	rakudo 52f9ea: OUTPUT«_block162␤»
04:27		masak	rakudo: sub infix:<+-*/>($a, $b)  { ( { $a + $b }, { $a - $b }, { $a * $b }, { $a / $b } ).pick(1).() }; say 5+-*/2;
04:27		p6eval	rakudo 52f9ea: OUTPUT«10␤»
04:27		masak	\o/
04:27		masak	you have to call the block before you return it.
04:28		masak	s/ it//
04:28		JimmyZ	是的，我刚刚发现了 ， 呵呵
04:28		masak	:)
04:30		JimmyZ	我刚刚在把perl6 介绍给同事
04:30		masak	nice! what did they think?
04:31			envi^home joined #perl6
04:31		JimmyZ	他们感觉非常强大和简单
04:31		JimmyZ	很有兴趣，呵呵
04:31		masak	excellent.
04:32		JimmyZ	rakudo: sub infix:<+-*/>($a, $b) { ( { $a + $b }, { $a - $b }, { $a * $b }, { $a / $b } )>>.() }; say 5+-*/2;
04:32		masak	JimmyZ: you are our Chinese representative. :)
04:32		p6eval	rakudo 52f9ea: OUTPUT«Tried to find null name␤  in 'infix:<+-*/>' at line 22:/tmp/lAzbNODIkJ␤  in main program body at line 22:/tmp/lAzbNODIkJ␤»
04:32		JimmyZ	呵呵
04:32		masak	您是我们在中国内地的代表。
04:32		JimmyZ	Thanks
04:32		masak	our pleasure :)
04:32		JimmyZ	这个代码有什么错误？
04:33			isBEKaml joined #perl6
04:33		masak	I don't know. It's related to the >>.(), to be sure.
04:33		JimmyZ	yes
04:33		masak	but I'm not sure if it should be allowed or no.
04:34		masak	rakudo: say ( { 1 + 1 }, { 2 + 2 } )>>.()
04:34		p6eval	rakudo 52f9ea: OUTPUT«Tried to find null name␤  in main program body at line 22:/tmp/ogz7FsNuwd␤»
04:34		masak	rakudo: say ( { 1 + 1 }, { 2 + 2 } )>>()
04:34		p6eval	rakudo 52f9ea: OUTPUT«Tried to find null name␤  in main program body at line 22:/tmp/ITMWg1XBhk␤»
04:34		* masak	submits rakudobug
04:34		JimmyZ	rakudo: say ( { 1 + 1 }, { 2 + 2 } ).>>.()
04:34		p6eval	rakudo 52f9ea: OUTPUT«===SORRY!===␤Confused at line 22, near "say ( { 1 "␤»
04:35		masak	not dot allowed before >> :)
04:35		JimmyZ	alpha: say ( { 1 + 1 }, { 2 + 2 } ).>>.()
04:35		p6eval	alpha 30e0ed: OUTPUT«Can only transform an Object to p6opaque␤in Main (file src/gen_setting.pm, line 324)␤»
04:35		JimmyZ	alpha: say ( 1,2 ).>>.say
04:35		p6eval	alpha 30e0ed: OUTPUT«1␤2␤11␤»
04:35		masak	huh.
04:35		JimmyZ	It's allowed
04:35		JimmyZ	by STD
04:35		masak	I stand corrected.
04:36		JimmyZ	std: say ( 1,2 ).>>.say
04:36		p6eval	std 32123: OUTPUT«ok 00:01 118m␤»
04:36		* masak	submits rakudobug
04:36		isBEKaml	std: say ( { 1 + 1 }, { 2 + 2 } )>>.()
04:36		p6eval	std 32123: OUTPUT«ok 00:01 118m␤»
04:36		isBEKaml	OHHAI,p6!
04:37		JimmyZ	It's easy to fix, I had fixed it in alpha :)
04:37		masak	isBEKaml: hi! \o/
04:37		isBEKaml	masak! good release?
04:38		masak	of Yapsi? yes, it was the easiest one so far, I think :)
04:38		masak	soon I'll be ready to make someone else try to be the release manager.
04:38		isBEKaml	:)
04:38		isBEKaml	Yapsi++
04:38		masak	the next big feature up ahead is definitely subs.
04:39		masak	a stepping-stone to that is first-class closures.
04:39		isBEKaml	yes. I can see that. for my part, I'm looking to get some rudimentary expression eval working.
04:39		masak	when we have subs, we can start adding operators the way they're supposed to be added.
04:39		masak	I'm reluctant to add in a lot of ops before that...
04:40		isBEKaml	not a lot, I said "rudimentary"
04:40		masak	:)
04:40		masak	also, I'd like to start exploring switching to STD as the frontend.
04:40		JimmyZ	that's a great idea :)
04:40		isBEKaml	yes, I looked at STD, then thought "we have a good syntax checker, why do we have to go all the way just to check syntax ?"
04:41		masak	I meant using STD to get the parse tree.
04:41		masak	(and better error messages, and a lot of error checking for free)
04:41		isBEKaml	I don't know we can do that. Didn't check fully. That thing was huge. ;)
04:42		masak	isBEKaml: there's a script called viv that'll give you a parse tree already.
04:42		isBEKaml	(almost hung up my editor until I disabled syntax and language features)
04:42		masak	we can use the same mechanism viv uses to get the parse tree.
04:43		masak	the big question is which format to use to deserialize it into Yapsi.
04:43		lue	STD spits out an AST, right? [in which case it's perfect for the beginnings of a gcc compiler]
04:43		isBEKaml	the answer would be "Whatever's comfortable. " knowing full well Whatever's a term here. :)
04:43		masak	lue: depends if you make a difference between "parse tree" and "AST".
04:44		isBEKaml	lue: I made that mistake once. :)
04:44		masak	isBEKaml: I think optimizing for ease-of-reading on the Perl 6 side is pretty sane. it'll be a big performance hit no matter what.
04:47		isBEKaml	masak: performance hit? as in, everytime we have to parse STD first?
04:48		masak	as in, there'll be a bottleneck shuttling the parse tree (in whatever form) from Perl 5 to Yapsi.
04:49		isBEKaml	hm, I see that now.
04:49		masak	rakudo: say ( 1,2 ).>>.say
04:49		p6eval	rakudo 52f9ea: OUTPUT«===SORRY!===␤Confused at line 22, near "say ( 1,2 "␤»
04:49		masak	especially if we read things using a grammar.
04:50			mfollett joined #perl6
04:50		isBEKaml	I already saw some slowness in blocks evaluation when doing infinite loop structure.
04:50		masak	yes, Yapsi is already a bit sluggish.
04:50		isBEKaml	it was just going on creating blocks. not just with 'loop' but with 'while', 'until' everything. Creating blocks seems to be slow.
04:51		masak	huh.
04:51		masak	there are a few linear scans in there; one in block calling and one in gotos.
04:52		isBEKaml	consider this: my $a = 3; while --$a { say 42 }; loop { say 24 }
04:52		masak	I've just been to lazy to write faster variants. but a single pre-scan + hash population might speed things up considerably, now that I think about it.
04:52		masak	and it's LHF. :)
04:52		isBEKaml	once that's parsed by yapsi, it'd print 42 twice and go doing the 24 prints infinitely. I'd expect it be faster from the first print.
04:53		masak	yapsi: my $a = 3; while --$a { say 42 }; loop { say 24 }
04:53		p6eval	yapsi:  ( no output )
04:53		masak	:/
04:53		isBEKaml	loop isn't in.
04:53		masak	oh right.
04:53		* masak	does 'while 1' locally
04:56		masak	it does about 2 iterations per second over here :)
04:56		isBEKaml	1 per second on my machine.
04:58		masak	lue: I just saw your patch on the .[1] ticket. thing is, I think it should be a 'fail', not a 'die'. though there's something wrong with either p6eval or with Rakudo's &say function.
05:01		* isBEKaml	looks at viv
05:16		lue	I'm trying to find the difference between fail and die, but google's stupid search engine won't let me :/
05:17		masak	you might try the spec instead.
05:18		lue	Well, I tend to use the web version of the spec for convenience, and it uses google... [I'm trying straight from the repo now]
05:19			mfollett left #perl6
05:22		lue	afk
05:25		sorear	someone said viv?
05:26		masak	yes. in the context of Yapsi.
05:26		masak	the big question mark right now is how to marshal data from STD to Yapsi efficiently.
05:27		sorear	I recommend writing a JSON::XS clone for Rakudo/Parrot
05:28		masak	does such a beast exist?
05:28		masak	it sounds like a fine idea, fwiw.
05:28			Schwern left #perl6
05:31		sorear	masak: definitely not at the Rakudo level; the best we have is JSON::Tiny, and it uses nqp-rx so will be a lot slower than a hypothetical C version
05:32			alester joined #perl6
05:32		masak	sorear: aye.
05:32		sorear	there might be something in the shadowy void of parrot/runtime/library
05:32		masak	sorear: I will investigate this venue further. thank you.
05:32		masak	there seems to be a JSON.pir in there, yes.
05:33			constant left #perl6
05:33		masak	and it runs!
05:33		masak	\o/
05:34		sorear	compilers/data_json is the corresponding reader
05:34		sorear	it seems to use PGE or NQP-rx, though
05:34		masak	oh.
05:34		sorear	<?PGE::util::die ''> # seems to be a tell
05:35		masak	aye, it uses PGE.
05:35		sorear	how is PGE for speed?
05:36		masak	pretty OK.
05:36			constant joined #perl6
05:36			constant left #perl6
05:36			constant joined #perl6
05:36		masak	might be a premature optimization to write something in PIR and expect it to be faster than PGE.
05:36		masak	I doubt one'd get many percent speedup from that.
05:37		sorear	does PGE even work?
05:38		sorear	I don't recall it having been touched since the last two complete Parrot compatibility breaks
05:38		masak	I don't know. haven't tried to run it for a while.
05:38		masak	but I'd expect it to, given that it has tests.
05:38		sorear	(symmetric PCC and immutable strings)
05:38		masak	nod
05:38		* masak	tries to run the PGE tests
05:39		masak	make library_tests, apparently
05:39			azert0x left #perl6
05:39		masak	yep, tests still pass.
05:41			patch joined #perl6
05:55			cotto left #perl6
05:56			JimmyZ left #perl6
06:07		masak	perl6: say substr("camel", 0|1, 2&3)
06:07		p6eval	rakudo 52f9ea: OUTPUT«all(any("ca", "am"), any("cam", "ame"))␤»
06:07		p6eval	..pugs: OUTPUT«all(VJunc any(VRef <Scalar::Proxy:0x2abd72eb26f1>,VRef <Scalar::Proxy:0x2abd72eb85f9>),VJunc any(VRef <Scalar::Proxy:0x2abd72eb6131>,VRef <Scalar::Proxy:0x2abd72eb9721>))␤»
06:07		masak	♥
06:08		masak	Rakudo++
06:08			alester left #perl6
06:20			Ross joined #perl6
06:28			Alias_ left #perl6
06:32		isBEKaml	masak: PGE?
06:32		isBEKaml	I'd be rather happy if I don't have to install a lot of deps for viv. ;)
06:33		masak	isBEKaml: it's the previous-generation grammar engine for Rakudo. it ships with Parrot.
06:33		masak	and no, I don't think we'll be using PGE. that feels a bit silly :)
06:34		isBEKaml	why? outdated grammar rules? :)
06:35		masak	outdated grammar engine.
06:35		masak	but I'd much rather employ something like JSON::XS, like sorear said.
06:36		isBEKaml	the json.pbc in parrot?
06:37		sorear	the json.pbc in parrot is a json emitter, not a json reader
06:37		sorear	the data_json.pbc in parrot's languages folder is a PGE-based json reader
06:37		isBEKaml	I see.
06:45			xinming_ is now known as xinming
06:50			molaf joined #perl6
07:04			masak left #perl6
07:12			isBEKaml left #perl6
07:35			chips joined #perl6
07:37			chips left #perl6
07:38			Elvin joined #perl6
07:38		Elvin	anyone have Perl free software?
07:38			Elvin left #perl6
07:40			xinming left #perl6
07:40			xinming joined #perl6
07:40			Su-Shee joined #perl6
07:40		Su-Shee	good morning.
07:40		araujo	perl ... free software ...
07:55			justatheory left #perl6
07:56			Guest23195 joined #perl6
07:59			Ross left #perl6
08:00			svetlins_ left #perl6
08:01			tadzik joined #perl6
08:01		tadzik	oh hai
08:08			karb joined #perl6
08:08			wamba joined #perl6
08:12			svetlins_ joined #perl6
08:13			snearch joined #perl6
08:23			Mowah joined #perl6
08:33			Ross joined #perl6
08:38		Quadrescence	anyone have Perl free software?
08:40		sorear	Parse error
08:40		tadzik	what do you mean?
08:40		tadzik	you mean do we have free software in Perl? Sure thing
08:45			svetlins_ left #perl6
08:45		_sri	in soviet russia Perl software frees you!
08:50		sorear	what is correct test form for when you commit something on purpose that breaks half the testsuite?
08:50			total joined #perl6
08:51		total	how to learn perl 6 ?
08:51		sorear	the book would not be a bad start
08:52		sorear	http://github.com/downloads/pe[…]ok/2010.08.a4.pdf
08:52		total	thx for reply...but i cant find any book abt perl6
08:53			gfx joined #perl6
08:53		total	dowloading
08:53		total	...
08:53		total	thx...
08:53			Siddy is now known as Trashlord
08:53		total	the best os for learning perl 6 is ?
08:53		total	i m totally noobie...
08:54		total	linux?
08:54		sorear	you should not let language dictate your choice of OS
08:54		sorear	unless you're completely new to computers, in which case, why are you diving right into programming?
08:55		total	i m really interest in computing..
08:55		total	but i dunno where to start ...
08:56		total	i m a student in civil...so....
08:56		total	as hobby as weel...
08:56		sorear	people in this channel use Linux, Mac OS X, and Windows
08:56		sorear	at least
08:56		sorear	so it's probably no object
08:56		moritz_	good morning
08:57		sorear	good morning moritz_
08:57		total	hi..
08:57		total	i using windows...
08:57		total	i had install the rakudo star...
08:58		total	its sth like terminal..???
08:58		total	right ?
08:59		moritz_	well, you have to open a terminal window to run the perl6.exe binary
09:01		total	i dowload from github...http://github.com/rakudo/star/downloads/
09:01		total	the msi version...
09:01			aloha left #perl6
09:02			svetlins_ joined #perl6
09:02			bacek left #perl6
09:04			total left #perl6
09:08			ruiwk joined #perl6
09:08		tadzik	yay newcomers
09:08		ruiwk	hi
09:09		ruiwk	i am new learner in programming world
09:09		tadzik	yeah, we can recognize you :)
09:09		ruiwk	perl 6 is the 1st languague i hope to learn ..
09:09		ruiwk	ha
09:10		tadzik	ruiwk: as you can probably see, the Rakudo Star msi is not really up to date, but I doubt it would matter to you as to the newcomer
09:12		ruiwk	will it better to use linux to learn perl6 ?
09:12		tadzik	In my humble opinion yes, but as sorear says, language shouldn't dictate your working environment
09:12		moritz_	most developers work on linux, you might get help easier
09:13		moritz_	(at least in here)
09:13		ruiwk	thx very much ...i will install 1 in the virtualbox..
09:14		ruiwk	u all always in this chat room ?
09:14		tadzik	usually
09:15		ruiwk	i m from malaysia..nice to meet u all.
09:15		sorear	it's a pretty nice international group here
09:15		sorear	there's always someone here, time zones and all
09:15		tadzik	people from various timezones, always someone to talk to :)
09:15		* sorear	US
09:16		ruiwk	u all student ?
09:16		sorear	many of us are, but I think it's a coincidence
09:17		sorear	programmers tend to be young, and fond of learning
09:17		karb	to make a positional parameter optional $pos?, to make a named parameter mandatory :$named!  .. Is there a reason to use two different character (?, !) for similar markups
09:17		dalek	niecza: cd6e7c1 | sorear++ | / (2 files):
09:17		dalek	niecza: A couple more steps toward nrx
09:17		dalek	niecza: review: http://github.com/sorear/niecz[…]eee51a841df95d232
09:17		dalek	niecza: 76cd7d9 | sorear++ | lib/ (2 files):
09:17		dalek	niecza: [nrx] Kernel-level code for initial nrx
09:17		dalek	niecza: review: http://github.com/sorear/niecz[…]ec79af3583fd2f8a2
09:17		dalek	niecza: 568bfa6 | sorear++ | / (3 files):
09:17		dalek	niecza: [nrx] fix non-regex build aspects (TESTS NOW BROKEN)
09:17		dalek	niecza: review: http://github.com/sorear/niecz[…]8f658826647f4c4ed
09:17		sorear	TimToady++ is most definitely not a student...
09:19		sorear	karb: unambiguity; ! and ? always force a specific meaning
09:22		ruiwk	*sorear how long u been learning programming ?
09:23			M_o_C joined #perl6
09:23		ruiwk	& wat abt mac osx system ?
09:23			aloha joined #perl6
09:23			bacek joined #perl6
09:24		sorear	ruiwk: 12 years, give or take
09:24			Trashlord left #perl6
09:25		sorear	mac osx is used by a couple people here, definitely not as many as windows or linux though
09:25		* sorear	leaves for the night
09:26			Trashlord joined #perl6
09:26		ruiwk	k...i dowloading the ubuntu ...
09:27		ruiwk	thx for all yo information & opinions ya...
09:31		karb	sorear: ! stands for NOT. And ? stands for boolean context. With this viewpoint :$named! nots the optionality, but I dont seem to find $pos? intuitive. Maybe I'll get used to it :)
09:32		moritz_	karb: beware of context. Sometimes a question mark is just a question mark
09:33		moritz_	in particular your quoted meanings of ? and ! only apply when they stand in front of something
09:34		karb	got it. but would $pos! not be more consistent?
09:35		moritz_	if you understand it as human-language punctuation after a word, ? for optional and ! for mandatory make much more sense
09:37		ruiwk	moritz: hi
09:37		karb	i see it now
09:38			dual left #perl6
09:40		* moritz_	gone for breakfast
09:44			snearch left #perl6
09:53			cibs joined #perl6
10:02		tadzik	http://perl6advent.wordpress.com/2009/12/21/ why doesn't this code run properly in Rakudo?
10:03		tadzik	reminds me of my Config::INI grammar fsckup over some Rakudo update
10:06			gfx left #perl6
10:13		x3nU	ą
10:13		x3nU	;
10:13		x3nU	oops
10:17			karb left #perl6
10:21			envi^home left #perl6
10:23			dual joined #perl6
10:27			rgrau_ joined #perl6
10:27			Trashlord left #perl6
10:28			Trashlord joined #perl6
10:29		smash	mornin' everyone
10:29			M_o_C left #perl6
10:29			M_o_C joined #perl6
10:30		tadzik	hello
10:38			snearch joined #perl6
10:45			zulon joined #perl6
10:48			rgrau_ left #perl6
10:54		ruiwk	hi...tadzik
10:57		tadzik	hi
10:59			M_o_C left #perl6
11:03			wamba left #perl6
11:03		ruiwk	sorry..
11:04		ruiwk	how did u start to learn perl ?
11:05		tadzik	Perl 5 or Perl 6? :)
11:05		ruiwk	after done googleing...
11:06		ruiwk	state tat...perl 5 is different from perl 6...
11:06		tadzik	I see... no reason... for so many... dots
11:10		ruiwk	i just start learning...ha..my idol is geohotz...
11:11		ruiwk	i m so curious,,,why by doing some tweaks...the thing totally change..so interesting
11:11			drbean left #perl6
11:11			drbean joined #perl6
11:14			meppl joined #perl6
11:23			smash left #perl6
11:42			TiMBuS joined #perl6
11:44			masonkramer left #perl6
11:44			masonkramer joined #perl6
11:54			zulon left #perl6
12:00			Axius joined #perl6
12:00			araujo left #perl6
12:01			araujo joined #perl6
12:11			jhuni left #perl6
12:13			whiteknight joined #perl6
12:16			envi^home joined #perl6
12:21			Axius left #perl6
12:23			jimi_hendrix joined #perl6
12:29			isBEKaml joined #perl6
12:41		Juerd	I'm killing svn.pugscode.org
12:41		Juerd	Whoever wants to look at it to make it behave and then maintain it, can perhaps get it back.
12:41			ruiwk left #perl6
12:45			ruiwk joined #perl6
12:48		ruiwk	how to install rakudo in ubuntu ...
12:49		isBEKaml	Juerd: The pugs repo is now fully migrated to git? Must have been some 3 days.
12:49		Juerd	isBEKaml: Sorry, I don't know.
12:49		Juerd	isBEKaml: I do know that svn was taking feather1's load over 20 again.
12:50		ruiwk	nvm...
12:50		ruiwk	thx for reply
12:50		Juerd	ruiwk: :)
12:50		Juerd	ruiwk: http://rakudo.org/how-to-get-rakudo does explain how to build from source
12:51		isBEKaml	Juerd: yeah, I saw that a few days but never knew that svn took that much load on the server.
12:51		isBEKaml	s/days/& ago/
12:51		Juerd	isBEKaml: It's just horribly broken in some way, and nobody wants to figure out exactly how.
12:52		isBEKaml	Juerd: 20 is just pure agony for an svn admin server.
12:52		ruiwk	i had dowload it...
12:52		isBEKaml	Juerd: I'd have liked to help but I'm clueless about svn admin servers. Sorry. :/
12:53		ruiwk	try d but dunno how to built from source...i m newbie...sorry
12:53		Juerd	ruiwk: How far did you get?
12:53		isBEKaml	ruiwk: have you untarred the downloaded tar file?
12:54		ruiwk	ya..
12:54		ruiwk	and follow the readme..
12:54		isBEKaml	ruiwk: did you download just rakudo or the parrot sources too? I mean, did you download Rakudo Star?
12:55		ruiwk	i dowload from here
12:55		ruiwk	http://github.com/rakudo/star/downloads/
12:55			mberends joined #perl6
12:55		Juerd	Hoi mberends
12:55		ruiwk	the tar.gz file
12:56		mberends	hoi Juerd: I'm able to come to amsterdam.pm on Tuesday! :-)
12:56		Juerd	Yay!
12:57		Juerd	Hm, killing SVN and SVNWeb made feather1 absolutely idle
12:57		Juerd	It has 2 GB unused RAM now.
12:57		Juerd	(of 2.5)
13:00		isBEKaml	ruiwk: what do you have right now?
13:00		isBEKaml	ruiwk: you can just follow the README. that says everything you need to do to properly build your Rakudo Star distribution.
13:01		ruiwk	i had done the aptitude install
13:01		isBEKaml	ruiwk: Especially the section "Building Rakudo Star". Just type those instructions, you should be done.
13:01		ruiwk	i shoud cd to the file 1st ?
13:02		isBEKaml	ruiwk: You have done the aptitude install? Well, that's good. But you don't exactly need libicu-dev package. Just go the directory where you have untarred the Rakudo Star package (cd command)
13:02		isBEKaml	and type those commands in the README.
13:02		ruiwk	thx...i try again..
13:04		isBEKaml	ruiwk: You... are... welcome. (You don't need to type out those dots. You got sticky keys on your keyboard? ) :)
13:05		ruiwk	wow its running  alrd thx very much
13:05		ruiwk	ha sorry
13:05		isBEKaml	ruiwk: that's good. Welcome to perl6! and enjoy your time programming in it.
13:05		ruiwk	the dots is typing habbit
13:06		ruiwk	thx perl 6 the 1st programming lg i wanna learn after googleing
13:07		isBEKaml	ruiwk: Not a problem. Just try to shake out that habit. Some find it pretty annoying on a text only screen. ;)
13:07		ruiwk	ok
13:08		ruiwk	isBEKaml:thx
13:09		isBEKaml	ruiwk++ (Newbie initiate)
13:09			orafu left #perl6
13:10			orafu joined #perl6
13:11			karb joined #perl6
13:32			araujo left #perl6
13:33			araujo joined #perl6
13:36			bjarneh left #perl6
13:38			Axius joined #perl6
13:43			zby_home joined #perl6
13:54			risou joined #perl6
14:02			snearch left #perl6
14:09			karb left #perl6
14:18		isBEKaml	phenny: tell masak yapsi fails on recent version of rakudo though it works on older 2010.07 and .08 versions. -->bin/yapsi  -e 'say 42'
14:18		phenny	isBEKaml: I'll pass that on when masak is around.
14:18		isBEKaml	Could not find sub &Nil
14:19		tadzik	ruiwk: my friend has made a ppa for R* on Ubuntu, if you're interested
14:19			hans__ joined #perl6
14:20		tadzik	alright, who's the grammar wizard around?
14:20		ruiwk	sorry wats tat ?
14:20		moritz_	isBEKaml: that means there's an old .pir file somewhere
14:20		moritz_	tadzik: usually pmichaud. I'm more of an apprentice :-)
14:20		isBEKaml	moritz_: I built the new yapsi and removed any old .pir from PERL6LIB.
14:20		tadzik	moritz_: I finally decided to learn grammars and actions, and see a strange issue
14:20		moritz_	tadzik: do tell
14:21		moritz_	did you read the book section?
14:21		tadzik	http://pb.rbfh.de/rFFzXgaBMXVn
14:21		tadzik	not yet
14:21		moritz_	well, do it
14:21		tadzik	so, this is day 21 from the Perl Advent Calendar
14:21		moritz_	it's not that long, and we wrote it for a reason :-)
14:21		tadzik	hold on, take a look at this
14:22		* moritz_	takes a look
14:22		tadzik	now it works, for I added \n to a header token
14:22		tadzik	after removing...
14:22		isBEKaml	moritz_: the funny thing is, it works fine if I just change the shebang to point to earlier version of Rakudo.
14:22		isBEKaml	moritz_: the shebang in bin/yapsi
14:22		tadzik	it no longer parses
14:22		tadzik	and the Advent version does not work due to it (there is no \n there)
14:23		pmichaud	good morning, #perl6
14:23		tadzik	now the thing is Q {} has the newline at the beginning. But adding \n on the top of, well, TOP, does not make it work
14:23		tadzik	good morning pmichaud
14:23		moritz_	it also probably needs
14:23		tadzik	oh well, it does
14:23		moritz_	for $match<question>.flat
14:24		tadzik	so the advent version needs fixes probably
14:24		moritz_	tadzik: do you have a wordpress account?
14:24		Juerd	http://juerd.nl/i/89a1f74d761c[…]07a12dff9d91b.png  # The effect of killing svn on feather1
14:24		tadzik	moritz_: yep
14:24		Juerd	That was memory. Here's load avg: http://juerd.nl/i/6cda7e2f6f4e[…]313c5c4c37296.png
14:25		moritz_	wow
14:25		tadzik	yes, works after those changes
14:25		moritz_	tadzik: what's your nickname on wordpress?
14:25		Juerd	And the cleaning up of the gigabytes of tempfiles it left behind: http://juerd.nl/i/a19681e1e223[…]5482dfc9398b0.png
14:25		tadzik	moritz_: ttjjss I think. At least that's the blag url of mine
14:25		* tadzik	checks
14:26		moritz_	what's the last svn revision of the pugs repo?
14:27			Axius left #perl6
14:28		* isBEKaml	thinks Juerd is probably the most happiest guy here of all: "cleaning up of the gigabytes of tempfiles" :)
14:28		tadzik	moritz_: yes, it's ttjjss
14:29		Juerd	isBEKaml: I'm not happy about this thing at all
14:30		isBEKaml	Juerd: your words I quoted seemed to convey that impression. :/
14:31		Juerd	I feel bad about disabling a useful service, and am using these charts to explain why I did feel the need to end it.
14:31		moritz_	Juerd: can you re-enable it for a minute so that I can do a last git-svn rebase, and then disable again?
14:31		Juerd	moritz_: Sure
14:32		moritz_	then the git copy can preserve full information
14:32		Juerd	moritz_: Could you do it yourself btw? You have sudo access.
14:32		moritz_	Juerd: when you tell me what you did to disable it...
14:32		Juerd	moritz_: I removed a symlink
14:32		Juerd	moritz_: a2ensite svn.pugscode.org should enable it.
14:33		pmichaud	svn? feather?
14:33		moritz_	Util: and how do I disable it again?
14:33		* pmichaud	feels like he's missed an important story/event.
14:33		Juerd	pmichaud: Dead.
14:33		moritz_	sorry, meant Juerd
14:33		Juerd	moritz_: a2dissite
14:34		moritz_	Juerd: thank you
14:34			risou_ joined #perl6
14:34		Juerd	moritz_: Now that you know how to do it, you can do it again in case you need it
14:35		pmichaud	are we moving pugs to github, then?
14:35		moritz_	yes
14:35		moritz_	I'm on it
14:35		pmichaud	are you getting a copy with full history, I hope?
14:35		moritz_	yes
14:36		pmichaud	okay.  having done two migrations of svn to github, let me know if you need any help :-)
14:36		tadzik	moritz_: could you C-f Question::Answer on http://perl6advent.wordpress.com/2009/12/21/ ? It dies while parsing on the current Rakudo, and I wonder what's wrong with this one
14:36		tadzik	In "has" declaration, typename Question::Answer must be predeclared (or marked as declarative with :: prefix) at line 41, near " @.answers"
14:36			risou left #perl6
14:36		tadzik	↑ that's a == SORRY ==
14:36		moritz_	pmichaud: since I already have a git-svn mirror, I'll just push that one
14:37		pmichaud	moritz_: was that a full clone, or one that was created using the -r option?
14:37		moritz_	pmichaud: that was full
14:37		pmichaud	okay, excellent.
14:37		moritz_	tadzik: I guess typed array attributes are NYI
14:37		moritz_	in current master
14:38		moritz_	removing the type constraint should help
14:39		tadzik	now it can't find sub Question::Answer, as it couldn't reach inside the class
14:39		pmichaud	try changing   "class Answer { "    to    "class Question::Answer { "
14:39		pmichaud	I know that rakudo master has difficulty with nested class and package names
14:40			Ross left #perl6
14:41		hans__	Hello, how to pass a NULL password to postgresql with MiniDBI?
14:41			Trashlord left #perl6
14:41			Ross joined #perl6
14:43			whiteknight left #perl6
14:44		pmichaud	Juerd: I'
14:44		pmichaud	Juerd: I'm sure I speak for many in saying "Thank you" for having supported the pugs repository for all of these years.
14:44			Trashlord joined #perl6
14:44		pmichaud	(and much more in Perl 6, for that matter)
14:44		moritz_	hans__: I don't know if it's possible... you likely need to dig into the source code tofind out
14:45		moritz_	indeed, Juerd++ Juerd++ Jurd += (1..2e6).pick
14:46			kuzuha joined #perl6
14:47			karb joined #perl6
14:48		isBEKaml	yes, Juerd++ : (1...2e6) { .say Juerd++ }
14:54			zulon joined #perl6
14:58			justatheory joined #perl6
15:03			s_mosher joined #perl6
15:04			Ross left #perl6
15:06		ruiwk	why the $git clone git://github.com/rakudo/rakudo.git
15:07		ruiwk	is not running when i paste it on terminal and wait for so long ?
15:07			Ross joined #perl6
15:08			timbunce joined #perl6
15:09			karb left #perl6
15:09			sftp joined #perl6
15:15			zulon left #perl6
15:15			ruiwk left #perl6
15:16			azert0x joined #perl6
15:16		hans__	method connect(Str $user, Str $password, Str $params, $RaiseError) { ... }    -  $password shoud be a postgresql Null pasword.
15:17		tadzik	iirc some Zavolaj examples use some fancy pir:: to produce NULL
15:17		tadzik	you might want to look there
15:17		hans__	If you do not plan to use password authentication you can omit this option. If no password is specified, the password will be set to null    - from http://www.postgresql.org/docs[…]l-createrole.html
15:22			karb joined #perl6
15:22		tadzik	duh, I wrote this whole actions class and my $match.perl is still returning this ugly Match.perl result
15:23		tadzik	http://pb.rbfh.de/32mc3REtkl7SV -- am I doing something wrong?
15:23		pmichaud	tadzik: maybe you're wanting $match.ast.perl ?
15:23		pmichaud	matches always return the match object
15:24		tadzik	ah, ast
15:27		tadzik	Also stringifying the results before make()ing them helped a lot
15:31			karb left #perl6
15:31			Su-Shee left #perl6
15:36		tadzik	hmm, the action methods are called on answers, but the resulting objects seem not to be put into the final object. http://pb.rbfh.de/288lue1INCFod -- debug info on line 42, result starting from 68. Can anyone take a look?
15:42		moritz_	http://github.com/perl6/Mu/ is now live, and intended as the successor of the pugs repo.
15:42		moritz_	if anybody wants commit access, please ping me
15:42		pmichaud	tadzik: :answers => $<answer>».ast
15:43		pmichaud	that looks.... weird
15:43		pmichaud	did you really want the key to be a Pair?  ;-)
15:43		pmichaud	I suspect you want to omit the colon.
15:43		moritz_	I'll look into setting up hugme for the 'perl6' team
15:43		moritz_	which has access to the Mu repo
15:43			karb joined #perl6
15:45		pmichaud	or you can do   :answers($<answer>».ast)
15:45		tadzik	oh, the typos! How boring life will be without you!
15:45		tadzik	pmichaud: thank you very much
15:45			Trashlord left #perl6
15:46		tadzik	...and now I'm missing .pretty :)
15:46		isBEKaml	moritz_: I think we'll have to change that url in the Mu README about commitbit.pugscode.org.
15:46		tadzik	moritz_: here I am
15:47			Mowah left #perl6
15:47			Trashlord joined #perl6
15:47		tadzik	oh, I'm alredy added. moritz_++
15:52		pmichaud	Can we perhaps move the spectests out of Mu?
15:52		pmichaud	either that, or we need some way for "make spectest" to retrieve just the tests, and not the entire Mu repo.
15:53		moritz_	pmichaud: github supports svn checkouts
15:54		pmichaud	also, I'm not sure that calling the repo "Mu" is a good name -- might be too many inadvertent puns with the Mu type
15:54		* moritz_	finds it fitting
15:54		* isBEKaml	too
15:54		pmichaud	I think some may find it confusing.
15:54			Trashlord left #perl6
15:54		pmichaud	It's often confusing when one (unfamiliar) term is used to refer to two entirely different things.
15:55		pmichaud	anyway, I've noted my opinion.
15:55		moritz_	pmichaud: do you have a suggestion for a better name?
15:55		tadzik	. o O ( once properly understood, Grammars and Actions are the most awesome thing I saw since... well, since Perl 6 )
15:56		pmichaud	moritz_: well, "pugs" is still a valid name.  :)
15:56		pmichaud	but no, I don't have an immediate idea for a replacement.
15:56		moritz_	pmichaud: but even more confusing, since it doesn't contain pugs anymore
15:56		pmichaud	how do I checkout a subtree via github?
15:56		pmichaud	the only references I can find are for checking out the entire repo
15:57		isBEKaml	you can just do svn co lik eyou do the spectests.
15:57		tadzik	doesn't git have something like 'subrepo'?
15:57		pmichaud	currently for the spectests I do:    svn co https://svn.pugscode.org/pugs/t/spec
15:57		isBEKaml	yes, svn co http://github.com/perl6/mu/t/spec should work, I think
15:57		pmichaud	hmmm
15:57		* pmichaud	tries
15:58		pmichaud	pmichaudorange:~$ svn co http://svn.github.com/perl6/mu/t/spec
15:58		pmichaud	svn: Server sent unexpected return value (500 Internal Server Error) in response to OPTIONS request for 'http://svn.github.com/perl6/mu/t/spec'
15:58		pmichaud	oh, "Mu" works.
15:58		pmichaud	ugh, you mean we have to have a capital 'M'?
15:58		pmichaud	*sigh*
15:59		isBEKaml	:)
15:59		tadzik	moritz_: I can has few suggestions about the G&A chapter?
15:59		pmichaud	ECASESENSITIVITYSUCKSINURLS
15:59		isBEKaml	ash_++ for telling me about the github specific git-svn mirrors.
15:59		moritz_	tadzik: sure
15:59		* pmichaud	now officially dislikes "Mu" as a repo name.
15:59		pmichaud	at least make it "mu"
16:00		TimToady	that would be the native type :)
16:00		pmichaud	oh, checking out "t/spec" didn't work.
16:00		pmichaud	it gives me the full repo.
16:00		moritz_	drat
16:01		pmichaud	oh, it didn't give me the full repo.
16:01		pmichaud	it just didn't give me t/spec
16:01		pmichaud	that's... weird.
16:01		pmichaud	pmichaudorange:~/spec$ ls
16:01		pmichaud	AUTHORS  docs  INSTALL.v6  LICENSE  README  READTHEM  READTOO  SLAVES  VICTUALS
16:01		* moritz_	renamed s/Mu/mu/
16:01		tadzik	moritz_: personally, I find JSON example pretty overwhelming. The grammar source alone is 3 pages long, the actions is a bit shorter, but I'd personally like to start with something more visible, less abstract than JSON. I liked the Perl6Advent Questions-Answers example, I think it'd be nice to include it there after fixing, pushing JSON a few pages further as a Real-World example
16:02			Trashlord joined #perl6
16:04		moritz_	tadzik: I fear (because it's much work) that that's the way to go
16:04		tadzik	moritz_: I think I can handle this a bit
16:04		tadzik	I was to fix Advent Calendar anyway
16:04		moritz_	who wrote the advent thing? Tene?
16:04		tadzik	no eye deer
16:05		tadzik	Tene, yes
16:05		moritz_	we have to ask for permission first
16:06		moritz_	Tene: may we use your grammar examples from the advent calendar in the "Using Perl 6" book?
16:06		pmichaud	the other reason for wanting the spectests in a separate repo is because we'll also want easy write support
16:06		pmichaud	(and svn commits to github still look a little... experimental)
16:07		moritz_	agreed, we don't want svn commits when we move to git
16:12			karb left #perl6
16:13		moritz_	any idea how to extract the t/spec subtree into a separate repo, and still preserve history?
16:13		moritz_	I've asked on #git, but they seem busy with another discussion
16:16		mberends	hello moritz_, pmichaud: how about hosting just the t/spec subtree in a downsized subversion repo?
16:16		isBEKaml	moritz_: raise a support issue on support.github.com. They are usually closed pretty fast.
16:17		moritz_	mberends: and who maintains that?
16:18		tadzik	moritz_: asking on #github was always helpful for me
16:18		isBEKaml	#github too. forgot about that. :|
16:19		pmichaud	when I created the rakudo repo, I extracted it as a subtree of the original svn
16:20		pmichaud	as opposed to converting the entire svn to git, and then trying to extract it from there.
16:20		moritz_	git filter-branch --subdirectory-filter t/spec # seems to help... still running
16:21		pmichaud	for reference, to create the rakudo repository, I did
16:22		pmichaud	git svn clone -A parrot-names.txt http://svn.perl.org/parrot/trunk/languages/perl6
16:22		pmichaud	which just grabbed the languages/perl6 directory (and all of its history)
16:23		pmichaud	(and parrot-names.txt was used to map svn commit identifiers to git email identifiers)
16:24		pmichaud	mberends: (downsized subversion)  yes, that's a possibility.  In some sense it would be nice if we could get to a single vcs tool (more)
16:24			s_mosher left #perl6
16:24		pmichaud	and also I think the pugs repo may have grown to the point where we ought to have separate repos for its various components (docs, tests, sources, etc.)
16:24		mberends	moritz_: if there is interest in having a subversion t/spec repo, I think I (or someone) can arrange where it is hosted. The experience of Juerd++ can surely help us plan such a thing.
16:25		pmichaud	personally, I think I'd like to see t/spec as a git repo
16:25		pmichaud	I'm not sure I see any big advantage to keeping it as subversion
16:25			Trashlord left #perl6
16:25		moritz_	I know a lot of "it would be nice" things. To me what really matters is that we'll end up with one more usable repos, with full history preserved
16:26			hans_ joined #perl6
16:26		moritz_	I can offer that with git, and I can offer to extract subtrees into separate repos
16:26		mberends	pmichaud: +1 to separating pugscode into components. It's logical and fairly obvious.
16:26		pmichaud	well, it may be logical to separate them, but I'm not sure we yet know what the logical separation should be :)
16:27		moritz_	if people want something else, they should do it *and* do the planning themselves; I don't feel qualified to do that (no offense to anybody intended)
16:29		moritz_	how should I name the spectest repo?
16:29		moritz_	'spec' would be convenient, but confusing
16:30		TimToady	these things are usually called "validation suites"
16:31		moritz_	TimToady: do you have a github account?
16:31		TimToady	no
16:31		moritz_	TimToady: then you should create one
16:31		TimToady	doubtless :)
16:32		TimToady	my inertial navigation system tends toward inertia...
16:33		pmichaud	'spectest' works for me
16:33		TimToady	is there anything to identify it as a p6 spectest?
16:33		pmichaud	it's in the perl6 account?
16:33		TimToady	ah
16:33		pmichaud	github.com/perl6/spectest
16:33		pmichaud	could even be 'spectests'
16:34			snearch joined #perl6
16:34		pmichaud	anyway, others will have to decide it for a bit -- I have errands that must be run :(
16:34		pmichaud	bbiaw
16:34			ruoso joined #perl6
16:35		moritz_	pushing to perl6/spectes now - can be renamed later on
16:35		moritz_	*spectest
16:36		tadzik	bbl
16:36			tadzik left #perl6
16:37			Trashlord joined #perl6
16:37			s_mosher joined #perl6
16:38		TimToady	"vets: Valid E$_ Test Suite" for slurp("unixdict.txt").words.grep: *.subst(/^e/,'')
16:39		TimToady	and because it vets your implementation
16:40		moritz_	anyway, I need to go shopping for a bit, will probably be back in about an hour or so
16:40		moritz_	if people have suggestions for subrepository names, feel free to share them
16:42		moritz_	also you can think about having a separate repo for the specs (which would make the perl6-language mail hook easier)
16:43			risou_ left #perl6
16:43			jimi_hendrix left #perl6
16:46			ruoso left #perl6
16:50		TimToady	Valid Enough Test Suite :)
16:50		TimToady	Very Excellent Test Suite
16:51			molaf left #perl6
16:51		TimToady	or VI Enough
16:51			Trashlord left #perl6
16:53		TimToady	Valid Essentials
16:54		TimToady	Validate Everything
16:56			Axius joined #perl6
16:56		mberends	Velociraptor Extreme Torture System
16:58		TimToady	except we're the accelerraptor
16:58		TimToady	or however that's spelled...
17:00		TimToady	VI Extreme Torture Suite
17:00		TimToady	except sounds like we're testing an editor
17:00			snearch left #perl6
17:00			ruoso joined #perl6
17:01		TimToady	Six Under eXamination
17:01		mberends	lol
17:02		TimToady	Wall's Really Obnoxious Xanadu
17:02			svetlins__ joined #perl6
17:03		TimToady	Six Encounters X-rays
17:04			svetlins_ left #perl6
17:05		TimToady	Six Is eXpected
17:06			Trashlord joined #perl6
17:06		TimToady	Six is Xenophobic
17:06		mberends	Six Ain't eXperimental (anymore)
17:07		TimToady	Six Is Generally Hard
17:10		mberends	Six Outwardly Feels Tricky While Actually Reads Easily
17:10		TimToady	:)
17:11		TimToady	but doesn't sound like a torture suite
17:12		TimToady	Six Candidate Acceptance Test
17:12		Juerd	mberends: Oh, hosting is not the issue. Feather is still available for that, provided that it is cared for by someone
17:13		Juerd	mberends: It's just that svn + svnweb littered huge (100MB+) temp files, were responsible for extreme loads, had memory leaks, and were completely unmaintained
17:14			Trashlord left #perl6
17:14		TimToady	or Six Core Acceptance Test
17:14		mberends	Juerd: thanks. If pugscode were somehow locked down to be readonly, I wonder if many of the massive overheads would stop happening.
17:16		Juerd	mberends: Not really. The majority of the issues with svnweb are common for all mod_perl crap: things stay in memory forever, and is multiplied by the number of apache processes over time, regardless of the number of users.
17:16		hans_	\
17:16		hans_	Z
17:16		TimToady	Really Obnoxious Torture System
17:16		Juerd	mberends: svn was mostly CPU hungry, and with very little commits I doubt it had anything to do with commits.
17:18		Juerd	mberends: As for my experience with setting svn up: none, whatsoever. I don't know who originally set up svn on feather.
17:18		TimToady	Rarely Obtained Feature List
17:19		isBEKaml	Light On LoadTESTs
17:20			hans_ left #perl6
17:21		isBEKaml	Six Ix Xanthippe
17:21		isBEKaml	:D
17:21		Juerd	http://juerd.nl/i/73ea6e3483eb[…]42df8ddab5d30.png  # It appears that svn was also responsible for the majority of traffic, with its gigantic diffs.
17:21		isBEKaml	Six Is Xanthippe
17:23		isBEKaml	I think Xanthippe was the name of Socrates' wife. Generally perceived evil figure in history, but also was partly responsible for driving Socrates towards philosophy! lol
17:24			Trashlord joined #perl6
17:24		TimToady	Sucks Or Rocks Reliably Yet
17:26		isBEKaml	Times Reliably On Load Lines
17:28		moritz_	Six Or Not Six, that's what the spectests answer
17:29		moritz_	http://github.com/perl6/spectests now contains the spectests; I haven't removed them from mu yet
17:29			Trashlord left #perl6
17:30		moritz_	I should do that now.
17:30			zulon joined #perl6
17:33		TimToady	Sux Or Rox Test Of Functionality
17:33			isBEKaml left #perl6
17:34		TimToady	I think I like sorry the best so far
17:35			tylercurtis joined #perl6
17:37		TimToady	Larry's Obnoxious Language Core Acceptance Test
17:38		moritz_	"you think this is cute today"
17:38		TimToady	that is wonderfully ambiguous, in the way of noun modifiers
17:38		TimToady	I bet jnthn++ would like it
17:39		moritz_	lol
17:50		dalek	rakudo: 29fcf8e | chromatic++ | src/ (4 files):
17:50		dalek	rakudo: [src] Updated deprecated string_from_literal calls.
17:50		dalek	rakudo:
17:50		dalek	rakudo: Parrot's (not all that) new approach is to use Parrot_str_new().
17:50		dalek	rakudo: review: http://github.com/rakudo/rakud[…]a3ab7204106c2c2a2
17:50		dalek	rakudo: 6b1b030 | chromatic++ | src/binder/bind.c:
17:50		dalek	rakudo: [binder] Hoisted several constant Parrot STRINGs.
17:50		dalek	rakudo: review: http://github.com/rakudo/rakud[…]227da5191d695ca4c
17:50		dalek	rakudo: e36452e | chromatic++ | src/ops/perl6.ops:
17:50		dalek	rakudo: [ops] Hoisted constant Parrot STRINGs.
17:50		dalek	rakudo: review: http://github.com/rakudo/rakud[…]ba0a93f159f7c21e2
17:51		ingy	o/
17:52		TimToady	trouble is, we want a word that doesn't occur too frequently in normal talk
17:53		TimToady	so maybe something more like "visor", Verify Implementation Sucks Or Rocks
17:53			molaf joined #perl6
17:53		TimToady	but the word itself isn't very funny
17:53		moritz_	otoh if we split off several repositories, the names should be as obvious as possible
17:53		TimToady	censor or sensor would be funnier
17:54			Axius left #perl6
17:55		moritz_	should I split off docs/ ?
17:55			tadzik joined #perl6
17:55			molaf left #perl6
17:57			Ross left #perl6
17:57			Trashlord joined #perl6
17:58			Ross joined #perl6
17:58			karb joined #perl6
17:59			zulon left #perl6
18:00		TimToady	valid: Verify All Language Implemenation Details
18:00		TimToady	*ment
18:00			rhebus joined #perl6
18:00		rhebus	Hello everyone
18:00		TimToady	everyone says hello back
18:00		rhebus	I've decided to jump in and learn Perl 6 using Rakudo* by doing the project euler problems
18:01		TimToady	cool
18:01		rhebus	but my naive implementation of problem 3 seems to be doing some int overflow
18:01		rhebus	http://projecteuler.net/index.[…]ion=problems&id=3
18:01		rhebus	(what's the largest prime factor of 600851475143)
18:02		rhebus	i will ask a question, just give me a minute :)
18:02		TimToady	hmm, well, rakudo doesn't implement Int as bigint yet, but overflows to Num
18:02		TimToady	but that should fit in a double
18:02		TimToady	and %% is defined on doubles, methinks
18:03		TimToady	rakudo: say 600851475143 %% 7
18:03		rhebus	what is %%?
18:03		p6eval	rakudo 52f9ea: OUTPUT«0␤»
18:03		TimToady	same as % == 0
18:03		TimToady	now that we have it, we find we use it all the time
18:03		rhebus	rakudo: say 6 %% 2
18:03		p6eval	rakudo 52f9ea: OUTPUT«1␤»
18:04			envi^home left #perl6
18:04		rhebus	rakudo: say 600851475143 %% 71
18:04		p6eval	rakudo 52f9ea: OUTPUT«1␤»
18:04			rindolf joined #perl6
18:04		rindolf	Hi all.
18:05		rhebus	(minor spoilers for that problem, sorry :) )
18:05		rhebus	but yeah, that looks good enough for me, thanks TimToady
18:05		rindolf	Some of the "make spectest" tests failed on my Mandriva 2010.1 x86-64 Linux laptop.
18:05		TimToady	we're more interested in the solution than the solution, as it were
18:06		TimToady	the interesting part is what the P6 code comes out looking like
18:06		TimToady	and whether it's intuitive to the reader
18:08		rhebus	well I like it so far
18:08		rhebus	but I haven't yet delved into hyper operators &c
18:08		TimToady	we hope those end up intuitive too
18:08		rhebus	tbh I'm just writing P5 in P6 atm
18:09		rhebus	"baby perl 6"
18:09		TimToady	that's officially okay
18:09		rhebus	can i have a certificate from the authorities to state that officially?
18:10		TimToady	if you're looking for someone with a little authority around here, I have about as little as anyone...
18:10		rhebus	:)
18:12			karb left #perl6
18:12		rhebus	the irony is that I'm running win64 so I *should* be able to represent 600851475143 natively. But I'm lazy and downloaded the .msi which I imagine is 32-bit.
18:12		rhebus	hooray for laziness
18:13		ingy	what does p6 call a lexical namespace?
18:13		tylercurtis	Well, Rakudo (last I looked at the promotion to Num code) converts anything that would overflow a 32-bit integer to Num, regardless of the size of your integers.
18:14		TimToady	ingy: lexpad or lexinfo, depending
18:14		TimToady	the pad is usually an active one for a particular call
18:15		TimToady	but sometimes it's used generically for the static table too
18:15			Axeman left #perl6
18:15		ingy	I'm considering namespace, scope or stash.
18:15		TimToady	but we're trying to switch over to calling those lexinfos
18:16		TimToady	well, all of those are rather overloaded
18:17		ingy	my meaning is a symbol hash that belongs to a function
18:17		ingy	a function object has a namespace object (just a hash)
18:18			toebu left #perl6
18:18		TimToady	yes, but "namespace" == "package" in parrotland currently
18:18		rindolf	Hi all. I'm getting svn: Server sent unexpected return value (403 Forbidden) in response to OPTIONS request for 'http://svn.pugscode.org/pugs/t/spec'
18:18		TimToady	and "stash" == "package" in P5land
18:18		ingy	does anyone else use lexpad?
18:19		ingy	I really wish naming was more standard...
18:19		TimToady	that's a p6ism, but historically didn't distinguish the name hash from the activation record it represents
18:19		ingy	it's a really annoying hurdle in acmeist life.
18:19		TimToady	actually "pad" is a p5ism for the whole structure, which has both elements
18:20		TimToady	well sometimes the common denominator is the empty set, and you have to just coin something new, like lexhash
18:21		ingy	nod
18:23		TimToady	though that makes the other into something like pkghash, which is ugly
18:23		TimToady	lstash and pstash would be how the old-timers woulda done it, which is how we got strange words like lvalue
18:24		rhebus	i particularly like the C standard which distinguishes between lvalues and modifiable lvalues >_<
18:24		TimToady	orz
18:25		ingy	well I'm sticking with namespace for now. Once you start following TimToady's lead it's a slippery slope :)
18:25		TimToady	I want the world, you can have the rest.  :)
18:26		ingy	rest is probably what I need most.
18:27		TimToady	a rest for ingy++ --> 𝄽
18:28		TimToady	at least it shows up in the log...
18:28		TimToady	assuming you're using a browser that does unicode for reals
18:30		rhebus	grr, project euler seems to be more about maths than programming :/
18:30		rhebus	or at least the first 4 problems are
18:30		TimToady	yes, that's they're stated purpose
18:31		TimToady	maybe you want to tackle rosettacode problems, though most of the low-hanging-fruit has solutions already
18:33		rhebus	ooh, hadn't seen that
18:33		TimToady	see also http://www.oreillynet.com/onla[…]ms_in_perl_6.html
18:34		TimToady	which have some proposed solutions in the test suite that you can check against
18:34		TimToady	or replace with better solutions :)
18:34		TimToady	the originals were written against pugs rather than rakudo
18:34		rhebus	ooh that is also good
18:36		sorear	good * #perl6
18:36		TimToady	o/
18:36		pmichaud	back again, for a bit
18:36		rhebus	aha they're also in http://github.com/perl6/perl6-examples
18:36		rindolf	pmichaud: hi.
18:37		TimToady	also in t/spec/integration, or wherever it ends up after this morning's svn meltdown
18:38		rindolf	TimToady: svn meltdown?
18:38		rindolf	TimToady: WDYM?
18:39		moritz_	http://github.com/perl6/specte[…]ster/integration/
18:40		pmichaud	(svn meltdown)  Juerd needed to disable svn on feather earlier today
18:40		pmichaud	which means svn.pugscode.org is no longer active
18:40		rindolf	pmichaud: ah.
18:41		Juerd	Is the github repository the main one now?
18:41		pmichaud	likely yes
18:41		moritz_	yes
18:42		pmichaud	we'll still have some refactorings to make, I think
18:42		moritz_	I've already split off the spectests
18:42		rindolf	moritz_: well, you need to update the rakudo build-system to reference it.
18:42		pmichaud	should we add TimToady as an owner of the perl6 organization?
18:42		pmichaud	rindolf: s/you/we/  :-)
18:42		rindolf	pmichaud: :-)
18:43		moritz_	I already have that locally; testing as we speak
18:43		moritz_	pmichaud: +1
18:43		pmichaud	TimToady: do you have a github account yet?
18:44		TimToady	working on it
18:46		moritz_	pmichaud, TimToady: would you prefer a separate repo for docs/ (or maybe docs/Perl6/Spec) ?
18:47		pmichaud	I think a separate repo makes sense... but I defer to TimToady on that.  He's the one that generally has to deal with it the most.
18:47		pmichaud	same goes for the other non-spectest parts of the pugs repo :)
18:48		TimToady	it says the account TimToady is already taken
18:48		pmichaud	yes, I saw that
18:49		pmichaud	said it was created in dec 2009
18:49		TimToady	did someone create it for me, I wonder...
18:49		TimToady	did I create it and not know it?
18:49		pmichaud	yeah, not sure if it's a squatter or what
18:49		dalek	rakudo: ff5b4c0 | moritz++ | / (2 files):
18:49		dalek	rakudo: partially update build system to deal with spectests in git
18:49		dalek	rakudo: review: http://github.com/rakudo/rakud[…]558b9d113ce9a4ab6
18:49		pmichaud	maybe request password recovery for the account and see what happens?
18:50		moritz_	missing: "make manifest" still uses the old repo
18:51		pmichaud	moritz_: I think I can correct that.
18:51		lue	wello there o/
18:53		sorear	Ay reason we don't just delete the 'perlsix' account?  It has zero repositories
18:54		pmichaud	maybe keep it so that someone else doesn't create one for nefarious purposes
18:55		moritz_	sorear: currently my only reason was "more important other things to do first"
18:55			plobsing joined #perl6
18:56		* sorear	seems to have the delete privelege on perlsix
18:56		pmichaud	sorear: yes, you were listed as an organization owner
18:56		pmichaud	(along with about seven others :)
19:00		TimToady	successful at resetting TimToady password, so I guess you can call me that; now for ssh token...
19:01		pmichaud	TimToady: added you as an organization owner for the perl6 account
19:01		pmichaud	"Owners have full access to all repositories and have admin rights to the organization."
19:02		sorear	you now have the ability to delete perl6.
19:02		sorear	use it wisely.
19:02		* sorear	blows away a 193MB pugs checkout
19:04		* TimToady	feels the power going to his head, but as his head is too big, it gets lost somewhere.
19:04		pmichaud	the current link for STD.pm6 is http://github.com/perl6/mu/raw[…]src/perl6/STD.pm6
19:04		pmichaud	or http://github.com/perl6/mu/blo[…]src/perl6/STD.pm6  if you want it with creamy github formatted goodness
19:04		TimToady	blob?
19:05		pmichaud	I don't create the urls -- that's all github.  :)
19:05		moritz_	Binary Large OBjection
19:05		lue	did pugs disappear permenantly, or just temporarily ?
19:05		pmichaud	.oO( OBject*ion* ? )
19:05		moritz_	pmichaud: :-)
19:05		lue	.oO(blob = Big Lump O' Binary)
19:05		sorear	github doesn't seem very good at formatting Perl6
19:06		sorear	lue: it didn't disappear, it moved to github
19:06		pmichaud	sorear: I'm sure they would accept patches :)
19:09		* diakopte1	backlogs a bit
19:09		sorear	seen diakopter
19:09		aloha	Sorry, I haven't seen diakopter.
19:11		lue	time to get rid of the pugs svn repo (and with it an unused commit bit)
19:12		sorear	What is docs/ in mu.git being used for?
19:12		sorear	moritz_, ping
19:12		moritz_	sorear: I'm about to delete it
19:12		pmichaud	(keep in mind we're still in the middle of a largely unplanned migration)
19:12		diakopte1	Juerd: I suspect you were wrong about "svn" causing the high load/mem usage.
19:13		pmichaud	oh, all of 'docs' became its own repo?
19:13		pmichaud	that feels.... wrongish
19:13		pmichaud	I would've wanted to have just the spec docs separated out
19:13		diakopte1	Juerd: rather, I suspect it was (as sorear mentioned the other day) the SVN::Web links continually being spidered by all the search engines.
19:14		moritz_	pmichaud: that can be done too
19:14		pmichaud	in general, I think docs should live with whatever it is they're documenting
19:14		pmichaud	indeed, perhaps what we really want is a perl6/spec  repository, that contains docs, tests, and STD.pm
19:15		diakopte1	Juerd: otoh, *finally* this is prompting what I was strongly suggesting 1 and 2 years ago
19:15		tadzik	hmm, I added <![ ]> at the end of the regex, but the matched string still has trailing spaces. Any ideas?
19:15			Trashlord left #perl6
19:15		diakopte1	(the very thing pmichaud just mentioned)
19:15		sorear	tadzik: use \s
19:15		tadzik	sorear: I used \h, but it didn't work eithger
19:15		tadzik	* either
19:16		sorear	and it's spelled <!before \h>
19:16		moritz_	backslash escapes don't work in char classes yet :()
19:16		tadzik	actually it is mentioned <![\h+]>
19:16		tadzik	oh
19:16		tadzik	sorear: so that would be <!before \h+> in my case?
19:16		pmichaud	tadzik: nopaste or reference to what you're trying to do?
19:16		sorear	<![\h+]> means "match if the next character is not \\, h, or +
19:16		sorear	"
19:17		pmichaud	sorear:  by spec, or by what rakudo currently has?
19:17		tadzik	well, it wasn't working for the plain space
19:17		sorear	pmichaud: by spec it
19:17		tadzik	oh wait, this is token, backtrackint is disabled. Bad, isn't it?
19:17		sorear	's \h +
19:17		sorear	tadzik: spaces in a token are ignored anyway
19:17		tadzik	nah, nothing new after changing it to regex
19:18		pmichaud	tadzik: nopaste url?
19:18		tadzik	well, I can just trim it in actions, but..
19:18		tadzik	http://pb.rbfh.de/2r4JjtYeSxCgh
19:19		lue	rakudo: say "hello   " ~~ /\H+/
19:19		p6eval	rakudo e36452: OUTPUT«hello␤»
19:19		tadzik	oh by the way, about line 16 -- if I change the last \h* to \h+ it dies with "Couldn't match the final ']'", although the case is not about the ], is this intentional?
19:20		pmichaud	if you change the last \h* to \h+, that rule will never match.
19:20		tadzik	I know, but it dies with a message about not finding the final ]
19:20		pmichaud	right, because of the ~ that occurs earlier
19:21		pmichaud	hmmmm
19:22		pmichaud	why not write it normally, in this case?
19:22		tadzik	normally?
19:22		pmichaud	^^ \h* '[' $<text>=[\w|\h]+ ']' ...
19:22		pmichaud	i.e., why use the ~ ?
19:22		tadzik	I just wanted to try it :)
19:22		tadzik	but it's not the issue, when I use proper \h* it works fine
19:23		pmichaud	there are some definite oddities with the failure mode for that operator
19:23		tadzik	now what I want to achieve is 'autotriming' keys via tuning regexes
19:23		tadzik	so it won't catch the trailing \h
19:23		pmichaud	trailing after the ']', you mean?
19:23		tadzik	no, in key
19:24		pmichaud	I don't understand "autotrimming" in this case
19:24		tadzik	so when I have keyval 'foo = bar' it is parsed to "foo " => "bar"
19:24		tadzik	foo has a trailing whitespace
19:24		tadzik	see the bottom of the paste
19:24		pmichaud	oh, your problem is that <-[;=]>+  also grabs whitespace
19:24		tadzik	yes
19:24		pmichaud	you can't fix that with anything that comes afterwards
19:24		tadzik	so I wanted to add <![\h+]> at the end
19:25		pmichaud	well, you can, if it's a regex
19:25		tadzik	that's why I changed it
19:25		pmichaud	what you really want is <!after>
19:25		pmichaud	<!before>  talks about what comes next, not about what you already scanned.
19:25		tadzik	<!after \h+>, yes?
19:25		pmichaud	yes, but !after is nyi in rakudo, iirc.
19:25		tadzik	I see. I'll have to read the synopsis on this one
19:25		TimToady	the + is useless
19:25		tadzik	oh
19:25		pmichaud	and yes, the + is useless.
19:26		tadzik	even better
19:26		tadzik	so what can I do in the current Rakudo?
19:26		rhebus	hmm I disagree with the interpretation of perl6-examples/99-problems/P02-scottp.pl
19:26			rindolf left #perl6
19:27		rhebus	The second-last box of <a b c d e> is <d e> not 'd'
19:27			rindolf joined #perl6
19:27		moritz_	then fix it :-)
19:27		rhebus	will do
19:27		moritz_	rhebus: do you have a github account? if yes, what's the nickname?
19:27		pmichaud	tadzik: [ \H && <-[;=]> ] ** \h+   # maybe
19:27		rhebus	I'll get one set up
19:27		tadzik	hmm
19:28		tadzik	I won't be able to read it after a few days I'm afraid. I'll go for .trim in actions, til Rakudo implements <!after>
19:28		Juerd	diakopte1: svnweb didn't run on svn.pugscode.org, and http://feather.perl6.nl/server-status did show that the processes that appeared high in top(1) were often on that hostname
19:28			Trashlord joined #perl6
19:28		rhebus	what's the github way of doing things? Do I fork it, commit to my copy, and send you the url?
19:29		Juerd	diakopte1: It's possible, of course, that these svn requests were done *by* svnweb, which at those moments apparently did behave.
19:29			smash joined #perl6
19:29		tadzik	thanks for help
19:29		moritz_	rhebus: the #perl6 way is that I give you commit access
19:29		pmichaud	rhebus: for perl6-examples?  you get a commitbit.  :)
19:29		smash	hello everyone
19:29		moritz_	pmichaud: I've added perl6-examples to the 'perl6' team
19:30		Juerd	diakopte1: Still, though, nobody volunteered to investigate or maintain it...
19:30		rhebus	my username is rhebus on github now
19:30		moritz_	rhebus: and you have commit access now
19:30		pmichaud	moritz_: +1
19:31		pmichaud	moritz_: so, "perl6" is the team for generic "we give you commit access to all of the common repos" ?
19:31		Juerd	moritz++: Thanks for moving everything to github and taking care of it, by the way :)
19:31		moritz_	pmichaud: yes
19:31		pmichaud	moritz_: excellent
19:31		smash	phenny: tell colomon http://gil.di.uminho.pt/users/[…]rakudo-bench.html updated with new test scripts
19:31		phenny	smash: I'll pass that on when colomon is around.
19:32		pmichaud	I still prefer the zero-based graphs.
19:33		tylercurtis	Can I (ekiru on github) have a perl6{*} commit bit?
19:33		moritz_	tylercurtis: sure, just a sec
19:33			Eevee joined #perl6
19:33		moritz_	tylercurtis: done
19:34		tylercurtis	moritz_++ thanks
19:34		tadzik	smash: why there are faster things in Slower Scripts?
19:34			timbunce left #perl6
19:34		tadzik	smash: graph shows as if they were faster than before
19:35		tadzik	can I override the \n in grammars?
19:35		smash	pmichaud: refresh
19:35		pmichaud	smash: much better, thanks
19:36		tadzik	smash: ah, slower means "longer running", I get it
19:36		pmichaud	now we need to re-do the iteration and recursion bench marks so they actually measure something meaningful :)
19:36		smash	tadzik: maybe the title isn't the best, that just meanos that on the right are the scripts that take longer to run
19:36		smash	tadzik: right
19:36		tylercurtis	And we need to figure out why pick-words.pl slowed down. :)
19:36		tadzik	is there a way to override the \n, as in token ws {} ?
19:36		moritz_	tadzik: not yet
19:37		tadzik	:(
19:37		tadzik	pmichaud: btw, what's the state of require?
19:37		pmichaud	tadzik: still nyi
19:37		* moritz_	has tried it, but couldn't get 'require <foo bar>;' to parse
19:38		smash	tadzik: i changed it to Fast/Slow scripts, maybe this way is more clear
19:38		pmichaud	smash: if I change the benchmark scripts for iteration and recursion, will that update all of the values in this graph?
19:39		smash	pmichaud: with the current script yes
19:39		smash	pmichaud: s/script/harness for the benchmarks/
19:40		tadzik	smash: how about something like lighter/heavier? That's a suprisingly difficult case I think :)
19:40		tadzik	it's hard not to disaumbiguate
19:40		tadzik	pmichaud: is there a chance for it in the near future?
19:40		tadzik	I know one how wishes to make testing and stuff :)
19:41		smash	tadzik: i like light/heavy scripts, thks for the tip
19:41			timbunce joined #perl6
19:41		rhebus	should I rename P02-scottp.pl to my name or leave it as is?
19:41		moritz_	rhebus: whatever you consider better
19:41		moritz_	or more convenient
19:42		rhebus	ok
19:42		rhebus	i think the first example is awkward and not the best example :/
19:42		tylercurtis	smash: does your harness manually distinguish between the light/heavy(or fast/slow or short-running/long-running or whatever) scripts? If so, could you perhaps move that information into the bench-scripts repo? That way, it's easier for new long scripts to be added.
19:42		tylercurtis	smash: ignore that if you automatically detect long-running scripts.
19:43		tadzik	moritz_: JSON example in the book becomes extremely helpful once I got comfortable with Actions
19:43		moritz_	tadzik: glad to hear
19:43		rhebus	last <A B C D> should be the same as <D> in P01, based on the original LISP interpretation of returning a 1-element list rather than a single element
19:44		rhebus	however in perl6 generally a 1-element sequence is not the right thing
19:44		lue	how do you modify a class's variables? Something like    class A { has $.b }; my A $c .= new(3); $c.b = 5    .
19:44		rhebus	maybe I should just leave it alone :/
19:44		smash	tylercurtis: when building the grphs it automatically chooses if the script goes to one grpah or another
19:44		tadzik	lue: they have to 'is rw'
19:44		tadzik	lue: so they can be lvalues
19:44		smash	the rest is exactly the same
19:44		tylercurtis	smash: good. :) That's even better.
19:45		tadzik	lue: or you can modify them from inside the class, with the private accessor ($!b = 'asd')
19:45		tylercurtis	rhebus: well, in Perl 6, generally, a single element is also treatable as a 1-element sequence.
19:45		lue	I was just thinking, it'd be better to not do what I wanted :)
19:45		moritz_	http://nopaste.snit.ch/23233 # I plan to send this to perl6-announce - any comments before I send it?
19:45		tylercurtis	rakudo: say 5.[0];
19:45		p6eval	rakudo e36452: OUTPUT«5␤»
19:45		smash	pmichaud: feel free to change the benchmark scripts, i'll re-run them and update the HTML page
19:45		pmichaud	smash: changed and pushed.
19:46		tadzik	moritz_: thereforE
19:46		tylercurtis	rakudo: say 5.map(* + 2); # curious
19:46		p6eval	rakudo e36452: OUTPUT«7␤»
19:46		rhebus	rakudo: say <5>.WHAT
19:46		p6eval	rakudo e36452: OUTPUT«Str()␤»
19:46		lue	(I can't seem to git clone the perl6/docs repository.)
19:46		rhebus	rakudo: say <5>[0].WHAT
19:46		pmichaud	moritz_: mention that people can get commit bits by asking on #perl6, also, instead of just email.
19:46		p6eval	rakudo e36452: OUTPUT«Str()␤»
19:47		tadzik	moritz_: to have commit -> to have commit access?
19:47		rhebus	rakudo: say <5 6>.WHAT
19:47		pmichaud	we can add more meta-committers
19:47		p6eval	rakudo e36452: OUTPUT«Parcel()␤»
19:47		moritz_	lue: I've deleted it again, after pmichaud convinced me it was a bad idea
19:47		rhebus	tylercurtis: it seems there's no distinction at all. curious
19:47		smash	pmichaud: pull'ed, running the benchmark now, i'll update the html page as soon as it finishes
19:47		moritz_	lue: docs/ is now still in the 'mu' repo, only the specs were extracted into a separate repo
19:48		pmichaud	tylercurtis: scalars are considered to be lists of one element if asked to act that way
19:49		pmichaud	thus  'abc'.join(' ')  becomes 'abc'
19:50		tylercurtis	pmichaud: right.
19:51		pmichaud	tadzik: I don't know when I (or someone else) will get around to 'require' again.
19:51			Guest78701 left #perl6
19:52		pmichaud	moritz_: message needs to go to p6l and p6c, also.  and probably p6u.
19:52		moritz_	pmichaud: ifyousayso
19:53		pmichaud	I don't see any other message edits for now.
19:57		tadzik	yay, Config::INI fixed
19:57		moritz_	ok, sending now
19:59		lue	there's no such thing as =head0 in Pod, right?
19:59			hans__ left #perl6
19:59		moritz_	that only exists in PseudoPod
20:00		moritz_	(git filter-branch)++
20:01		smash	moritz_: i add a commit bit on pugs, can i have one for the new ones ?
20:01			bartolin joined #perl6
20:01		TimToady	note that <A> is just ('A')
20:02		pmichaud	smash: github id?
20:02		moritz_	smash: you're also 'smash' on github?
20:02		TimToady	smashz I think
20:02		pmichaud	smash: added (as 'smashz')
20:02		pmichaud	(I already had the page open from an earlier session :)
20:02		smash	pmichaud, moritz_: my id on github s smashz, thank you
20:03		moritz_	fwiw I'm now moving perl6.org to a separate repo now too
20:03		smash	moritz_++ # pugs repo revamp
20:03		bartolin	hi #perl6
20:03		pmichaud	good idea
20:03		lue	could I get a commit bit as well? github id is lue
20:03		pmichaud	lue: added
20:03		moritz_	lue: sure
20:03		tadzik	bartolin: hi
20:03		moritz_	done
20:04		pmichaud	lue: interesting gravatar image :)
20:04			pjcj left #perl6
20:05		pmichaud	that's a lot of 42s
20:07		lue	42 42's to be exact :)
20:07		smash	pugs, reminds me haskell, which reminds me the damn transpose function.. hmmm
20:07		pmichaud	that's a bit square.  :)
20:08		* pmichaud	chuckles at the phrase "bit square" :)
20:08		tadzik	mind me adding Config::INI back to R*? I got it all fixed
20:08		pmichaud	tadzik: +1
20:09		pmichaud	afk, shopping
20:09			svetlins__ left #perl6
20:10		bartolin	rakudo:  class Dog { has Int $.age }; my $dog = Dog.new( age => 'zzz' ); say $dog.age.WHAT;
20:10		p6eval	rakudo e36452: OUTPUT«Str()␤»
20:11		tadzik	hah
20:11		bartolin	a newbie's question: why doesn't "has Int $.age" prevent setting age to 'zzz'?
20:11		bartolin	is this not yet implemented?
20:11		tylercurtis	bartolin: rakudobug.
20:11		tylercurtis	It seems like I recall it having worked properly at one time, but I may be incorrect.
20:11		bartolin	tylercurtis: a known one, probably?
20:12		* tylercurtis	looks.
20:13			jaldhar_ joined #perl6
20:13		moritz_	I'm pretty sure it's known and submitted to RT
20:13		moritz_	we have failing tests for it
20:14		bartolin	moritz_: ahh, thanks (tylercurtis++ too)
20:14			bluescreen joined #perl6
20:14			rindolf left #perl6
20:15			bluescreen is now known as Guest58080
20:16		rhebus	hmm, perl6-examples/99-problems/P08-viklund.pl fails for me with "Could not find sub &infix:<compress>"
20:16			_jaldhar_ joined #perl6
20:16			_jaldhar left #perl6
20:16		rhebus	which is a shame because it looks very impressive to a perl6 newbie
20:17		tadzik	x3nU was to fix perl6-examples, wasn't he? :P
20:19			jaldhar_ left #perl6
20:19			jaldhar joined #perl6
20:22		x3nU	tadzik: i can't do that on my own, also i was bit busy recently ;d
20:23			_jaldhar_ left #perl6
20:25			tadzik left #perl6
20:26			svetlins__ joined #perl6
20:27		TimToady	why is perl6.org labeled JavaScript?
20:27		moritz_	because it contains some .js files?
20:28		Tene	moritz, tadzik: you can use anything I've ever published online for anything you want, without restrictions.
20:28		Tene	moritz_: oops, I forgot your _
20:28		moritz_	Tene++
20:29		* dukeleto	is very excited to see specs and spectests repos on github!
20:29		* moritz_	is very stressed
20:30		* TimToady	votes moritz_++ as one of the Ten Best-Stressed Men
20:31		moritz_	TimToady: do you want src/perl6/ also moved? if yes, what repo name? 'grammar' maybe?
20:31		* lue	thinks    make stresstest    deserves #2
20:32			tadzik joined #perl6
20:32		tylercurtis	moritz_++ # For endangering his mental health for the rest of us. Try not to go crazy, though.
20:32		tadzik	Tene++
20:32		moritz_	I'll got to bed very soon :-)
20:32		tadzik	seen masak
20:32		aloha	masak was last seen in #perl6 13 hours 57 mins ago saying "but I'd much rather employ something like JSON::XS, like sorear said.".
20:33		TimToady	druther have something like std
20:33		tadzik	rhebus: iirc perl6-examples were written for the alpha branch of Rakudo, and most of them probably don't work now
20:34		rhebus	I'll see what i can do to fix that
20:34		rhebus	but I don't know perl 6 yet
20:34			BillN1VUX joined #perl6
20:35		moritz_	TimToady: will do
20:35		tadzik	is there any working template engine?
20:35		moritz_	TimToady: but probably won't get around to it tonight
20:35		tadzik	html I mean
20:35		moritz_	tadzik: I think szabgab++ fixed HTML-Template in the masak/web repo
20:37		TimToady	wow, I haz 1 followers!!!
20:37		lue	how do I put restrictions on a parameter? Things like   sub a($a when >=0) # although that doesn't seem to work
20:37		moritz_	sub a($x where { $_ >= 0 }) { ... }
20:38		lue	aaah. thankyou
20:38		TimToady	rakudo: sub a($x where * >= 0) {...} # does this parse yet?
20:38		p6eval	rakudo e36452: OUTPUT«===SORRY!===␤Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 22␤»
20:38		TimToady	std: sub a($x where * >= 0) {...} # does this parse yet?
20:38		p6eval	std 32123: OUTPUT«Potential difficulties:␤  $x is declared but not used at /tmp/_7FJoYF9Ir line 1:␤------> sub a(⏏$x where * >= 0) {...} # does this parse␤ok 00:01 119m␤»
20:40		Juerd	moritz_: Er, I never did maintain the pugs repository
20:40		rhebus	tadzik: are they expected to work once rakudo* implements more features? should I include some examples which work now?
20:40		tadzik	rhebus: it's about them I think, not about Rakudo
20:41		moritz_	Juerd: I had the impression you maintained the apache part of it
20:41		rhebus	tadzik: so I should fix them then?
20:41		tadzik	rhebus: would be great
20:41		rhebus	see what I can do
20:41		tadzik	poke moritz_ or someone for a commit bit
20:41		rhebus	got one already
20:41		moritz_	Juerd: if not, you still deserve thanks for maintaining and providing feather
20:42		Juerd	moritz_: I've prevented feather from crashing, which in some cases meant trying to make apache behave, or at least killing it when it didn't.
20:42		Juerd	(It has crashed, several times, under the severe loads)
20:42			Guest58080 left #perl6
20:42		moritz_	well, that's some form of maintenance
20:42		Juerd	Heh
20:42		moritz_	not the one you usually do if you are being paid for it
20:43		moritz_	but still
20:43		Juerd	I'm afraid that the "community maintained" part of feather never got very popular, except for its first year of existence :)
20:43		Juerd	A lot of people have done maintenance tasks on feather
20:43		TimToady	it turned out to be a bit too anarchical
20:43		Juerd	TimToady: Certainly.
20:44		Juerd	But then, originally it was a box for compiling pugs and running irssi
20:44		Juerd	It's still used for irssi and compiling (not pugs anymore), so it's still true to its initial purpose :)
20:45		Juerd	9 irssis :)
20:46		TimToady	now I just have to git learned
20:46			s_mosher left #perl6
20:47		moritz_	pushing to perlt/std as we speak
20:48		moritz_	http://github.com/perl6/std
20:48		TimToady	I seen it
20:48		TimToady	thanks
20:49		moritz_	you're welcome
20:49			Ross left #perl6
20:50			Ross joined #perl6
20:50		* moritz_	-> sleep
20:50		TimToady	zzz --> moritz_++
20:51		TimToady	I suppose that should be ==>
20:53			timbunce left #perl6
20:55		lue	Why is it that typing 'use A::B' in a certain file (let's call it X.pl) doesn't work when the file containing A::B isn't in A/B.pm, instead being in the same place as X.pl
20:55			CoCol joined #perl6
20:58		TimToady	to the first approximation, because that's not how it works in Perl 5
20:58		TimToady	used modules don't autoadd themselves to the search path
20:58		tylercurtis	lue: because that's what import is for, no?
20:59		tylercurtis	Or, wait.. what do you mean by "instead being in the same place as X.pl"?
20:59			timbunce joined #perl6
20:59		lue	typing `ls ./' would result in listing X.pl and C.pm [the file containing module A::B]
21:00		tylercurtis	lue: well, how would Perl 6 know that C.pm contains A::B?
21:00		moritz_	lue: if you wrote a Perl 6 compiler, how would it know that A::B can be found in C.pm?
21:01		dukeleto	Juerd++ # irssi :)
21:03		lue	I want it to search all the .pm files in ./ for A::B. What else are module declarations for? [I guess that's not happening though]
21:05		TimToady	eventually we'll move more toward a model where you register your modules in the database, and then it can track things like that
21:05			Mowah joined #perl6
21:05		TimToady	in any case, the official library is supposed to work like that, and your local code could work like that, or more like the current PERL6LIB search model
21:05		TimToady	but that isn't gonna just search a bunch of files on spec for you
21:06		TimToady	that would be slow and error prone
21:08		lue	When I had A/B.pm to experiment, use A::B didn't work in the REPL.
21:08		tylercurtis	Was A/B.pm in a directory in your @*INC?
21:10		lue	in ./
21:10		moritz_	and was . in @*INC?
21:10		mathw	o/
21:10		lue	rakudo: say @*INC
21:10		p6eval	rakudo ff5b4c: OUTPUT«lib/home/p6eval/.perl6/lib/home/p6eval//p​1/lib/parrot/2.7.0-devel/languages/perl6/lib.␤»
21:10		rhebus	does the reduce metaoperator work on user operators?
21:11		TimToady	supposed to
21:11		rhebus	I've been having a go but I can't work it out: http://pastebin.com/kchCQev7
21:11		rhebus	the line labelled #BOOM doesn't work
21:11		TimToady	in rakudo you have to make them 'our' though
21:11		moritz_	rhebus: in rakudo it only works if you declare it with 'our'
21:11		TimToady	which is a known bug
21:11		rhebus	aha
21:11		moritz_	rakudo: our sub infix:<abc>($a, $b) { $a + $b + 1 }; say [abc] 1, 2, 3
21:11		p6eval	rakudo ff5b4c: OUTPUT«8␤»
21:11		lue	I'm having an amazingly difficult time figuring out what is IMO a very confusing system for modules.
21:12		TimToady	still the bogus design that came from parrot assuming it was re-implementing Perl 5
21:12		rhebus	should I update the 99-examples to use "our" or should I leave it given it's a known bug
21:12		moritz_	update please
21:12		tylercurtis	lue: What is confusing you?
21:12		TimToady	our will still work later, since it does install a lexical alias
21:12		moritz_	(as long as the workaround is still valid Perl 6)
21:13		rhebus	and should I comment that it's a bug workaround?
21:13		* moritz_	-> really sleep :-)
21:13		TimToady	zzz ==> moritz_++ really
21:13		rhebus	I'm not the person to ask if it's valid perl 6, I'm doing the 99 examples to learn it...
21:14		TimToady	our is valid, if suboptimal in the long run
21:14		TimToady	especialy if it causes global collisions
21:15		lue	I just can't seem to get anything working. I do A/B.pm like you're 'supposed to', and then I can't use sub C, even after attaching `is export' to sub C and putting a redundant 'module A::B' at the top.
21:15		rhebus	I think that's worth noting in a comment
21:16			Guest23195 left #perl6
21:16		tylercurtis	lue: can you show us your code example? "module A::B; our sub C () is export {...}" should make C available to any scope that does "use A::B;".
21:17		lue	...
21:17		* lue	is in stunned silence
21:18			bluescreen joined #perl6
21:18		lue	You mean to say all I had to do was add `our' in front of sub C !? *sobs to self*
21:18			bluescreen is now known as Guest62968
21:18		TimToady	it's supposed to work with 'my' too, but as the other conversation was pointing out, rakudo is still kinda busted that way
21:19		lue	It solves why my experimenting failed, but it doesn't solve the originial problem of there being only OWTDI when it comes to modules.
21:23		tylercurtis	If @*INC works like Perl 5's @INC, you can stick callables in there (or maybe objects with an INC method?), and do whatever bizarre import magic you want with it.
21:26		tylercurtis	Or, at least, it should for require. Not sure about use.
21:26		* tylercurtis	hasn't written much Perl 5 and is just basing this on perldoc -f require.
21:28			Italian_Plumber joined #perl6
21:29			CoCol left #perl6
21:32		Tene	lue: you don't need 'our'
21:32		Tene	sec, I'll post
21:32		lue	I just don't like the A/B.pm setup in a source code repository, and it shouldn't be the only way to do it. [after all, this *is* Perl 6 :)]
21:33		Tene	lue: http://gist.github.com/565509
21:33		Tene	lue: that example works fine for me.  How does it differ from what you say doesn't work?
21:34		lue	Hm. I wonder why it didn't work before.
21:38		lue	Grr. Not even ./A::B.pm will work with "use A::B".
21:38			hercynium joined #perl6
21:41		Tene	whaaat?  I don't follow what that means.
21:41		Tene	Oh.
21:41		Tene	No, of course not.
21:42		Tene	I don't think that that would be *unreasonable* to add to the spec.
21:43		Tene	Having multiple possible locations for modules within a single search path would require a specced order to search them in, and possible confusion there, but that's doable.
21:44		Tene	You'd have to specify the behavior when there *are* multiple locations.  Would there be a warning?  Would it be an error?  Is A/B::C/D.pm allowed?  A::B/C/D.pm?  Only trailing /s can be replaced with ::s?
21:44		Tene	This isn't anything that anyone involved with Perl 6 has addressed yet, or asked for.  I think it would be a nice addition to the spec.
21:45		lue	afk, but let me say this: imo, the use of modules in Perl 6 is not very Perl 6 right now, because There Is Only One Way To Do It.
21:46		rhebus	using ::s is a Good Thing, because it abstracts away the notion of a path separator
21:46		TimToady	for the official library, there will be only one way to do it, insofar as once something is official, it must be considered immutable; to change you must give it a different longname (which includes version)
21:47		TimToady	the local stuff can be looser, of course
21:48		TimToady	but I suspect some people would like their user view to work much like the official library writ small
21:49		TimToady	so every use is essentially a database query rather than a path search
21:49		TimToady	path searching just doesn't scale very well, and is somewhat fragile
21:49		Tene	lue: if you want to do something else, you certainly can do it, there's just only one module loading style specified ATM.  You could certainly write a lue_special_use 'A::B' that did what you wanted.
21:50			tadzik left #perl6
21:50		Tene	lue: I do recommend that you put together an initial proposal with a few open questions and submit to p6l
21:54			Ross left #perl6
21:54			zby_home left #perl6
21:55			Ross joined #perl6
22:00		TimToady	I dislike that specs/ and spectests/ are not very tab-friendly
22:00		TimToady	so whatever we rename it to should probably not start with "spec"
22:01			Chillance_ joined #perl6
22:01		diakopte1	testspec?
22:01		diakopte1	vs spec?
22:01		TimToady	"valid" starts with a rare letter
22:01		TimToady	or something like, "inspect"
22:01		TimToady	Implementation Neutral SPEC Tests
22:01			synth joined #perl6
22:01		diakopte1	heh
22:02		diakopte1	Implementation Neutered
22:02		TimToady	or "respect", "prospect", etc
22:03		tylercurtis	testsuite or official-test-suite if we want a boring name?
22:03		TimToady	validation is more officialese
22:05		TimToady	and I kinda liked Validate All Language Implementation Details
22:06		TimToady	but then I also liked Sucks Or Rocks Reliably Yet
22:07		TimToady	is the only way to rename a repo to fork it?
22:07		diakopte1	Backronyms Are Cool, Kinda, Rendering Order Naturally Yet Mnemonically
22:08		TimToady	++
22:08		TimToady	so we should first pick a word like TORTURE and then figure out what it means
22:09		tylercurtis	TimToady: Going to the repo, click Admin, type in a different name in the Repository Name place and click Rename also works, I think (I haven't tried it, but I'm looking at the text field next to the button that says Rename).
22:09		TimToady	.oO(torcher)
22:09		rhebus	rakudo: <1 2>.map: { $_ xx 2 }
22:09		p6eval	rakudo ff5b4c:  ( no output )
22:09		TimToady	just seems like this would be better done sooner than later
22:09		TimToady	if we can come up with a suitable name
22:10		tylercurtis	rakudo: .say for <1 2>.map: { $_ xx 2};
22:10		p6eval	rakudo ff5b4c: OUTPUT«1␤1␤2␤2␤»
22:10		tylercurtis	rhebus: ^^
22:10		rhebus	rakudo: my @l = <1 2>.map: { $_ xx 2 }; @l.perl.say
22:10		p6eval	rakudo ff5b4c: OUTPUT«["1", "1", "2", "2"]␤»
22:10		rhebus	hmm
22:11		tylercurtis	rakudo: my $l = <1 2>.map: { $_ xx 2}; say $l.perl;
22:11		p6eval	rakudo ff5b4c: OUTPUT«(("1", "1"), ("2", "2"))␤»
22:11		diakopte1	what are we naming again?
22:11		rhebus	that's what I'm after
22:12		rhebus	tylercurtis: what do those round brackets mean? is it a flat list or not?
22:12		TimToady	it means it's a parcel that hasn't decided flatness yet
22:12		TimToady	that's a lazy decision
22:13		tylercurtis	rakudo: say (<1 2>.map: { [$_ xx 2] }).perl
22:13		rhebus	when might it not be flat?
22:13		TimToady	in slice context (not yet impl) it would turn into subsequences
22:13		p6eval	rakudo ff5b4c: OUTPUT«(["1", "1"], ["2", "2"])␤»
22:13		tylercurtis	The above also works if you really want the result to not be flat.
22:13		rhebus	no, i want the result flut
22:13		rhebus	*flat
22:13		TimToady	most listy functions will assume flat by default
22:14		rhebus	I'm confused, have I suceeded or not?
22:14			Ross left #perl6
22:14		rhebus	here's my attempt at P12: http://pastebin.com/TzK9EZv2
22:14		tylercurtis	"my @a = <1 2>.map: { $_ xx 2 };" flattened.
22:14		TimToady	that would depend on what you want to do with the value
22:15		TimToady	rakudo: my $l = <1 2>.map: { $_ xx 2}; say $l.flat.perl
22:15		p6eval	rakudo ff5b4c: OUTPUT«("1", "1", "2", "2")␤»
22:15		rhebus	it's shorter and neater than the existing P12 but it's of ambiguous flatness
22:16		TimToady	that's fine, the ~ will flatten
22:17		TimToady	most operators that don't want a multidimensional argument will flatten
22:17		TimToady	you'd generally use slice to force it the other way
22:17		rhebus	I think it's best to explicitly flatten
22:17			timbunce left #perl6
22:17		rhebus	so I'll do that
22:18			sahadev joined #perl6
22:18		rhebus	tada: http://pastebin.com/34FaLSav
22:18		diakopte1	were we naming the testsuite dir?
22:19		diakopte1	niecza: say 3
22:19		p6eval	niecza 568bfa6: OUTPUT«3␤»
22:19		diakopte1	O_O
22:19		diakopte1	cool
22:20		TimToady	you can just say 'when Array'
22:20		TimToady	since $_ is already set
22:20		TimToady	well, not as the arg to ??!! though
22:21		diakopte1	std: say 4 ?? !! 4
22:21		p6eval	std 32123: OUTPUT«===[0mSORRY!===[0m␤Found ?? but no !! at /tmp/MRT9qt2wvc line 1 (EOF):␤------> say 4 ?? !! 4⏏<EOL>␤Confused at /tmp/MRT9qt2wvc line 1 (EOF):␤------> say 4 ?? !! 4⏏<EOL>␤Parse failed␤FAILED 00:01 116m␤»
22:22		diakopte1	nits are always discoverable with a powerful enough microscope
22:22		TimToady	in any case, you can always leave the $_ off the front of $_.method
22:23			aloha left #perl6
22:23			bacek left #perl6
22:23		rhebus	TimToady: ++ http://pastebin.com/nHeaMMZy
22:24		TimToady	that should work
22:24		rhebus	it does read nicer than ?? !!
22:24		rhebus	it does work
22:24			M_o_C joined #perl6
22:25		rhebus	and it's nicer than the other solution :)
22:25		TimToady	once you start getting used to smartmatching, things like .isa start looking like a code smell...er, smelling like an eyesore...er...
22:26		rhebus	don't look at me in that tone of voice
22:26		TimToady	:)
22:27		rhebus	perl 6 is really really nice so far
22:27		* TimToady	bows on behalf of a whole bunch o' folks
22:28		* mathw	is always glad to hear somebody say that
22:28		rhebus	I've known of it for a while, and heard Damian Conway do a talk at london.pm raving about it
22:29		rhebus	but today's the first day I've dived into it
22:29			BillN1VUX left #perl6
22:30		mathw	Perl 6 tends to get people to want to rave about it
22:30		diakopte1	I presume the raving was intelligible? :)
22:30		TimToady	we won't ask if it was mad, we already know the answer
22:30		rhebus	as intelligible as Damian ever gets
22:35			crythias joined #perl6
22:35			bacek joined #perl6
22:35			aloha joined #perl6
22:37			svetlins__ left #perl6
22:37		rhebus	is there no infix !%% ?
22:38		rhebus	rakudo: say 4 !%% 3
22:38		p6eval	rakudo ff5b4c: OUTPUT«===SORRY!===␤Infix !% is deprecated in favor of infix %% at line 22, near "% 3"␤»
22:38		TimToady	rakudo: say 4 % 3
22:38		rhebus	:/
22:38		p6eval	rakudo ff5b4c: OUTPUT«1␤»
22:38		rhebus	oh
22:38		rhebus	er
22:38		rhebus	yes
22:38		rhebus	that error message is confusing, but you have a very good point
22:38		TimToady	in fact, %% used to be named !%, till we discovered that was confusing
22:39		TimToady	hmm
22:39		TimToady	rakudo: say 4 ![%%] 3
22:39		p6eval	rakudo ff5b4c: OUTPUT«1␤»
22:40		rhebus	heh, it's nice to write an exercise and get pretty much exactly the same as what somebody else has got
22:41		rhebus	a lot of these 99 problems are only actually difficult in a LISPy environment
22:42		rhebus	they'd be a lot trickier when having to decompose conses rather than just indexing arrays
22:43			bjarneh joined #perl6
22:44			bjarneh left #perl6
22:44		rhebus	i got 99 problems but lisp ain't one
22:45			jhuni joined #perl6
22:45			hercynium left #perl6
22:49		rhebus	does rakudo do tail-call? are recursive examples a sensible idea?
22:50			svetlins__ joined #perl6
22:56			Italian_Plumber left #perl6
22:57		TimToady	p6 is specced to assume tail calls are optimized, but I doubt rakudo does it yet
22:57			xinming left #perl6
22:59			xinming joined #perl6
23:02			meppl left #perl6
23:14		rhebus	right I've pushed my examples to github, time for bed
23:16			rhebus left #perl6
23:17			Mowah left #perl6
23:30			sahadev left #perl6
23:31			M_o_C left #perl6
23:31			svetlins__ left #perl6
23:32			svetlins__ joined #perl6
23:37			cotto joined #perl6
23:37			svetlins__ left #perl6
23:44			masonkramer left #perl6
23:44			masonkramer_ joined #perl6
23:45			jedai joined #perl6
23:53		tylercurtis	Has Perl 5 stolen ... yet?
23:54		TimToady	they have a problem: ... already means something else
23:54		TimToady	it's a variant of scalar ..
23:55		TimToady	I'm thinking "roast"  repo of all spec tests (or some such)
23:55			svetlins joined #perl6
23:56		TimToady	can be used as both a noun and a verb
23:57			svetlins left #perl6
23:58		tylercurtis	http://perldoc.perl.org/perlop[…]ada-Yada-Operator They seem to have "!!!", but it's named "...".
23:58			svetlins joined #perl6
23:58		TimToady	that's different, where a term is expected
23:58		TimToady	or were you asking about p6's yada?
23:58		TimToady	rather than series
23:59		* tylercurtis	was unclear.
23:59		tylercurtis	I meant yada, yes.
23:59		TimToady	spectests/ is now roasts/
23:59		TimToady	er, roast/
23:59			svetlins left #perl6

← Previous day | Channels | #perl6 index | Today | Next day → | Search | Google Search | Plain-Text | plain, newest first | summary

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs

Powered by ilbot, built with the Perl programming language. Provided by Moritz Lenz. Please direct suggestions and bug reports to moritz on irc.freenode.net or moritz 'at' faui2k3 dot org.

Imprint