Camelia, the Perl 6 bug

IRC log for #perl6, 2010-09-04

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time 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/Result​s.html?Query=(Subject%20LIKE%20'lhf'​)%20AND%20Status%20!%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 + $^b }
04:20 JimmyZ yes
04:21 masak with the extra curly braces, you get two levels of curlies. and it doesn't work.
04:21 JimmyZ rakudo: say ~(my @Fibonacci := (0, 1, * + * ... 900));
04:21 p6eval rakudo 52f9ea: OUTPUT«0 1 1 2 3 5 8 13 21 34 55 89 144 233 377 610␤»
04:22 JimmyZ masak: my fault :0
04:22 masak no sweat. now you know :)
04:23 JimmyZ masak: yes, thanks ;)
04:24 \xF0 nice
04:25 * masak really likes Haskell's </> operator for FilePath. http://www.haskell.org/ghc/docs/6.​6.1/html/libraries/filepath/System​-FilePath-Posix.html#v%3A%3C%2F%3E
04:25 masak maybe we should steal it.
04:25 JimmyZ rakudo: sub infix:<+-*/>($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​/perl6/book/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/niecza/commit/c​d6e7c1883dd0eccc8c6ff2eee51a841df95d232
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/niecza/commit/7​6cd7d9098e590d41569ee0ec79af3583fd2f8a2
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/niecza/commit/5​68bfa6c13a45b19a824fbf8f658826647f4c4ed
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/89a1f74d7​61c1dafe4807a12dff9d91b.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/6cda7e2f6​f4e940a9c2313c5c4c37296.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/a19681e1e​223cd10c285482dfc9398b0.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/8.​3/interactive/sql-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 pmichaud@orange:~$ 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 pmichaud@orange:~/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/73ea6e348​3eb7e949c942df8ddab5d30.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/rakudo/commit/2​9fcf8e7d323336a1e03c59a3ab7204106c2c2a2
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/rakudo/commit/6​b1b0302d0e5942cc6b7c49227da5191d695ca4c
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/rakudo/commit/e​36452e26d6083c88b2103dba0a93f159f7c21e2
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.​php?section=problems&amp;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/onlamp/blo​g/2006/12/99_problems_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/spect​ests/tree/master/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/rakudo/commit/f​f5b4c0d076d609483fcccf558b9d113ce9a4ab6
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/r​aw/master/src/perl6/STD.pm6
19:04 pmichaud or http://github.com/perl6/mu/b​lob/master/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/use​rs/smash/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:␤  $x is declared but not used at /tmp/_7FJoYF9Ir line 1:␤------> [32msub a([33m⏏[31m$x where * >= 0) {...} # does this parse[0m␤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«[31m===[0mSORRY![31m===[0m␤Found ?? but no !! at /tmp/MRT9qt2wvc line 1 (EOF):␤------> [32msay 4 ?? !! 4[33m⏏[31m<EOL>[0m␤Confused at /tmp/MRT9qt2wvc line 1 (EOF):␤------> [32msay 4 ?? !! 4[33m⏏[31m<EOL>[0m␤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/per​lop.html#Yada-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

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs