Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2012-08-04

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 cognominal_ joined #perl6
00:03 Tedd1 joined #perl6
00:07 tokuhiro_ joined #perl6
00:24 cognominal__ joined #perl6
00:27 whiteknight joined #perl6
00:40 cognominal joined #perl6
01:00 cognominal joined #perl6
01:06 UncleFester6 joined #perl6
01:33 ponbiki joined #perl6
02:37 Psyche^ joined #perl6
02:39 leprevost joined #perl6
02:55 orafu joined #perl6
03:20 bluescreen10 joined #perl6
03:28 dalek roast: d0cf38e | coke++ | S05-interpolation/regex-in-variable.t:
03:28 dalek roast: pugs fudge
03:28 dalek roast: review: https://github.com/perl6/roast/commit/d0cf38e5a9
03:30 gongyiliao joined #perl6
03:31 Timbus r: sub ♞ {"knight"}
03:31 p6eval rakudo 1f662c: OUTPUT«===SORRY!===␤Missing block␤at /tmp/QYTg3e9hBc:1␤»
03:31 Timbus not allowed?
03:31 Timbus n: sub ♞ {"knight"}; say ♞;
03:32 p6eval niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![31m===[0m��Malformed block at /tmp/3WDDbuhVVO line 1:�------> [32msub [33m�[31m♞ {"knight"}; say ♞;[0m��Parse failed��»
03:32 Timbus guess so
03:32 dalek roast: 1b99898 | coke++ | S06-other/main.t:
03:32 dalek roast: niecza fudge
03:32 dalek roast: review: https://github.com/perl6/roast/commit/1b998989bb
03:37 [Coke] r: sub postfix<♞> { "knight" };
03:37 p6eval rakudo 1f662c: OUTPUT«===SORRY!===␤Missing block␤at /tmp/fdJ_TSY3Zl:1␤»
03:38 benabik r: sub postfix:<♞>($x) { $x ~ "knight" }; say "white "♞
03:38 p6eval rakudo 1f662c: OUTPUT«white knight␤»
03:39 timotimo what exactly does fudge mean? is it different from "skip" or "xfail"?
03:39 benabik r: sub term:<♞>() { "knight" }; say ♞
03:39 p6eval rakudo 1f662c: OUTPUT«===SORRY!===␤Cannot add tokens of category 'term'␤at /tmp/ELsIQfALdE:1␤»
03:39 benabik I think "fudge" is for things that fail only in some implementations.
03:40 timotimo OK
03:40 timotimo oh my, it's already light out. time to go to bed >_>
03:40 timotimo good * #perl6
03:41 Timbus r: macro circumfix:«♞ ♞»($text) { return "kn" ~ "i" x $text.chars ~ "ghts" }; macro circumfix:«☀ ☀»($text) { return "d" ~ "a" x $text.chars ~ "ze" }; say "arabian " ~ ♞'♞♞♞♞'♞; say "like arabian " ~ say "Like arabian " ~ ☀'☀☀☀☀☀'☀; say "more often than not, are hotter than hot"
03:41 p6eval rakudo 1f662c: OUTPUT«arabian kniiiights␤Like arabian daaaaaze␤like arabian True␤more often than not, are hotter than hot␤»
03:41 Timbus aw
03:43 Timbus is it possible to get macros to take non-strings
03:47 birdwindupbird joined #perl6
03:47 Timbus r: macro circumfix:«♞ ♞»($text) is parsed / <-[♞]>+ / { return "kn" ~ "i" x $text.chars ~ "ghts" }; say ♞♘♘♘♞
03:47 p6eval rakudo 1f662c: OUTPUT«===SORRY!===␤Missing block␤at /tmp/IvVjMgDvki:1␤»
03:48 Timbus d'oh
03:51 ingy joined #perl6
04:05 jaldhar joined #perl6
04:07 Stocks29 joined #perl6
04:31 kaare_ joined #perl6
04:33 telex joined #perl6
04:48 thou joined #perl6
05:11 JimmyZ_ joined #perl6
05:15 moritz jnthn: (re documenting metamodel in perl6/doc) yes, reasonable. Go ahead :-)
05:17 moritz good morning everybdoy
05:18 gongyiliao joined #perl6
05:19 JimmyZ_ moritz: good morning
05:22 sorear good morning moritz.
06:13 dalek doc: 45054e8 | moritz++ | lib/objects.pod:
06:13 dalek doc: [object] self, private methods, submethods
06:13 dalek doc: review: https://github.com/perl6/doc/commit/45054e89f4
06:14 moritz r: class A { has $.x; submethod BUILD() { } }; A.new(:x<3>).x.say
06:14 p6eval rakudo 1f662c: OUTPUT«Any()␤»
06:39 SamuraiJack joined #perl6
07:06 cognominal joined #perl6
07:30 GlitchMr joined #perl6
07:59 tokuhiro_ joined #perl6
08:02 nodmonkey joined #perl6
08:19 masak good antenoon, #perl6
08:20 moritz \o masak
08:20 brrt joined #perl6
08:22 * masak back-clogs
08:23 moritz r
08:24 Timbus arr!
08:24 moritz array!
08:24 moritz r: say Array.^mro
08:24 p6eval rakudo 1f662c: OUTPUT«Array() List() Iterable() Cool() Any() Mu()␤»
08:24 moritz r: say Parcel.^mro
08:25 p6eval rakudo 1f662c: OUTPUT«Parcel() Cool() Any() Mu()␤»
08:25 * moritz thought Parcel was Iterable
08:26 Timbus i oh thought you were txting in pirate. "r me hrties. wnt 2 plndr l8r?"
08:27 moritz akshually it was just a mostly dead ssh connection where I didn't see what I typed
08:28 masak r: $_ = "mostly dead ssh connection"; s[ea] = "rea"; .say
08:28 p6eval rakudo 1f662c: OUTPUT«mostly dread ssh connection␤»
08:30 sorear o/ masak
08:30 JimmyZ_ r: s[ea] = "rea"; .say
08:30 p6eval rakudo 1f662c: OUTPUT«No such method 'subst' for invocant of type 'Any'␤  in block  at /tmp/htGuR5us8T:1␤␤»
08:30 masak sorear: \o
08:31 JimmyZ_ r: $_ = ""; s[ ] = "rea"; .say
08:31 p6eval rakudo 1f662c: OUTPUT«rea␤»
08:32 JimmyZ_ r: $_ = ""; s[] = "rea"; .say
08:32 p6eval rakudo 1f662c: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter ] (must be quoted to match literally) at line 2, near " = \"rea\"; "␤»
08:33 JimmyZ_ std: $_ = ""; s[ ] = "rea"; .say
08:33 p6eval std f43a358: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Null pattern not allowed at /tmp/VK1jegzuR7 line 1:â�¤------> [32m$_ = ""; s[ [33mâ��[31m] = "rea"; .say[0mâ�¤    expecting quantifierâ�¤Parse failedâ�¤FAILED 00:00 43mâ�¤Â»
08:34 tadzik good morning
08:35 moritz \o tadzik
08:36 thelazydeveloper joined #perl6
08:38 sorear .oO( there's only one? also, that's quite some Hubris you have there )
08:38 brrt left #perl6
08:39 moritz sorear: only one what? /me misses some context
08:40 sorear 01:36 -!- thelazydeveloper [~lazydevel@unaffiliated/thelazydeveloper] has joined #perl6
08:41 * masak comes crashing in through the window
08:41 masak moritz: grammar!
08:41 * masak runs off
08:41 moritz now we can just hope for impatience
08:41 sorear I'm missing the context
08:44 * moritz too
08:44 * masak .oO( Perl 6 Coding Context )
08:45 masak by the way, first and second prize winners have gotten their prizes now, if not all their book deliveries yet.
08:45 moritz \o/
08:46 masak the p6cc2011 was awesome. the judging was a bit slow, but that's my fault.
08:46 masak moritz++ helped with a second set of eyeballs, moral support, and occasional necessary reminders.
08:46 moritz and much of the review for t1 and t2 :-)
08:47 masak oh, yes. moritz++
08:47 masak also, p6cc2012 will be *awesome*.
08:49 moritz r: class A { has $.x = 5; submethod BUILD() { } }; A.new(:x<3>).x.say
08:49 p6eval rakudo 1f662c: OUTPUT«5␤»
08:52 * masak finds he used "awesome" about both p6cc2011 and p6cc2012, but felt he meant different things...
08:52 moritz MTA: More Than Awesome
08:53 masak :P
08:55 mucker joined #perl6
09:01 masak I'm implementing Lambda Calculus. planning to do a small writeup about it. http://github.com/masak/lambda-calculus/commits/master
09:01 masak but now I have a case of a left-recursive rule, and I'm thinking about how best to handle it.
09:02 masak here's what the Wikipedia page has to say about it: "Applications are assumed to be left associative: M N P may be written instead of ((M N) P)"
09:03 masak so both of these create the AST `app[ app[M, N], P ]`.
09:03 moritz with a quantifier :-)
09:03 masak heh :)
09:03 masak yes, the thought crossed my mind.
09:03 masak and then just have the action method create the right AST.
09:03 moritz right
09:03 masak moritz++
09:03 moritz or
09:04 moritz build a parser generator that correctly handles left recursion :-)
09:04 masak yeah.
09:04 moritz or take an existing one :-)
09:04 masak I was wondering if maybe the new LTM-y bits would correctly handle left recursion.
09:04 moritz you can make bison emit its jump table
09:04 moritz recursion terminates longest prefixes
09:04 masak but there's a potential infinity in there which I guess is a bit too much even for our LTM.
09:04 moritz so, no
09:04 masak ah, yes.
09:05 masak does that go for mutual recursion as well? that must be a hoot to compute.
09:05 moritz mutual recursion also terminates longest prefixes
09:05 * masak head spins
09:06 moritz well, it would be much harder to have it not terminate LTM :-)
09:06 moritz because it would mean the NFA/DFA builder would need to know how to turn recursion into quantifiers
09:06 masak ah well. it's declarative, and everyone knows declarations don't require any computation :P
09:06 moritz now you can just have a hash with all method names that have already been visited (when doing the declarative prefix calculations)
09:07 masak ah; yes.
09:07 masak hm.
09:07 moritz I'd say that's one of the simpler parts of LTM implementation :-)
09:07 masak I can't quite get that to fit, though.
09:07 masak you have to remove stuff from that hash as you go up the call tree.
09:08 masak oh, I guess it works out.
09:08 arnsholt I think the "standard" way to avoid left-recursion in CFGs is to refactor the grammar
09:08 moritz masak: you just have to copy the hash for each new branch
09:09 arnsholt There's a weakly equivalent (same language, but different parse tree) Greibach normal form for all CFGs
09:09 dalek doc: c6ac771 | moritz++ | lib/objects.pod:
09:09 dalek doc: [objects] inheritance, object construction
09:09 dalek doc: review: https://github.com/perl6/doc/commit/c6ac77175f
09:09 moritz ah yes, I remember that fun in the formal languages class :-)
09:09 masak oh, so much to learn... :)
09:10 moritz reviews of the commit above would be very welcome
09:10 moritz is it understandable, is it correct?
09:11 * masak reviews
09:12 masak s/the parents' method/the parents classes' method/, perhaps?
09:13 masak s/of the name/of that name/ -- and this doesn't address "longnames".
09:13 masak s/The order by which/The order in which/
09:13 masak s/\(mro\)/(MRO)/
09:14 sorear sleep&
09:14 masak 'night, sorear.
09:15 masak "All calls to public method are "virtual" in the C++ sense" -- nitpick: is it the call that is virtual? or the public method?
09:15 masak "Point.new( x => 1, y  => 2);" -- unbalanced whitespace inside parens.
09:16 mberends joined #perl6
09:16 mberends o/
09:16 masak mberends! \o/
09:16 masak s/\(ie/(i.e./
09:17 * mberends had time to read irclogs for a change :-)
09:18 moritz \o mberends
09:18 dalek doc: 2c69e87 | moritz++ | lib/objects.pod:
09:18 dalek doc: fixes by masak++
09:18 dalek doc: review: https://github.com/perl6/doc/commit/2c69e87601
09:19 masak ss/However this is considered poor practi<(s)>e/c/
09:19 * masak loves <( )>
09:20 masak moritz: maybe worth pointing out that the $!attr binding in BUILD isn't particular to BUILD, or even submethods
09:20 masak anyway, moritz++. nice, clear explanation.
09:20 Khisanth joined #perl6
09:21 masak oh, and I'd perhaps go with s/much harder/harder/ in the end. :)
09:26 brrt joined #perl6
09:27 arnsholt Oooh. <( )> is pretty spiffy
09:29 masak moritz, arnsholt: ended up refactoring my l-recursion into a quantifier :) thanks.
09:32 * JimmyZ_ likes <( )> too.
09:41 brrt left #perl6
09:41 brrt joined #perl6
09:44 masak rn: grammar G { regex TOP { ^ $ { make "OH HAI" } } }; say G.parse('').ast
09:44 p6eval rakudo 1f662c, niecza v19-15-g051783d: OUTPUT«OH HAI␤»
09:45 masak hm.
09:45 PacoAir joined #perl6
09:46 jnthn moritz++ # taking my scribblings in objects.pod and adding loads more to them :)
09:47 jnthn good morningish
09:47 masak good mornington, jnthn.
09:47 masak rn: grammar G { regex TOP { ^ <foo> $ { make $<foo>.ast } }; regex foo { foo { make "OH HAI" } } }; say G.parse('foo').ast
09:47 p6eval rakudo 1f662c, niecza v19-15-g051783d: OUTPUT«OH HAI␤»
09:48 JimmyZ_ good morning jnthn
09:50 masak rn: grammar G { regex TOP { ^ <foo> $ { make $<foo>.ast } }; regex foo { [foo|bar] { make "OH HAI" } } }; say G.parse('foo').ast
09:50 p6eval rakudo 1f662c, niecza v19-15-g051783d: OUTPUT«OH HAI␤»
09:54 dalek doc: 12121ef | jimmy++ | lib/objects.pod:
09:54 dalek doc: fixed typo.
09:54 dalek doc: review: https://github.com/perl6/doc/commit/12121ef7e0
10:02 * arnsholt starts reading up on S26
10:13 u_u joined #perl6
10:20 xinming joined #perl6
10:31 brrt joined #perl6
10:31 spider-mario joined #perl6
10:36 M_o_C joined #perl6
10:43 masak hehe. the 'λx.M N' abbreviation associates to the left ('(λx.M) N'), but the 'λxyz.N' abbreviation associates to the right ('λx(λy(λz.N))').
10:44 masak er, sorry. 'λx.(λy.(λz.N))'
10:45 shachaf λxyz.N isn't really "associating"
10:46 masak but left-associations translate into for loops, and right-associations translate into reverse for loops. the body of the for loop does a new 'make' with $/.ast somewhere inside its new argument.
10:46 masak shachaf: oh?
10:46 shachaf Oh, you weren't done.
10:46 masak shachaf: there's a variable and an expression. two things. it associates.
10:46 shachaf What would it mean for it to associate to the left?
10:47 masak ...heh.
10:48 arnsholt shachaf: For some operators, whether you go from right to left or left to right matters
10:48 masak yeah, I guess there's something else going on here than associativity.
10:48 masak but it felt just like associativity. strange.
10:49 moritz it's a bit like with multiple prefix operators
10:49 shachaf arnsholt: Sure. But there's no operator here, really.
10:49 moritz !+$foo can *only* mean !(+$foo)
10:49 moritz because (!+)($foo) doesn't make sense
10:49 masak right.
10:49 masak so it's just a kind of nesting.
10:49 moritz right
10:49 moritz shachaf++
10:50 crab2313 joined #perl6
10:51 masak this is very clean and makes me happy: https://github.com/masak/lambda-calculus/commit/aae9c586973222ee4b0980c9f1c15b23ce805cfe
10:51 moritz ooh, nice use of make and $/.ast to avoid a variable to hold the intermediate AST
10:51 masak hm, which is nicer, 'reverse $<variable>[0..*-2]', or '$<variable>[reverse 0..*-2]'?
10:52 masak moritz: yeah, that's what I meant above.
10:52 moritz the first
10:52 dalek rakudo/inliner: 25d7c81 | jnthn++ | src/Perl6/Optimizer.pm:
10:52 masak why?
10:52 dalek rakudo/inliner: A further allomorphy tweak, fixing one more test that the inliner regressed.
10:52 dalek rakudo/inliner: review: https://github.com/rakudo/rakudo/commit/25d7c818a7
10:52 dalek rakudo/inliner: 72075f2 | jnthn++ | src/Perl6/Actions.pm:
10:52 dalek rakudo/inliner: Don't consider things that use pseudo-packages to be inlinable.
10:52 dalek rakudo/inliner: review: https://github.com/rakudo/rakudo/commit/72075f2eca
10:52 moritz it has the advantage of working
10:52 masak huh?
10:52 masak oh!
10:52 moritz 0..*-2 is a WhateverCode
10:52 masak right.
10:52 masak sometimes I have too high expectations on *
10:53 shachaf Hah, "Lambda/Calculus.pm"
10:53 jnthn .oO( When you wish up on a *... )
10:53 moritz well, it does do many things right intuitively
10:53 moritz masak: in fact, (0..^*-2).reverse works
10:53 masak 'course, I could go '$<variable>[*-2 ... 0]' ;)
10:54 masak or, hm.
10:54 moritz I don't think you can
10:54 masak no.
10:54 moritz because the series is evaluated first
10:54 masak rn: my @a = <a b c d e f g>; .say for @a[*-2 ... 0]
10:54 p6eval rakudo 1f662c: OUTPUT«Not enough positional parameters passed; got 0 but expected 1␤  in block  at /tmp/M0eDqs6Y7Y:1␤␤»
10:54 p6eval ..niecza v19-15-g051783d: OUTPUT«Unhandled exception: Need more items on the LHS␤  at /home/p6eval/niecza/lib/CORE.setting line 1414 (die @ 5) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3147 (get-next-closure @ 11) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3191 (ANON @ 17) ␤  a…
10:54 masak right.
10:54 masak the series grabs the whatevs.
10:55 whiteknight joined #perl6
10:57 dalek doc: 4d4236e | moritz++ | lib/objects.pod:
10:57 dalek doc: [objects] roles, role application
10:57 dalek doc: review: https://github.com/perl6/doc/commit/4d4236ec43
10:58 moritz jnthn: in your talks on roles, do you have a better example than DrunkenGymnast for role conflicts?
10:58 * moritz would like to steal one :-)
10:59 masak what? clearly, that's the best example possible! :P
11:01 masak here's one: role VScrollable { method scroll($dy) { ... } }; role HScrollable { method scroll($dx) { ... } }; class MyWindow does VScrollable does HScrollable {}
11:02 jnthn moritz: I recall doing one where it was about booking a train ticket, and there was a $.line (as in, which railway line) but it conflicting with a $.line related to "line on the invoice"
11:06 moritz masak: hm, nice
11:06 GlitchMr joined #perl6
11:07 bluescreen10 joined #perl6
11:11 JimmyZ joined #perl6
11:34 GlitchMr perl6: printf '% %'
11:34 p6eval rakudo 1f662c: OUTPUT«'%' is not a valid sprintf format␤  in sub sprintf at src/gen/CORE.setting:2079␤  in sub printf at src/gen/CORE.setting:2084␤  in block  at /tmp/jZhTBBbewa:1␤␤»
11:34 p6eval ..niecza v19-15-g051783d: OUTPUT«Unhandled exception: invalid format specifier␤  at /home/p6eval/niecza/lib/CORE.setting line 1286 (sprintf @ 4) ␤  at /home/p6eval/niecza/lib/CORE.setting line 1287 (printf @ 5) ␤  at /tmp/RNA187pMhC line 1 (mainline @ 3) ␤  at /home/p6eval/niecza/lib/CORE…
11:35 gongyiliao joined #perl6
11:35 GlitchMr But... Perl 6 Synopsis allows things like '% %'
11:35 GlitchMr I guess it's bug in synopsis
11:38 birdwindupbird joined #perl6
11:41 dalek doc: 3a0a898 | moritz++ | lib/objects.pod:
11:41 dalek doc: roles and stubs
11:41 dalek doc: review: https://github.com/perl6/doc/commit/3a0a8986bd
11:43 masak GlitchMr: yes, I would say so.
11:45 masak GlitchMr: though C prints '%' on printf("% %");
11:45 masak so maybe we should allow it after all.
11:46 moritz perl 5 also prints %
11:56 crab2313 joined #perl6
12:08 dalek nqp: 6589656 | jnthn++ | src/QAST/Operations.nqp:
12:08 dalek nqp: Add a way for ops evaluating to native types to attach that information to nodes.
12:08 dalek nqp: review: https://github.com/perl6/nqp/commit/6589656170
12:08 brrt joined #perl6
12:09 dalek rakudo/inliner: 4a658ee | jnthn++ | src/ (4 files):
12:09 dalek rakudo/inliner: Fix return type check handling when inlining.
12:09 dalek rakudo/inliner: review: https://github.com/rakudo/rakudo/commit/4a658ee913
12:09 dalek rakudo/inliner: 57697e2 | jnthn++ | src/Perl6/Actions.pm:
12:09 dalek rakudo/inliner: Attach return types to nqp::op nodes.
12:09 dalek rakudo/inliner: review: https://github.com/rakudo/rakudo/commit/57697e2798
12:15 MayDaniel joined #perl6
12:19 _fall0ut_ joined #perl6
12:43 masak nr: class A { class B {}; method foo { say B.new } }; A.new.foo
12:43 p6eval niecza v19-15-g051783d: OUTPUT«B.new(...)␤»
12:43 p6eval ..rakudo 1f662c: OUTPUT«B.new()␤»
12:43 masak nr: class A::B { class C {}; method foo { say C.new } }; A::B.new.foo
12:43 p6eval niecza v19-15-g051783d: OUTPUT«C.new(...)␤»
12:43 p6eval ..rakudo 1f662c: OUTPUT«C.new()␤»
12:48 masak folks, Crypt::Game is now called Game::Crypt. and Hanoi::Game is called Game::Hanoi. all tests still pass and all CLIs still run.
12:48 masak I believe the code got a little better for it, too. japhb++
12:49 user joined #perl6
12:49 JimmyZ masak++
12:50 brrt joined #perl6
12:52 masak I've received *lots* of good crypt feedback from Keving Coyler, the winner of July's open challenge.
12:53 masak I will trawl through it during the weekend, as I find slots of time.
12:58 tokuhiro_ joined #perl6
12:58 dalek rakudo/inliner: 3bfb7a6 | jnthn++ | src/ (4 files):
12:58 dalek rakudo/inliner: Very minimal implementation of the 'soft' pragma, but enough that we'll be able to use it to make wrap.t happy.
12:58 dalek rakudo/inliner: review: https://github.com/rakudo/rakudo/commit/3bfb7a68d6
13:04 * masak hugs vim digraphs
13:09 dalek rakudo/nom: 5cd3686 | jnthn++ | src/Perl6/Ops.pm:
13:09 dalek rakudo/nom: Mark some Perl 6 ops as inlinable.
13:09 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/5cd36869f8
13:09 dalek rakudo/nom: 958a3ea | jnthn++ | src/ (2 files):
13:09 dalek rakudo/nom: We will keep around the QAST tree of inlinable things now, not some string thingy.
13:10 dalek joined #perl6
13:10 moritz ooh, inliner merged?
13:11 jnthn Yeah
13:11 masak yay.
13:11 masak jnthn++
13:11 moritz \o/
13:11 jnthn It's able to inline more stuff
13:11 jnthn Also
13:11 jnthn r: say 2 +< 63
13:11 p6eval rakudo 1f662c: OUTPUT«0␤»
13:11 jnthn > say 2 +< 63
13:11 jnthn 18446744073709551616
13:11 jnthn Those kinds of things are mixed.
13:12 moritz \o/
13:12 jnthn It's less stupid about literal allomorphy, or whatever it's called. :)
13:12 jnthn s/mixed/fixed/
13:13 jnthn oh, dang, I needed to bump NQP_REVISION
13:14 moritz I like to call it stupomorphy :-)
13:14 dalek rakudo/nom: 5f2af91 | jnthn++ | tools/build/NQP_REVISION:
13:14 dalek rakudo/nom: Bump NQP_REVISION to get improvements needed by the inliner.
13:14 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/5f2af91226
13:22 boni joined #perl6
13:24 mberends joined #perl6
13:27 dalek nqp: e401db5 | jnthn++ | src/6model/serialization.c:
13:27 dalek nqp: Re-enable re-use of strings on the string heap; it makes a fairly notable difference to the output size of CORE.setting compilation, and a small but certainly worthwhile difference to CORE.setting compilation time. Causes no issues here; hopefully not for anyone else now either.
13:27 dalek nqp: review: https://github.com/perl6/nqp/commit/e401db56c3
13:29 boni left #perl6
13:31 dalek rakudo/nom: 44cb610 | jnthn++ | tools/build/NQP_REVISION:
13:31 dalek rakudo/nom: Get string sharing optimization for faster builds.
13:31 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/44cb6102f7
13:31 dalek rakudo/nom: 5e1b9a8 | jnthn++ | docs/ChangeLog:
13:31 dalek rakudo/nom: ChangeLog updates.
13:31 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/5e1b9a8d61
13:31 jnthn away for a bit
13:34 cognominal_ joined #perl6
13:35 leprevost joined #perl6
13:49 JimmyZ hmm, I didn't get a difference to the output size of CORE.setting compilation
13:51 JimmyZ sorry, wrong nqp branch
13:53 awwaiid joined #perl6
13:53 GlitchMr [\+] ^Inf
13:54 GlitchMr I like Perl 6 :)
13:54 masak :)
13:54 GlitchMr I like how this list is lazy
13:55 masak well, it turns out it's much easier to have infinite lists if they're lazy... :)
13:55 GlitchMr > ([\+] ^Inf)[^20]
13:55 GlitchMr 0 1 3 6 10 15 21 28 36 45 55 66 78 91 105 120 136 153 171 190
13:55 GlitchMr :)
13:55 GlitchMr (well, ok, I could just type [\+] ^20, but whatever
13:55 masak the \ was chosen because it makes like a little triangle with the [
13:59 GlitchMr > 'ł' gt 'z'
13:59 GlitchMr True
13:59 GlitchMr Well, I guess that 'ł' is indeed bigger basing on codepoint
13:59 masak yes.
13:59 masak if you want locale-specific sorting, you have to specify that.
13:59 GlitchMr How?
13:59 masak through some means not specified by the Perl 6 spec.
14:00 masak that sounds like a CPAN module to me.
14:01 GlitchMr > sort 'a' .. 'z', <ą ć ę ł ń ó ś ź ż>
14:01 GlitchMr a b c d e f g h i j k l m n o p q r s t u v w x y z ó ą ć ę ł ń ś ź ż
14:01 GlitchMr This isn't really useful
14:01 masak depends on your use case.
14:02 GlitchMr I was doing it wrong all along
14:02 GlitchMr I shouldn't use sort with names
14:02 masak well, you should.
14:03 masak just not the default string sort.
14:04 GlitchMr But whatever, it's hard to determine what should be first: > sort 'a' .. 'z', <ą ć ę ł ń ó ś ź ż>
14:04 GlitchMr oops
14:04 GlitchMr sorry for paste
14:04 GlitchMr ß or ś for example
14:05 GlitchMr What should be first? I've no idea
14:05 masak that's why you need a locale.
14:05 mucker hi ... what does <!before ...> mean ?
14:06 GlitchMr Well, in case of ß or ś those characters belong to different languages
14:06 masak r: my $polish-alphabet = 'aąbcćdeęfghijklłmnńoópqrsśtuvwxyzźż'; sub polishly($c) { $polish-alphabet.index($c) }; say ('a' .. 'z', <ą ć ę ł ń ó ś ź ż>).sort(&polishly)
14:06 p6eval rakudo 1f662c: OUTPUT«a ą ć ę ł ń ó ś ź ż b c d e f g h i j k l m n o p q r s t u v w x y z␤»
14:06 masak \o/
14:06 masak GlitchMr: yes. they belong to different languages.
14:06 JimmyZ r: say 6.9 / 7.7
14:06 p6eval rakudo 1f662c: OUTPUT«0.896104␤»
14:06 masak mucker: it's a zero-width look-ahead.
14:08 Timbus did uh
14:08 Timbus did that actually sort
14:08 mucker i basically want to match any characters before END .. so i'm thinking the regex should look like <!before END <.+>>
14:08 masak oh ye of little faith :)
14:08 GlitchMr <!before > is (?! ) in Perl 5 (if I think correctly)
14:08 masak r: my $polish-alphabet = 'aąbcćdeęfghijklłmnńoópqrsśtuvwxyzźż'; sub polishly($c) { $polish-alphabet.index($c) }; say ('b', 'c', 'a', 'd' .. 'z', <ą ć ę ł ń ó ś ź ż>).sort(&polishly)
14:08 p6eval rakudo 1f662c: OUTPUT«a ą ć ę ł ń ó ś ź ż b c d e f g h i j k l m n o p q r s t u v w x y z␤»
14:08 masak \o/
14:09 masak oh, but it still puts the weird chars wrong :/
14:09 Timbus so they all just come before 'b'
14:09 GlitchMr That sort is weird
14:09 masak no :(
14:09 * masak can't see immediately what's wrong
14:10 masak r: my $polish-alphabet = 'aąbcćdeęfghijklłmnńoópqrsśtuvwxyzźż'; sub polishly($c) { $polish-alphabet.index($c) }; say ('a' .. 'z', <ą ć ę ł ń ó ś ź ż>)>>.&polishly
14:10 p6eval rakudo 1f662c: OUTPUT«0 2 3 5 6 8 9 10 11 12 13 14 16 17 19 21 22 23 24 26 27 28 29 30 31 32 1 4 7 15 18 20 25 33 34␤»
14:11 masak that looks... right.
14:12 GlitchMr n: my $polish-alphabet = 'aąbcćdeęfghijklłmnńoópqrsśtuvwxyzźż'; sub polishly($c) { $polish-alphabet.index($c) }; say ('b', 'c', 'a', 'd' .. 'z', <ą ć ę ł ń ó ś ź ż>).sort(&polishly)
14:12 p6eval niecza v19-15-g051783d: OUTPUT«d e f g h i j k l m n o p q r s t u v w x y z ą ć ę ł ń ó ś ź ż a b c␤»
14:12 GlitchMr huh?
14:12 masak there's something I'm missing here.
14:12 GlitchMr r: ('b', 'c', 'a', 'd' .. 'z', <ą ć ę ł ń ó ś ź ż>).flat.sort: {'aąbcćdeęfghijklłmnńoópqrsśtuvwxyzźż'.index($_)}
14:12 p6eval rakudo 1f662c:  ( no output )
14:13 GlitchMr r: my $polish-alphabet = 'aąbcćdeęfghijklłmnńoópqrsśtuvwxyzźż'; sub polishly($c) { $polish-alphabet.index($c) }; say ('b', 'c', 'a', 'd' .. 'z', <ą ć ę ł ń ó ś ź ż>).flat.sort(&polishly)
14:13 p6eval rakudo 1f662c: OUTPUT«a ą b c ć d e ę f g h i j k l ł m n ń o ó p q r s ś t u v w x y z ź ż␤»
14:13 masak GlitchMr++
14:13 masak yeah. there was a .flat missing.
14:13 masak I'm surprised it's needed.
14:14 GlitchMr I guess that Rakudo sorted by first character in list for some reason
14:14 sneex joined #perl6
14:14 GlitchMr karma GlitchMr
14:14 aloha GlitchMr has karma of 12.
14:15 GlitchMr Good enough I guess :)
14:15 Timbus r: say ('b', 'c', 'a', 'd' .. 'z', <ą ć ę ł ń ó ś ź ż>).perl
14:15 p6eval rakudo 1f662c: OUTPUT«("b", "c", "a", "d".."z", ("ą", "ć", "ę", "ł", "ń", "ó", "ś", "ź", "ż"))␤»
14:15 masak that looks right.
14:15 Timbus just m,aking sure
14:16 GlitchMr It's not Perl 5 where lists are automatically flattened when possible
14:17 GlitchMr In Perl 5, my %hash = (key => do_something()) always calls do_something in list context
14:17 GlitchMr Potentially causing warnings about missing value for key
14:19 GlitchMr Usually scalar do_something() or [do_something()] or {do_something()} is what you want
14:23 cognominal joined #perl6
14:23 cognominal joined #perl6
14:30 Chillance joined #perl6
14:31 tokuhiro_ joined #perl6
14:31 GlitchMr perl6: my hash = <a b c d>; print "%a{}{}[]{}";
14:31 p6eval rakudo 5e1b9a: OUTPUT«===SORRY!===␤Malformed my␤at /tmp/gHUsgsPaoi:1␤»
14:31 p6eval ..niecza v19-15-g051783d: OUTPUT«[31m===[0mSORRY![31m===[0m��Malformed my at /tmp/3SQaGq_ZML line 1:�------> [32mmy[33m�[31m hash = <a b c d>; print "%a{}{}[]{}";[0m��Parse failed��»
14:31 GlitchMr perl6: my %hash = <a b c d>; print "%hash{}{}[]{}";
14:31 p6eval rakudo 5e1b9a: OUTPUT«a b c     d»
14:31 p6eval ..niecza v19-15-g051783d: OUTPUT«a        b␤c     d␤»
14:32 GlitchMr std: my %hash = <a b c d>; print "%hash{}{}[]{}";
14:32 p6eval std f43a358: OUTPUT«ok 00:00 42m␤»
14:32 Chat3354 joined #perl6
14:39 nwc10 joined #perl6
14:39 nwc10 what is the 'soft' pragma?
14:41 jnthn nwc10: See 2889 on in S06
14:42 masak S06:2889 # creates a link in the irclogs
14:42 nwc10 thanks
14:42 brrt joined #perl6
14:43 nwc10 gosh, I think I must have read that at some time, maybe years ago
14:43 nwc10 as I remember this phrase: Be aware that this may turn your optimizer into more of a "pessimizer".
14:44 nwc10 is it Christmas yet? :-/
14:45 jnthn It's *August* :P
14:49 mucker is this a right regex  <!before END <.+>> ??
14:49 mucker before END match anychars
14:49 jnthn That doesn't look like it's doing what you wish.
14:50 mucker :(
14:50 jnthn In fact, it really ain't
14:50 jnthn What's an example of the string you're trying to match?
14:51 mucker "foo bar baz END" ... I want to match "foo bar baz "
14:51 masak sounds more like you want to set up a goal.
14:52 jnthn Dunno goal matching is gonna work out too well for this.
14:52 jnthn Maybe...
14:52 masak r: say "foo bar baz END" ~~ / '' ~ 'END' [\w+ % [\s+]] /
14:52 p6eval rakudo 5e1b9a: OUTPUT«Unable to parse , couldn't find final 'END'␤  in any FAILGOAL at src/stage2/QRegex.nqp:862␤  in regex  at /tmp/a1AOYX8xFc:1␤  in method ACCEPTS at src/gen/CORE.setting:9884␤  in block  at /tmp/a1AOYX8xFc:1␤␤»
14:52 jnthn r: say "foo bar baz END" ~
14:52 p6eval rakudo 5e1b9a: OUTPUT«===SORRY!===␤Confused␤at /tmp/hakFlunXjw:1␤»
14:52 jnthn argh
14:52 masak r: say "foo bar baz END" ~~ / '' ~ 'END' [<!before END> \w+ % [\s+]] /
14:52 p6eval rakudo 5e1b9a: OUTPUT«Unable to parse , couldn't find final 'END'␤  in any FAILGOAL at src/stage2/QRegex.nqp:862␤  in regex  at /tmp/xQxbJyjH5m:1␤  in method ACCEPTS at src/gen/CORE.setting:9884␤  in block  at /tmp/xQxbJyjH5m:1␤␤»
14:52 jnthn r: say "foo bar baz END" ~~ /.+? <before END>/
14:52 p6eval rakudo 5e1b9a: OUTPUT«q[foo bar baz ]␤ before => q[]␤␤»
14:53 jnthn r: say "foo bar baz END" ~~ /.+? <?before END>/
14:53 p6eval rakudo 5e1b9a: OUTPUT«q[foo bar baz ]␤␤»
14:53 jnthn That may be one way
14:53 masak yeah, but it backtracks a lot. :/
14:53 jnthn Relies on backtracking though.
14:53 masak (forwards)
14:53 mucker Ah ! thnkx was confused about the before's syntax
14:53 nwc10 left #perl6
14:54 jnthn Curiously, it is declarative.
14:54 masak not so curious, I think.
14:54 jnthn oh, hmm, maybe anyway
14:54 masak in the best case, its contents is just more DFA :)
14:54 jnthn But yeah, it ain't so efficient.
14:55 mucker also whats the p6 equivalent of [^\d]
14:56 masak <-[\d]>
14:57 jnthn \D
14:57 masak heh.
14:58 masak suggestion for a bot: p5re-to-p6re-bot.
14:58 masak (maybe with a better name) :)
14:58 GlitchMr Hmmm... that sounds interesting
14:58 GlitchMr Perhaps I should implement this as part of my IRC bot
14:59 masak good luck :)
14:59 GlitchMr Well, if I would parse Perl 5 regexpes first :P.
14:59 GlitchMr Regexpes like: /$a)/ could mean anything
15:00 masak I could imagine implementing something like that with someone as a weekend hackathon thing.
15:03 GlitchMr I mean, Perl is already too clever for its good when it sees any variables in RegExp
15:03 GlitchMr /$length[2345]/
15:03 GlitchMr ^ what this does?
15:04 GlitchMr And what this does?
15:04 GlitchMr /$length[23]/
15:05 masak the main task of such a bot would not be to get it right each time, but to be informative.
15:06 GlitchMr I mean, $length[2345] is variable $length followed by character class [2345]
15:06 GlitchMr $length[23] is variable $length[23]
15:06 masak for example, both of the parentheses-to-indices systems in p5 and p6 are statically knowable, but in different ways.
15:07 masak GlitchMr: oh? what's the difference? just the size of the number?
15:07 GlitchMr Yes
15:07 masak what's the rule?
15:07 GlitchMr I've no idea
15:09 GlitchMr I read about this in... perldiag
15:09 GlitchMr Ambiguous use of %c{%s[...]} resolved to %c%s[...]
15:12 u_u joined #perl6
15:12 GlitchMr "Interpolation in patterns has several quirks: $| , $( , $) , @+ and @- are not interpolated, and constructs $var[SOMETHING] are voted (by several different estimators) to be either an array element or $var followed by an RE alternative."
15:13 GlitchMr Voted?
15:13 GlitchMr VOTED?
15:13 GlitchMr I know about DWIM, but isn't this going too far?
15:14 brrt1 joined #perl6
15:14 masak wrong channel to be upset about this :)
15:16 GlitchMr I wonder how it is possible that Perl 5 uses like everything EXCEPT for spaces for DWIM.
15:17 masak "Perl 5 users" isn't one coherent group with just one opinion.
15:18 huf the solution is simple! let's give : and {/} yet another meaning!
15:18 GlitchMr Now I wonder how it is voted :P.
15:19 nodmonkey joined #perl6
15:23 brrt joined #perl6
15:26 GlitchMr http://paste.uk.to/c6569207
15:26 GlitchMr Why I have so pointless utilities on my PC...
15:28 cognominal_ joined #perl6
15:34 brrt joined #perl6
15:46 SamuraiJack joined #perl6
15:50 birdwindupbird How can one get AST for some statements?
15:50 masak could you provide more of the context in which you're asking this question?
15:54 birdwindupbird Sorry. I am trying to golf why LEAVE phaser do not fire in multi sub. I want to see AST of such sub. I am trying 'perl6 --target=past -e "my $k = 1; $k++". But get 'Method '__dump' not found for invocant of class 'Any'. May be i am doing it wrong way?
15:57 masak last I heard, --target=past was never really updated for the nom/6model branches.
15:57 masak so it's inaccessible except for perhaps really simple stuff, if that.
15:59 birdwindupbird Ok. May be there are other means to look into AST?
16:00 snearch joined #perl6
16:00 masak no. the feature doesn't seem to be a priority among those who can implement it (jnthn++ and pmichaud++, basically).
16:01 jnthn Well, thing is that phasers are (a) declarative and (b) fired by something in perl6.ops.
16:01 masak which I somewhat understand. it's not on a critical path.
16:01 jnthn I'm doubtful that even if --target=past worked it would yield too many clues.
16:02 jnthn r: multi foo() { 1; LEAVE { say 'bai' } }; foo();
16:02 p6eval rakudo 5e1b9a:  ( no output )
16:02 jnthn r: multi foo(:$x) { 1; LEAVE { say 'bai' } }; foo();
16:02 p6eval rakudo 5e1b9a:  ( no output )
16:02 MayDaniel joined #perl6
16:02 jnthn r: multi foo(:$x) { 1; ENTER { say 'hi' } }; foo();
16:02 p6eval rakudo 5e1b9a: OUTPUT«hi␤»
16:03 masak submitted on 2012-06-07 by moritz++: https://rt.perl.org/rt3/Ticket/Display.html?id=113548
16:05 birdwindupbird jnthn: Thanks. I am just trying to find my way into improvement of perl6. My steps seem incompetent and naive, i know. :)
16:06 masak birdwindupbird++
16:07 birdwindupbird First "plus-plus". It's so inspiring...
16:08 GlitchMr karma birdwindupbird
16:08 aloha birdwindupbird has karma of 1.
16:08 jnthn Yeah, I've little idea why we've problems with LEAVE phasers (I guess UNDO and KEEP also) in multis...it's a curious bug.
16:09 jnthn I wonder if it's anything to do with the callframe re-use that happens in enter_multi_dispatch op.
16:22 dalek nqp/toqast: 2637cf1 | jnthn++ | src/NQPQ/World.pm:
16:22 dalek nqp/toqast: Create scopes as QAST::Block.
16:22 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/2637cf1e3d
16:22 dalek nqp/toqast: 8ef56a3 | jnthn++ | src/NQPQ/Actions.pm:
16:22 dalek nqp/toqast: Update a couple other block creations to use QAST.
16:22 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/8ef56a31d3
16:22 dalek nqp/toqast: 18cec82 | jnthn++ | src/NQPQ/ (2 files):
16:22 dalek nqp/toqast: Update comp_unit to start using QAST::CompUnit, and tackle some of the other bits of comp_unit updating.
16:22 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/18cec8216a
16:22 dalek nqp/toqast: d579e84 | jnthn++ | src/NQPQ/World.pm:
16:22 dalek nqp/toqast: Update compile-time binding of lexicals.
16:23 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/d579e8410e
16:23 dalek nqp/toqast: a05823a | jnthn++ | src/NQPQ/Actions.pm:
16:23 dalek nqp/toqast: Pull in CTXSAVE override (will update the default one in HLL::Actions after QAST migration).
16:23 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/a05823abf7
16:27 xinming joined #perl6
16:28 dalek nqp/toqast: 476dd8e | jnthn++ | src/NQPQ/Actions.pm:
16:28 dalek nqp/toqast: Correct :hll copy-pasto.
16:28 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/476dd8e222
16:28 dalek nqp/toqast: 8440713 | jnthn++ | src/NQPQ/Actions.pm:
16:28 dalek nqp/toqast: Update argument list processing for QAST.
16:28 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/8440713c28
16:28 dalek nqp/toqast: 10f4753 | jnthn++ | src/NQPQ/Actions.pm:
16:28 dalek nqp/toqast: Update numeric literal handling. Now with no setting/regex lib loading, nqp::say(42) works.
16:28 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/10f4753f6e
16:57 phenny joined #perl6
17:07 cognominal joined #perl6
17:11 dalek nqp/toqast: b9f1e11 | jnthn++ | src/NQPQ/Actions.pm:
17:11 dalek nqp/toqast: Add an EXPR that produces QAST.
17:11 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/b9f1e11787
17:11 dalek nqp/toqast: a5e4e27 | jnthn++ | src/NQPQ/Grammar.pm:
17:11 dalek nqp/toqast: Update most of the operators to use :op instead of :pirop/:pasttype.
17:11 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/a5e4e27baa
17:11 dalek nqp/toqast: c945388 | jnthn++ | src/NQPQ/Actions.pm:
17:12 dalek nqp/toqast: Couple of fixes so if/unless now work again.
17:12 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/c945388cb0
17:12 dalek nqp/toqast: cd9fe1d | jnthn++ | src/QAST/Operations.nqp:
17:12 dalek nqp/toqast: Add an nqp::isfalse to go with nqp::istrue; NQP wants it.
17:12 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/cd9fe1dd46
17:12 dalek nqp/toqast: 375dfd0 | jnthn++ | src/NQPQ/ (2 files):
17:12 dalek nqp/toqast: Get setting/module loading to work again.
17:12 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/375dfd0b60
17:12 dalek nqp/toqast: 753dc29 | jnthn++ | src/NQPQ/Actions.pm:
17:12 dalek nqp/toqast: Get strings working. With this, we pass a few of the simple tests in t/nqp again.
17:12 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/753dc29e0e
17:41 cognominal_ joined #perl6
17:59 jaldhar joined #perl6
18:15 thou joined #perl6
18:15 jaldhar joined #perl6
18:15 moritz nr: say (\1).WHAT
18:15 p6eval rakudo 5e1b9a, niecza v19-15-g051783d: OUTPUT«Capture()␤»
18:15 moritz nr: say (\"foo").uc
18:15 p6eval rakudo 5e1b9a: OUTPUT«No such method 'uc' for invocant of type 'Capture'␤  in block  at /tmp/rZ0YnP39Td:1␤␤»
18:15 p6eval ..niecza v19-15-g051783d: OUTPUT«Unhandled exception: Unable to resolve method uc in type Capture␤  at /tmp/zGzqinNY8A line 1 (mainline @ 4) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3929 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3930 (module-CORE @ 564) ␤  at /home…
18:17 moritz nr: say Capture ~~ Positional
18:17 p6eval rakudo 5e1b9a: OUTPUT«False␤»
18:17 p6eval ..niecza v19-15-g051783d: OUTPUT«True␤»
18:17 moritz nr: say Capture ~~ Associative
18:17 p6eval rakudo 5e1b9a: OUTPUT«False␤»
18:17 p6eval ..niecza v19-15-g051783d: OUTPUT«True␤»
18:17 * moritz agrees with niecza here
18:17 moritz pmichaud, jnthn: is there a good reason that Capture is neither Positional nor Associative in Rakudo?
18:18 moritz it implements at_pos and at_key
18:19 moritz nr: say (\(1, a => 2)).list.perl
18:19 p6eval rakudo 5e1b9a: OUTPUT«(1,).list␤»
18:19 p6eval ..niecza v19-15-g051783d: OUTPUT«(1, ).list␤»
18:19 moritz nr: say (\(1, a => 2)).hash.perl
18:19 p6eval rakudo 5e1b9a: OUTPUT«EnumMap.new("a", 2, )␤»
18:19 p6eval ..niecza v19-15-g051783d: OUTPUT«{"a" => 2}.hash␤»
18:19 masak yeah, I agree with Niecza too.
18:20 masak just as Match is Positional & Associative
18:20 moritz nr: say Match ~~ Associative
18:20 p6eval rakudo 5e1b9a: OUTPUT«False␤»
18:20 p6eval ..niecza v19-15-g051783d: OUTPUT«True␤»
18:20 moritz :-)
18:22 cj joined #perl6
18:23 moritz nr: say (\ a => 1).perl
18:23 p6eval niecza v19-15-g051783d: OUTPUT«"a" => 1␤»
18:23 p6eval ..rakudo 5e1b9a: OUTPUT«===SORRY!===␤You can't backslash that at line 2, near " a => 1).p"␤»
18:25 masak r: my $q = quasi { {{{$a}}} }
18:25 p6eval rakudo 5e1b9a: OUTPUT«===SORRY!===␤Variable $a is not declared␤at /tmp/ly_Tyc8lal:1␤»
18:26 masak r: my $a; my $q = quasi { {{{$a}}} }
18:26 p6eval rakudo 5e1b9a:  ( no output )
18:26 masak hm. that probably should compile but not run.
18:26 masak r: my $a; my $q = quasi { {{{$a}}} }; say $q.^name
18:26 u_u joined #perl6
18:26 p6eval rakudo 5e1b9a: OUTPUT«AST␤»
18:27 moritz r: my $a; my $q = quasi { {{{$a}}} }; say $q.perl
18:27 p6eval rakudo 5e1b9a: OUTPUT«AST.new()␤»
18:28 mikemol joined #perl6
18:28 moritz ok, here's a weirdness
18:28 moritz capture literals are made with \
18:28 moritz s/capture/Capture/
18:29 moritz and then there are Parcel and Capture parameters
18:29 moritz and they use | and \
18:29 moritz but it's the | that makes the Capture
18:29 moritz isn't that... totally confusing?
18:31 masak yes.
18:31 masak I've never thought about that before.
18:32 moritz I realize that there's an analogy to | on the caller side
18:32 moritz f(|$c) interpolates whatever is in |$c in the argument list
18:32 moritz and   sub f(|$c) { } puts whatever comes in into $c
18:38 moritz nr: say (\(1, a => 2)).elems
18:38 p6eval rakudo 5e1b9a, niecza v19-15-g051783d: OUTPUT«1␤»
18:38 moritz erm, what?
18:38 moritz why does it count only positional elems?
18:39 dalek doc: 447afc8 | moritz++ | lib/Capture.pod:
18:39 dalek doc: document Capture
18:39 dalek doc: review: https://github.com/perl6/doc/commit/447afc88e5
18:40 masak moritz: because that's what the list nature does?
18:40 TimToady because it's a Capture, not a Parcel
18:41 TimToady Capture is after named args have been pulled out separately
18:42 jnthn I suspect Capture not doing Positional and Associative is an accident.
18:42 masak I have a question about unquotes. do they run ASAP on surrounding block entry, or ALAP when evaluating the quasiquote?
18:42 * masak guesses ALAP
18:42 masak ...in resonance with how qq strings work.
18:43 TimToady I'd think ALAP would be more useful
18:43 masak yeah.
18:43 masak that's probably what people expect.
18:43 masak I mean, you might want to synthetically create more ASTs in the macro body.
18:43 masak and then unquote them.
18:44 * jnthn expects ALAP too
18:44 moritz nr: say (\(1, 2, b => 3)).hash
18:44 p6eval niecza v19-15-g051783d: OUTPUT«{"b" => 3}␤»
18:44 p6eval ..rakudo 5e1b9a: OUTPUT«EnumMap.new("b", 3, )␤»
18:44 * moritz doesn't understand "Capture is after named args have been pulled out separately"
18:44 moritz the named args are still in the Capture, no?
18:45 TimToady but they are no longer part of masak's "list nature"
18:47 moritz so far I've thought of Hash.elems and List.elems in equal terms
18:47 cognominal joined #perl6
18:48 moritz but I guess one could say that Hash.elems coerces the hash to list and then obtains the number of elements, at least notionally
18:48 tadzik \o/
18:48 TimToady nr: 'foobar' ~~ /(...) $<bar>=[...]/; say $/.elems
18:48 p6eval rakudo 5e1b9a, niecza v19-15-g051783d: OUTPUT«1␤»
18:48 TimToady same deal
18:49 * moritz nods
18:49 masak tadzik: \o/
18:49 moritz \o tadzik
18:50 masak so. ALAP.
18:50 masak next question: on which hook, conceptually, is the code that does the unquote AST stiching-in hung?
18:51 masak at parse time there's static quasis. but the thing that gets returned from a macro or assigned to a variable is an incarnated quasi.
18:52 masak on which hook does the code hang that incarnates a static quasi?
18:53 masak hm. I guess it's just a question of code generation, actually.
18:53 jnthn Isn't it just when we hit the place wher the quasi is in the code?
18:53 masak indeed.
18:54 moritz just like with strings with interpolation, no?
18:54 masak yes, just so.
18:54 * masak goes to find the piece of code that generates quasis today
18:54 moritz ok, making Capture does Positional does Associative causes some spectest fallout
18:55 moritz http://perlpunks.de/paste/show/501d701b.584.228
18:55 moritz t/spec/integration/error-reporting.rakudo seems to be caused by the inliner, not by that patch
18:56 moritz hash.t and splicte.rakudo hung
19:02 moritz huh
19:03 moritz make t/spec/S32-array/splice.t   say "Non-zero wait status: 11"
19:03 moritz ./perl6 -Ilib t/spec/S32-array/splice.rakudo; echo $?
19:03 moritz 0
19:04 masak r: macro foo { my $a = {}; quasi { say $a.WHICH } }; foo; foo
19:04 p6eval rakudo 5e1b9a: OUTPUT«Hash|-1115445535␤Hash|-1108920089␤»
19:04 moritz r: say Match.^mro
19:04 p6eval rakudo 5e1b9a: OUTPUT«Match() Capture() Cool() Any() Mu()␤»
19:04 moritz ah, and the hang in t/spec/S02-types/hash.t is for using a Match object as a hash index
19:05 moritz which tries to slice because it's a Positional
19:05 moritz and then loops, because the iteration still returns Match
19:05 fgomez joined #perl6
19:06 moritz might be solved with special-casing
19:07 moritz r: say 1; ␤␤␤ +Any
19:07 p6eval rakudo 5e1b9a: OUTPUT«1␤use of uninitialized value of type Any in numeric context  in block  at /tmp/epqfuAfH1b:1␤␤»
19:07 masak yeah. the unquote stitching code should go here: https://github.com/rakudo/rakudo/blob/nom/src/core/AST.pm#L11
19:08 moritz jnthn: the line number for the warning is wrong; that's rather new, and causes a test failure in error-reporting.t
19:08 masak right now all that happens there is that the appropriate context gets set.
19:09 moritz probably needs to generate code that does the unquoting
19:09 moritz or splicing
19:09 moritz or... /me confuses all the terms right now
19:09 jnthn moritz: Hmm...curious.
19:10 moritz jnthn: does the default optimization level inline stuff?
19:11 jnthn Yeah
19:11 jnthn But I'm not sure what it'd inline there
19:11 masak moritz: well, there is already code generated to call that method. I think that's all that's needed.
19:11 jnthn Oh...
19:11 moritz jnthn: the prefix:<+>?
19:11 jnthn The + I guess
19:11 moritz jnthn: --optimize=0 fixes the line number
19:11 jnthn --optimize=1 too?
19:11 moritz with =1 it's also correct
19:11 cognominal joined #perl6
19:13 jnthn moritz: Does https://gist.github.com/3259432 help?
19:13 p6rd joined #perl6
19:13 moritz p6rd exited after 3548m26.170s with "Terminated."
19:14 masak amy idea why?
19:14 moritz jnthn: will test
19:14 masak any*
19:15 jnthn Given that it was terminated, I blame Arnold Schwarzenegger
19:15 moritz masak: it might have leaked memory and exceeded the ulimit
19:15 moritz masak: or could be some memory corruption
19:15 moritz p6rd: next
19:15 p6rd parrot 2012-08-21, rakudo 2012-08-23, niecza 2012-08-27
19:16 tadzik oh, I get to release Rakudo during the YAPC
19:17 jnthn ooh :)
19:17 jnthn Do it in a lightning talk ;-)
19:17 tadzik tee-hee
19:17 tadzik well, if you bring your PC...
19:17 moritz jnthn: huh, your patch applied with offset -50 lines, and then it complained "Symbol '$inlined' not predeclared in inline_immediate_block"
19:17 jnthn wtf!
19:17 jnthn It was meant to go in inline_call
19:18 masak git... making a mistake? surely not!
19:18 mberends joined #perl6
19:18 jnthn I know...it normally gits everything right.
19:19 moritz i applied with patch(1), not git apply
19:19 moritz but I think I see where it should go
19:19 masak nice to see some other implementation of *anything* saying "It's mainly an architectural change with little visible benefit, but it will be a big improvement in terms of flexibility and modularity inside the compiler.": http://hackage.haskell.org/trac/ghc/blog/newcg-update
19:19 arnsholt jnthn: Are you sure there's enough time to do spectests and everything during a lightning talk? =)
19:19 jnthn arnsholt: Probably not even on my main dev machine, actually...
19:20 arnsholt Yeah. I might be able to get it done if I do TEST_JOBS=8 on my work machine
19:21 moritz I could let you remote into my 24core $work machine :-)
19:21 moritz but it's windows, and doesn't have a working C compiler :(
19:22 jnthn moritz: "Great, another re-write!"
19:22 jnthn s/moritz/masak/
19:25 arnsholt Oooooh, I could do it on our work big machine
19:25 arnsholt TEST_JOBS=32  >:D
19:26 jnthn o.O
19:26 arnsholt What can I say, we do comp.ling. Lots of data is our business =)
19:27 masak jnthn: yeah, can't those Haskell guys just finish one compiler according to spec and *release* the damn thing!? :P
19:27 moritz jnthn: your patch fixes the warning line number
19:27 arnsholt jnthn: Don't ask how much memory there is ;)
19:27 jnthn moritz: asbra! I can commit it, unless you already did.
19:27 moritz jnthn: go ahead
19:28 masak arnsholt: TEST_JOBS=32, sure, but t/spec/S32-io/IO-Socket-INET.t will still be the bottleneck :)
19:28 moritz arnsholt: 8GB... per core?
19:28 arnsholt moritz: Yeah =)
19:28 dalek rakudo/nom: 7e1b380 | jnthn++ | src/Perl6/Optimizer.pm:
19:28 dalek rakudo/nom: Don't lose line numbers when inlining; moritz++ for reporting.
19:28 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/7e1b380a25
19:29 jnthn Hm, this means we inline calls to + :D
19:29 arnsholt Oh, and maybe we could just put the bottleneck test first in spectest.data, masak =)
19:29 jnthn r: say 32 * 8
19:29 p6eval rakudo 5e1b9a: OUTPUT«256␤»
19:29 masak arnsholt: that might be a good idea anyway...
19:29 jnthn That's bigger than my darn SSD on thi smachine!
19:30 tadzik hehe, same here
19:30 arnsholt Well, it's a shared machine. I don't have it all to myself, really =)
19:30 * moritz can just offer 92GB for 24 cores. Should still be enough for ~30 parallel jobs
19:30 moritz easily
19:37 tadzik masak: Game::Crypt doesn't seem to pass its tests
19:38 tadzik 59 modules ok, 49 not ok (108 total) on new rakudo
19:40 masak tadzik: huh. I ran them locally, and they passed...
19:40 * masak runs them again
19:40 benabik joined #perl6
19:41 masak tadzik: all tests pass here. do I need to update my nom?
19:42 * masak does so
19:45 flussence joined #perl6
19:48 fgomez joined #perl6
19:48 masak tadzik: I now have an up-to-date installed nom. tests still pass.
19:48 moritz ==> Testing Game::Crypt
19:48 moritz t/crypt.t .. 1/? Could not find symbol '&NoExitThere'
19:48 masak :/
19:49 moritz masak:  do you have local modifications? or local modifications in Adventure::Engine?
19:49 moritz or unpushed commits?
19:50 japhb Or a pony?
19:50 masak no, there's still just the one commit there.
19:50 masak (in the Adventure::Engine repository)
19:51 masak I have X::Adventure::NoExitThere defined at line 143 of Adventure/Engine.pm
19:51 masak I have no idea what would be wrong on your setups. tadzik, moritz, suggestions welcome.
19:52 moritz masak: did you precompile Adventure::Engine?
19:52 masak no.
19:52 * moritz did
19:53 masak if precompilation breaks Adventure::Engine, then that's very bad.
19:54 moritz that seems to be exactly what happens
19:54 moritz removing the .pir files, t/crypt.t runs fine
19:55 * masak submits rakudobug
19:55 moritz might be an instance of https://rt.perl.org/rt3/Ticket/Display.html?id=112626
19:56 * masak looks
19:58 masak yes, that looks very similar.
20:06 masak well, maybe not.
20:06 masak that one is about declarations colliding which shouldn't be.
20:06 masak mine is about declarations missing.
20:06 masak of course, the underlying cause could still be the same.
20:07 masak I could golf and submit a separate rakudobug if that's deemed worthwhile.
20:07 masak or golf and add to #112626
20:10 masak g'ah! it's 2012, and precompilation breaks my modules in Rakudo. :(
20:11 MayDaniel joined #perl6
20:12 birdwindupbird joined #perl6
20:13 jnthn masak: We didn't *have* the serialization bit of precompilation until 2012.
20:14 jnthn And we all know how many people clamber to fix the issues in that area.
20:16 masak ...not many? just a guess.
20:17 * jnthn suspects the number of people who comitted to serialization.c is somewhere around 1. :)
20:17 masak jnthn: sorry about the overt frustration. I just want Perl 6 to be a reliable platform to build software on.
20:18 masak if these breakages happened in 2008, I wouldn't bat an eyelash.
20:19 jnthn Then try to find ways to increase the bus number in areas where it's too low to have enough people to get to "reliable" as quickly as you'd like.
20:20 masak yeah.
20:20 awwaiid joined #perl6
20:20 benabik joined #perl6
20:23 nwc10 joined #perl6
20:23 nwc10 jnthn++ # nom now uses only 56% of the RAM that the last release needed to compile the setting (on "my" machine)
20:24 nwc10 at this rate it will be using negative RAM by about October
20:25 jnthn Uh...I'm not sure extrapolation works like that, but awesome :)
20:26 * awwaiid uses the negative-ram to open up some faster-than-light time travel and goes back in time to now to by himself another coffee
20:27 awwaiid jnthn++ # thanks for the coffee
20:27 jnthn \o/
20:27 jnthn nwc10: That's...quite a big drop.
20:27 nwc10 yes. I wonder if my measurement is duff for some reason.
20:27 jnthn The ChangeLog quotes a value around 35%
20:27 flussence I came to ask a stupid question, but I ended up fixing it by making my code nicer...
20:28 nwc10 oh, that was a drop from the last release all the way to HEAD
20:28 dalek nqp/toqast: 729ce4d | jnthn++ | src/NQPQ/Actions.pm:
20:28 dalek nqp/toqast: Fix lexical variables.
20:28 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/729ce4d6d5
20:28 dalek nqp/toqast: c115641 | jnthn++ | src/NQPQ/Actions.pm:
20:28 dalek nqp/toqast: Fix empty blocks.
20:28 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/c115641f6b
20:28 dalek nqp/toqast: 0758db1 | jnthn++ | src/NQPQ/Grammar.pm:
20:28 dalek nqp/toqast: Spell defor correctly.
20:28 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/0758db1db4
20:28 dalek nqp/toqast: 50bd8cc | jnthn++ | src/NQPQ/Actions.pm:
20:28 dalek nqp/toqast: Eliminate many uses of viviself.
20:28 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/50bd8ccb2d
20:28 dalek nqp/toqast: bfd58e5 | jnthn++ | src/NQPQ/ (2 files):
20:28 dalek nqp/toqast: A bunch more PAST::Var => QAST::Var, fixing simple sub declarations along the way.
20:28 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/bfd58e5cb8
20:29 nwc10 but top says
20:29 nwc10 Mem: 2881M Active, 2534M Inact, 586M Wired, 100M Cache, 399M Buf, 1814M Free
20:29 nwc10 so it's not swapping
20:29 jnthn Nice
20:29 jnthn Well, the string de-dupe thing I did today could easily have helped.
20:29 nwc10 OK, to be fair that's top when I'm not running it - let me log in twice
20:34 nwc10 This is while running:
20:34 nwc10 96290 nicholas       1 118    0   905M   835M CPU0   0   5:18 100.00% perl6
20:35 nwc10 size just peaked at 10055M before it exited
20:35 nwc10 er, 1005M
20:35 TimToady whew
20:35 jnthn !
20:35 nwc10 so it's under a Gb now on (at least one) x86_64 FreeBSD machine
20:35 jnthn I don't think we've ever needed 10055M :)
20:36 nwc10 maybe on a system with 512 bit pointers
20:36 nwc10 actually, only 256 bit pointers for the most recent release
20:38 sorear good * #perl6
20:38 nwc10 it seems to be a very good * today
20:38 tadzik good star sorear
20:39 * jnthn hopes we can shave some more off yet :)
20:39 jnthn o/ sorear
20:39 nwc10 me too. but you have to hit diminishing returns at some point, don't you? :-(
20:41 jnthn Yeah
20:41 masak sorear! \o/
20:42 * nwc10 hopes someone else can check his figures aren't insanely wrong
20:44 dalek nqp/toqast: b1a033c | jnthn++ | src/NQPQ/Actions.pm:
20:44 dalek nqp/toqast: Eliminate multitype usage.
20:44 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/b1a033c63e
20:44 dalek nqp/toqast: 8f9d295 | jnthn++ | src/NQPQ/Actions.pm:
20:44 dalek nqp/toqast: Toss hopefully unrequired namespace call.
20:45 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/8f9d295a98
20:45 dalek nqp/toqast: 263e374 | jnthn++ | src/NQPQ/World.pm:
20:45 dalek nqp/toqast: s/subid/cuid/
20:45 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/263e374152
20:45 dalek nqp/toqast: e8e11ea | jnthn++ | src/NQPQ/World.pm:
20:45 dalek nqp/toqast: Fix method setup, which gets basic classes working again.
20:45 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/e8e11eae24
20:47 flussence I did stuff: https://github.com/flussence/perl6-Test-Corpus
20:48 flussence (so that I could be increasingly lazy writing tests)
20:50 masak flussence: nice!
20:50 masak flussence++
20:51 flussence .oO( now I just need to test it... )
20:52 nwc10 left #perl6
20:52 masak flussence: IO::Path has a .basename public attribute.
20:54 masak flussence: also, I'd write `Callable $test-block` as `&test` if I were you.
20:55 tadzik errr
20:55 flussence I'll make the Callable change, but I'm trying to make this work in rakudo+niecza simultaneously and niecza hasn't caught up on IO yet.
20:55 tadzik star: use JSON::Tiny; my $json = "\n [1, 2, 3"; say (from-json $json).perl
20:55 p6eval star 2012.07: OUTPUT«Any␤»
20:56 tadzik star: use JSON::Tiny; my $json = "[1, 2, 3"; say (from-json $json).perl
20:56 p6eval star 2012.07: OUTPUT«Unable to parse array, couldn't find final ']'␤  in any FAILGOAL at src/stage2/QRegex.nqp:831␤  in regex array at /home/p6eval/star/lib/parrot/4.6.0/languages/perl6/lib/JSON/Tiny/Grammar.pm:8␤  in regex TOP at /home/p6eval/star/lib/parrot/4.6.0/languages/perl6/lib/J…
20:56 tadzik star: use JSON::Tiny; my $json = "[1, 2, 3]"; say (from-json $json).perl
20:56 p6eval star 2012.07: OUTPUT«[1, 2, 3]␤»
20:56 tadzik moritz: is that intentional?
20:56 tadzik as in: leading \n breaks stuffs
20:56 dalek nqp/toqast: 24fb046 | jnthn++ | src/NQPQ/Actions.pm:
20:56 dalek nqp/toqast: Somewhat fix self, attribute access.
20:56 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/24fb0469ac
20:57 flussence ...and s/&/$/ breaks rakudo too, wtf
20:58 flussence t/expand.t .. Nominal type check failed for parameter '&func'; expected Callable but got Sub instead # https://github.com/flussence/Text-Tabs-Wrap/blob/spec-rewrite/t/expand.t
20:59 cognominal jnthn, what is a WVal in QAST?
20:59 jnthn World Value
20:59 jnthn As in, something we know at compile time.
21:00 jnthn flussence: Did you write something like Sub &foo?
21:01 flussence I'm trying to think why I wrote Callable there in the first place...
21:02 jnthn Note that Callable &foo is like Callable[Callable] $foo
21:02 flussence I guess I could just get rid of the type entirely, the & should be enough
21:02 jnthn Rigth
21:02 jnthn Having the type *and* & is probably messing it up for you
21:02 tadzik my %old = from-json slurp $old-results;
21:02 tadzik my %new = from-json slurp $old-results;
21:02 tadzik it looked so sensemaking...
21:02 cognominal thx jnthn++
21:03 masak r: say Sub ~~ Callable
21:03 p6eval rakudo 7e1b38: OUTPUT«True␤»
21:05 tadzik http://i.imgur.com/2lskK.png \o/
21:05 tadzik that may come in handy
21:06 jnthn tadzik: ooh!
21:06 jnthn tadzik++
21:07 tadzik it'll be more awesome. Stay tuned :)
21:07 tadzik speaking of which, which feather is a good place to put some resource-heavy operations to be running once in a while?
21:09 cognominal What is the 'soft' pragma?
21:10 jnthn cognominal: S06 explains it in some detail
21:10 cognominal ha, thx
21:10 jnthn It's about managing the interaction of wrap and inlining.
21:11 cognominal yup, I see that
21:12 dalek emmentaler: 34e1831 | tadzik++ | difftool:
21:12 dalek emmentaler: Add difftool for nice results comparing
21:12 dalek emmentaler: review: https://github.com/tadzik/emmentaler/commit/34e18311a1
21:14 RobiX joined #perl6
21:15 benabik joined #perl6
21:32 benabik joined #perl6
21:36 MayDaniel joined #perl6
21:53 cognominal joined #perl6
21:57 tadzik I like how it takes ~5 seconds to install some stuff with panda :)
21:59 * masak beats the big drum faster to hasten the Singularity
22:03 sorear o/ masak
22:04 masak \o
22:08 tadzik masak: I fixed Yapsi build and compiler.t, but there are things beyond my understanding there and in runtime.t
22:09 tadzik (https://github.com/tadzik/yapsi)
22:10 masak tadzik: oh, ok. thanks for the heads-up.
22:10 masak I might have a look tomorrow.
22:10 tadzik okay
22:11 * jnthn hasn't written a module for quite a while
22:11 jnthn Maybe I should try my hand at it sometime :)
22:11 tadzik random idea from today: Text::Colors
22:11 tadzik or ::Colours
22:11 tadzik something along Term::ANSIColor, but able to produce terminal stuff, maybe also HTML text and so
22:11 tadzik like a common interface for colo(u)ring text
22:12 jnthn tadzik++ # awareness of correct spelling as well as American spelling of words
22:12 jnthn ;-)
22:12 tadzik :)
22:13 tadzik that reminds me of one of Tim Minchin's pieces
22:14 masak reminds me of guerilla editing of Wikipedia.
22:14 jnthn .oO( gorilla editing of Wikipedia )
22:14 masak https://secure.wikimedia.org/wikipedia/meta/wiki/Guerilla_spelling_campaigns
22:14 jnthn oO( use GORILLA_EDITING; )
22:16 masak the best part of that page? it is categorized both as "Humour" and "Humor" :P
22:16 tadzik :D
22:22 tadzik No such method 'MATCH' for invocant of type 'Integer'
22:22 tadzik that sounds awful
22:25 jnthn Indeed. :/
22:25 jnthn How'd you get it?
22:26 tadzik use Grammar::Tracer on Yapsi.pm
22:26 tadzik gone away when I used <?after> instead of a hack that was there
22:27 tadzik https://github.com/tadzik/yapsi/blob/bf9cf97406df530/lib/Yapsi.pm#L31 here
22:28 jnthn r: say "ab" ~~ /<?{ $/.CURSOR.pos > 0 }> ./
22:28 p6eval rakudo 7e1b38: OUTPUT«q[b]␤␤»
22:29 jnthn oh, but using Grammar::Tracer
22:30 tadzik aye
22:35 benabik joined #perl6
22:37 masak tadzik++ # de-hacking Yapsi
22:38 tadzik masak: I made it pass all tests and things :)
22:38 tadzik well, almost
22:39 masak tadzik: you rock
22:39 tadzik well, almost
22:40 tadzik I finally stumbled upon something that looks like being not a change in Perl 6 itself and not a rakudo de-regression :)
22:45 tadzik yep, all tests pass now
22:45 jnthn Pass all the tests! \o/
22:46 tadzik REmove all the warnings!
22:46 tadzik (but one)
22:47 tadzik masak: pull request sent
22:52 tadzik now I feel like sleeping. Good knight #perl6!
22:53 jnthn 'night, tadzik
22:54 LadySainthilaire joined #perl6
22:55 LadySainthilaire hi
22:56 masak hilairious.
23:02 sftp joined #perl6
23:09 masak good knight, #perl6.
23:11 * jnthn gets some rest also
23:24 cognominal joined #perl6
23:35 sivoais joined #perl6
23:52 rvchangue joined #perl6

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

Perl 6 | Reference Documentation | Rakudo