Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2012-07-22

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 dalek doc: 02886c1 | raiph++ | lib/glossary.pod:
00:00 dalek doc: delete extraneous space
00:00 dalek doc: review: https://github.com/perl6/doc/commit/02886c137b
00:00 dalek doc: 7179ca6 | Matthew++ | lib/glossary.pod:
00:00 dalek doc: Merge pull request #3 from raiph/patch-1
00:00 dalek doc:
00:00 dalek doc: delete extraneous space
00:00 dalek doc: review: https://github.com/perl6/doc/commit/7179ca6bdd
00:02 raiph gfldex: the article u linked provides a very complex path in comparison to the trivial thing i did
00:04 raiph i just went to the repo, picked a file, clicked edit, edited, wrote change comment, clicked save. done.
00:05 raiph lue++ gfldex++
00:08 benabik github makes these things easy
00:17 crab2313 joined #perl6
00:20 gfldex was toqast merged tonight?
00:20 gfldex or is my brain running at halve speed? :)
00:21 timotimo wasn't toqast at 7/11 steps two days ago?
00:22 gfldex it feels like startup speed went up quite a bit
00:22 timotimo on my end, toqast isn't merged into nom yet
00:29 timotimo oh, actually it was step 9 of 11
00:29 timotimo http://6guts.wordpress.com/2012/07/20/the-rakudo-move-to-qast-progressing-nicely/ <- i guess you have already read this
00:35 dalek roast: 89bfdc5 | (Solomon Foster)++ | S05-interpolation/regex-in-variable.t:
00:35 dalek roast: Remove extraneous | from a Niecza skip line.
00:35 dalek roast: review: https://github.com/perl6/roast/commit/89bfdc5389
00:35 dalek roast: 18f3c2b | (Solomon Foster)++ | packages/Test/Util.pm:
00:35 dalek roast: Add #OK to silence Niecza warning.
00:35 dalek roast: review: https://github.com/perl6/roast/commit/18f3c2bdb2
00:50 cognominal joined #perl6
00:57 cognominal joined #perl6
01:06 [Coke] +# 07/21/2012 - rakudo++ (22933); niecza (89.76%); pugs (40.14%)
01:07 raiph benabik: and especially if you don't know git. with the web ui i just used u don't have to know anything other than create an account, go to an url, navigate to the file to change, click edit, edit in an ok editor, write a change comment, click save.
01:08 raiph benabik: no need to have a clue about git, version control, forks, pull requests, etc., etc.
01:09 raiph s/ in an ok editor//
01:29 adu joined #perl6
01:32 fridim_ joined #perl6
01:43 benabik joined #perl6
01:45 whiteknight joined #perl6
01:49 simcop2387 joined #perl6
01:54 larks joined #perl6
02:02 colomon joined #perl6
02:07 robinsmidsrod joined #perl6
02:11 orafu joined #perl6
02:24 drbean joined #perl6
02:36 Psyche^ joined #perl6
02:39 * araujo tests if perl6 (nqp) already compiles
02:40 jerome joined #perl6
02:40 drbean joined #perl6
02:54 kurahaupo joined #perl6
02:54 drbean joined #perl6
03:08 drbean joined #perl6
03:23 drbean joined #perl6
03:35 drbean joined #perl6
03:47 drbean joined #perl6
04:01 kaare_ joined #perl6
04:03 drbean joined #perl6
04:03 mucker joined #perl6
04:15 drbean joined #perl6
04:20 preflex joined #perl6
04:30 drbean joined #perl6
04:42 cognominal_ joined #perl6
04:44 drbean joined #perl6
04:55 icwiener joined #perl6
05:43 yeltzooo7 joined #perl6
05:43 SHODAN joined #perl6
05:43 masak joined #perl6
05:43 thelazydeveloper joined #perl6
05:43 pochi joined #perl6
05:43 Vlavv joined #perl6
05:43 integral joined #perl6
05:43 jlaire joined #perl6
05:43 [particle] joined #perl6
05:43 HarryS joined #perl6
05:43 hoelzro|away joined #perl6
05:43 arnsholt joined #perl6
05:43 estrai_ joined #perl6
05:43 buubot_backup joined #perl6
05:43 sporous joined #perl6
05:43 colomon joined #perl6
05:43 kaare_ joined #perl6
05:43 icwiener joined #perl6
05:43 daniel-s joined #perl6
05:43 mikemol joined #perl6
05:43 aloha joined #perl6
05:43 sivoais joined #perl6
05:43 phenny joined #perl6
05:43 lue joined #perl6
05:43 larks joined #perl6
05:43 orafu joined #perl6
05:43 cognominal_ joined #perl6
05:43 dju joined #perl6
05:43 Ulti joined #perl6
05:43 charsbar____ joined #perl6
05:43 pjcj joined #perl6
05:43 carlin joined #perl6
05:43 flussence joined #perl6
05:43 cj joined #perl6
05:43 geekosaur joined #perl6
05:43 Celelibi joined #perl6
05:43 rurban joined #perl6
05:43 sudokode joined #perl6
05:43 Timbus joined #perl6
05:43 BooK_ joined #perl6
05:43 telex joined #perl6
05:43 broquaint joined #perl6
05:43 Ex joined #perl6
05:43 benabik joined #perl6
05:43 simcop2387 joined #perl6
05:43 clkao joined #perl6
05:43 diakopter joined #perl6
05:43 literal joined #perl6
05:43 takesako_ joined #perl6
05:43 Juerd joined #perl6
05:43 rjbs joined #perl6
05:43 hanekomu joined #perl6
05:43 odoacre joined #perl6
05:43 jtpalmer joined #perl6
05:43 au joined #perl6
05:43 kcwu_ joined #perl6
05:43 Yappocall joined #perl6
05:43 sunnavy joined #perl6
05:43 hugme joined #perl6
05:43 krunen joined #perl6
05:43 ashleydev joined #perl6
05:43 felher joined #perl6
05:43 japhb joined #perl6
05:43 revdiablo joined #perl6
05:43 Helios joined #perl6
05:43 Maddingue joined #perl6
05:43 mj41 joined #perl6
05:43 lorn joined #perl6
05:43 frettled joined #perl6
05:43 mhasch joined #perl6
05:43 Gothmog_ joined #perl6
05:43 ingy joined #perl6
05:43 szbalint joined #perl6
05:43 `patch` joined #perl6
05:43 cxreg joined #perl6
05:43 eiro joined #perl6
05:43 PerlJam joined #perl6
05:43 sjn joined #perl6
05:43 Grrrr joined #perl6
05:43 DarthGandalf joined #perl6
05:43 eiro_ joined #perl6
05:43 stepnem joined #perl6
05:43 REPLeffect joined #perl6
05:43 araujo joined #perl6
05:43 Exodist joined #perl6
05:43 c1sung joined #perl6
05:43 baest joined #perl6
05:43 Tene joined #perl6
05:43 pnu joined #perl6
05:43 kst joined #perl6
05:43 gfldex joined #perl6
05:43 lumi_ joined #perl6
05:43 jnthn joined #perl6
05:43 mathw joined #perl6
05:43 sjohnson joined #perl6
05:43 renormalist joined #perl6
05:43 pmichaud joined #perl6
05:43 jfried joined #perl6
05:43 crazedpsyc joined #perl6
05:43 shachaf joined #perl6
05:43 awwaiid_ joined #perl6
05:43 Util_ joined #perl6
05:43 sorear joined #perl6
05:43 snarkyboojum joined #perl6
05:43 ranguard_ joined #perl6
05:43 jevin_ joined #perl6
05:43 zb joined #perl6
05:43 patspam joined #perl6
05:43 Khisanth joined #perl6
05:43 pyrimidine joined #perl6
05:43 prammer joined #perl6
05:43 silug joined #perl6
05:43 mikec_ joined #perl6
05:43 erkan joined #perl6
05:43 fgomez joined #perl6
05:43 skids joined #perl6
05:43 mtk joined #perl6
05:43 cotto joined #perl6
05:43 adu joined #perl6
05:43 preflex joined #perl6
05:43 drbean joined #perl6
05:43 mucker joined #perl6
05:43 wolong joined #perl6
05:43 risou_ joined #perl6
05:43 timotimo joined #perl6
05:43 imarcusthis joined #perl6
05:43 topologist joined #perl6
05:43 ivan`` joined #perl6
05:43 ilogger2 joined #perl6
05:43 moritz joined #perl6
05:43 tomaw joined #perl6
05:43 rhr joined #perl6
05:43 Guest7858 joined #perl6
05:43 bonsaikitten joined #perl6
05:43 spacebat_ joined #perl6
05:43 felipe joined #perl6
05:43 Patterner joined #perl6
05:43 xinming joined #perl6
05:43 cosimo joined #perl6
05:43 betterworld joined #perl6
05:43 libertyprime joined #perl6
05:43 ponbiki joined #perl6
05:43 itz joined #perl6
05:43 lestrrat joined #perl6
05:43 nuba joined #perl6
05:43 p6eval joined #perl6
05:43 PZt joined #perl6
05:43 BinGOs joined #perl6
05:43 nebuchadnezzar joined #perl6
05:43 rafl joined #perl6
05:43 eternaleye joined #perl6
05:43 TimToady joined #perl6
05:43 avar joined #perl6
05:43 [Coke] joined #perl6
05:43 _jaldhar joined #perl6
05:43 huf joined #perl6
05:43 alvis joined #perl6
05:43 Woodi joined #perl6
05:43 sirrobert joined #perl6
05:43 FireFly joined #perl6
05:43 Pleiades` joined #perl6
05:43 aindilis joined #perl6
05:43 plobsing joined #perl6
05:43 dalek joined #perl6
05:43 mattp_ joined #perl6
05:43 Bucciarati joined #perl6
05:43 mjflick joined #perl6
05:43 tadzik joined #perl6
05:43 BlueT_ joined #perl6
05:44 moritz \o
05:44 HarryS joined #perl6
05:49 daniel-s joined #perl6
05:52 robinsmidsrod joined #perl6
05:52 moritz phenny: tell raiph that he now has direct commit access to perl6/doc on github. Welcome!
05:52 phenny moritz: I'll pass that on when raiph is around.
05:58 adu \p
05:58 adu o/
06:00 daniel-s joined #perl6
06:08 JimmyZ_ joined #perl6
06:08 JimmyZ_ good afternoon
06:09 JimmyZ_ clear
06:17 moritz good clear afternon, JimmyZ_
06:18 dalek doc/type-graph: 28f5a6c | (Geoffrey Broadwell)++ | type-graph.txt:
06:18 dalek doc/type-graph: First snapshot: human- and computer-parseable map of Perl6's type graph
06:18 dalek doc/type-graph: review: https://github.com/perl6/doc/commit/28f5a6c8e5
06:19 japhb *yawn*
06:19 japhb Good *, all
06:20 moritz japhb: my approach would have been to parse the graph out of the .pod files
06:25 moritz but yours i probably easier to get started with
06:27 JimmyZ_ \o
06:31 japhb moritz, I thought about that, but figured A) this method would likely be more complete for a while yet, and B) after I finish cleaning the raw data, the next stage of my plan may require a fair amount of hand-tuning anyway.
06:33 japhb I plan to feed the massaged data to graphviz to produce not just an overall visualization of the type graph, but also type-specific ones that show just the relevant parts, so the mini-map images can be inserted into the HTML docs.
06:34 Tene very nice
06:34 japhb The hand-tuning will come in the form of layout hints, node coloring, etc. to make it more intuitive.
06:35 japhb Tene, Glad to know my idea seems sane to someone *else*.  ;-)
06:37 japhb BTW, anyone interested in reviewing/sanity-checking the current snapshot data is most welcome to do so.  :-)
06:42 moritz at some point we need the data in htmlify
06:43 birdwindupbird joined #perl6
06:43 moritz so that we can generate HTML files with methods from superclasses and roles
06:47 muixirt joined #perl6
06:47 muixirt good morning
06:50 japhb moritz, agreed
06:50 japhb OK, time for me to trundle off to slumber.
06:50 japhb Night, all.
06:50 moritz night japhb
06:50 japhb &
07:00 SamuraiJack joined #perl6
07:04 GlitchMr joined #perl6
07:11 DarthGandalf joined #perl6
08:07 MayDaniel joined #perl6
08:11 mucker joined #perl6
08:26 birdwind1pbird joined #perl6
08:49 dalek doc: 28f5a6c | (Geoffrey Broadwell)++ | type-graph.txt:
08:49 dalek doc: First snapshot: human- and computer-parseable map of Perl6's type graph
08:49 dalek doc: review: https://github.com/perl6/doc/commit/28f5a6c8e5
08:49 dalek doc: e1eaa4c | moritz++ | / (2 files):
08:49 dalek doc: add Perl6::TypeGraph, wihch reads type-graph.txt and constructs a -- you have already guessed it -- type graph
08:49 dalek doc: review: https://github.com/perl6/doc/commit/e1eaa4c5cc
08:50 GlitchMr perl6: say AST.new ~~ Mu
08:50 p6eval niecza v19-13-g442e075: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤Undeclared name:â�¤        'AST' used at line 1â�¤â�¤Unhandled exception: Check failedâ�¤â�¤  at /home/p6eval/niecza/boot/lib/CORE.setting line 1402 (die @ 5) â�¤  at /home/p6eval/niecza/src/STD.pm6 line 1147 (P6.comp_unit @ 37) â�¤  at /home/p6…
08:50 p6eval ..rakudo 5b56cf: OUTPUT«True␤»
08:51 GlitchMr Why "AST" doesn't expand anything?
08:52 JimmyZ_ joined #perl6
08:55 moritz what should it expand to?
08:55 GlitchMr I don't know, but it looks that AST is Mu
08:55 GlitchMr But perhaps it's because rakudo
08:56 moritz well, any type is Mu
08:56 GlitchMr It also seems that AST is Any
08:56 moritz there#s no contradiction there, since Any ~~ Mu
08:57 moritz s/'#'/'/
08:58 GlitchMr perl6: /[]/
08:58 p6eval rakudo 5b56cf: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter ] (must be quoted to match literally) at line 2, near "/"␤»
08:58 p6eval ..niecza v19-13-g442e075: OUTPUT«[31m===[0mSORRY![31m===[0m��Null pattern not allowed at /tmp/3DR3cs5YlW line 1:�------> [32m/[[33m�[31m]/[0m��Parse failed��»
08:59 GlitchMr std: /[]/
08:59 p6eval std 3b49ac8: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Null pattern not allowed at /tmp/3nhfgzahuu line 1:â�¤------> [32m/[[33mâ��[31m]/[0mâ�¤    expecting regex atomâ�¤Parse failedâ�¤FAILED 00:00 41mâ�¤Â»
08:59 GlitchMr I guess that rakudo could have better error message, but whatever
08:59 moritz known NYI
09:01 masak lol good morning, #perl6
09:01 masak well, mid-antenoon.
09:01 moritz lolitsmasak!
09:02 * moritz has waffels with bjørnebær for breakfast
09:04 masak whoa.
09:05 * masak bolts to the store to buy björnbär
09:11 GlitchMr sub postfix:<¶>($a) { "$a\n" }
09:11 GlitchMr I have feeling I abuse mutable Perl 6 grammar
09:13 GlitchMr But I guess it's not as bad as infix:[] thing I've done once
09:22 moritz masak: visit me, and we'll make waffles together. And I'm sure we'll find some tasty fruits or berries too :-)
09:23 moritz and whipped cream, or vanilla ice cream
09:25 * moritz adds free wifi and macro discussions to sweaten the deal
09:28 * moritz is a bit proud of https://github.com/perl6/doc/commit/e1eaa4c5cc6a2cab24a4d656419cdda8243d1f63 -- not much code to nice effect
09:34 moritz next step: MRO computation
09:34 sergot joined #perl6
09:34 sergot hi o/
09:35 moritz \o sergot
09:43 fridim_ joined #perl6
09:44 MayDaniel joined #perl6
09:48 kaare_ joined #perl6
10:01 jnthn ohhai o/
10:10 jnthn moritz: (MRO) there's an NQP implementation of C3 handy
10:10 jnthn moritz: src/Perl6/Metamodel/C3MRO.pm
10:10 moritz jnthn: yes, I'm looking at that right now
10:10 jnthn ah :)
10:11 cognominal_ joined #perl6
10:26 whiteknight joined #perl6
10:27 moritz r: for [1, 2], [3, 4] -> @a { .say for @a }
10:27 p6eval rakudo 5b56cf: OUTPUT«1␤2␤3␤4␤»
10:27 zby_home_ joined #perl6
10:30 moritz r: my @b = [1, 2], [3, 4]; for @b -> @a { .say for @a unless @a =:= @b[0] }
10:30 p6eval rakudo 5b56cf: OUTPUT«===SORRY!===␤Missing block␤at /tmp/gMMtwpPqyx:1␤»
10:31 moritz r: my @b = [1, 2], [3, 4]; for @b -> @a { unless @a =:= @b[0] { .say for @a } }
10:31 p6eval rakudo 5b56cf: OUTPUT«1␤2␤3␤4␤»
10:31 moritz hm, that part doesn't work that easily in p6, it seems
10:34 pupoque joined #perl6
10:39 tokuhiro_ joined #perl6
10:46 pmurias joined #perl6
10:46 jnthn r: my @b = [1, 2], [3, 4]; for @b -> $a { unless $a =:= @b[0] { .say for @a } }
10:47 p6eval rakudo 5b56cf: OUTPUT«===SORRY!===␤Variable @a is not declared␤at /tmp/f2Omg_LZ8O:1␤»
10:47 jnthn r: my @b = [1, 2], [3, 4]; for @b -> $a { unless $a =:= @b[0] { .say for @$a } }
10:47 p6eval rakudo 5b56cf: OUTPUT«1␤2␤3␤4␤»
10:47 jnthn r: my @b = [1, 2], [3, 4]; for @b -> $a { unless $a === @b[0] { .say for @$a } }
10:47 p6eval rakudo 5b56cf: OUTPUT«3␤4␤»
10:48 moritz r: say Str.^mro
10:48 p6eval rakudo 5b56cf: OUTPUT«Str() Cool() Any() Mu()␤»
10:50 moritz r: say Match.^mro
10:50 p6eval rakudo 5b56cf: OUTPUT«Match() Capture() Cool() Any() Mu()␤»
10:51 cognominal__ joined #perl6
10:52 dalek doc: b81241b | moritz++ | / (2 files):
10:52 dalek doc: port C3 linearization from nqp
10:52 dalek doc: review: https://github.com/perl6/doc/commit/b81241b119
10:53 moritz Match seems to be about the only class in core that uses MI
10:54 moritz so, not many test cases
10:55 xinming joined #perl6
10:59 dalek doc: 791b30a | moritz++ | lib/Perl6/Type (2 files):
10:59 dalek doc: move Type to a separate file, re-style it as Perl6::Type
10:59 dalek doc: review: https://github.com/perl6/doc/commit/791b30a0f1
11:23 JimmyZ joined #perl6
11:24 JimmyZ what does TypeGraph do ?
11:25 moritz JimmyZ: computes method resolution order for classes
11:26 moritz JimmyZ: the goal is to create HTML pages for a class which contains method from superclasses and roles
11:26 JimmyZ got it, that's great
11:27 moritz it's going be some work :/
11:28 jnthn moritz++ # working on things that need some work
11:29 dalek nqp/toqast: f6a5c44 | jnthn++ | src/QAST/Compiler.nqp:
11:29 dalek nqp/toqast: Fix a couple of small code-gen bugs.
11:29 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/f6a5c44d55
11:29 dalek nqp/toqast: 9133c44 | jnthn++ | src/QAST/Compiler.nqp:
11:29 dalek nqp/toqast: Fix to unicode string handling.
11:29 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/9133c4425a
11:29 moritz windows-- just ejected a CD I was copying from (and I didn't tell it to eject it), and then complained that the device isn't ready for reading
11:29 jnthn o.O
11:29 jnthn fail
11:30 crab2313 joined #perl6
11:32 moritz oh "good" old windows XP
11:32 crab2313 joined #perl6
11:32 moritz if a windows explorer window dies, the window manager itself dies
11:36 PacoAir joined #perl6
11:47 seldon joined #perl6
12:16 masak of course. it's "integrated", it's good for you.
12:17 jnthn Of course, who'd want to work with a system where everything was derivative...
12:17 benabik My favorite part was always that the graphics toolkit was part of the kernel, so bad calls could bring down the entire system.
12:23 dalek nqp/toqast: e106957 | jnthn++ | src/QAST/Compiler.nqp:
12:23 dalek nqp/toqast: Restore capture_lex on block entry.
12:23 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/e106957f4b
12:23 dalek rakudo/toqast: b2e068f | jnthn++ | src/core/Regex.pm:
12:23 dalek rakudo/toqast: Fix a pirop sig that led to invalid code gen.
12:23 dalek rakudo/toqast: review: https://github.com/rakudo/rakudo/commit/b2e068f972
12:23 dalek rakudo/toqast: cb9e7de | jnthn++ | src/QPerl6/Actions.pm:
12:23 dalek rakudo/toqast: Fix rather over-eager whatever currying (it was currying stuff like $a ?? * !! $b).
12:23 dalek rakudo/toqast: review: https://github.com/rakudo/rakudo/commit/cb9e7de8e7
12:23 jnthn Seems I now have a QAST-compiled CORE.setting that vaguely works.
12:25 masak \o/
12:25 * masak pulls recent stuffs and rebuilds
12:25 jnthn Well, you don't get that without a lot of knowing what to do.
12:25 masak I just want to re-run the spectests.
12:25 masak I plan to address the failing macros tests now.
12:26 jnthn Right, but they won't run with the QAST-compiled CORE.setting yet
12:26 jnthn Which is fine, just saying. :)
12:28 masak rightright
12:29 aindilis joined #perl6
12:29 masak aindilis: hi.
12:30 masak oh, that was a re-join.
12:35 jnthn Spectesting with the QAST-compiled CORE.setting looks fairly good. A few issues.
12:36 moritz $ ack '\bPAST(?!::QAST)' src/QPerl6/|wc -l
12:36 moritz 64
12:38 masak kudos for even remembering the look-ahead syntax of Perl 5 regexes...
12:38 masak I never do. have to 'perldoc perlre' *every* time.
12:38 jnthn moritz: Note that the bottom of Actions has a bunch of PAST::Node usages that stay for the time being too.
12:38 jnthn (the regex section)
12:39 moritz masak: negative look-ahead is the only one I remember reliably
12:40 moritz oh, and it seems much of that is commented out
12:40 jnthn wtf, chmod hangs :)
12:40 moritz $ ack '^[^#]*\bPAST(?!::QAST)' src/QPerl6/|wc -l
12:40 moritz 33
12:41 jnthn ah, exception related...
12:44 MayDaniel joined #perl6
12:44 drbean joined #perl6
12:46 timotimo how hard will it be for rakudo to use type information/native types to run "as fast as C"? or how long will i have to wait until thath appens?
12:46 * masak .oO( mnemonic for negative look-ahead in Perl 5: '?!' is like the end of 'wtf?!', what you'd say if you found the thing there )
12:47 masak timotimo: question is so vague as to be almost meaningless.
12:47 timotimo that's what i feared :|
12:47 masak timotimo: http://www.perl.com/pub/2001/06/27/ctoperl.html
12:48 masak funnily enough, that article ends by referring to the Perl 6 design.
12:50 jnthn timotimo: It's difficult from many angles. Using native type info to generate better code is possible; we do that to a limited extent already in Rakudo. That's the kind of code that a JIT compiler can actually do quite a bit with, but Parrot has no JIT, so it'd seem being on a VM that has one is a pre-req for starting to get that sorta code faster. Even then I suspect it'd not be "as fast as C".
12:51 masak Occaml people sometimes boast that their code is faster than C because they do so much good type inference.
12:52 timotimo my int @a is dim(1000); - reminds me of basic :)
12:53 timotimo hm, i suppose i'm spoiled by pypy in this regard, which comes with a pretty impressive jit
12:53 JimmyZ timotimo: Parrot has a plan to get a jit too :)
12:54 timotimo JimmyZ: do you have a guess how long it'll take until the first bits of jitted code would run properly?
12:56 JimmyZ timotimo: I'm sure It'll take a long time until somebodies continue coding to it
12:57 timotimo :|
13:02 timotimo is thorough type analysis of perl6 programs feasible?
13:03 timotimo i've looked at friends who do coq and agda with a bit of envy. i don't understand much of it, but would i be able to do similar checking of type constraints etc. in perl6?
13:05 timotimo i don't even know if what i'm asking makes any sense
13:08 jnthn Perl 6 is gradually typed, meaning that the compiler will do what it can with the information you give it. It may be able to make extra inferences also.
13:08 moritz but there are certain type errors it can (almost?) never complain about
13:09 moritz like  sub f(MyT $x) { $x.mymethod ]
13:09 moritz erm
13:09 moritz like  sub f(MyT $x) { $x.mymethod }
13:09 moritz is valid even if MyT doesn't define a 'mymethod'
13:09 timotimo right, because you could put a subtype in there
13:09 timotimo or maybe an instance got a role on it
13:09 moritz correct
13:10 jnthn Yeah, method dispatch is very late bound.
13:10 moritz though one might toy with the idea of a pragma that makes that warn at compile time
13:10 timotimo i'd very much like such a thing
13:11 timotimo i'll probably have to spend more time unit testing anyway, but i think it would be a great help
13:11 timotimo if my interpreter (or any analysis program) explodes in my face rather than my code exploding in the end user's face, i consider that a win
13:13 moritz well, Perl 6 is better suited for static analysis than many other dynamic languagues, but it's still a dynamic-ish language
13:13 moritz r: sub f(Int $x) { }; my Str $a; f($a)
13:13 p6eval rakudo 5b56cf: OUTPUT«===SORRY!===␤CHECK FAILED:␤Calling 'f' will never work with argument types (Str) (line 1)␤    Expected: :(Int $x)␤»
13:18 JimmyZ_ joined #perl6
13:31 masak r: macro foo { my $x = "OH HAI"; quasi { $x } }; my $result = foo; say $result
13:31 p6eval rakudo 5b56cf: OUTPUT«OH HAI␤»
13:31 masak jnthn: the above is what fails on the toqast branch currently.
13:32 masak (hm, we don't have toqast wired up to p6eval yet, do we? is it worth bothering to do that, or is the merge very soon?)
13:32 jnthn Probably a week or so away.
13:32 masak might be worth it, then.
13:32 masak I could do it if moritz++ or someone reminds me where stuff is.
13:32 jnthn Well, to me that's "soon" :)
13:32 thelazydeveloper joined #perl6
13:32 masak I meant like tomorrow or something.
13:33 masak jnthn: getting 'Cannot assign a non-Perl 6 value to a Perl 6 container'. assuming that's to $result, because to $x seems harmless.
13:33 masak r: macro foo { say 'before'; my $x = "OH HAI"; say 'after'; quasi { $x } }; my $result = foo; say $result
13:33 p6eval rakudo 5b56cf: OUTPUT«before␤after␤OH HAI␤»
13:33 masak er :)
13:34 masak oh yes, of course.
13:34 masak 'after' in the test file (t/spec/S06-macros/macros-d1.rakudo) happens long before the tests even run.
13:34 masak and the tests run up until number 6.
13:34 moritz masak: on feather1, you need to run 'su - p6eval' to get into the right user account, then set up a directory there that builds rakudo-on-toqast
13:35 moritz (and you need root privs to do that,, which I can give you, if you don't have them already)
13:38 masak jnthn: since this worked before toqast, I'm assuming the error is to be found in the generated QAST. am I thinking right?
13:39 jnthn masak: That or a mistake in the bits of Actions.pm that related to macro/AST handling.
13:39 jnthn s/AST/quasi/
13:40 masak oh, ok. I'll start looking there, then.
13:40 jnthn masak: There's a thingy in World too that does those lexical fixups
13:40 masak ah, yes.
13:41 dalek rakudo/toqast: e242a9a | jnthn++ | src/QPerl6/Actions.pm:
13:41 dalek rakudo/toqast: Re-implement handling of exceptions thrown inside exception handlers.
13:41 dalek rakudo/toqast: review: https://github.com/rakudo/rakudo/commit/e242a9a571
13:42 jnthn OK, it seems make spectest now doesn't hang when we build CORE.setting using QAST. So I think it's time to eliminate qperl6.
13:42 masak yay
13:49 moritz masak: remember --ll-exception
13:49 moritz might or might not be useful for you
13:52 JimmyZ_ joined #perl6
13:58 moritz nqp::isnull(foo) returns 1
13:59 JimmyZ__ joined #perl6
14:01 masak moritz: thanks for the reminder.
14:03 dalek rakudo/toqast: feda23a | jnthn++ | / (17 files):
14:03 dalek rakudo/toqast: Fold QPerl6::* back into Perl6::*, thus eliminating qperl6. This means that the perl6 that is now built is using QAST, as is the setting, test run, etc.
14:03 dalek rakudo/toqast: review: https://github.com/rakudo/rakudo/commit/feda23a2c3
14:06 jnthn Now to merge latest nom changes into toqast. :)
14:07 moritz is the branch now far enough that the remaining test failures are due to missing new patches?
14:07 fgomez joined #perl6
14:08 moritz jnthn: I hope you know about git-rerere
14:08 moritz git config --global rerere.enabled 1
14:08 moritz and then all your merge conflict resolutions will be recorded automatically
14:09 moritz and replayed when you need to do the same merge again
14:09 fgomez joined #perl6
14:11 moritz make: *** No rule to make target `src/Perl6/Ops.pm', needed by `blib/Perl6/Ops.pbc'.  Stop
14:12 Chillance joined #perl6
14:12 jnthn Did you re-configure?
14:12 jnthn Or did I screw it up somehow?
14:13 * jnthn is a little surprised it would fail in such a way elsewhere and work for him...
14:13 jnthn moritz: Hm, I didn't know about that...though only two files had conflicts.
14:13 moritz jnthn: you just didn't commit src/Perl6/Ops.pm I guess
14:13 moritz erm, git add I mean
14:13 jnthn oh...
14:14 moritz thought the build doesn't care what's under version control
14:14 moritz +though
14:14 jnthn hmm...so why it doesn't show up as missing in my git status :/
14:14 jnthn s/missing/untracked/
14:14 moritz so it succeed on your machine
14:14 jnthn yeah, makes sense.
14:15 moritz over-eager gitignore maybe?
14:19 nightlord joined #perl6
14:19 jnthn Hmm, I tossed some stuff from it that seemed redundant/possible, but no help
14:20 moritz the Perl6/*
14:20 moritz line looks suspicious to me
14:20 jnthn Yeah
14:20 jnthn I removed that.
14:21 * jnthn tests the merge
14:27 jnthn Hmm. Just merging latest in gets 274 more tests
14:27 * masak learns about the "${home}" syntax in Perl 5 -- again
14:27 masak that feels so wrong, but I can see the use for it :)
14:28 dalek Heuristic branch merge: pushed 87 commits to rakudo/toqast by jnthn
14:28 moritz eval: my $ x = 3; $x
14:28 buubot_backup moritz: 3
14:28 moritz eval: my $x = 3; $ x
14:28 buubot_backup moritz: 3
14:28 moritz eval: my $x = 3; "$ x"
14:28 buubot_backup moritz: 3
14:28 moritz eval: my $x = 3; "$ {x}"
14:28 buubot_backup moritz: 3
14:29 moritz eval: my $x = 3; "$ {x }"
14:29 buubot_backup moritz: 3
14:29 moritz eval: my $x = 3; "$ { x }"
14:29 buubot_backup moritz: 3
14:29 timotimo eval is perl5?
14:29 buubot_backup timotimo: ERROR: Search pattern not terminated or ternary operator parsed as search pattern at (eval 20) line 1.
14:29 timotimo heh. right.
14:29 moritz timotimo: yes
14:30 jnthn moritz: Should have the missing file now.
14:31 muixirt eval: say (1+2)/3
14:31 buubot_backup muixirt: ERROR: Undefined subroutine &main::say called at (eval 20) line 1.
14:31 muixirt eval: print (1+2)/3
14:31 buubot_backup muixirt: 30.333333333333333
14:32 timotimo huh. how does that work?
14:32 muixirt eval: (print (1+2))/3
14:32 buubot_backup muixirt: 30.333333333333333
14:33 moritz timotimo: the 0.3333333333 is the return value of the expression
14:33 moritz print returns 1
14:36 timotimo oooh, that's amusing
14:38 moritz first test failure in toqast is t/spec/S02-types/native.t
14:38 moritz that's much better than before
14:39 jnthn Yeah
14:39 jnthn The overall failure list is fairly small by now.
14:40 jnthn There's some triage to go yet.
14:40 jnthn But it's not a crazy amount.
14:40 masak (not crazy)++
14:50 masak jnthn: will toqast be 'make install'able later today? I might wait setting p6eval up until then. easier; less changing the build scripts.
14:51 jnthn masak: Should be already.
14:52 jnthn masak: oh, maybe needs an addition to the Makefile to copy the new Ops.pbc
14:52 jnthn I'd be surprised if it needs any more than that.
14:55 masak yay
14:56 masak this always amused me:
14:56 masak eval: print 04.5
14:56 buubot_backup masak: 451
14:56 masak eval: 04.5
14:56 buubot_backup masak: 45
14:56 masak eval: 044.5
14:56 buubot_backup masak: 365
14:57 masak now *that*'s high-end crazy :)
14:57 GlitchMr joined #perl6
14:57 masak nr: say 0o44.5
14:57 p6eval niecza v19-13-g442e075: OUTPUT«[31m===[0mSORRY![31m===[0m��Two terms in a row at /tmp/o_PpuKJVeg line 1:�------> [32msay 0o44[33m�[31m.5[0m��Parse failed��»
14:57 p6eval ..rakudo 5b56cf: OUTPUT«===SORRY!===␤Confused␤at /tmp/r5m45JgVSc:1␤»
14:58 masak std: 0o44.5
14:58 p6eval std 3b49ac8: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Confused at /tmp/byXo8hORkm line 1:â�¤------> [32m0o44.[33mâ��[31m5[0mâ�¤    expecting any of:â�¤        dotty method or postfixâ�¤        postcircumfixâ�¤  postfix_prefix_meta_operatorâ�¤   postopâ�¤Parse failedâ�¤FAILED 00:00 41mâ�¤Â»
14:58 masak fairy nuff.
14:58 masak octal literals are only for integers.
14:58 masak (and that's what happens in the Perl 5 case, too. guess what '.' also means in Perl 5.)
14:58 * masak grins
14:58 timotimo concatenation?
14:59 timotimo if anyone ever uses that in production, wow.
14:59 masak right.
14:59 dalek rakudo/toqast: db138e7 | jnthn++ | src/core/Mu.pm:
14:59 dalek rakudo/toqast: Fix a proto's signature.
14:59 dalek rakudo/toqast: review: https://github.com/rakudo/rakudo/commit/db138e78cf
15:00 dalek nqp/toqast: ca9306c | jnthn++ | src/QAST/Operations.nqp:
15:00 dalek nqp/toqast: Fix 2-arg rindex, thus fixing the rindex test failures in Rakudo.
15:00 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/ca9306c7ff
15:16 SamuraiJack joined #perl6
15:19 felher nom: .say for (<a b> Z <1 2>)
15:19 p6eval rakudo 5b56cf: OUTPUT«a␤1␤b␤2␤»
15:19 felher nom: .say for (<a b> Z <1 2>).tree
15:19 p6eval rakudo 5b56cf: OUTPUT«Not enough positional parameters passed; got 1 but expected between 3 and 4␤  in method new at src/gen/CORE.setting:5022␤  in method tree at src/gen/CORE.setting:5278␤  in block <anon> at /tmp/h_PuKBUYXL:1␤␤»
15:19 felher nom: .say for (<a b> Z <1 2>).lol
15:19 p6eval rakudo 5b56cf: OUTPUT«ListIter.new()␤»
15:20 felher How do i get 'a' and '1' the first iteration and 'b' and '2' the second iteration? It seems i keep forgetting that :/
15:21 moritz r: for <a b> Z <1 2> -> $a, $b { say "$a, $b" }
15:21 p6eval rakudo 5b56cf: OUTPUT«a, 1␤b, 2␤»
15:28 felher moritz: yeah, i know that. But i want them in one variable. So that i can pass them to other functions via |$var, and not via $var1, $var2, $var3, $var4.
15:28 dalek nqp/toqast: bfbf20a | jnthn++ | src/QRegex/NFA.nqp:
15:28 dalek nqp/toqast: Get NFA builder to recognize/handler qastnode.
15:28 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/bfbf20ac43
15:28 jnthn walk &
15:29 Ex joined #perl6
15:29 moritz r: .say for (<a b> Z <1 2>).tree
15:29 p6eval rakudo 5b56cf: OUTPUT«Not enough positional parameters passed; got 1 but expected between 3 and 4␤  in method new at src/gen/CORE.setting:5022␤  in method tree at src/gen/CORE.setting:5278␤  in block <anon> at /tmp/PSfccntYeY:1␤␤»
15:30 benabik joined #perl6
15:30 moritz that looks like a regression in .tree
15:33 moritz still no accepted perl 6 talks at YAPC::EU :(
15:34 felher moritz: okay. I'll wait until that is fixed, than. Thanks :)
15:44 felher Well, maybe i can fix it myself instead of waiting for someone to fix it. But wouldn't have too much faith in that. :)
15:45 sander_ joined #perl6
15:46 sander_ hi everybody
15:46 felher o/
15:46 pmurias moritz: the accepted list is not yet final?
15:47 * pmurias doesn't imply it is
15:47 moritz pmurias: well, http://act.yapc.eu/ye2012/news/881 sounds like the accepted list would be final by second-to-last friday
15:53 moritz felher: spectesting a fix now
15:54 sander_ compiled the recent version of rakudo star. Now I can run scripts only when I pass the absolute path to perl6, otherwise it says "Error while reading from file: Unable to open filehandle from path '$ARG'"
15:54 sander_ i.e. the absolute path of the script
15:55 sander_ even perl6 ./numerize.p6 doesn't work
15:55 moritz sander_: did you run 'make install' ?
15:56 felher moritz++ :)
15:57 sander_ yes, but then I moved the install dir to another location and created a symlink from orig. in-memory /tmp location to the new
15:57 sander_ guess I have to compile anew, right?
15:58 moritz sander_: you might have to, rakudo and parrot aren't relocatable
15:59 moritz but you can give a --prefix= at configure-time
15:59 sander_ that was my idea too but thought a symlink will do
15:59 felher moritz: oh, the problem is $flattens, which got introduced earlier this month, right?
16:00 crab2313 joined #perl6
16:00 moritz felher: no, the problem was that MapIter.new now wants positional arguments, used to be named args
16:00 felher moritz: ah, okay. :)
16:00 moritz felher: i guess pmichaud++'s last list refactoring changed that
16:00 moritz and that we didn't have enough test coverage to catch the .tree case
16:01 dalek rakudo/nom: d0dc2d3 | moritz++ | src/core/IO/Socket/INET.pm:
16:01 dalek rakudo/nom: explicitly set encoding on sockets
16:01 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d0dc2d33d0
16:01 dalek rakudo/nom: e09fbe6 | moritz++ | src/core/IO/Socket/INET.pm:
16:01 dalek rakudo/nom: Merge remote branch 'remotes/origin/parrot-iocleanup1' into nom
16:01 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/e09fbe6870
16:01 dalek rakudo/nom: cd31947 | moritz++ | src/core/List.pm:
16:01 dalek rakudo/nom: fix List.tree
16:01 dalek rakudo/nom:
16:01 dalek rakudo/nom: it got broken by the last API change of MapIter.new
16:01 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/cd31947243
16:01 moritz oops, I didn't mean to push that all
16:02 dalek rakudo/nom: 6db12d6 | moritz++ | src/core/IO/Socket/INET.pm:
16:02 dalek rakudo/nom: Revert "Merge remote branch 'remotes/origin/parrot-iocleanup1' into nom"
16:02 dalek rakudo/nom:
16:02 dalek rakudo/nom: This reverts commit e09fbe68707707004d1111dba7b0b6fb9e100541, reversing
16:02 dalek rakudo/nom: changes made to 5b56cfcf9286e70b9a77a73369b80ad98ad10786.
16:02 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/6db12d6e97
16:02 moritz moritz--
16:05 GlitchMr joined #perl6
16:05 fridim_ joined #perl6
16:07 moritz felher: fixed
16:07 felher \o/ moritz: thanks :)
16:08 dalek roast: eec04b7 | moritz++ | S02-lists/tree.t:
16:08 dalek roast: test iteration for parcel-of-parcel.tree
16:08 dalek roast: review: https://github.com/perl6/roast/commit/eec04b7a6c
16:10 felher moritz: why did you only fix the first two tree methods? The third seems to still use named arguments instead of positionals?
16:11 moritz felher: oh, I didn't spot that
16:14 felher :)
16:26 MayDaniel_ joined #perl6
16:27 sander_ moritz: compiled rakudo star anew (having deleted really everything prior to that) -- error persists
16:35 vmspb joined #perl6
16:36 sander_ unless I absolutize either the path to perl6 interpreter or to the script to be run I get the following error
16:36 sander_ Error while reading from file: Unable to open filehandle from path 'numerize.p6' current instr.: 'nqp;HLL;Compiler;_block2002' pc 35779 (src/stage2/gen/NQPHLL.pir:15079) (src/stage2/gen/NQPHLL.pm:1438)
16:36 sander_ [...]
16:41 sander_ my 'perl6' command was
16:41 sander_ #!/bin/sh
16:41 whiteknight joined #perl6
16:42 sander_ cd ~/D/programing/rakudo-star/install/bin
16:42 sander_ ./perl $@
16:42 masak './perl6', surely?
16:43 sander_ yes, masak, typo
16:43 sander_ but now I changed the sh script to
16:43 sander_ ~/D/programing/rakudo-star/install/bin/perl6 $@
16:43 sander_ and now it works
16:45 sander_ so I wonder if my perl6 interpreter, i.e. the very ELF binary, has problems with being started from its location
16:55 moritz sander_: no
16:55 moritz sander_: the problem is that if you 'cd' in the script, then the relative paths to the script stops working
16:59 proller_ joined #perl6
16:59 proller joined #perl6
16:59 sander_ oh
17:03 xinming joined #perl6
17:03 dalek rakudo/nom: 3ac8c77 | moritz++ | src/core/List.pm:
17:03 dalek rakudo/nom: fix another misuse of MapIter.new, felher++
17:03 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/3ac8c77fd7
17:04 felher :)
17:18 moritz .oO( it would be a fehler to spell felher fehler )
17:18 moritz phenny: de en "fehler"?
17:18 phenny moritz: "error" (de to en, translate.google.com)
17:19 felher moritz: indeed :)
17:19 masak FEHL.
17:23 sander_ moritz: thx, what a noob mistake
17:25 sander_ n8 or tschüss ...
17:27 rurban joined #perl6
17:27 masak heh, 'n8'.
17:28 masak nr: sub german_for($number) { <null eins zwei drei vier fünf sechs sieben acht neun zehn>[$number] }; say "n", german_for 8
17:28 p6eval rakudo 3ac8c7, niecza v19-13-g442e075: OUTPUT«nacht␤»
17:31 moritz except that it's 'zwo' instead of 'zwei' where I come from :-)
17:31 masak I remember.
17:49 dalek rakudo/toqast: bec9a8e | jnthn++ | tools/build/Makefile.in:
17:49 dalek rakudo/toqast: Install the new Perl6::Ops PBC.
17:49 dalek rakudo/toqast: review: https://github.com/rakudo/rakudo/commit/bec9a8e269
17:51 quietfanatic joined #perl6
17:55 japhb moritz, thank you for all the type graph work!
17:55 japhb And you should be proud ... you whipped that out pretty darn quickly, and it's still (quite) good code.  :-)
17:56 * japhb has so far absorbed the parser, and is about to start reading the MRO code
17:57 masak what "type graph work" is this? it sounds interesting.
18:00 japhb masak, see commits in perl6/doc over the last half day or so.
18:00 japhb moritz++ # oops, forgot.  ;-)
18:02 masak oh, ok.
18:02 rurban joined #perl6
18:08 moritz I think the next step is a topo sort
18:09 moritz so that traversal of the .pod files can be done in Mu-first order
18:10 snearch joined #perl6
18:10 moritz masak, that's your speciality, no? :-)
18:10 masak 呵呵
18:11 * masak .oO( "ask him over there, he's the one knows how to do it." )
18:12 jnthn moritz: If you know the MRO length of all the things you can always cheat and just do them in increasing order of MRO length... :)
18:12 jnthn iiuc what you're dong, anyway
18:12 jnthn *doing
18:13 moritz jnthn: I'm not sure about the MRO for roles
18:14 masak here's how you do a topo sort: DFS down the graph. as you *go back up*, the first time you see every node, emit it.
18:14 moritz so, post-order traversal
18:15 quietfanatic If you have to do that every time you dispatch, you're gonna get slow code.
18:15 tadzik hello #perl6
18:15 * quietfanatic is jumping into a conversation he knows nothing about again
18:15 jnthn moritz: roles don't show up in the MRO because we don't visit them during method resolution :)
18:16 moritz jnthn: yes, but I need to sort them too
18:16 jnthn quietfanatic: Relax, the conversation is about p6doc :)
18:16 quietfanatic aaah
18:16 moritz to determine in which order to process the docs
18:16 jnthn moritz: OK, then just considering MRO won't cut it :)
18:16 jnthn hi tadzik
18:16 quietfanatic Okay.  Generally you don't have to process docs as part of a function dispatch mechanism :)
18:17 masak tadzik! \o/
18:18 moritz I need to handle the 'Any' default somehow
18:24 dalek doc: 008d651 | moritz++ | / (2 files):
18:24 dalek doc: add Any as default superclass
18:24 dalek doc: review: https://github.com/perl6/doc/commit/008d651108
18:44 dalek evalbot: e1d7f66 | masak++ | / (2 files):
18:44 dalek evalbot: added 'toqast' target
18:44 dalek evalbot: review: https://github.com/perl6/evalbot/commit/e1d7f6613e
18:45 masak evalbot control restart
18:45 p6eval joined #perl6
18:45 jnthn toqast: say "no, not toast"
18:46 masak hm.
18:46 masak oh!
18:46 * masak pulls server-side
18:47 masak evalbot control restart
18:47 p6eval joined #perl6
18:47 jnthn toqast: say "no, not toast"
18:47 flussence toqast: pizza?
18:48 masak sigh :)
18:48 masak I'm new at this, apparently.
18:48 masak moritz: halp!
18:53 sorear good * #perl6
18:53 masak r: say "this still works, I hope."
18:53 p6eval rakudo 3ac8c7: OUTPUT«this still works, I hope.␤»
18:53 masak phew.
18:53 masak sorear! \o/
18:53 jnthn o/ sorear
18:55 dalek evalbot: ff944c6 | masak++ | sync.pl:
18:55 dalek evalbot: updated sync.pl with toqast
18:55 dalek evalbot: review: https://github.com/perl6/evalbot/commit/ff944c68ca
18:55 * sorear tries to log in to feather3's p6eval screen, fails
18:56 p6eval joined #perl6
18:56 moritz sorear: a 'git pull' would have been enough :-)
18:56 moritz toqast: say 1
18:56 moritz r: say 1
18:57 p6eval rakudo 3ac8c7: OUTPUT«1␤»
18:57 masak moritz: I think I did everything by the book. feel free to review it if you have time.
18:58 moritz masak: why did you change rakudo's safe mode from RESTRICTED back to SAFE?
18:58 masak I did no such thing.
18:58 moritz oh
18:58 moritz I might have read the diff wrong
18:59 masak no, https://github.com/perl6/evalbot/commit/e1d7f6613e61c47e3691d1b058ee609851d4634f does look worrying.
18:59 moritz masak: looks like you accidentally reverted some older commits :(
19:00 masak here's what I did. I made a commit as p6eval, realized I was the wrong user, copied things over to my own home directory on feather, and committed there.
19:00 masak seems so, yes.
19:00 moritz oh yes
19:00 masak I'll straighten things out immediately.
19:00 moritz copying stuff is how you get an old state into your sources
19:00 masak right.
19:00 dalek evalbot: c03db41 | masak++ | / (2 files):
19:00 dalek evalbot: Revert "added 'toqast' target"
19:00 dalek evalbot:
19:00 dalek evalbot: This reverts commit e1d7f6613e61c47e3691d1b058ee609851d4634f.
19:00 dalek evalbot: review: https://github.com/perl6/evalbot/commit/c03db412ea
19:02 dalek nqp/toqast: 4f9fa15 | jnthn++ | src/QAST/ (2 files):
19:02 dalek nqp/toqast: Fix if/unless with pointy blocks taking the condition as an argument.
19:02 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/4f9fa15062
19:02 dalek nqp/toqast: 69b0f82 | jnthn++ | src/QAST/Operations.nqp:
19:02 dalek nqp/toqast: while and until with pointy.
19:02 dalek nqp/toqast: review: https://github.com/perl6/nqp/commit/69b0f8255c
19:02 dalek rakudo/toqast: 709b09f | jnthn++ | src/core/Numeric.pm:
19:02 dalek rakudo/toqast: Correct another wrong proto.
19:02 dalek rakudo/toqast: review: https://github.com/rakudo/rakudo/commit/709b09fda9
19:03 dalek evalbot: 7222a04 | masak++ | / (2 files):
19:03 dalek evalbot: added 'toqast' target
19:03 dalek evalbot:
19:03 dalek evalbot: This time with correctness.
19:03 dalek evalbot: review: https://github.com/perl6/evalbot/commit/7222a04a03
19:04 masak evalbot control restart
19:04 p6eval joined #perl6
19:04 masak r: say "OH HAI"
19:05 moritz the first execution always takes ages
19:05 moritz I have no idea why
19:05 * masak .oO( the bot is cold and needs to warm up )
19:05 moritz r: say 1
19:05 p6eval joined #perl6
19:05 moritz t: say 1
19:05 moritz cat: /home/p6eval/toqast-inst/rakudo-revision: No such file or directory
19:06 masak there was when I checked. hm.
19:06 p6eval toqast : OUTPUT«Can't exec "./toqast-inst/bin/perl6": No such file or directory at lib/EvalbotExecuter.pm line 200.␤exec (./toqast-inst/bin/perl6 --setting=RESTRICTED /tmp/7xwHaMw8Nz) failed: No such file or directory␤Lost connection to server irc.freenode.org.␤»
19:06 masak $ toqast-inst/bin/perl6 -e 'say "OH HAI"'
19:06 masak OH HAI
19:07 masak wrong CWD?
19:07 moritz that's feather 1
19:07 moritz t: say 1
19:07 p6eval joined #perl6
19:08 masak "that's feather 1"? are there more feathers I should have been aware of when making this addition? :/
19:08 moritz masak: that's what sync.pl is about
19:08 nodmonkey joined #perl6
19:08 moritz masak: there are three virtual maschines on the same physical host
19:09 masak ah.
19:09 moritz masak: the feather you usually log in to is feather1. It's where we build the compilers for p6eval
19:09 moritz and p6eval runs on feather3
19:09 masak I see.
19:09 moritz so sync.pl rsync's stuff over
19:09 moritz r: say 'OH HAI'
19:09 p6eval rakudo 3ac8c7: OUTPUT«OH HAI␤»
19:09 masak I didn't run sync.pl
19:09 moritz t: say 'me too'
19:09 p6eval toqast bec9a8: OUTPUT«./toqast-inst/bin/perl6: error while loading shared libraries: libparrot.so.4.5.0: cannot open shared object file: No such file or directory␤»
19:09 masak clearly something is building, though.
19:09 moritz masak: which parrot does toqast-inst use?
19:09 moritz I ran it
19:10 moritz and build.pl runs it
19:10 crab2313 /quit
19:11 masak moritz: same parrot as nom uses.
19:11 masak I didn't change it in any way.
19:11 moritz masak: so --with-parrot and separate --prefix did work in the end?
19:13 masak no, I added a 'cp -r' to the end of rebuild-toqast.pl
19:13 jnthn Uhm. How has Buf.setbuf ever worked?
19:13 jnthn *subbuf
19:13 jnthn It creates a new byte buffer with a string in instead of a byte buffer.
19:14 moritz masak: parrot isn't relocatable
19:14 jnthn Oh...maybe because QAST knows to box to a Perl 6 Str, not a Parrot String, and Parrot String implements elements...
19:15 masak moritz: the toqast/bin/perl6 executable works from the ~ dir of p6eval.
19:15 moritz masak: that's all fine, but it doesn't work when rsync'ed over to feather3
19:16 moritz jnthn: it just needs a conversion from parrot binary String to ByteBuffer
19:16 moritz jnthn: like infix:<~>(Buf, Buf) does it
19:17 moritz jnthn: want me to patch it?
19:17 jnthn moritz: Trying a patch now.
19:17 jnthn moritz: Was mostly wondering how on earth we got away with it more than how to patch it :)
19:17 masak moritz: I take your word for it. so clearly I did something wrong in setting up all this. I am at a loss how to fix it, unfortunately.
19:18 moritz masak: compile another parrot for toqast-inst{1,2}/
19:21 masak how?
19:21 jnthn For anybody wanting to help with toqast: I know why repeat.t is busted and know the three leftover S05 fails are all the same root cause. Analysis/golfing/fixes for any of the other remaining failures is most welcome.
19:22 moritz masak: git clone parrot; perl Configure --prefix=$prefix --optimize && make install
19:23 masak ah, as part of rebuild-toqast.pl?
19:23 moritz yes
19:23 * masak does it
19:23 UncleFester6 joined #perl6
19:23 moritz (not the git cloning, just the git pulling, if ncessary)
19:25 masak ok.
19:26 daxim joined #perl6
19:26 UncleFester6 r: our @a; say @a.perl; our %h; say %h.perl
19:26 p6eval rakudo 3ac8c7: OUTPUT«Any␤Any␤»
19:27 UncleFester6 r: our @a =()
19:27 p6eval rakudo 3ac8c7: OUTPUT«No such method 'STORE' for invocant of type 'Any'␤  in block <anon> at /tmp/LpfyDUlKRB:1␤␤»
19:27 moritz UncleFester6: known problem :(
19:27 UncleFester6 r: our %h = ()
19:27 p6eval rakudo 3ac8c7: OUTPUT«No such method 'STORE' for invocant of type 'Any'␤  in block <anon> at /tmp/GAorSSASi2:1␤␤»
19:28 UncleFester6 moritz: this is the basis, I think of RT 102876 - if there is another relevant ticket maybe they should be merged
19:28 dalek evalbot: 974082c | masak++ | build-scripts/rebuild-toqast.pl:
19:28 dalek evalbot: instructions to rebuild parrot for toqast
19:28 dalek evalbot: review: https://github.com/perl6/evalbot/commit/974082caed
19:28 dalek doc: e7b836c | moritz++ | / (2 files):
19:28 dalek doc: topological sorting of types
19:28 dalek doc: review: https://github.com/perl6/doc/commit/e7b836cd5f
19:28 masak evalbot control restart
19:28 p6eval joined #perl6
19:28 moritz you only need to restart evalbot for actual code changes to the bot
19:28 masak I need to focus on crypt now, unfortunately.
19:29 masak ok.
19:29 moritz t: say 1
19:29 p6eval toqast bec9a8: OUTPUT«./toqast-inst/bin/perl6: error while loading shared libraries: libparrot.so.4.5.0: cannot open shared object file: No such file or directory␤»
19:34 dalek evalbot: 9906490 | moritz++ | build-scripts/rebuild-toqast.pl:
19:34 dalek evalbot: [toqast rebuild] list form of system() does not interpret shell meta chars
19:34 dalek evalbot: review: https://github.com/perl6/evalbot/commit/9906490430
19:42 birdwindupbird joined #perl6
19:46 moritz /home/p6eval/toqast/install/bin/nqp --vmlibs=perl6_group,perl6_ops  --target=pir --output=src/gen/perl6-actions.pir --encoding=utf8 \ src/Perl6/Actions.pm
19:46 moritz Null PMC access in find_method('operations')
19:47 UncleFester6 moritz: the other ticket is 100560 and neither of the tickets note that 'our @a; say @a.perl' gives any which seems useful to mention.
19:47 jnthn The Rakudo toqast branch also needs the toqast branch of NQP
19:48 moritz jnthn: I know. evalbot needs to know too :/
19:50 moritz UncleFester6: yes, seems useful to add
19:54 UncleFester6 r: our @a; my $x = shift @a;
19:54 p6eval rakudo 3ac8c7: OUTPUT«Cannot call 'shift'; none of these signatures match:␤:(@a)␤␤  in block <anon> at /tmp/KwEJ4wnr8k:1␤␤»
19:54 UncleFester6 r: my @a; my $x = shift @a;
19:54 p6eval rakudo 3ac8c7:  ( no output )
19:55 UncleFester6 r: our @a; @a[3] = 12;
19:55 p6eval rakudo 3ac8c7:  ( no output )
19:55 UncleFester6 that's ok  ...
19:57 UncleFester6 based on that last example with shift I propose merging 100560 and 102876 and adding the shift example ... I don't think I can do it ...
19:57 fridim_ joined #perl6
19:59 moritz masak: oh
20:00 moritz masak: the problem might be the cp -r instead of changing the symlink
20:01 masak :/
20:01 masak well, I never got --prefix to work.
20:01 masak no idea why.
20:02 moritz because the bird's prefix was at odds with the one you gave to Configure.pl
20:02 masak hm.
20:03 moritz how about we screw this complicated setup, and do a simple compile without a custom prefix, and rsync it all over?
20:04 moritz the two-install-dirs-setup was from the times where we built the cmopiler on the same machine as we ran evalbot
20:04 moritz now rsync does the snapshotting for us
20:06 masak sounds fine to me.
20:06 masak sorry for messing it up in the first place.
20:06 masak it looked easy, but it was complicated.
20:06 moritz no problem
20:06 moritz that's true for a lot of problems :-)
20:07 masak :)
20:11 dalek evalbot: 978c680 | moritz++ | / (2 files):
20:11 dalek evalbot: simplfy toqast rebuild
20:11 dalek evalbot: review: https://github.com/perl6/evalbot/commit/978c680231
20:14 dalek evalbot: 92190fc | moritz++ | build-scripts/rebuild-toqast.pl:
20:14 dalek evalbot: [toqast] fix thinko
20:14 dalek evalbot: review: https://github.com/perl6/evalbot/commit/92190fc819
20:16 dalek evalbot: 1ebd56b | moritz++ | / (2 files):
20:16 dalek evalbot: more thinkos
20:16 dalek evalbot: review: https://github.com/perl6/evalbot/commit/1ebd56bdfa
20:17 bruges joined #perl6
20:24 moritz evalbot control restart
20:24 p6eval joined #perl6
20:26 crab2313 joined #perl6
20:31 crab2313 joined #perl6
20:35 UncleFester6 r: our (@a)
20:35 p6eval rakudo 3ac8c7: OUTPUT«use of uninitialized value of type Mu in string context  in any <anon> at src/gen/BOOTSTRAP.pm:104␤␤use of uninitialized value of type Mu in string context  in any <anon> at src/gen/BOOTSTRAP.pm:104␤␤===SORRY!===␤error:imcc:Multiple declarations of lexical '@a'␤␤   i…
20:35 UncleFester6 known bug but golf
20:37 jnthn UncleFester6++
20:38 jnthn Add to ticket; once I get toqast landed I'll make a pass over this our scoping stuff
20:38 moritz t: say 3
20:38 p6eval toqast : OUTPUT«3␤»
20:38 moritz toqast: say 'loliamhere'
20:38 p6eval toqast : OUTPUT«loliamhere␤»
20:39 jnthn r: sub foo(Int $a) { CATCH { say "oops" } }; foo(~"lol")
20:39 p6eval rakudo 3ac8c7: OUTPUT«oops␤Nominal type check failed for parameter '$a'; expected Int but got Str instead␤  in sub foo at /tmp/z7VxGgBXGV:1␤  in block <anon> at /tmp/z7VxGgBXGV:1␤␤»
20:39 jnthn t: sub foo(Int $a) { CATCH { say "oops" } }; foo(~"lol")
20:39 p6eval toqast : OUTPUT«Nominal type check failed for parameter '$a'; expected Int but got Str instead␤␤»
20:39 jnthn Well, that's one thing fixed in toqast :)
20:40 tadzik \o/
20:41 jnthn masak++, moritz++ for setting this up. It'll be useful in the coming days as spectests and module space are triaged. :)
20:41 masak moritz++ \o/
20:41 masak t: macro foo { my $a = "it's us, the space puppies!"; quasi { $a } }; my $result = foo; say $result
20:41 p6eval toqast : OUTPUT«Cannot assign a non-Perl 6 value to a Perl 6 container␤␤»
20:42 masak jnthn: there, that's what I wanted to show :P
20:42 jnthn t: macro foo { my $a = "it's us, the space puppies!"; quasi { $a} }; say foo.WHAT
20:42 p6eval toqast : OUTPUT«Can only use get_what on a SixModelObject␤␤»
20:42 jnthn Well, there's another clue. :)
20:42 jnthn t: macro foo { my $a = "it's us, the space puppies!"; quasi { $a} }; say DUMP(foo)
20:42 p6eval toqast : OUTPUT«(null)␤»
20:43 jnthn t: say (quasi { 1 }).WHAT
20:43 p6eval toqast : OUTPUT«AST()␤»
20:43 jnthn Hm, not so obvious then :)
20:44 jnthn t: macro foo { my $a = "it's us, the space puppies!"; quasi { $a} }; say nqp::isnull(foo)
20:44 p6eval toqast : OUTPUT«No such method 'gist' for invocant of type 'Integer'␤␤»
20:44 jnthn t: macro foo { my $a = "it's us, the space puppies!"; quasi { $a} }; say nqp::p6box_i(nqp::isnull(foo))
20:44 p6eval toqast : OUTPUT«1␤»
20:44 masak something goes wrong in the macro application?
20:45 jnthn Guessing so
20:47 masak then I'll know where to look.
20:53 Johnson joined #perl6
21:00 tadzik eww, no Perl 6 talks on the YAPC?
21:10 lue hello world o/
21:12 jnthn tadzik: I've not had mine accepted or rejected yet.
21:13 tadzik hello lue!
21:14 * tadzik checks how toqast awesome got over the weekend
21:25 pupoque joined #perl6
21:29 kurahaupo joined #perl6
21:48 UncleFester6 r: our $a = 12; { our $a =2;}; say $a
21:48 p6eval rakudo 3ac8c7: OUTPUT«2␤»
21:48 masak aye.
21:48 UncleFester6 looking at RT 111284  and wondering if this isn't redeclaration
21:48 masak it's never redeclaration if it's in another scope.
21:49 tadzik suddenly, I have a feeling that I should stop using any sort of real name and real connection on the internet
21:49 UncleFester6 but it kept the value from the inner scope after that scope is gone.  Was it really declared in a different scope?
21:50 masak tadzik: why?
21:50 tadzik I mean, LinkedIn guesses people you may know, from, I don't know, your company, your gmail contacts, their gmail contacts, facepalm, whatever
21:50 masak UncleFester6: before we go any further, do you know what 'our' variables are?
21:50 tadzik and it's quite well at figuring out people
21:50 tadzik to the point that it just listed one of the bartenders I know in "People you may know"
21:50 tadzik How on earth does it know that
21:51 tadzik makes me want to crawl under my bed and change my name and face
21:51 masak tadzik: people will buy your privacy by the kilo whether you're selling it or not. welcome to the 21st century.
21:51 UncleFester6 masak: my knowledge is based on perl5 but if you still think I am heading in the wrong direction I will take your word for it for now
21:51 masak UncleFester6: I didn't mean to discourage you, just to gauge your knowledge.
21:52 masak UncleFester6: basically, 'our' = 'my' + storage in package scope.
21:52 masak if that makes any sense.
21:52 UncleFester6 in my example weren't both our declarations in the same package scope?
21:53 masak yes, correct.
21:53 masak that's why it overwrote the 12.
21:53 UncleFester6 overwrite is fine.  allowing the redeclaration is what I found confusing ...
21:54 UncleFester6 r: our $a; our $a
21:54 p6eval rakudo 3ac8c7: OUTPUT«===SORRY!===␤Redeclaration of symbol $a␤at /tmp/riUERDeBje:1␤»
21:57 UncleFester6 Interesting in perl5  even with strict 'our $a; our $a' is allowed.
21:58 UncleFester6 warnings catches it though
22:00 lue std: our $a; our $a;
22:00 p6eval std 3b49ac8: OUTPUT«ok 00:00 41m␤»
22:01 moukeddar joined #perl6
22:01 benabik r: our $a; our $a
22:01 p6eval rakudo 3ac8c7: OUTPUT«===SORRY!===␤Redeclaration of symbol $a␤at /tmp/KyPq1bELiX:1␤»
22:01 UncleFester6 perl5 with strict + warnings complains about my original example
22:01 lue hm. Seems there's a bug in rakudo.
22:11 gfldex i'm toying around with my own implementation for search in doc
22:11 gfldex i'm making great progress: Firefox had a problem and crashed. We'll try to restore your tabs and windows when it restarts.
22:12 lue r: class A { state $b; };
22:12 p6eval rakudo 3ac8c7: OUTPUT«(signal SEGV)»
22:13 lue should I file in RT or in Github issues?
22:13 benabik rakudo does bugs in RT
22:13 benabik Yeah, Rakudo doesn't even have gh issues enabled.
22:13 masak lue: nice!
22:13 masak lue: are you submitting it, or shall I?
22:13 lue thought so (there was talk of moving a while ago, so I wasn't sure).
22:13 MayDaniel joined #perl6
22:13 lue masak: you could if you want, it's been a while since I've used RT :)
22:13 masak oh, you're submitting it. good.
22:14 masak oh, ok.
22:14 masak I'll do it, then.
22:14 benabik lue: There's been resistance to upgrading RT, much less moving somewhere else.
22:14 lue .oO(who am I to deprive masakbot of his life's purpose?)
22:14 masak benabik: I believe that was me, yes :)
22:14 lue benabik: yeah, I remember that part of the discussion too :)
22:14 * masak submits rakudobug
22:18 lue masak: actually, that "hm. seems there's a bug" line was for mishandling our $a; our $a; above :) [ unless it's std that's wrong here ]
22:22 masak lue: I don't know. for once I agree with Rakudo.
22:22 masak r: my $a; my $a
22:22 p6eval rakudo 3ac8c7: OUTPUT«===SORRY!===␤Redeclaration of symbol $a␤at /tmp/BFy2EKcS3p:1␤»
22:22 benabik n: our $a; our $a
22:22 p6eval niecza v19-13-g442e075: OUTPUT«Potential difficulties:â�¤  Useless redeclaration of variable $a (see line 1) at /tmp/bQjhMFTpYD line 1 (EOF):â�¤------> [32mour $a; our $a[33mâ��[31m<EOL>[0mâ�¤â�¤Â»
22:22 masak though the spec says that that should be OK, IIRC.
22:22 lue std: my $a; my $a;
22:22 p6eval std 3b49ac8: OUTPUT«Potential difficulties:â�¤  Useless redeclaration of variable $a (see line 1) at /tmp/eE2NODalG8 line 1:â�¤------> [32mmy $a; my $a[33mâ��[31m;[0mâ�¤ok 00:00 41mâ�¤Â»
22:22 lue std: our $a; our $a;
22:22 p6eval std 3b49ac8: OUTPUT«ok 00:00 41m␤»
22:23 lue I personally can't see the reason for allowing redeclaration of our variables (unless there's some package-scoping-related scenario I'm not thinking of)
22:23 masak lol, I blogged! \o/ http://strangelyconsistent.org/blog/july-22-2012-playing-the-hanoi-game
22:28 UncleFester6 lue: I still think that there should be consistency with below
22:28 UncleFester6 r:our $a = 12; { our $a =2;}; say $a
22:28 UncleFester6 r: our $a = 12; { our $a =2;}; say $a
22:28 p6eval rakudo 3ac8c7: OUTPUT«2␤»
22:28 masak there *is* consistency.
22:28 lue masak++ good post
22:28 masak r: my $a = 12; { my $a = 2 }; say $a
22:28 p6eval rakudo 3ac8c7: OUTPUT«12␤»
22:28 masak neither the 'my' or the 'our' variant is a redeclaration.
22:28 UncleFester6 in perl5 with strict and warnings my example is identified as a redeclaration just like 'our $a; our $a'
22:29 lue our $a; our $a;    still seems... interesting (unless, again, there's a good reason for allowing this but not    my $a; my $a;    )
22:29 masak UncleFester6: well, Perl 5 has slightly different semantics for 'our'.
22:29 masak eval: { our $a = 12 }; $a
22:29 buubot_backup masak: 12
22:30 masak r: { our $a = 12 }; say $a
22:30 p6eval rakudo 3ac8c7: OUTPUT«===SORRY!===␤Variable $a is not declared␤at /tmp/48NluKkbrV:1␤»
22:30 UncleFester6 have to go for now sorry ...
22:34 masak o/
22:46 lue I'm very interested in taking Adventure::Engine for a spin after July :)
22:46 masak \o/
22:47 * lue wonders how hard it would be to implement a "hamster wheel" room...
22:51 cognominal__ joined #perl6
22:54 moukeddar joined #perl6
22:55 tadzik Files=683, Tests=24822, 1332 wallclock secs
22:55 tadzik that's qast
22:56 tadzik if panda runs, I'm trying emmentaler
23:00 tadzik yep, panda runs
23:01 masak \o/
23:01 tadzik I have a feeling it's slightly faster, but there's a high chance that it's just my imagination :)
23:02 masak no, it's slightly faster. the Q in QAST is more aerodynamic than the P in PAST.
23:02 tadzik true
23:02 timotimo the little wiggly thing on the Q is actually what the QAST uses to move forwards
23:02 tadzik and it has a spoiler, which makes it look faster too
23:02 tadzik oh, that can't be it
23:02 timotimo no? not like bacteria?
23:03 tadzik that'd be a snail-type movement then, no?
23:03 tadzik doesn't fit :)
23:03 masak I voted for using a D, which has great lift power, too. but people thought that idea was really DAST.
23:03 masak timotimo: flagellae.
23:03 masak phenny: "flagellum"?
23:03 phenny masak: "flagellum" (en to en, translate.google.com)
23:04 tadzik phenny: "dast"?
23:04 phenny tadzik: "dast" (de to en, translate.google.com)
23:04 tadzik phenny: en pl "dast"?
23:04 phenny tadzik: "Dast" (en to pl, translate.google.com)
23:04 masak hm, plural appears to be "flagella".
23:04 timotimo masak: thanks, i just looked it up on wikipedia, too :)
23:04 masak Intelligent-Design people love flagella. or did a few years ago, at least.
23:05 timotimo they did?
23:05 timotimo didn't notice any of that
23:05 timotimo don't really know if i should give them any more of my time ...
23:05 masak probably not.
23:05 masak well, unless you happen to like being frustrated.
23:06 timotimo sure it can be amusing every once in a while, when they come up with crocoducks or claim monkeys still existing disproves evolution, but then you notice that those people are quite a big and powerful (or at least very vocal) group in 'merica and then you just get sad
23:06 sorear masak: "whip" (classical latin to en, sorear)
23:06 tadzik this is when "catholic poland" suddenly feels *so* sane...
23:07 lue wasn't their argument that it (flagella) couldn't have come in a simpler, earlier form, ∴ !evolution?
23:07 timotimo it seems many arguments they make are quite a lot like that, yeah
23:07 sorear basically, eys
23:07 tadzik this is also when I'm close to coming up with arguments which make people hate me, so I guess I'd just shut up for a while :)
23:08 sorear and it's unfalsifiable since flagella don't leave fossils!
23:08 * sorear hugs tadzik
23:08 lue (but then a scientist came along and proved its earlier form was as a stinger of sorts. It was part of a court case, at least what I'm thinking of)
23:08 tadzik sorear: thanks, I feel better
23:09 jnthn tadzik: (panda works) yay!
23:10 jnthn tadzik: Will be happy to see how the cheese holds up.
23:10 tadzik yeah, the smoketests seem to be going along nicely so far
23:10 tadzik well, NativeCall doesn't work, for one
23:10 timotimo https://en.wikipedia.org/wiki/Evolution_of_flagella <- this article suggests a former version of the flagellum might have been used to inject toxin into other cells, so you are probably remembering correctly, lue
23:11 timotimo and with this, i'm done with intelligent design for today
23:11 jnthn tadzik: There's still some spectests that fail too; it's entirely possible NativeCall relies on one of them.
23:11 tadzik could be
23:11 cognominal__ joined #perl6
23:11 jnthn tadzik: Alternatively, it may rely on recent arnsholt++ work not merged from nqp master into toqast, perhaps.
23:11 jnthn (Didn't check if there is any.)
23:11 lue (watched a video about it a case with that as a part of it in class long ago) The scientist came to court with a somewhat-taken-apart mousetrap as a tie clip as an analogy
23:12 lue (I'm done with ID too now)
23:12 benabik My biggest argument against intelligent design is that most biological design doesn't seem too intelligent.  :-D
23:12 tadzik :P
23:12 tadzik I mean, fingers on feet! Gah!
23:13 quietfanatic Hey, my feetgers are very useful sometimes.
23:13 crab2313 joined #perl6
23:15 tadzik they lift furniture sometimes, but they usually don't
23:16 lue .oO(that whole caffeine isn't good for you thing was a bit of a slip-up :P)
23:18 timotimo everything in the universe is trying to kill us!
23:18 masak 'night, #perl6
23:18 timotimo 'night masak :)
23:19 tadzik 'night
23:19 tadzik good knight masak
23:27 tadzik 49 modules ok, 57 not ok (106 total)
23:27 tadzik Testing took 25 minutes and 36 seconds
23:28 tadzik uploaded to http://tjs.azalayah.net/index.html
23:28 tadzik so about 5 regressions
23:30 jnthn tadzik: Do you know which are the regressions? (more)
23:31 jnthn I suspect if one is NativeCall, anything that depends on it will also be regressed.
23:32 jnthn tadzik: I count 4 modules that say they failed because of a dependency on NativeCall.
23:32 jnthn tadzik: Those + native call = 5. That's...surprisingly good if true.
23:38 tadzik URI also fails
23:38 tadzik and thus LWP and Bailador
23:40 tadzik I need a tool that'll do the diff between two results.json
23:40 tadzik I'll tackle that tomorrow
23:41 tadzik or maybe today
23:41 jnthn tadzik: Great, thanks.
23:41 tadzik well, the tea is still warm anyway
23:41 jnthn tadzik: URI is regex heavy and we have 3 test fails in S05.
23:41 tadzik the bad part is that I don't think I've saved the old results.json
23:41 tadzik I'll rebuild and re-run then
23:42 tokuhiro_ joined #perl6
23:42 jnthn tadzik: OK. I'm gonna sleep soon; feel free to gist me the failure modes of NativeCall and URI or triage it a bit...it may boil down to one of the failing spectests though.
23:43 jnthn The majority of things being OK is very good news.
23:47 timotimo lue: http://www.youtube.com/watch?v=8wOKuFu217E <- this would be a video about the flagellum and the mousetrap tie clip, from NOVA
23:54 jnthn 'night, #perl6
23:55 tadzik good knight jnthn
23:57 crab2313 joined #perl6
23:57 lichtkind joined #perl6
23:58 lichtkind woot #perl6

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

Perl 6 | Reference Documentation | Rakudo