Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2014-05-18

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:14 BenGoldberg joined #perl6
00:16 fhelmberger joined #perl6
00:22 dalek perl6-roast-data: d6c0fbf | coke++ | / (6 files):
00:22 dalek perl6-roast-data: today (automated commit)
00:22 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/d6c0fbf913
00:22 dalek perl6-roast-data: d38223d | coke++ | / (6 files):
00:22 dalek perl6-roast-data: today (automated commit)
00:22 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/d38223dd37
00:24 rurban joined #perl6
00:32 dj_goku joined #perl6
00:55 hoverboard joined #perl6
01:00 Sqirrel_ joined #perl6
01:10 Marverick joined #perl6
01:10 Marverick left #perl6
01:11 klapperl_ joined #perl6
01:27 rurban joined #perl6
01:54 Su-Shee_ joined #perl6
02:17 Sqirrel joined #perl6
02:28 rurban joined #perl6
03:11 rurban joined #perl6
03:40 xragnar_ joined #perl6
03:40 atroxaper joined #perl6
03:44 rindolf joined #perl6
03:53 SamuraiJack__ joined #perl6
03:58 kurahaupo joined #perl6
04:02 kurahaupo_mobile joined #perl6
04:03 kurahaupo left #perl6
04:07 kurahaupo_mobile joined #perl6
04:07 bjz_ joined #perl6
04:08 kurahaupo__ joined #perl6
04:11 bjz joined #perl6
04:20 darutoko joined #perl6
04:28 dayangkun joined #perl6
04:34 xenoterracide__ joined #perl6
04:58 hoverboard joined #perl6
05:37 darutoko joined #perl6
05:39 masak morning, #perl6
05:49 FROGGS morning
06:22 takesako_____ joined #perl6
06:24 zby_home_ joined #perl6
07:06 anaeem1 joined #perl6
07:07 anaeem1 joined #perl6
07:07 rurban joined #perl6
07:08 denis_boyun_ joined #perl6
07:08 anaeem1__ joined #perl6
07:32 kaare_ joined #perl6
07:39 rurban joined #perl6
07:41 kurahaupo joined #perl6
07:48 isBEKaml joined #perl6
07:50 kivutar joined #perl6
07:59 lizmat joined #perl6
08:00 jnthn morning, #perl6
08:04 lizmat good UGT, #perl6!
08:04 lizmat from the PLPW
08:08 FROGGS joined #perl6
08:08 nwc10 good UGT, *
08:09 nwc10 please could someone be fixing the test count in the failing NQP test
08:09 nwc10 s/53/54/ IIRC
08:09 sergot_ morning ! o/
08:09 nwc10 t/nqp/19-file-ops.t
08:10 FROGGS I am about to
08:10 FROGGS t/nqp/19-file-ops.t                 (Wstat: 0 Tests: 54 Failed: 2)
08:10 FROGGS Failed tests:  36, 54
08:10 nwc10 they all pass for me, but then the wrong test count correctly causes the test to fail
08:11 dalek nqp: 65e4e41 | (Tobias Leich)++ | t/nqp/19-file-ops.t:
08:11 dalek nqp: fix test plan, nwc10++
08:11 dalek nqp: review: https://github.com/perl6/nqp/commit/65e4e4143f
08:12 FROGGS hmmm, 36 still fails :/
08:15 nwc10 works on "my" machine. Not sure why
08:17 fhelmberger joined #perl6
08:18 matt____ joined #perl6
08:19 matt____ left #perl6
08:23 denis_boyun___ joined #perl6
08:29 vendethiel joined #perl6
08:36 woolfy joined #perl6
08:37 woolfy Just did my presentation on marketing Perl for the  Polish Perl Workshop.  117 slides in 20 minutes (10 seconds per slide).  Went well.  Liz said I was just a bit hyper.  :-)
08:37 woolfy I mentioned Perl 6 several times.
08:38 sergot_ woolfy: that was really motivating!!! \o/
08:39 woolfy Thanks...   was it?  really?
08:39 sergot_ it was :)
08:39 woolfy \o/
08:40 tadzik yeah, it was great :)
08:40 woolfy Hehe, sergot asked me to not mention his name.  Of course I did mention his name.  D'oh.
08:40 tadzik I now have 40-ish slides for 20 minutes
08:40 tadzik that should be fine :)
08:40 lizmat tadzik: and a live demo?
08:41 sergot_ You have very interesting stuff there tadzik.
08:41 woolfy tadzik: maybe you will fall asleep halfway?  We will keep you awake!!!
08:41 molaf joined #perl6
08:41 tadzik lizmat: yeah, there'll be a short live demo
08:41 rurban joined #perl6
08:41 tadzik but some of the slides are just the pictures of The Good, The Bad and The Ugly
08:41 tadzik and memes
08:42 sergot_ tadzik: with xbox cotroller I hope :)
08:42 tadzik of course :)
08:42 lizmat memememememememememememememememememe (beaker impression)
08:42 lizmat .oO( isn't one troller enough? )
09:04 kaare__ joined #perl6
09:07 Rotwang joined #perl6
09:12 telex joined #perl6
09:13 anaeem1 joined #perl6
09:13 tadzik TIL: entering "gode" instead of "doge" in google images in unpleasant
09:14 vike joined #perl6
09:15 vendethiel tadzik: french says sorry
09:18 anaeem1 joined #perl6
09:20 tadzik oooh
09:20 tadzik this explains so many things
09:35 spider-mario joined #perl6
09:36 bbkr_ joined #perl6
09:37 bbkr_ joined #perl6
09:42 rurban joined #perl6
09:49 spider-mario joined #perl6
09:55 FROGGS jnthn: I hit a spesh bug, and I think that is the one tadzik++ already have seen
09:55 FROGGS instead of throwing this: Missing or wrong version of dependency 'src/Perl6/Grammar.nqp'
09:56 FROGGS I get a segfault in: MVM_interp_run (tc=tc@entry=0x6035e0, initial_invoke=initial_invoke@entry=0x7ffff7a03222 <toplevel_initial_invoke>, invoke_data=<optimized out>)
09:56 FROGGS at src/core/interp.c:4292
09:56 FROGGS 4292                if (IS_CONCRETE(check) && STABLE(check) == want_c) {
09:56 FROGGS (check is NULL)
09:57 jnthn Are you using MoarVM HEAD?
09:58 tadzik oh yes, that's what I was getting too
09:58 FROGGS jnthn: yes
09:58 jnthn OK
09:58 jnthn I assume you weren't doing the Panda install?
09:58 FROGGS weird is, I made clean in v5, but still get this
09:58 FROGGS jnthn: no panda involved
09:59 FROGGS so I am removing my Perl5/ in install..
09:59 jnthn ah, ok
10:01 dalek joined #perl6
10:01 FROGGS yeah, it continues now.... so is MoarVM+spesh messing with --libpath? should be very unlikely
10:02 FROGGS Error while compiling op setcodeobj: QAST::Block with cuid cuid_1_1400407321.74398 has not appeared
10:02 FROGGS make: *** [lib/Perl5/if.moarvm] Fehler 1
10:02 FROGGS :o(
10:03 FROGGS so, that seems to come from something that changed in rakduo/nqp/moarvm since This is perl6 version 2014.04-36-g87a6833 built on MoarVM version 2014.04-11-g4749729
10:03 jnthn oh, so that wasn't a regression introduced by my attempted refactors after all?
10:04 nwc10 bisect!
10:04 FROGGS jnthn: at least does not look like...
10:04 FROGGS but wait, I was able to build that yesternight
10:04 nwc10 (it is not obvious to me how to automate a reliable bisect of the threesome)
10:05 FROGGS to me neither
10:05 nwc10 as moarvm is supposed to be backwards compatible, and also nqp, I think the start is to hold moarvm and nqp constant and attempt earlier rakudo revisions until a good point is found
10:05 nwc10 then keep rakudo and moarvm constant and bisect nqp
10:06 nwc10 and then moarvm
10:10 xinming joined #perl6
10:12 FROGGS jnthn: stripping '$*ACTIONS := %*LANG<Perl5-actions>;' from Perl5.nqp triggers the setcodeobj problem
10:15 dmol joined #perl6
10:44 rurban joined #perl6
10:48 denis_boyun_ joined #perl6
10:49 takesako_____ joined #perl6
10:53 masak hm, I think we shouldn't start with the question "what's the percentage of *all* the CPAN modules that are pure?"
10:55 * FROGGS .oO( 0%? )
10:56 FROGGS shouldn't?
10:56 masak maybe start with the smaller question "how do the dependency trees look for these popular modules (like DBIx::Class, Moose, Mojolicious, App-cpanminus, Plack)?"
10:56 * masak just found https://metacpan.org/lab/dependencies?module=Moose
10:57 masak "(...Moo, DateTime, Dist-Zilla, Dancer, App-perlbrew, Try-Tiny, Catalyst, Test-Simple...)"
10:57 FROGGS yeah
10:58 masak I'm picturing the resulting graphs. they're pretty.
10:58 masak yes, I will do this.
10:58 masak tadzik++ # "We're good at libraries, but not at products"
10:59 FROGGS masak: http://deps.cpantesters.or​g/?module=DBIx%3A%3AClass
10:59 masak FROGGS++
11:00 woolfy PLPW is heavy with Perl 6...
11:00 masak \o/
11:04 masak ooh -- http://deps.cpantesters.org/static/purity.html
11:04 masak so deps.cpantesters.org can already show this information.
11:05 masak http://deps.cpantesters.org/​?module=Moose&amp;pureperl=1
11:14 moritz \o
11:15 lizmat moritz o/
11:16 masak moritz! \o/
11:16 masak moritz: we just saw tadzik demo Space Invaders!
11:16 masak er, tadzik++
11:17 moritz tadzik++
11:19 FROGGS hi moritz
11:20 masak "we are not a spam company. [...] we send over 1e8 emails per day." :P
11:20 flussence good *, * :)
11:24 masak *, flussence
11:24 masak huh -- the hackathon topic list has only Perl 6 things.
11:24 dalek rakudo/nom: 4ea6b62 | (Elizabeth Mattijsen)++ | docs/ChangeLog:
11:24 dalek rakudo/nom: Mention optimizations in grep/first and friends
11:24 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/4ea6b623d4
11:26 tadzik :)
11:27 flussence I've made a nice, self-contained, readable, installable... nondeterministic threading crash :D
11:28 flussence (try running `TEST_CORPUS_THREADED=1 panda install Test-Corpus` if you have nothing better to do :)
11:30 flussence (but be warned one of the failure modes is eating all your memory)
11:35 jnthn om nom nom memory
11:35 tadzik Sterods slides uploaded http://feather.perl6.nl/~tjs/talks/plpw14/
11:36 masak suggestion: move throws_like from Test::Util to Test
11:36 vendethiel tadzik++
11:36 pippo joined #perl6
11:46 flussence tadzik++
11:46 rurban joined #perl6
11:53 nnunley_ joined #perl6
11:59 takesako_____ joined #perl6
12:02 nwc10 FROGGS: ooh, curious. On "my" other machine: not ok 36 - seekfh with invalid whence fails
12:03 nwc10 (passes on x86_64 Linux, fails on pp64 Linux)
12:10 FROGGS hmmm
12:10 FROGGS weird
12:10 colomon tadzik++ # slides
12:19 masak moritz: would you be fine with moving throws_like into Test.pm ?
12:19 masak huh. Test::Util isn't on http://modules.perl6.org/
12:20 masak oh, it's in t/spec!
12:21 berekuk joined #perl6
12:22 masak colomon: Test::Junkie could probably be updated to use jnthn++'s new IO::Notification stuff.
12:23 masak (removing the sleep .5 in the core of the module)
12:23 colomon intended for [Coke] ?
12:23 * colomon tries to remember if he ever had anything to do with Test::Junkie
12:27 masak colomon: https://github.com/colomon/Test-Junkie/
12:37 colomon ah, yes, before we got the idea of moving modules that desperately needed a bug fix to perl6 ownership.  :)
12:37 colomon does IO::Notification work everywhere?
12:40 masak dunno
12:40 jnthn colomon: On MoarVM we use the libuv abstractions for it, which should work in a range of place,s though htere rae quickr
12:40 jnthn s
12:40 jnthn there are quirks
12:40 jnthn oops
12:40 colomon by everywhere, I meant, moar, jvm, parrot
12:40 masak hm, is there a way to indicate "import module M, and also pass all of its exports along as exports from this module"?
12:41 jnthn colomon: oh
12:41 jnthn colomon: No, then just Moar.
12:41 lizmat r: use lib "t/spec/packages"; use Test::Tap; ok 1
12:41 camelia rakudo-moar 4ea6b6: OUTPUT«===SORRY!===␤Could not find Test::Tap in any of: t/spec/packages, /home/p6eval/.perl6/2014.04-261-g4ea6b62/lib, /home/p6eval/rakudo-inst-1​/languages/perl6/site/lib, /home/p6eval/rakudo-inst-1/​languages/perl6/vendor/lib, /home/p6eval/rakudo-inst-1/languag…»
12:41 camelia ..rakudo-jvm 4ea6b6: OUTPUT«(timeout)»
12:41 camelia ..rakudo-parrot 4ea6b6: OUTPUT«===SORRY!===␤Could not find Test::Tap in any of: t/spec/packages, /home/p6eval/.perl6/2014.04-261-g4ea6b62/lib, /home/p6eval/rakudo-inst-1/lib/parrot/​6.1.0-devel/languages/perl6/site/lib, /home/p6eval/rakudo-inst-1/lib/pa​rrot/6.1.0-devel/languages/perl…»
12:41 jnthn colomon: None of the async stuff works on Parrot
12:41 jnthn colomon: We can do it on JVM, but not yet ported, afaik
12:43 atroxaper joined #perl6
12:44 takesako_____ joined #perl6
12:44 xenoterracide__ joined #perl6
12:44 colomon p6: say $*VM.name
12:44 camelia rakudo-moar 4ea6b6: OUTPUT«moar␤»
12:44 camelia ..niecza v24-109-g48a8de3: OUTPUT«Unhandled exception: Unable to resolve method name in type Any␤  at /tmp/tmpfile line 1 (mainline @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4595 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4596 (module-CORE @ 576) ␤  …»
12:44 camelia ..rakudo-jvm 4ea6b6: OUTPUT«jvm␤»
12:44 camelia ..rakudo-parrot 4ea6b6: OUTPUT«parrot␤»
12:45 colomon so we can easily do VM specific code now...
12:46 masak r: say "Hi from a $*VM VM!"
12:46 camelia rakudo-jvm 4ea6b6: OUTPUT«Hi from a jvm VM!␤»
12:46 camelia ..rakudo-moar 4ea6b6: OUTPUT«Hi from a moar VM!␤»
12:46 camelia ..rakudo-parrot 4ea6b6: OUTPUT«Hi from a parrot VM!␤»
12:46 masak :D
12:48 rurban joined #perl6
12:49 * vendethiel would really use labels here :)
12:51 lizmat r: say $*PERL.perl
12:51 camelia rakudo-jvm 4ea6b6: OUTPUT«{"name" => "rakudo", "compiler" => {"build-date" => "2014-05-18T12:12:02Z", "name" => "rakudo", "codename" => "", "ver" => "2014.04-261-g4ea6b62", "release-number" => ""}}␤»
12:51 camelia ..rakudo-moar 4ea6b6: OUTPUT«{"compiler" => {"ver" => "2014.04-261-g4ea6b62", "codename" => "", "build-date" => "2014-05-18T12:21:04Z", "release-number" => "", "name" => "rakudo"}, "name" => "rakudo"}␤»
12:51 camelia ..rakudo-parrot 4ea6b6: OUTPUT«{"name" => "rakudo", "compiler" => {"release-number" => "", "build-date" => "2014-05-18T12:27:00Z", "codename" => "", "ver" => "2014.04-261-g4ea6b62", "name" => "rakudo"}}␤»
12:51 lizmat for some reason, the current implementation of $*PERL has a compiler sub-hash (object)
12:52 lizmat is that something we want to keep?   or should the fields in $*PERL<compiler> be moved up one level?
12:52 lizmat (like I specced yesterday)
12:55 atroxaper joined #perl6
12:56 masak lizmat: well, $*PERL<compiler><ver> feels kinda different from $*PERL<ver>...
12:57 jnthn yeah, it's language spec version vs. rakudo version
12:57 rurban joined #perl6
12:57 masak lizmat: spec/S17-OLD.pod has been sitting still and doing nothing since the day I resurrected it, on 2013-11-03. can I remove it now?
12:59 dalek specs: ecfddec | masak++ | S17-concurrency.pod:
12:59 dalek specs: [S17] spec :expires named for .uniq
12:59 dalek specs: review: https://github.com/perl6/specs/commit/ecfddece39
13:00 masak lizmat: the spec for uniq/squish makes it seem that the :expires named has 1 second as a default, but the Rakudo implementation shows that that's not the case. can I change the spec to look like 1 is not the default?
13:01 jnthn masak: yeah, I don't think having a default there is useful
13:03 * moritz back from celbrating 200 years Norway
13:06 masak happy 200 years without Sweden, Norway! \o/
13:07 masak jnthn: oh -- there's nothing to fix. the thing in the spec is just a usage example.
13:11 jnthn oh :)
13:11 BenGoldberg joined #perl6
13:11 moritz masak: if it confused you, maybe it'll confuse others too
13:13 masak moritz: maybe.
13:13 masak moritz: I'm just used to S32 giving signatures, I think.
13:13 masak lizmat++ # teaching people Perl 6
13:14 dalek specs: 94cb8d8 | masak++ | S17-concurrency.pod:
13:14 dalek specs: [S17] improve the description of squish/uniq :expires
13:14 dalek specs: review: https://github.com/perl6/specs/commit/94cb8d8b1b
13:17 masak how was it again -- can I match a named argument against a positional parameter?
13:17 masak and under what conditions?
13:17 tadzik r: [0..2, 2..0].perl.say # that's... unexpected
13:17 camelia rakudo-{parrot,jvm,moar} 4ea6b6: OUTPUT«[0, 1, 2]␤»
13:18 masak tadzik: no, it's not.
13:18 masak tadzik: backwards ranges are *empty*. that's a *feature*. ;)
13:19 masak r: say [0..2, reverse 0..2].perl
13:19 jnthn use le sequence operator if you want to go the other way
13:19 camelia rakudo-{parrot,jvm,moar} 4ea6b6: OUTPUT«[0, 1, 2, 2, 1, 0]␤»
13:19 jnthn r: say 2...0
13:19 camelia rakudo-{parrot,jvm,moar} 4ea6b6: OUTPUT«2 1 0␤»
13:19 masak tadzik: if this is surprising to you, you haven't felt the pain of accidentally reversing a range and having it contain things.
13:19 tadzik oh
13:20 tadzik alrighty
13:22 masak m: sub numcmp ($x, $y) { return $x <=> $y }; say numcmp(x=>2, y=>7)
13:22 camelia rakudo-moar 4ea6b6: OUTPUT«Not enough positional parameters passed; got 0 but expected 2␤  in sub numcmp at /tmp/UNnhoXhp6N:1␤  in block  at /tmp/UNnhoXhp6N:1␤␤»
13:22 moritz we got rid of that feature.
13:22 masak ladies and gentlebots, we have either a spec fossil, or a fairly serious NYI.
13:22 moritz spec fossil
13:22 masak moritz: right, so fossil.
13:22 * masak removes it
13:22 moritz it used to be a feature, but was removed
13:23 masak there's a guy sitting next to me who was very disappointed that he couldn't get this example to run. :/
13:24 moritz tell him to put a : before the parameters
13:24 moritz and that $x and $y are lexicals, and thus *private*, including the amname
13:24 dalek specs: 25eb879 | masak++ | S06-routines.pod:
13:24 dalek specs: [S06] remove fossil
13:24 dalek specs:
13:24 dalek specs: We no longer map named arguments to positional parameters.
13:24 dalek specs: review: https://github.com/perl6/specs/commit/25eb879f36
13:24 moritz s/am//
13:25 jnthn Yeah, that decision dates back some years now...certainly fossil
13:26 masak *nod*
13:28 woolfy1 joined #perl6
13:28 takesako_____ joined #perl6
13:29 masak guy next to me solved it with multis, the named multi delegating to the positional one.
13:29 berekuk joined #perl6
13:29 masak but he's also disappointed. in his opinion, S06 was right before I removed the fossil :)
13:30 masak the best I can offer him is a maybe-future module that restores the old behavior, but that's not much consolation now.
13:30 masak (he seems to have a Python background, and Python does the named-to-positional mapping)
13:31 jnthn It works great until you improve/correct a parameter name, release a new version of your module, and then find you broke folk's code...
13:31 jnthn *folk's
13:33 masak aye.
13:33 masak jnthn: I remember us walking along a Russian river and discussing this :)
13:41 Sqirrel joined #perl6
13:41 jnthn aye
13:43 masak grrr
13:43 jnthn ?
13:43 masak turns out .squish(:expires) is hard to write tests for.
13:43 jnthn I'll bet
13:43 masak because it's so *boring*
13:43 jnthn We could use a TestScheduler
13:43 * masak tries again, in a better mood :)
13:43 masak jnthn: I'll get to the TestScheduler, I promise.
13:44 masak after I implement .squish(:expires)
13:44 masak yes, that works better.
13:45 masak turns out I only need two tests.
13:45 jnthn No, no, you supply it, not promise it :P
13:49 kaare_ joined #perl6
13:51 btyler joined #perl6
13:54 masak m: say get
13:54 camelia rakudo-moar 4ea6b6: OUTPUT«There were three men came out of the West␤»
13:54 masak m: say get(2)
13:54 camelia rakudo-moar 4ea6b6: OUTPUT«No such method 'get' for invocant of type 'Int'␤  in sub get at src/gen/m-CORE.setting:14233␤  in block  at /tmp/kt24kd3jSJ:1␤␤»
13:55 masak m: my @lines := gather { take get; take get; take get }; say "OH HAI"; say @lines.elems
13:55 camelia rakudo-moar 4ea6b6: OUTPUT«OH HAI␤3␤»
13:57 masak guy-next-to-me just ran into the fact that lines() isn't lazy :/
13:57 masak https://rt.perl.org/Public/​Bug/Display.html?id=112716
13:57 masak but he wrote a gather which a loop in it, and he seems happy.
13:57 jnthn masak: huh, TimToady fixed that quite a while ago, I thought...
13:58 moritz masak: https://gist.github.com/mo​ritz/08274d7e01ef4ff2c244 for the guy next to you
13:59 moritz how to write a primitive binder, part 1 :-)
14:00 jnthn m: for lines() { say $*IN.ins }
14:00 camelia rakudo-moar 4ea6b6: OUTPUT«0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0​␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0​␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤0␤»
14:00 jnthn Hm
14:00 moritz $*ARGS, not $*IN
14:00 moritz m: for lines() { say $*ARGS.ins }
14:00 camelia rakudo-moar 4ea6b6: OUTPUT«===SORRY!===␤Dynamic variable $*ARGS not found␤»
14:00 moritz m: for lines() { say $*ARGFILES.ins }
14:00 camelia rakudo-moar 4ea6b6: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤11␤12␤13␤14␤15␤16␤17␤​18␤19␤20␤21␤22␤23␤24␤25␤26␤27␤28␤29␤30␤31␤32␤33␤3​4␤35␤36␤37␤38␤39␤40␤41␤42␤43␤44␤45␤46␤47␤48␤49␤»
14:00 jnthn Ah, yes
14:01 jnthn masak: And perl6-m -e ".say for lines()" works and would not if it weren't lazy
14:01 jnthn masak: So I suspect some other misunderstanding
14:01 raiph joined #perl6
14:01 moritz or ancient-ish rakudo version
14:02 masak m: class C { has $.foo }; .say for C.^attributes
14:02 camelia rakudo-moar 4ea6b6: OUTPUT«Mu $!foo␤»
14:02 masak jnthn: the guy ran some process thing and saw that it read all the lines of the file up front.
14:03 masak jnthn: do you have a good example of creating an attribute dynamically through the MOP?
14:03 guru joined #perl6
14:03 masak I'm talking to dexter++, by the way :)
14:06 jnthn masak: Yes, thus why I'm asking what was done ;)
14:06 jnthn Remember assignment is mostly eager
14:06 jnthn masak: timo's ADT module may well do that
14:08 masak oh right, that's where it was.
14:09 masak jnthn: the code was `my @w := $fh.lines`
14:10 masak timotimo++ # ADT
14:13 jnthn masak: I still don't see the issue
14:13 jnthn >perl6-m -e "my $fh = open('README.md'); my @b := $fh.lines; say $fh.ins; say @b[0]; say $fh.ins"
14:13 jnthn 0
14:13 jnthn # Rakudo Perl 6
14:13 jnthn 1
14:13 jnthn masak: How did he ascertain it read ahead?
14:14 masak jnthn: strace
14:14 masak jnthn: you can come over here and look if you want :)
14:16 masak (he did.)
14:16 atroxaper joined #perl6
14:18 fhelmberger joined #perl6
14:18 masak ...turns out it was a sinking issue with the last statement of the program.
14:20 darutoko- joined #perl6
14:21 skids joined #perl6
14:25 zakharyas joined #perl6
14:28 vendethiel moritz++ # cool gist
14:30 ggoebel111115 joined #perl6
14:32 dalek nqp: b85f70e | (Tobias Leich)++ | tools/build/MOAR_REVISION:
14:32 dalek nqp: bump moar rev for bigint fixes
14:32 dalek nqp: review: https://github.com/perl6/nqp/commit/b85f70eda9
14:35 dalek rakudo/nom: 70ca03e | (Tobias Leich)++ | tools/build/NQP_REVISION:
14:35 dalek rakudo/nom: bump nqp/moar rev for bigint fixes
14:35 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/70ca03edeb
14:38 vendethiel FROGGS: will that make a huge difference ?
14:44 dalek roast: a4f8732 | (Tobias Leich)++ | S32-num/rat.t:
14:44 dalek roast: unfudge test about bigints, RT 121527
14:44 dalek roast: review: https://github.com/perl6/roast/commit/a4f873212e
14:45 masak moritz: yes, cool gist. I think it went mostly unappreciated in this case, though :)
14:46 FROGGS vendethiel: fixing a bug always makes a difference :o)
14:46 vendethiel definitely underappreciated!
14:48 moritz The Underappreciated 10-Minute Hack!
14:49 moritz FROGGS++
14:49 dalek rakudo/eleven: 8bb2c46 | (Tobias Leich)++ | src/core/terms.pm:
14:49 dalek rakudo/eleven: assign to PROCESS.WHO instead of bind, so we can "temp" things
14:49 dalek rakudo/eleven: review: https://github.com/rakudo/rakudo/commit/8bb2c46ade
14:51 dalek rakudo/nom: 834a415 | (Tobias Leich)++ | src/core/terms.pm:
14:51 dalek rakudo/nom: assign to PROCESS.WHO instead of bind, so we can "temp" things
14:51 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/834a4152ff
14:51 FROGGS more free karma \o/
14:53 Alina-malina joined #perl6
14:54 masak hm, Heroku seems to have undocumented support for Perl 5.
14:55 dalek panda/depends: bdbe638 | (Tobias Leich)++ | lib/Panda/Bundler.pm:
14:55 dalek panda/depends: do not try to jsonify match objects
14:55 dalek panda/depends: review: https://github.com/tadzik/panda/commit/bdbe6389ac
14:59 masak I would like to do Perl 6 in Heroku, but I'm not 100% sure where to start. still researching :)
15:03 masak aha.
15:04 masak we need to be a "Third-Party Buildpack". https://devcenter.heroku.com/a​rticles/third-party-buildpacks
15:04 masak miyagawa++ # for the Perl 5 buildpack
15:05 masak so it seems that the proximal goal is to create such a buildpack for Rakudo Moar.
15:07 jnthn yay :)
15:08 masak wow, this looks pretty easy, actually.
15:09 Pleiades` joined #perl6
15:11 masak mainly need to write a setup script like this one: https://github.com/miyagawa/heroku-bu​ildpack-perl/blob/master/bin/compile
15:12 masak and one can try it out using `heroku create myapp --buildpack https://github.com/some/buildpack.git` while experimenting.
15:12 masak yep, this feels eminently doable.
15:13 lizmat hmmm... seems #118601 is fixed  :-)
15:13 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=118601
15:14 dalek roast: 064ab9a | (Elizabeth Mattijsen)++ | S32-str/sprintf.t:
15:14 dalek roast: Unfudge #118601 test
15:14 dalek roast: review: https://github.com/perl6/roast/commit/064ab9abf9
15:14 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=118601
15:16 takesako_____ joined #perl6
15:16 molaf_ joined #perl6
15:17 dalek roast: 09d0bc5 | (Tobias Leich)++ | S01-perl-5-integration/basic.t:
15:17 dalek roast: skip code that tries to "use v5"
15:17 dalek roast: review: https://github.com/perl6/roast/commit/09d0bc559b
15:20 tadzik somebody claimed today that you can't fork more than 7 times in rakudo or it will crash. I checked. I can run 10000 forked rakudo processes and generate a load of 3600 without crashing :)
15:21 flussence were they using an OS without a real fork()? :)
15:22 dalek rakudo/nom: 3216e73 | (Tobias Leich)++ | src/Perl6/ (2 files):
15:22 dalek rakudo/nom: allow to switch to Perl 5 module by switching to a foreign statementlist
15:22 dalek rakudo/nom:
15:22 dalek rakudo/nom: This makes "use v5" work, though this is not the ideal patch for slangs
15:22 dalek rakudo/nom: implemented in Perl 6 itself. We only do it this way for now to have
15:22 dalek rakudo/nom: something at all. jnthn++ for reviewing
15:22 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/3216e7334a
15:22 tadzik maybe :)
15:23 masak rn: say sprintf "%x %x", 301281685344656640, 301281685344656669
15:23 camelia rakudo-moar 4ea6b6: OUTPUT«42e5e18b84c9d00 42e5e18b84c9d00␤»
15:23 camelia ..rakudo-{parrot,jvm} 4ea6b6, niecza v24-109-g48a8de3: OUTPUT«42e5e18b84c9d00 42e5e18b84c9d1d␤»
15:24 masak lizmat: doesn't look fixed on moar :/
15:25 lizmat does the test file "S32-str/sprintf.t" work for you ?
15:25 moritz masak: note that camelia's rakudo version doesn't yet include FROGGS++'s patch
15:26 FROGGS true
15:26 lizmat I pulled, rebuild from scratch and got a TODO passed on that file
15:26 lizmat that's all I know  :-)
15:26 nebuchadnezzar joined #perl6
15:27 FROGGS perl6-m -e 'say sprintf "%x %x", 301281685344656640, 301281685344656669'42e5e18b84c9d00 42e5e18b84c9d1d
15:27 FROGGS LF missing, but yeah, one get the idea :o)
15:28 masak ok, FROGGS++ lizmat++ then :)
15:29 atroxaper joined #perl6
15:51 woolfy1 What?  What?  says lizmat.  Answer from jnthn: huh?  oh...  hmmm...  What?  What?   to which lizmat responds: exactly, "WHAT???"
15:51 woolfy1 WAT?
15:51 woolfy1 lizmat told me I should say "WAT?"
15:52 tadzik I'd like some kind of DESTROY :/
15:52 woolfy1 sergot_ listens to all this and smiles (what else can you do)
15:53 woolfy1 but we have chocolate cookies
15:53 sergot_ Yes, we have! :)
15:53 kaare_ joined #perl6
15:53 woolfy1 tadzik: I can walk over to you and destroy you, if you really want!
15:54 tadzik woolfy1: can I hire you? I have a lot of objects that need destroying at semi-random moments :)
15:55 woolfy1 Hmm. I am offered a job at destroying tadzik's notebook, than he tells me I also should destroy the notebooks of FROGGS and jnthn and lizmat and sergot
15:57 tadzik why do you think yours is safe :o
15:57 tadzik captain sinking with her ship, or something?
16:01 takesako_____ joined #perl6
16:02 lizmat joined #perl6
16:02 masak wow, cool. http://evening-atoll-4752.herokuapp.com/ -- my first Heroku app deployed.
16:02 masak it's kinda cool to have something be deployed on `git push`.
16:02 masak (disappointment warning: this is Ruby/Sinatra, not Perl 6)
16:02 jnthn aww :P
16:02 masak :)
16:02 masak but I don't believe the step to Perl 6 will be that big.
16:03 jnthn .oO( To be Frank, I'm disappoint... :P )
16:03 masak need to have it install Moar, nqp, Rakudo, panda, and Bailador.
16:03 woolfy joined #perl6
16:04 hoverboard joined #perl6
16:06 xfix std: null
16:06 camelia std ec3d853: OUTPUT«[31m===[0mSORRY![31m===[0m�Undeclared routine:� 'null' used at line 1�Check failed�FAILED 00:01 120m�»
16:06 xfix std: NULL
16:06 camelia std ec3d853: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Undeclared name:â�¤    'NULL' used at line 1â�¤Check failedâ�¤FAILED 00:01 120mâ�¤Â»
16:07 xfix Hm, I wonder, could it be possible to have a special error for null?
16:07 tgt joined #perl6
16:08 tadzik why?
16:08 tadzik it is an undeclared name :)
16:09 masak xfix: why do you feel the need for a special error?
16:09 * vendethiel just crashed his perl5 by typing (-v)
16:10 xfix std: undef
16:10 camelia std ec3d853: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Unsupported use of undef as a value; in Perl 6 please use something more specific:â�¤  an undefined type object such as Any or Int,â�¤   :!defined as a matcher,â�¤        Any:U as a type constraint,â�¤    Nil as the absense of an expected valu…»
16:10 xfix Because undef has a special error.
16:11 masak yes. because undef has a history in Perl 5.
16:11 masak NULL and null don't.
16:11 xfix std: if (True) { } else if (False) {}
16:11 camelia std ec3d853: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Please use 'elsif' at /tmp/JYhJLEy2BT line 1:â�¤------> [32mif (True) { } else if [33mâ��[31m(False) {}[0mâ�¤Confused at /tmp/JYhJLEy2BT line 1:â�¤------> [32mif (True) { } else if [33mâ��[31m(False) {}[0mâ�¤    expecting …»
16:11 xfix else if doesn't exist in Perl 5.
16:12 tadzik heh
16:12 xfix I personally think that errors should help moving from other languages.
16:12 flussence p5: ping
16:12 flussence eval ping
16:12 flussence eval: ping?
16:12 vendethiel ENOP5
16:12 flussence :(
16:13 moritz p5eval is down due to lack of hardware by its maintainer, I've heard
16:13 masak xfix: drop it. there's no use case for someone coming from another language and being confused because Perl 6 doesn't have a custom error message for it. :)
16:14 mtk0 joined #perl6
16:14 FROGGS joined #perl6
16:15 xfix std: <?php phpinfo(); ?>
16:15 camelia std ec3d853: OUTPUT«ok 00:01 121m␤»
16:15 xfix Hm, I guess this helps with polyglots.
16:16 xfix (if someone is crazy enough to make polyglot with PHP)
16:17 flussence std: <!DOCTYPE HTML 2.0> <P/Hello world/>
16:17 camelia std ec3d853: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Preceding context expects a term, but found infix > instead at /tmp/PNotHxzXjj line 1:â�¤------> [32m<!DOCTYPE HTML 2.0> <P/Hello world/[33mâ��[31m>[0mâ�¤Undeclared name:â�¤    'P' used at line 1â�¤Undeclared routine:â�¤ 'world' …»
16:17 flussence ack, caught...
16:17 xfix You probably meant <P/Hello world/
16:17 xfix If you want to use SHORTTAGS.
16:17 flussence it's been a while since I actually wrote SGML...
16:18 flussence std: <foo/bar/>
16:18 camelia std ec3d853: OUTPUT«ok 00:01 121m␤»
16:18 flussence std: <foo/bar baz/>
16:18 camelia std ec3d853: OUTPUT«ok 00:01 121m␤»
16:19 flussence LTA ttiar error?
16:19 flussence std: <foo> <bar>
16:19 camelia std ec3d853: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Bogus term at /tmp/drCGnGo7EH line 1 (EOF):â�¤------> [32m<foo> <bar>[33mâ��[31m<EOL>[0mâ�¤Undeclared routine:â�¤        'bar' used at line 1â�¤Parse failedâ�¤FAILED 00:01 123mâ�¤Â»
16:19 flussence r: <foo> <bar>
16:20 masak flussence: no, because the second '<' is read as an operator.
16:20 camelia rakudo-jvm 834a41: OUTPUT«(timeout)»
16:20 camelia ..rakudo-{parrot,moar} 834a41: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfileâ�¤Two terms in a rowâ�¤at /tmp/tmpfile:1â�¤------> [32m<foo> <bar>[33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        argument listâ�¤        postfixâ�¤        infix stopper…»
16:20 flussence r-pm gets it :)
16:20 masak don't really know how.
16:21 vendethiel LTM ?
16:21 moritz LTM, I guess
16:21 vendethiel *the magical word*
16:21 moritz there's a 'term' infix for generating the error message
16:21 * masak waves the magical LTM explanation
16:21 flussence r-m: <foo> < bar>
16:21 camelia rakudo-moar 834a41: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/6hQZJjKlcHâ�¤Two terms in a rowâ�¤at /tmp/6hQZJjKlcH:1â�¤------> [32m<foo> < bar>[33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        argument listâ�¤        postfixâ�¤        infix stopperâ�¤ …»
16:21 moritz and it parses more than the infix operator itself
16:21 xfix flussence: Also, you are probably aware of that, but practically to HTML web browser (no, W3C Validator is not web browser) implements SGML SHORTTAG.
16:22 xfix (well, technically emacs-w3 supports SGML SHORTTAG, but that's all)
16:22 flussence I wonder if that Amaya thing does it "correctly"...
16:23 xfix No.
16:24 xfix Even W3C doesn't care.
16:24 xfix Also, these days many browsers implement HTML5 parsing, which forbids SGML SHORTTAG support.
16:25 xfix And SGML SHORTTAG support is incompatible with XHTML1 Appendix C, which recommends using syntax like <br /> (with SGML SHORTTAG support, it's parsed as <br>&gt;).
16:26 Psyche^_ joined #perl6
16:27 FROGGS m: class Foo { method bar { my regex baz { . }; "a" ~~ /<baz>/ } }; say Foo.bar
16:27 camelia rakudo-moar 834a41: OUTPUT«Type check failed in binding <anon>; expected 'Foo' but got 'Cursor'␤  in regex baz at /tmp/DSUtikOSXt:1␤  in method INDRULE at src/gen/m-CORE.setting:13063␤  in method ACCEPTS at src/gen/m-CORE.setting:13136␤  in method bar at /tmp/DSUtikOSXt:1␤…»
16:27 FROGGS why?? ó.ò
16:27 itz joined #perl6
16:27 jnthn It probably took the surrounding class as the cursor type for the regex...
16:28 xfix Appendix C shouldn't have been written IMO. It allows writting XHTML with text/html, which is just wrong. And this HTML is invalid HTML that just happens to work in many web browsers.
16:28 jnthn I guess for my scoped ones it maybe should not do that.
16:32 xfix It also annoys me that Perl 6 website claims to be XHTML from doctype, but sends itself as text/html, and if it would be parsed as actual XHTML, the browser would simply fail to parse the website.
16:33 dalek specs: b7de316 | (Elizabeth Mattijsen)++ | S28-special-names.pod:
16:33 dalek specs: More precisely spec $?PERL / $*PERL
16:33 dalek specs: review: https://github.com/perl6/specs/commit/b7de316d28
16:34 anaeem1_ joined #perl6
16:34 moritz xfix: you're welcome to implement content negotiation that serves it as application/xml+xhtml if the browser accepts it
16:35 xfix That's not the thing.
16:35 moritz that's exactly the thing.
16:35 xfix The website uses &nbsp;, and none of browsers load entities for XML documents.
16:35 xfix The validator may claim that &nbsp; is valid, but browsers won't download DTD just to get entities.
16:36 moritz you can only serve xhtml with the proper content-type, because $older_browsers
16:36 moritz xfix: so what?
16:36 moritz xfix: they have integrated the (now immutable) DTD into the html parser
16:36 moritz that's a valid optimization.
16:37 dalek rakudo/nom: 0a6c98c | (Elizabeth Mattijsen)++ | / (7 files):
16:37 dalek rakudo/nom: Properly implement $*PERL according to spec
16:37 dalek rakudo/nom:
16:37 dalek rakudo/nom: This introduces the Perl and Compiler classes
16:37 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0a6c98c05d
16:37 xfix Hm, ok.
16:37 xfix Well, I guess I confused it with something else.
16:39 SamuraiJack__ joined #perl6
16:39 xfix <link rel="icon" href="/favicon.ico" type="favicon.ico" />
16:39 xfix There is also that.
16:39 xfix type="favicon.ico" is not valid relation type, but XHTML validator depends on DTD which cannot tell that.
16:39 moritz wha' abou' it?
16:39 xfix Not that this causes parse errors, but it's still invalid XHTML.
16:40 moritz what's invalid?
16:40 xfix type="favicon.ico"
16:40 moritz what's invalid?
16:40 woolfy left #perl6
16:40 xfix favicon.ico is not MIME type.
16:41 dalek rakudo/nom: 0a2f0cd | (Elizabeth Mattijsen)++ | docs/ChangeLog:
16:41 dalek rakudo/nom: Mention $*PERL upgrade
16:41 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0a2f0cd95b
16:41 moritz and the DTD say it must be?
16:42 xfix CDATA
16:42 xfix Which means anything.
16:42 beastd joined #perl6
16:42 dalek rakudo/nom: 57a08ad | (Tobias Leich)++ | src/Perl6/Actions.nqp:
16:42 dalek rakudo/nom: fix my scoped regexes in classes
16:42 dalek rakudo/nom:
16:42 dalek rakudo/nom: Regexes tried to install itself into the class, and then exploded by getting a Cursor instead
16:42 dalek rakudo/nom: of the type of the class.
16:42 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/57a08ad7c6
16:44 moritz so, what's invalid?
16:45 xfix http://www.w3.org/TR/html401/​struct/links.html#adef-type-A
16:45 dalek roast: 6fac9f4 | (Elizabeth Mattijsen)++ | S17-supply/interval.t:
16:45 dalek roast: We're using Perl 6 here!
16:45 dalek roast: review: https://github.com/perl6/roast/commit/6fac9f488f
16:46 SamuraiJack__ joined #perl6
16:46 xfix Proper MIME types have `/` character. For example, `text/html`.
16:46 flussence I'll be travelling for the next 4h or so; I'll go through and WHATWG-HTMLify the source in the meantime to fix any lingering problems :)
16:46 itz is there any sort of beer meeting on Monday in Prague?  I don't see anything on the website apart from a food event on the Tuesday night
16:47 moritz xfix: "This attribute gives an advisory hint as to the content type of the content available at the link target address." doesn't say it must be a content type
16:47 jnthn itz: I didn't see anything on the website either
16:47 moritz xfix: only that it provides a hint towards the content type
16:48 xfix <moritz>21 xfix: "This attribute gives an advisory hint as to the content type of the content available at the link target address." doesn't say it must be a content type
16:48 xfix As for definition of "advisory hint", it means that it doesn't have to be specified, not that you may write anything here.
16:49 moritz xfix: where is that defined?
16:50 xfix http://www.w3.org/TR/html401/​types.html#type-content-type
16:50 moritz I read "advisory" as "the UA shouldn't consider it definitive, but rather use it only to decide whether to download the resource, and take the content type from that second request as definitive"
16:51 xfix To be honest, DTD defines it as  %ContentType;, except  %ContentType; is alias for CDATA.
16:51 xfix Anyway, I wonder, would it be possible to move from XHTML sent as text/html to HTML5 for Perl 6 webpage?
16:52 moritz it's a spec. If the spec wants it to be a content type, it could just state clearly that it needs to be a content type
16:52 jnthn dinner &
16:52 moritz since it doesn't, authors have some liberties
16:52 moritz xfix: what would be the benefit?
16:54 xfix Sure, if you think this way, let's all write <META http-equiv="Content-Type" content="I don't know">.
16:54 moritz if the content type is the main concern, fix the content type
16:54 xfix DTD allows that.
16:55 IllvilJa joined #perl6
16:59 dalek perl6.org: a763f62 | (Konrad Borowski)++ | includes/header:
16:59 dalek perl6.org: Remove type="favicon.ico"
16:59 dalek perl6.org:
16:59 dalek perl6.org: It's not valid according to XHTML standard, and doesn't mean anything.
16:59 dalek perl6.org: Therefore, it can be easily removed.
16:59 dalek perl6.org: review: https://github.com/perl6/p​erl6.org/commit/a763f627ab
16:59 xfix karma Konrad Borowski
17:00 moritz xfix: that's not quite true. If user agents understand it, it has value.
17:00 xfix Yes, it's represented in DOM.
17:00 moritz and no, the xhtml standard does not forbid it.
17:00 xfix But it doesn't mean anything.
17:00 moritz it means whatever the communication partners agree on that it means
17:01 xfix Well, XHTML expects content type here (according to DTD). While it's true that %ContentType is aliased to CDATA, it's still content type.
17:01 * moritz gives up
17:01 dalek perl6.org: 3f0dcf6 | flussence++ | includes/header:
17:01 dalek perl6.org: Change doctype to WHATWG-HTML
17:01 dalek perl6.org: review: https://github.com/perl6/p​erl6.org/commit/3f0dcf65d6
17:01 dalek perl6.org: 5993415 | flussence++ | includes/header:
17:01 dalek perl6.org: Fix cosmetic errors in <head>
17:01 dalek perl6.org:
17:01 dalek perl6.org:   - Add correct mime type to favicon <link>
17:01 dalek perl6.org:   - Remove redundant <style> type attribute
17:01 dalek perl6.org:   - Fix a template variable name typo (but it was unused anyway)
17:01 dalek perl6.org: review: https://github.com/perl6/p​erl6.org/commit/599341528d
17:01 * flussence &
17:02 xfix <link rel="icon" href="/favicon.ico" type="favicon.ico" />
17:02 xfix At this point I wonder why I even bother.
17:02 xfix favicon.ico is not MIME type.
17:03 dalek perl6.org: 75b4f5d | flussence++ | includes/header:
17:03 dalek perl6.org: Fix duplicate line typo
17:03 dalek perl6.org: review: https://github.com/perl6/p​erl6.org/commit/75b4f5de0b
17:09 xfix On 25/Dec/2010, Perl 6 website had 4636 unique visits. Interesting...
17:10 lizmat joined #perl6
17:10 colomon joined #perl6
17:11 IllvilJa joined #perl6
17:12 masak dinner &
17:20 dalek perl6.org: acea17d | (Konrad Borowski)++ | source/index.html:
17:20 dalek perl6.org: Use alt which accurately describes download image
17:20 dalek perl6.org: review: https://github.com/perl6/p​erl6.org/commit/acea17daff
17:25 HiriB joined #perl6
17:32 HiriB left #perl6
17:46 takesako_____ joined #perl6
17:55 amkrankruleuen joined #perl6
18:14 plobsing joined #perl6
18:44 anaeem1 joined #perl6
18:45 itz joined #perl6
18:48 molaf joined #perl6
18:51 takesako_____ joined #perl6
19:04 prevost joined #perl6
19:10 takesako_____ joined #perl6
19:15 xilixy joined #perl6
19:15 xilixy left #perl6
19:21 timotimo much hackathon?
19:21 tadzik timotimo: http://feather.perl6.nl/~tjs/talks/plpw14/#/16
19:22 tadzik hackathon's over; I'm on the train home, everyone else is dinnering afaik
19:23 berekuk joined #perl6
19:29 anaeem1 joined #perl6
19:35 bjz_ joined #perl6
19:42 denis_boyun joined #perl6
19:51 timotimo i take it the conference is over as well?
19:51 timotimo i spent the day renovating the future apartment of friends
19:51 tadzik yes, it's over
19:52 tadzik it was really good
19:52 tadzik better than the last year's PLPW
19:52 tadzik we're evolving \o/
19:52 timotimo way cool :)
19:53 tadzik and a new PM group has formed and volunteered to organize the next year's workshop in Gdańsk
19:53 timotimo we started out thinking we'd be replacing the wallpaper in one room and paint over the existing wallpaper in all other rooms
19:53 tadzik so, awesome :)
19:53 timotimo now we've taken the wallpaper off of the walls in three different rooms and one room is still to be done
19:53 tadzik and the steroids talk was well received, I like that :)
19:53 timotimo yays :)
19:53 timotimo moar is only going to get better for stuff :)
19:54 tadzik it is getting better all the time :)
19:54 tadzik so, no wallpaper after all? :)
19:54 timotimo we're only keeping half of the kitchen and three quarters of the bedroom and we are not doing any of the ceilings
19:56 timotimo it would have been over really quickly if the wallpaper would have come off in one layer
19:56 tadzik :D
19:56 timotimo but there was quite a lot of residue we had to remove "manually" with spatulas
19:56 timotimo and the people who did the renovation last time(s) weren't very good at it
19:56 timotimo there were many WTF per hour
19:56 tadzik sounds like an extra layer of abstraction has been solving problems before :)
19:57 timotimo yeah, at least one party was contempt with just painting over the existing layer
19:57 timotimo in some parts it seemed like there were three layers of (distinct) paint
19:57 timotimo and there was wallpaper and paint behind the radiators
19:57 tadzik well, that was your initial plan too :)
19:57 timotimo you're right. we got a spark of sanity, though
19:58 timotimo also, in the kitchen there was actually a layer of paint below the wallpaper
19:58 timotimo which is just ... wtf?
19:58 tadzik so, soemone was content with just wallpapering over the existing paint :P
19:58 tadzik so, the apartment is a bit bigger now, no? :P
19:59 timotimo aye, we've certainly won a few cubic centimeters
19:59 timotimo but the paint that was under the wallpaper had been applied directly to the wall
20:01 timotimo you know anchor bolts, right?
20:01 timotimo what do you do with them if you no longer need the hole the screw used to be in?
20:02 timotimo apparently someone thought it would be a great idea to leave the anchor bolts in and fill them with filling mass stuff
20:02 tadzik uh
20:02 tadzik well, I just leave them be, I'll probably want to hang something in the future
20:02 tadzik and they don't look all that bad, at least in the quantity that I have them
20:02 timotimo you have hundreds? :)
20:03 tadzik no, I may have like, 10 ;)
20:03 timotimo OK
20:03 timotimo well, at least you wouldn't leave them in and fill them up with putty
20:03 tadzik nah
20:03 timotimo you'd keep them or pull them out before filling the hole
20:03 tadzik unless I wanted to hide them from a landloard or something
20:03 timotimo anyway, i'm going to take a shower and unwind my stack at the local hackspace
20:03 timotimo that makes no sense :)
20:04 tadzik well, you'd still want to pull them out probably :)
20:04 tadzik unless they were stuck or something, I dunno
20:04 tadzik the idea of being forbidden from keeping a dog by a landlord is insane enough to me
20:04 tadzik and having it written down in a contract that a gecko is okay
20:07 dwarring joined #perl6
20:15 xenoterracide__ joined #perl6
20:19 fhelmberger joined #perl6
20:25 timotimo %)
20:28 awwaiid are hyper-operators (allowed to) run in parallel if the implementation wants?
20:29 skids As long as the results are ordered correctly.
20:29 skids Side effects need not be ordered.
20:30 awwaiid cool. that is the programmer's problem for depending on side-effects in such a world, eh?
20:31 skids Well, it's the programmer's fault for relying on side-effect ordering when the specification says you cannot.
20:32 skids I don't generally hate on side effects to the extent that the heavy FP folks do, personally.
20:33 timotimo m: (1, 2, 3, 4, 5, 6)>>.&say
20:33 camelia rakudo-moar 57a08a: OUTPUT«6␤4␤2␤5␤3␤1␤»
20:34 timotimo we actually shuffle around the values if you do a hyperop, even though we don't parallelize yet
20:35 timotimo just to make extra sure the programmer doesn't rely on the ordering
20:35 timotimo though it is still deterministic in our implementation
20:38 awwaiid timotimo: why did you do &say instead of (1,2,3,4)>>.say ?
20:39 timotimo oh
20:39 timotimo i forgot there's a say method :)
20:39 timotimo you can .& any sub in scope
20:39 awwaiid I can do 3.^methods, but I don't see "+" as a method -- is there a way to see that through introspection?
20:39 awwaiid I'm going to give a talk or two on perl6 and I'm brushing up :)
20:39 skids + is a sub.
20:39 timotimo you can go through SETTING::.pairs
20:43 skids or &infix:<+>.candidates.say if you already know the name.
20:45 * skids tries to fathom how to go about debugging a moar "inconsistent bind result"
20:46 timotimo well, if you go through the SETTING::.pairs (or is it CORE::?) you can look at the signatures and find anything that matches your type or parent types
20:46 skids timitimo: SETTING::.pairs worked for me.
20:48 awwaiid what do I do with SETTING::.pairs? I see some stuff about Mu that isn't too helpful
20:49 awwaiid I like that &infix:<+>.candidates.say, it's pretty interesting
20:49 skids .oO(HOWTO curried introspection)
20:49 awwaiid &infix:<+>.candidates>>.perl>>.say even :)
20:50 berekuk joined #perl6
20:52 awwaiid I like to know some of this basic introspection stuff, makes it easier to poke around on the REPL without looking at documentation or source :)
20:53 dwarring r: &infix:<+>.candidates>>.perl>>.say
20:53 camelia rakudo-parrot 57a08a: OUTPUT«sub infix:<+> (Int:D $x, Date:D $d) { #`(Sub|-6805760002719140261) ... }␤sub infix:<+> (Duration:D $a, Duration:D $b) { #`(Sub|-6805760002719122110) ... }␤sub infix:<+> (Duration:D $a, Real $b) { #`(Sub|-6805760002719105715) ... }␤sub infix:<+> (In…»
20:53 camelia ..rakudo-jvm 57a08a: OUTPUT«sub infix:<+> (Int:D $x, Date:D $d) { #`(Sub|780530416) ... }␤sub infix:<+> (Duration:D $a, Duration:D $b) { #`(Sub|2044305003) ... }␤sub infix:<+> (Duration:D $a, Real $b) { #`(Sub|1107153240) ... }␤sub infix:<+> (Instant:D $a, Duration:D $b) { #`(Su…»
20:53 camelia ..rakudo-moar 57a08a: OUTPUT«sub infix:<+> (Int:D $x, Date:D $d) { #`(Sub|72812848) ... }␤sub infix:<+> (Duration:D $a, Duration:D $b) { #`(Sub|72570432) ... }␤sub infix:<+> (Duration:D $a, Real $b) { #`(Sub|72570144) ... }␤sub infix:<+> (Instant:D $a, Duration:D $b) { #`(Sub|72…»
20:57 skids m: &infix:<+>.assuming(3).candidates.say # ?
20:57 camelia rakudo-moar 57a08a: OUTPUT«sub CURRIED (*@pos, *%named) { #`(Sub|140328309456552) ... }␤»
21:01 dwarring r: &infix:<+>.candidates>>.signature>>.say
21:01 camelia rakudo-{parrot,jvm,moar} 57a08a: OUTPUT«:(Int:D $x, Date:D $d)␤:(Duration:D $a, Duration:D $b)␤:(Duration:D $a, Real $b)␤:(Instant:D $a, Duration:D $b)␤:(Instant:D $a, Real:D $b)␤:(Complex:D \a, Real \b --> Complex)␤:(Int \a, Rational \b)␤:(Rational \a, Rational \b)␤…»
21:02 john3213 joined #perl6
21:04 * skids wonders at the '&bark:(Dog)' syntax in S06 priming section.
21:07 john3213 left #perl6
21:10 skids I can too adverb that.  Stop being such an adverb! :-)
21:12 awwaiid hah
21:15 awwaiid what is Int:D ?
21:15 skids A defined Int.
21:16 skids As opposed to Int:U which onli matches (Int) (and children classes)
21:16 awwaiid interesting. Is ':' part of the name directly, or does ':D' mean 'defined-thing' always?
21:16 awwaiid (feel free to link me somewhere if this is documented)
21:16 skids I think that syntax is in signatures only.
21:18 awwaiid ah
21:19 skids http://perlcabal.org/syn/S12.h​tml#Abstract_vs_Concrete_types
21:20 skids So I guess there is more to it than signatures.
21:23 timotimo (a happy int)
21:23 * vendethiel still thinks :D should be the default, if :_ exists
21:24 timotimo :_?
21:25 timotimo oh, i see
21:26 Vlavv joined #perl6
21:28 timotimo In standard Perl 6, Int is generally assumed to mean Int:_, except for invocants, where the default is Int:D. - that's not how rakudo does it
21:29 skids m: 3.new(:VALUE(2)).say # case in point.
21:29 camelia rakudo-moar 57a08a: OUTPUT«0␤»
21:30 timotimo that's not what i meant, though
21:31 timotimo you can call any method on the type object as long as they don't try to $
21:31 timotimo access $! variables
21:45 timotimo jnthn, can i reach you at the moment?
21:50 skids Oddly, FROGGS++ patch for bigint literals fixes rakudo but not nqp, unless something got strangely hosed with my builds.
22:00 hoverboard joined #perl6
22:03 lizmat joined #perl6
22:14 camelia joined #perl6
22:18 lizmat joined #perl6
22:19 sergot_ joined #perl6
22:19 huf joined #perl6
22:19 nwc10_ joined #perl6
22:19 ivanshma` joined #perl6
22:19 berekuk joined #perl6
22:19 molaf joined #perl6
22:19 IllvilJa joined #perl6
22:19 nebuchadnezzar joined #perl6
22:19 skids joined #perl6
22:19 xinming joined #perl6
22:19 dalek joined #perl6
22:19 xragnar joined #perl6
22:19 Su-Shee joined #perl6
22:19 klapperl_ joined #perl6
22:19 cognominal joined #perl6
22:19 PZt joined #perl6
22:19 bowtie joined #perl6
22:19 Timbus joined #perl6
22:19 Khisanth joined #perl6
22:19 ivan`` joined #perl6
22:19 raydiak joined #perl6
22:19 lestrrat joined #perl6
22:19 Util joined #perl6
22:19 tadzik joined #perl6
22:19 apejens joined #perl6
22:19 64MAAITRZ joined #perl6
22:19 sunnavy joined #perl6
22:19 TimToady joined #perl6
22:19 simcop2387 joined #perl6
22:19 xalbo joined #perl6
22:19 risou joined #perl6
22:19 tokuhirom joined #perl6
22:19 yogan_ joined #perl6
22:19 erdic joined #perl6
22:19 flussence joined #perl6
22:19 atrodo joined #perl6
22:19 rurban_ joined #perl6
22:19 timotimo joined #perl6
22:19 cooper joined #perl6
22:19 sorear joined #perl6
22:19 Juerd joined #perl6
22:19 xfix joined #perl6
22:19 bcode joined #perl6
22:19 geekosaur joined #perl6
22:19 [particle] joined #perl6
22:19 segomos joined #perl6
22:19 pochi joined #perl6
22:19 breinbaas joined #perl6
22:19 stux joined #perl6
22:19 Ulti joined #perl6
22:19 bonsaikitten joined #perl6
22:19 FOAD joined #perl6
22:19 zacts joined #perl6
22:19 nhayashi joined #perl6
22:19 17SAADFTQ joined #perl6
22:19 ilogger2 joined #perl6
22:19 go|dfish joined #perl6
22:19 17SAACR98 joined #perl6
22:19 nwc10 joined #perl6
22:20 zacts_ joined #perl6
22:21 FROGGS joined #perl6
22:23 breinbaas joined #perl6
22:24 dalek rakudo/nom: 4a8eb3d | (Elizabeth Mattijsen)++ | src/core/ (2 files):
22:24 dalek rakudo/nom: Fix advent test fail caused by grep optimization
22:24 dalek rakudo/nom:
22:24 dalek rakudo/nom: MMD candidates for grep and friends where selected on Callable and Regex,
22:24 dalek rakudo/nom: instead of the more correct Callable:D and Regex:D.  This caused grep
22:24 dalek rakudo/nom: expecting to smartmatch on Callable type to fail.
22:24 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/4a8eb3debf
22:25 woolfy joined #perl6
22:27 LLamaRider joined #perl6
22:31 timotimo i don't know enough about the innards of Array.pm to build a proper assign_pos for the typed arrays
22:31 timotimo what is p6bindattrinvres for, for example?
22:31 thou joined #perl6
22:33 BenGoldberg joined #perl6
22:34 timotimo for assign_pos, i'll always have to create a container with p6scalarfromdesc?
22:40 dalek perl6-roast-data: 69cb7d5 | coke++ | / (6 files):
22:40 dalek perl6-roast-data: today (automated commit)
22:40 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/69cb7d5067
22:45 timotimo well, it seems like i can just copy the way the supertype's assign_pos works
22:45 timotimo though in that case it may already have that implementation inherited from Array itself
22:50 timotimo p6: my Int @foo; say @foo.WHAT;
22:50 64MAAITRZ rakudo-{parrot,jvm,moar} 57a08a: OUTPUT«(Array[Int])␤»
22:50 64MAAITRZ ..niecza v24-109-g48a8de3: OUTPUT«Rebuild in progress␤»
22:50 camelia rakudo-{parrot,moar} 57a08a: OUTPUT«(Array[Int])␤»
22:50 camelia ..niecza v24-109-g48a8de3: OUTPUT«(Array)␤»
22:50 camelia ..rakudo-jvm 57a08a: OUTPUT«Rebuild in progress␤»
22:50 timotimo p6: my int @foo; say @foo.WHAT;
22:50 64MAAITRZ rakudo-{parrot,jvm,moar} 57a08a: OUTPUT«(Int)␤»
22:50 64MAAITRZ ..niecza v24-109-g48a8de3: OUTPUT«[31m===[0mSORRY![31m===[0m��Malformed my at /tmp/tmpfile line 1:�------> [32mmy[33m�[31m int @foo; say @foo.WHAT;[0m��Parse failed��»
22:50 camelia rakudo-{parrot,jvm,moar} 57a08a: OUTPUT«(Int)␤»
22:50 camelia ..niecza v24-109-g48a8de3: OUTPUT«[31m===[0mSORRY![31m===[0m��Malformed my at /tmp/tmpfile line 1:�------> [32mmy[33m�[31m int @foo; say @foo.WHAT;[0m��Parse failed��»
22:50 timotimo LTA error, IMHO
22:58 dalek perl6.org: ae390ec | flussence++ | / (16 files):
22:58 dalek perl6.org: Replace LHF anon divs with semantic counterparts
22:58 dalek perl6.org: review: https://github.com/perl6/p​erl6.org/commit/ae390ec81f
22:58 flussence (couldn't find anything else to nitpick :( )
22:59 flussence (did remember to test IE8, don't panic)
22:59 timotimo cool :)
23:00 timotimo can we get a meta generator and generation date+time?
23:01 lizmat joined #perl6
23:08 woolfy joined #perl6
23:08 timotimo flussence: http://perl6.org/about/ - do you see the "if you want to contribute" text getting split into two pieces, one of which is located in an awkward spot near the top?
23:08 timotimo http://perl6.org/documentation/ - holy fuck, this looks b0rked
23:08 timotimo oh, full refresh fixed it
23:08 timotimo never mind in that case
23:11 woolfy joined #perl6
23:12 flussence yeah, that's the one downside
23:12 flussence the files are static, so they get cached "properly" :)
23:12 timotimo aye
23:12 woolfy joined #perl6
23:17 dalek perl6.org: 628cfa8 | flussence++ | README.md:
23:17 dalek perl6.org: Fix indents so code block renders correctly
23:17 dalek perl6.org: review: https://github.com/perl6/p​erl6.org/commit/628cfa8aea
23:17 hoverboard joined #perl6
23:21 xiaomiao joined #perl6
23:26 pippo left #perl6
23:27 woolfy joined #perl6
23:29 tadzik wow, Poznaniers are still drinking :D
23:38 jsn_ joined #perl6
23:39 jsn_ (sorry for the off-topic spam: new AI project) https://drive.google.com/folderview?id=0Bx​OraO_JvOXkeTgyLXQ2dmRHYkU&amp;usp=sharing
23:47 woolfy left #perl6

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

Perl 6 | Reference Documentation | Rakudo