Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2013-06-02

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 colomon_phone He's checked in obviously...
00:00 diakopter just tell them you're on that reservation
00:00 diakopter I'll be right down to knock some heads
00:01 raiph joined #perl6
00:02 raiph timotimo: aiui moarvm has adequate gc and jit stories
00:03 raiph (re pypy as alternate to moarvm)
00:13 colomon_phone diameter , we're good now
00:21 * tadzik grumbles at delayed flight
00:22 tadzik although it's nice that they refer to passengers as "you guys"
00:22 tadzik it has this friendly touch
00:30 lue colomon_phone: does your phone use autocorrect? :)
00:30 colomon_phone Yes
00:31 japhb_ auto-do-the-wrong-thing
00:31 stevan_ joined #perl6
00:32 lue .oO[ Autocorrect isn't stupid. It's just optimized for a small, narrow subset of common language. If the "auto" wasn't there, there'd be no complaints :) ]
00:33 geekosaur .oO { autoincorrect }
00:40 dalek perl6-roast-data: 4510272 | coke++ | / (4 files):
00:40 dalek perl6-roast-data: today (automated commit)
00:40 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/45102728af
00:41 TimToady .oO(autocorrupt)
00:43 dalek roast: a53aa97 | coke++ | S32-num/int.t:
00:43 dalek roast: pugs fudge
00:43 dalek roast: review: https://github.com/perl6/roast/commit/a53aa97e5a
00:49 tadzik the problem, imho, is that it assumes you are wrong and, well, autocorrects
00:49 tadzik not autosuggests
01:06 stevan_ joined #perl6
01:08 adu joined #perl6
01:14 dalek rakudo-star-daily: e9e085f | coke++ | log/ (5 files):
01:14 dalek rakudo-star-daily: today (automated commit)
01:14 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/e9e085f62b
01:43 swuecho joined #perl6
02:02 snearch_ joined #perl6
02:17 timotimo i can't build moarvm because apparently apr.h is missing
02:18 timotimo (and installing the -dev package for libapr1 doesn't help)
02:18 timotimo but now it's time to get some sleep
02:22 FROGGS joined #perl6
02:22 FROGGS o/
02:25 geekosaur timotimo: may need to tell it where apr lives, on my Mint the header is /usr/include/apr-1.0/apr.h
03:00 * pmichaud arrives at the doubletree
03:00 TimToady welcome!
03:08 census joined #perl6
03:11 diakopter hi
03:11 Teratogen howdy
03:11 Teratogen what is happening in Austin?
03:11 diakopter could someone who's met tadzik before come with me to the airport to pick him up at 1 a.m
03:12 japhb_ timotimo, join #moarvm, and then we can figure out build problems there ....
03:12 Teratogen I would, but I am in Tempe, Arizona
03:12 census hi diakopter!
03:13 census when does the conference begin?  tomorrow?
03:14 pmichaud diakopter: I can do that
03:14 diakopter pmichaud: awesome
03:14 pmichaud conference begins monday, hackathons and tutorials tomorrow
03:14 diakopter well, today
03:14 pmichaud well, yes, there were some today
03:14 pmichaud they start for *me* tomorrow :)
03:15 pmichaud diakopter: meet when/where to go to airport?
03:15 diakopter lemme track his flight; sec
03:16 mikemol joined #perl6
03:23 pmichaud diakopter: I'm going to run to the store quickly; will check back on airport time
03:36 japhb_ pmichaud, is there currently a link somewhere that always points to the *newest* Rakudo Star tarball/msi?
03:40 preflex_ joined #perl6
03:42 adu joined #perl6
03:47 danaj joined #perl6
03:51 arlinius joined #perl6
04:00 woolfy joined #perl6
04:10 woolfy left #perl6
04:10 pmichaud japhb_: (link) not at present, no.
04:10 pmichaud we can certainly create one.
04:16 japhb_ joined #perl6
04:22 sorear hihi.
04:26 Psyche^ joined #perl6
04:35 raiph joined #perl6
04:35 Targen joined #perl6
04:44 ggoebel joined #perl6
04:55 kaare_ joined #perl6
05:10 atroxaper joined #perl6
05:16 rindolf joined #perl6
05:34 sqirrel joined #perl6
06:03 ggoebel joined #perl6
06:05 tadzik hey
06:06 tadzik I'm at doubletree, but the wrong one apparently
06:06 tadzik this one is 6070 pr 6017 or something
06:07 tadzik could someone drop diakopter a call or something? I think we missed each other
06:08 drbean joined #perl6
06:10 `patch` joined #perl6
06:12 flussence joined #perl6
06:13 hoelzro joined #perl6
06:18 _sri joined #perl6
06:18 pecastro joined #perl6
06:21 Ulti joined #perl6
06:21 espent joined #perl6
06:21 atroxaper joined #perl6
06:21 espent_ joined #perl6
06:29 zby_home joined #perl6
06:29 dmol1 joined #perl6
06:32 bruges joined #perl6
06:32 aindilis` joined #perl6
06:38 tadzik ok, I'm found anf fine :)
06:39 TimToady \o/
06:39 rindolf Hi all.
06:39 atroxaper Hello!
06:40 rindolf joined #perl6
06:49 Ulti joined #perl6
06:53 rindolf joined #perl6
07:08 rindolf joined #perl6
07:13 cooper joined #perl6
07:13 cooper joined #perl6
07:33 denisboyun_ joined #perl6
07:42 rindolf joined #perl6
07:47 cognominal joined #perl6
08:06 smace joined #perl6
08:16 rindolf joined #perl6
08:32 stevan_ joined #perl6
08:48 rindolf joined #perl6
08:51 xilo joined #perl6
09:00 tgt joined #perl6
09:02 ggoebel joined #perl6
09:06 rindolf joined #perl6
09:24 Guest1337 joined #perl6
09:38 rindolf joined #perl6
09:44 moritz \o
09:52 nwc10 o/
10:00 cognominal joined #perl6
10:00 spider-mario joined #perl6
10:08 spider-mario joined #perl6
10:08 dalek doc: 2814352 | GlitchMr++ | lib/Int.pod:
10:08 dalek doc: expmod and is-prime aren't operators.
10:08 dalek doc: review: https://github.com/perl6/doc/commit/28143523cc
10:08 tomyan joined #perl6
10:14 rindolf joined #perl6
10:25 moritz anyway, back to the topic of those operators that want their types known at compile time
10:25 moritz what's the use case?
10:25 moritz beyond a desparate case of premature optimization, that is
10:35 timotimo yes, premature all the optimizations!
10:36 huf pedobear the programmer?
10:38 smace joined #perl6
10:39 moritz timotimo: no honestly, is there a use case except optimization?
10:40 timotimo exploding even sooner at compile time than before is the only one i can think of
10:41 moritz I'm not convinced, for the following reasons:
10:41 moritz 1) it seems we have many performance issues not related to dispatching at all; I think we're attacking from the wrong angle here, and not attacking the real culprit
10:43 moritz 2) such a feature should probably belong to the caller, not to the callee
10:43 timotimo that seems sensible, indeed.
10:43 moritz the callee doesn't care whether it got the right types at run time or a compile time; it's the caller that cares (in terms of optimization)
10:46 moritz 3) we have many operators take (Any, Any) and then coerce; so in nearly all cases, such a trait would be a noop
10:47 moritz *that take
10:57 rindolf joined #perl6
11:03 testtest joined #perl6
11:06 testtest left #perl6
11:12 census joined #perl6
11:43 rindolf joined #perl6
11:45 ztt_ joined #perl6
11:49 smace joined #perl6
11:59 atroxaper Does anybody know where i can find .fmt implementation?
12:01 tadzik various places in src/core
12:01 tadzik (assuming you mean rakudo)
12:01 smace joined #perl6
12:02 tadzik atroxaper: https://gist.github.com/tadzik/5693435
12:05 rindolf joined #perl6
12:09 atroxaper tadzik: Oh... It's great way! Thank you.
12:10 kivutar joined #perl6
12:11 colomon joined #perl6
12:23 colomon o/
12:23 arnsholt \o
12:23 arnsholt tadzik: You in the lobby or your room? =)
12:27 tgt joined #perl6
12:28 tadzik arnsholt: still in the room
12:28 tadzik arnsholt: are you in the lobby? I'll be there in 2 minutes
12:31 arnsholt tadzik: Still in the room
12:32 arnsholt colomon's just jumped in the shower, but I'll drag him down to the lobby as soon as he's out
12:32 rindolf joined #perl6
12:32 arnsholt I need some breakfast, as well
12:34 tadzik okay, I'm sitting near the reception, will wait for you guys
12:36 arnsholt Cool. Sounds like the shower's stopped, so not too long now, I think
12:37 tadzik heh
12:37 tadzik I feel like ine Big Brother :D
12:38 colomon putting on shoes...
12:41 * tadzik can't wait
12:44 domidumont joined #perl6
12:54 lizmat joined #perl6
12:57 lizmat morning #perl6!
13:00 rindolf joined #perl6
13:03 woolfy joined #perl6
13:21 espadrine joined #perl6
13:29 daniel-s joined #perl6
13:34 tomyan joined #perl6
13:36 FROGGS[mobile] joined #perl6
13:36 FROGGS[mobile] morning
13:36 grondilu morning
13:39 kivutar joined #perl6
13:45 woolfy left #perl6
13:46 crab2313 joined #perl6
13:46 dmol joined #perl6
13:47 thou joined #perl6
13:49 crab2313 joined #perl6
13:52 kaare_ joined #perl6
14:02 Chillance joined #perl6
14:03 pmichaud good morning, #perl6
14:04 nwc10 is it cooler than Frankfurt?
14:05 pmichaud so far, yes.
14:05 pmichaud 20degC at the moment (at 09h00)
14:05 dukeleto joined #perl6
14:06 dukeleto o/
14:06 dukeleto github.com is down, in case y'all didn't notice
14:06 dukeleto push/pull and website is borked
14:06 nwc10 http://perl5.git.perl.org/ is up
14:06 pmichaud today's high is expected to be 32degC
14:09 dmol joined #perl6
14:10 pmichaud "14:00 UTC We are investigating issues with our database server"  -- status.github.com
14:10 leont joined #perl6
14:12 ajr joined #perl6
14:13 anocelot Wait, github is down?  I'm in the git training today.  :D
14:13 anocelot So it is.  LOL
14:17 colomon joined #perl6
14:18 colomon o/
14:22 birdwindupbird joined #perl6
14:29 FROGGS[mobile] joined #perl6
14:30 denisboyun_ joined #perl6
14:35 FROGGS joined #perl6
14:37 aborazmeh joined #perl6
14:39 pmichaud anyone at the hackathon venue yet?
14:40 FROGGS pmichaud: yes
14:40 FROGGS 8 ppl or so
14:40 pmichaud I'll head that way then.
14:40 FROGGS \o/
14:41 FROGGS the TCC building is on the north side of the roadworks (so between the roadworks and the I.35, but directly near the roadworks)
14:41 FROGGS I walked a bit to figure that out
14:41 pmichaud no problem, I'll be driving my car to the parking lot :)
14:41 pmichaud already got my map
14:42 anocelot That's good, because the UT website is down for maint.
14:43 FROGGS anocelot: ya :/
14:43 FROGGS but no problem, so I had the chance to see some squirrels
14:44 anocelot FROGGS: They're pretty acrobatic here.
15:00 japhb_ joined #perl6
15:04 * jnthn makes it to the hackathon o/
15:05 FROGGS \o/
15:05 FROGGS it is more some of slideathon, ehh? :o)
15:05 jnthn well, my slides have hacking dependencies :P
15:05 FROGGS maybe I'll use the night more for slide rather than this time
15:06 FROGGS[mobile] joined #perl6
15:08 * masak rezzes o/
15:08 masak hello from the pre-YAPC::NA hackathon, Austin, TX! \o/
15:10 zby_home_ joined #perl6
15:13 colomon jnthn: Not sure how we beat you here...
15:18 anocelot Looks like github is back up now.
15:24 zakalwe joined #perl6
15:28 lizmat joined #perl6
15:29 woolfy joined #perl6
15:32 * moritz waves to the hackathoneers
15:33 colomon o/
15:33 lizmat moritz /o
15:33 colomon http://71.10.146.50:3001/report # new look, not sure it's an improvement
15:34 FROGGS ohh, all these colors :o)
15:35 moritz colomon: I'd propose to leave out the question marks with black background
15:35 * moritz should fix MiniDBI
15:37 colomon moritz: FROGGS is trying to find the CPAN smoker page to steal... he's got it.
15:37 FROGGS colomon: http://www.cpantesters.org/distro/S/SDL.html#SDL-2.540
15:37 colomon FROGGS++
15:38 FROGGS but maybe someone should just play a bit with the html and make a design
15:42 colomon I'd very gratefully accept contributions...
15:42 * colomon is busy stealing colors
15:42 colomon there's a good bit of additional information that should probably go up, too.
15:43 rindolf joined #perl6
15:44 TimToady UT maps are back up too
15:44 TimToady though you can just find TCC on google maps
15:45 colomon http://71.10.146.50:3001/report
15:45 colomon updated, duller colors.
15:45 colomon err, less insane colors
15:45 colomon hidden question marks
15:45 labster joined #perl6
15:47 FROGGS colomon: I'm giving it a try
15:47 FROGGS (after coffee)
15:51 anocelot TimToady: Yeah; that's what I eventually did.  :)
15:52 masak colomon++ # http://71.10.146.50:3001/report
15:53 masak I might get around to fixing some of my modules today.
15:53 masak first priority is my talks, though.
15:55 masak colomon: could I donate a new yellow to you? I know yellow is really hard to get right in these kinds of status diagrams.
15:55 colomon masak: I'd love it.
15:55 masak ok, hold on.
15:59 dalek rakudo/nom: af2b6ce | (Elizabeth Mattijsen)++ | src/Perl6/ModuleLoader.nqp:
15:59 dalek rakudo/nom: A temporary fix for faster module loading
15:59 dalek rakudo/nom:
15:59 dalek rakudo/nom: *All* directories and *all* types were being checked, whereas always only the
15:59 dalek rakudo/nom: first was being used.  This quick fix short-circuits search as soon as the
15:59 dalek rakudo/nom: first candidate is found.
15:59 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/af2b6cef0f
15:59 dalek rakudo/nom: 0020ef7 | lizmat++ | src/Perl6/ModuleLoader.nqp:
15:59 dalek rakudo/nom: Merge pull request #145 from lizmat/fastermoduleload
15:59 dalek rakudo/nom:
15:59 dalek rakudo/nom: A temporary fix for faster module loading
15:59 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0020ef749f
15:59 masak colomon: suggest red => #d22, yellow => #ff5, green => #1c1. also suggest the yellow have black text, #000. the other two can have white text, I think.
15:59 masak (tested in inkscape against the same grey background as the page has.)
15:59 colomon okay, will try that once I've got my current change tested.  :)
15:59 masak cool.
15:59 dalek rakudo/nom: 0f88d34 | (Elizabeth Mattijsen)++ | src/ (2 files):
15:59 dalek rakudo/nom: Fix for autoviv Typed hash problem
15:59 dalek rakudo/nom:
15:59 dalek rakudo/nom: As suggested by jnthn.  I think there is a ticket for this, but do not know
15:59 dalek rakudo/nom: its number.
15:59 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0f88d34e30
15:59 dalek rakudo/nom: 7d5480e | lizmat++ | /:
15:59 dalek rakudo/nom: Merge pull request #146 from lizmat/autovivtypedhash
15:59 dalek rakudo/nom:
15:59 dalek rakudo/nom: Fix for autoviv Typed hash problem
15:59 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/7d5480ea1c
16:01 tadzik I get build failures in nqp, ./nqp --target=pir --output=src/stage2/gen/NQPP5QRegex.pir \
16:01 tadzik src/stage2/gen/NQPP5QRegex.nqp
16:01 tadzik exists_keyed_str() not implemented in class 'ResizableStringArray'
16:01 * colomon has royally broken his code
16:02 lizmat fwiw, I'm utilizing by brand new commit bit to merge my own pull requests  :-)
16:02 masak oh noes what have we done
16:03 masak j/k, lizmat++ :D
16:03 tadzik :)
16:04 dalek rakudo/nom: 88dae22 | (Elizabeth Mattijsen)++ | src/core/ (2 files):
16:04 dalek rakudo/nom: Fix for "exists" handling on typed hashes
16:04 dalek rakudo/nom:
16:04 dalek rakudo/nom: Made in cooperation with jnthn++ and moritz++
16:04 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/88dae22ecf
16:04 dalek rakudo/nom: 6973abb | (Elizabeth Mattijsen)++ | src/core/Hash.pm:
16:04 dalek rakudo/nom: Suggested change by jnthn
16:04 dalek rakudo/nom:
16:04 dalek rakudo/nom: See http://irclog.perlgeek.de/perl6/2013-05-28#i_7124130 .  Spectested ok.
16:04 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/6973abb4cc
16:04 dalek rakudo/nom: 99c82d3 | lizmat++ | src/core/ (2 files):
16:04 dalek rakudo/nom: Merge pull request #147 from lizmat/existstypedhash
16:04 dalek rakudo/nom:
16:04 dalek rakudo/nom: Fix for "exists" handling on typed hashes
16:04 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/99c82d3732
16:08 rindolf joined #perl6
16:11 dalek joined #perl6
16:13 gdey joined #perl6
16:13 masak gdey: \o
16:14 gdey masak:hello
16:14 masak gdey: again, let me know as soon as you get stuck; either IRL or here. :)
16:14 gdey Will do thanks!
16:14 moritz it seems lizmat++ has commit access to rakudo. \o/
16:15 colomon \o/
16:15 masak I was convinced RT had a ticket about .perl infinite-regressing on things like 'my $a = []; $a.push($a); $a.perl' -- but I looked, twice, carefully, and I cannot find it.
16:15 masak anyone else want to have a go -- or should I just create a Justin Case ticket about it?
16:16 moritz I'm also convinced that we have a ticket for that
16:17 moritz I'll try to find it
16:17 masak I searched for '.perl' -- 50 tickets, but not that one. then I searched for 'infinite' -- 34 tickets, but not that one.
16:17 * masak searches in his mailbox, too
16:18 tadzik /me has a "what idiot wrote that" moment looking at https://github.com/tadzik/panda/blob/master/lib/Panda/Ecosystem.pm#L51
16:18 tadzik and I'm quite sure it was me
16:19 colomon tadzik: one of the great joys of being a programmer
16:19 tadzik yeah
16:19 raiph joined #perl6
16:19 dalek rakudo/nom: d670aaa | (Elizabeth Mattijsen)++ | src/Perl6/ModuleLoader.nqp:
16:19 dalek rakudo/nom: Oops, we had a diff + in there that wasn't supposed to be there
16:19 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d670aaa0e1
16:19 tadzik does anyone around know how ./configure --prefix usually works, and can explain it to me?
16:22 hoelzro tadzik: how do you mean?
16:23 vmspb joined #perl6
16:23 tadzik hoelzro: well, as far as I understand it, what you put in --prefix somehow affects the source code itself
16:23 tadzik I wonder how does that work, istr there's a config.h or so?
16:23 hoelzro right
16:23 moritz masak: I can't find it either :(
16:24 * masak submits rakudobug
16:24 hoelzro if I were writing C, I would probably write prefix >> config.h
16:24 tadzik I'm trying to make --prefix work for panda, so I need to alter bin/panda
16:24 hoelzro it's a bit trickier for a language like Perl =/
16:24 hoelzro but you could have a Config.pm or something I guess
16:24 tadzik I can just put ###CHOCOLATECAKE on top of the file, and then s/// it
16:24 masak r: my $a = []; $a.push($a); say $a.perl; say "alive"
16:24 camelia rakudo 99c82d: OUTPUT«(timeout)»
16:24 tadzik well, Config.pm may not work, because I also need to alter the module loading process
16:25 tadzik I need panda to be entirely self-contained, so installing new modules or new versions will never confuse it
16:25 tadzik so I'll be putting all its deps along with it
16:25 moritz masak: https://rt.perl.org/rt3/Ticket/Display.html?id=58922
16:25 * masak looks
16:26 moritz masak: it seems that [Coke]++ has migrated to be a spec issue (which it also is)
16:26 moritz and then rejected it
16:26 masak oh!
16:26 rindolf joined #perl6
16:26 moritz maybe re-open it, and stall it until the spec issue is resolved
16:27 masak it's an issue on roast; I should probably migrate it to specs.
16:27 moritz oh
16:27 * masak does that
16:27 masak oh; already done.
16:27 masak https://github.com/perl6/specs/issues/6
16:31 dalek rakudo/nom: 4c6ef79 | labster++ | src/core/IO.pm:
16:31 dalek rakudo/nom: make dir() lazy; written while waiting for plane to YAPCNA
16:31 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/4c6ef790b9
16:31 dalek rakudo/nom: 9557da2 | (Brent Laabs)++ | src/core/IO.pm:
16:31 dalek rakudo/nom: Merge pull request #161 from labster/nom
16:31 dalek rakudo/nom:
16:31 dalek rakudo/nom: make dir() lazy; written while waiting for plane to YAPCNA
16:31 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/9557da29eb
16:32 * masak commented on https://github.com/perl6/specs/issues/6
16:32 tadzik labster++
16:32 tadzik nice one
16:32 labster I did make it safely to the hackathon :)
16:32 moritz labster++
16:33 colomon labster++
16:33 denisboyun_ joined #perl6
16:35 moritz jnthn: I have a fix for JSON::Path; want to give me a commit bit, or do you prefer pull requests?
16:36 jnthn commkt bit
16:37 colomon rn: say $Inf
16:37 jnthn moritz: you got one
16:37 camelia niecza v24-55-g7942a09: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤Variable $Inf is not predeclared at /tmp/i13FYgoTks line 1:â�¤------> [32msay [33mâ��[31m$Inf[0mâ�¤â�¤Unhandled exception: Check failedâ�¤â�¤  at /home/p6eval/niecza/boot/lib/CORE.setting line 1502 (die @ 5) â�¤  at /home/p6eval/niecz…
16:37 camelia ..rakudo 99c82d: OUTPUT«Inf␤»
16:37 moritz jnthn: tusen takk; fix pushed
16:37 jnthn takk *you*
16:38 masak :)
16:38 moritz var så gud
16:38 dalek nqp/rak-jvm-support: cd35271 | jonathan++ | src/vm/parrot/6model/reprs/NativeCall.c:
16:38 dalek nqp/rak-jvm-support: Add missing gc_mark registration.
16:38 dalek nqp/rak-jvm-support: review: https://github.com/perl6/nqp/commit/cd35271c0c
16:38 masak you people attempting Scandinavian -- it's so cute!
16:39 tadzik takk for i dag
16:39 moritz masak: understanding is quite OK, speaking... well. Writing is really hard for me though
16:40 masak tadzik: impressive.
16:41 smace joined #perl6
16:41 masak moritz: I have the same with German.
16:41 rindolf joined #perl6
16:41 moritz colomon: can you please nopaste/gist the test output for DateTime::Format? I just got no test failures, and wonder why http://71.10.146.50:3001/project/DateTime::Format shows it red
16:42 colomon moritz: give me a minute or two
16:42 moritz and the last commit was May 06, so it's not a recent fix
16:42 tadzik my german showoff is "Ich besischtige die sehenswuerdigkeiten", or something
16:42 moritz s/sch/ch/but otherwise correct
16:42 tadzik I'm good at pronouncing it :)
16:42 moritz so you are :-)
16:44 * masak can parse that :)
16:44 masak "I'm checking the sights", right?
16:45 moritz yes
16:46 konundra joined #perl6
16:46 pmichaud lizmat:  Hash[TKey,TVal].new   # create hash with TKey keys and TVal values
16:47 pmichaud for example:  Hash[Int, Book]   # create hash with Int keys and Book values
16:47 masak rn: class Book {}; say Hash[Int, Book].new
16:47 camelia niecza v24-55-g7942a09: OUTPUT«Unhandled exception: System.NullReferenceException: Object reference not set to an instance of an object␤  at Builtins.InvokeSub (Niecza.P6any obj, Niecza.Variable[] pos) [0x00000] in <filename unknown>:0 ␤  at Niecza.Kernel.ToComposable (Niecza.STable arg…
16:47 camelia ..rakudo 99c82d: OUTPUT«().hash␤»
16:47 colomon moritz: when I run the tests by hand I get all successes as well.
16:47 masak rakudo accepts that :)
16:47 pmichaud Array[Int]
16:48 pmichaud if I say  "my Book %hash{Int}"
16:48 moritz colomon: then it might be interesting why it thinks the tests fail
16:48 colomon moritz: but when I try to install it with panda, it fails.
16:48 pmichaud that desugars to    Hash[Int, Book]
16:49 colomon moritz: https://gist.github.com/colomon/5694107
16:50 Guest1337 joined #perl6
16:51 FROGGS colomon: https://gist.github.com/FROGGS/624b75ab0d331015b35c
16:51 colomon FROGGS: https://github.com/colomon/SmokeResults is my github repo for this.  :)
16:51 colomon public/css has the style sheet
16:52 moritz colomon: might be a precompilation issue :(
16:53 moritz indeed, fails when precompiled
16:55 moritz seems like it might be not too bad to reduce to a minimal example, because it doesn't depend on any other modules
16:56 moritz and Text::Emotion is another module that fails its tests when precompiled, and passes otherwise
16:56 moritz no other prrequs necessary
16:56 masak :/
16:57 masak oh, and that's a single 50-line module.
16:57 masak should be eminently golfable.
16:57 masak https://github.com/MattOates/Text--Emotion/blob/master/lib/Text/Emotion/Scorer.pm
16:57 masak wut, it has a '1;' in the end!?
16:57 masak :D
16:58 * masak makes a pull request removing it
16:59 masak other things that look wrong in that module: 'package Text::Emotion' and then, inside that, 'class Text::Emotion::Scorer'...
17:00 lizmat r: class Baz { method new (*@p, *%n) { say @p; say %n } }; say Baz[Int,Str].new
17:00 camelia rakudo 99c82d: OUTPUT«===SORRY!===␤Type Baz cannot accept type arguments␤»
17:01 pmichaud r: class Baz[::TKey, ::TVal] { };  say Baz[Int,Str].new
17:01 camelia rakudo 99c82d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Unable to parse class definitionâ�¤at /tmp/n4iXbNemZD:1â�¤------> [32mclass Baz[33mâ��[31m[::TKey, ::TVal] { };  say Baz[Int,Str].[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or termâ�¤        prefix or meta-prefixâ�¤    â€¦
17:02 colomon FROGGS: I think I'm going to swap the two reds
17:04 pmichaud r: say Hash[Int,Int].new
17:04 camelia rakudo 99c82d: OUTPUT«().hash␤»
17:06 masak r: my %h = Hash[Int, Int]; %h{"OH"} = "HAI"; say %h.perl
17:06 camelia rakudo 99c82d: OUTPUT«Cannot look up attributes in a type object␤  in method pairs at src/gen/CORE.setting:7097␤  in method list at src/gen/CORE.setting:6828␤  in method STORE at src/gen/CORE.setting:6959␤  in block  at /tmp/QDEAjEYe9b:1␤␤»
17:06 woolfy left #perl6
17:06 masak r: my %h = Hash[Int, Int].new; %h{"OH"} = "HAI"; say %h.perl
17:06 camelia rakudo 99c82d: OUTPUT«("OH" => "HAI").hash␤»
17:09 rindolf joined #perl6
17:09 pmichaud Hash[TVal, TKey}
17:09 pmichaud Hash[TVal, TKey]
17:11 colomon http://71.10.146.50:3001/report # new look thanks to FROGGS++
17:12 masak ...sorry everyone who is not attending the hackathon; IRC lines are bound to be incomprehensible...
17:13 colomon FROGGS: I'm thinking maybe centering instead of left: 25% for the table?
17:13 TimToady are you trying to say that the rest of them are comprehensible?
17:13 FROGGS colomon: if you know how to center a table cross platf^Wbrowser
17:13 masak TimToady: point taken.
17:14 colomon FROGGS: there's no modern equivalent of align="center"  ?
17:15 zby_home joined #perl6
17:18 masak colomon: for text or images?
17:18 colomon masak: tables
17:18 colomon I've got something that works for Chrome, anyway.
17:18 colomon reload the report.  :)
17:18 masak colomon: http://www.w3.org/Style/Examples/007/center
17:18 masak colomon: url?
17:19 labster masak: http://71.10.146.50:3001/report
17:21 masak colomon: did you try my color suggestions?
17:21 broone joined #perl6
17:21 colomon masak: I let FROGGS redesign the colors.
17:21 sorear greetings
17:21 masak oh, ok.
17:21 colomon masak: so, no.  sorry.
17:21 colomon (I mean, I didn't even try them.)
17:21 masak fwiw, I think my suggestion is better. :)
17:21 colomon masak++
17:21 masak it's in the backlog if you're curious.
17:23 masak also, what do the letters "P", "T", and "B" mean?
17:23 pmichaud r: my @array = 3,4,5;  say @array.perl;   say [3,4,5].perl
17:23 camelia rakudo 9557da: OUTPUT«Array.new(3, 4, 5)␤[3, 4, 5]␤»
17:23 colomon Prereq, Tests, Build
17:23 colomon the stage it failed at
17:24 colomon I'm going to add a key in a minute or two.
17:25 colomon except some of the Ps are Passed
17:25 masak :/
17:25 colomon Is there an easy way to get a check mark or something?
17:26 masak
17:26 masak https://en.wikipedia.org/wiki/Tick_(check_mark)
17:27 jnthn masak: Yes, but is there a tick in a box?
17:27 pmichaud .u 2611
17:27 yoleaux U+2611 BALLOT BOX WITH CHECK [So] (☑)
17:27 colomon I'm thinking outside the box.  ;)
17:27 rjbs and don't forget...
17:27 rjbs .u 2327
17:27 yoleaux U+2327 X IN A RECTANGLE BOX [So] (⌧)
17:27 pmichaud .u 237b
17:27 yoleaux U+237B NOT CHECK MARK [So] (⍻)
17:27 dalek rakudo/nom: 9cf3a88 | labster++ | src/core/IO.pm:
17:27 dalek rakudo/nom: keep lines() with limit from reading an extra line
17:27 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/9cf3a88b60
17:28 rjbs I really hate 237B.
17:28 rjbs It often seems to be rendered almost indistinguishably from a check mark.
17:28 rjbs pmichaud: See you in Austin?
17:28 pmichaud rjbs: I'm here now!  and having fun
17:29 rindolf joined #perl6
17:29 rjbs I need to put U+20E0 onto my keyboard map.
17:29 rjbs pmichaud: Cool!
17:30 tomyan joined #perl6
17:30 pmichaud rjbs: we're all now amused by looking up U+237B
17:31 pmichaud and mostly saying "that's just wrong"
17:31 masak s/amused/horrified/ :)
17:32 sorear rjbs: looking forward to seeing you
17:33 masak +1
17:33 rjbs I am looking forward to being seen, especially if I am eating something delicious when it happens.
17:33 * rjbs hopes to make it to hear about MOARVM.
17:33 rjbs I forget what its competitionw as...
17:34 dalek roast: 2004f14 | labster++ | S16-filehandles/io.t:
17:34 dalek roast: unfudge rakudo: lines() with limit
17:34 dalek roast: review: https://github.com/perl6/roast/commit/2004f14224
17:34 rjbs Oh, man, p6 on JVM is opposite Tim's Size Me!
17:35 rjbs Looks like I'll almost certainly get to moar, though.
17:41 dalek rakudo/nom: 21a9579 | (Elizabeth Mattijsen)++ | src/core/Any.pm:
17:41 dalek rakudo/nom: Support for subscript adverb :delete in combination with :kv :p :k :v
17:41 dalek rakudo/nom:
17:41 dalek rakudo/nom: As described in S02:2514
17:41 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/21a957937d
17:41 dalek rakudo/nom: d0ac9a7 | (Elizabeth Mattijsen)++ | / (16 files):
17:41 dalek rakudo/nom: Merge remote-tracking branch 'upstream/nom' into deleteadverbkvpkv
17:41 dalek rakudo/nom:
17:41 dalek rakudo/nom: Conflicts:
17:41 dalek rakudo/nom: src/core/Any.pm
17:41 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d0ac9a78de
17:41 dalek rakudo/nom: d292900 | (Elizabeth Mattijsen)++ | src/core/Any.pm:
17:41 dalek rakudo/nom: We don't need no Mu
17:41 dalek rakudo/nom:
17:41 dalek rakudo/nom: After discussing this with jnthn at PLPW
17:41 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d2929008ff
17:41 dalek rakudo/nom: 12b13e5 | lizmat++ | src/core/Any.pm:
17:41 dalek rakudo/nom: Merge pull request #158 from lizmat/deleteadverbkvpkv
17:41 dalek rakudo/nom:
17:41 dalek rakudo/nom: Support for subscript adverb :delete in combination with :kv :p :k :v
17:41 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/12b13e5254
17:43 moritz .u U+237B
17:43 yoleaux U+237B NOT CHECK MARK [So] (⍻)
17:44 colomon New version, with checks and key: http://71.10.146.50:3001/report
17:44 tadzik yoleaux: seen daxim?
17:44 tadzik seen daxim
17:44 tadzik pff
17:44 moritz .seen daxim
17:44 yoleaux I saw daxim 29 May 2013 13:22Z in #perl6: <daxim> r: say ('foo.bar.baz' ~~ /(\w+)+/).list.perl
17:44 tadzik ah, thanks moritz
17:44 masak .seen masak
17:44 yoleaux You're right there.
17:45 masak hrm. :)
17:46 masak yoleaux: does that mean you're at the hackathon, too? :P
17:47 atroxaper night! gl hf gg (as starcrafters says)
17:49 dalek rakudo/nom: 2fe2537 | (Elizabeth Mattijsen)++ | src/core/List.pm:
17:49 dalek rakudo/nom: Make List.gist not warn on undefined values
17:49 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/2fe2537aa1
17:49 dalek rakudo/nom: 1dace73 | (Elizabeth Mattijsen)++ | src/core/List.pm:
17:49 dalek rakudo/nom: .gist is better for defined values than .Str
17:49 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/1dace737f9
17:49 dalek rakudo/nom: 7869a67 | (Elizabeth Mattijsen)++ | src/core/List.pm:
17:49 dalek rakudo/nom: Always just use .gist to create the final .gist
17:49 dalek rakudo/nom:
17:49 dalek rakudo/nom: As suggested by Moritz++
17:49 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/7869a67fbb
17:49 dalek rakudo/nom: b2072f5 | lizmat++ | src/core/List.pm:
17:49 dalek rakudo/nom: Merge pull request #152 from lizmat/ListgistAnyok
17:49 dalek rakudo/nom:
17:49 dalek rakudo/nom: Make List.gist not warn on undefined values
17:49 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/b2072f578c
17:50 spider-mario joined #perl6
17:53 colomon Okay, that's enough fiddling with the smoke results.  Next project to hack on...
18:00 fgomez joined #perl6
18:02 masak lunch &
18:02 jnthn lunch &
18:04 TimToady lunch &
18:04 lizmat lunch &
18:05 colomon lunch !
18:05 felher lunch :
18:06 sorear felher: you here?
18:07 sorear luch &
18:07 dalek panda/prefix: cdf0c1a | tadzik++ | b (2 files):
18:07 dalek panda/prefix: Add --prefix to bootstrap.
18:07 dalek panda/prefix:
18:07 dalek panda/prefix: Now panda and its dependencies are self-contained separately, so you
18:07 dalek panda/prefix: actually can update some panda dependencies without breaking panda
18:07 dalek panda/prefix: itself.
18:07 dalek panda/prefix:
18:07 dalek panda/prefix: Awesome!
18:07 dalek panda/prefix: review: https://github.com/tadzik/panda/commit/cdf0c1ab85
18:08 felher sorear: nope, but I though i'll join you in spirit and get something to eat, too. Even though it's (a bit past) dinner time here in germany :)
18:08 colomon joined #perl6
18:11 moritz it's not past dinner time until about 6am :-)
18:13 smace joined #perl6
18:17 felher moritz: ah, fair enough :)
18:23 ggoebel joined #perl6
18:30 FROGGS joined #perl6
18:40 rindolf joined #perl6
18:41 rindolf joined #perl6
18:43 rindolf joined #perl6
18:44 rindolf joined #perl6
18:44 stevan_ joined #perl6
18:48 woolfy1 joined #perl6
18:54 ajr joined #perl6
19:02 rindolf joined #perl6
19:08 telex joined #perl6
19:12 ggoebel joined #perl6
19:14 nwc10 it's all gone very quiet. Is everyone sniffing whiteboard pens? :-)
19:25 woolfy1 the group of Perl 6 developers that are at YAPC::NA went out to lunch and heated debate just an hour ago.  They can return within the hour, or it may take longer...
19:25 woolfy1 Big group, by the way, many of them quite vocal on this channel...
19:29 lue .oO(heated debate?)
19:33 mucker joined #perl6
19:36 mucker http://themucker.github.io/plu/ perl on the browser !
19:39 moritz wtf does that do? it froze my browser :-)
19:39 colomon FROGGS: http://acmeism.org/
19:39 TimToady we're back now, so the noise should go back up :)
19:39 moritz \o/
19:40 * jnthn back too :)
19:40 moritz mucker: does that try to compile perl to JS?
19:40 mucker moritz: yes. emscripten kinda like
19:41 mucker http://dg.cx/jsperl/ but I the build files are on github
19:41 moritz that explains why it's so slow :-)
19:43 mucker the interesting thing, (I haven't added it yet) you can use open and other file ops
19:44 Ultali joined #perl6
19:45 moritz and where can it open those files?
19:45 moritz from local paths?
19:46 * moritz kills his browser after several minutes of hanging
19:46 mucker currently you can add files in tmp
19:46 mucker but later I add tutorial.txt or something similiar
19:46 * masak is back, but hacking slides
19:46 sorear hiii
19:47 sorear wb masak ;)
19:47 mucker moritz: does the execute button hang ?
19:47 timotimo we should emscripten parrot to js, it will be vastly faster than on C, because js has jits!
19:48 dalek panda: 096063e | tadzik++ | bin/panda:
19:48 dalek panda: Uncomment some code which I commented out for unknown reasons
19:48 dalek panda: review: https://github.com/tadzik/panda/commit/096063ec77
19:48 masak sorear: everything's interconnected. category theory is warping my brain ;)
19:49 mucker masak: sounds like LSD :P
19:50 timotimo https://rt.perl.org/rt3/Ticket/Display.html?id=118263 - so apparently if i derive a class form a role, the role will be "mixed into" an empty class and that's why the error message is so opaque
19:51 colomon Is there somewhere out there with hints on building rakudo/jvm-support?
19:52 tadzik yeah
19:52 tadzik you need nqp on branch rak-jvm-support
19:53 tadzik and build both with ConfigureJVM.pl
19:54 nwc10 and java (1.)7
19:54 nwc10 and if the JVM SEGVs, you need a newer JVM with that bug fixed
19:54 nwc10 pwnage!
19:54 colomon there's no lazy man's --gen-nqp option for ConfigureJVM, eh?
19:55 s1n joined #perl6
19:55 tadzik no eye deer
19:55 colomon doesn't seem to be.
19:56 census joined #perl6
19:56 dagurval joined #perl6
19:57 dagurval r: "/tmp/a.pm6".spurt("sub abc() is export { }"); require("/tmp/a.pm6") <abc>; abc();
19:57 camelia rakudo b2072f: OUTPUT«===SORRY!===␤Error while compiling block  (source text: "\"/tmp/a.pm6\".spurt(\"sub abc() is export { }\"); require(\"/tmp/a.pm6\") <abc>; abc();"): Error while compiling op call: Error while compiling block : Error while compiling op call: Null PMC access in find_…
19:57 atroxaper left #perl6
19:58 dagurval Null PMC is always a bug right? :)
19:58 jnthn yes
19:58 dagurval is this a known one?
19:58 sorear evil code :)
19:59 jnthn Probably accidental code.
19:59 jnthn It probably wanted '...' around the sub
19:59 jnthn Not "..." which interpolates the closure
19:59 jnthn But that shouldn't caues an error like this
19:59 sorear r: "foo { }"
19:59 camelia rakudo b2072f:  ( no output )
20:00 sorear r: sub abc() is export
20:00 camelia rakudo b2072f: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Missing blockâ�¤at /tmp/Fh7RFUXTXA:1â�¤------> [32msub abc() is export[33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or termâ�¤        prefix or meta-prefixâ�¤        new name to be definedâ�¤Â»â€¦
20:00 sorear r: eval "sub abc() is export { }"
20:00 camelia rakudo b2072f: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Missing blockâ�¤at eval_0:1â�¤------> [32msub abc() is export [33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or termâ�¤        prefix or meta-prefixâ�¤        new name to be definedâ�¤Â»
20:04 dalek panda: 914450e | tadzik++ | bin/panda.bat:
20:04 dalek panda: Re-add panda.bat which I accidentally removed when working on panda2
20:04 dalek panda: review: https://github.com/tadzik/panda/commit/914450e3a2
20:04 dalek panda: e0474ee | tadzik++ | lib/Panda/Installer.pm:
20:04 dalek panda: Fix GH-36
20:04 dalek panda: review: https://github.com/tadzik/panda/commit/e0474ee191
20:06 masak there's a lot of bubbling enthusiasm in this room. I realize I'd been missing the hackathon feeling a bit :)
20:07 * moritz missses it
20:08 sorear :)
20:08 colomon :)
20:09 avuserow anything bite-sized going on that I could help a bit with?
20:09 * avuserow is around but not hacking actively yet
20:11 sorear t/nqp/24-module.t ............... All 3 subtests passed
20:11 sorear jnthn: rak-jvm-support is not totally clean tests, expected?
20:11 jnthn sorear: yes, there' sthat one, and then a handful of regex test failures
20:11 jnthn The regex ones are probably not so hard to hunt
20:12 sorear how about I make a pass at that?
20:12 jnthn Go for it. iirc, some of them are related to newline stuff
20:13 sorear nod
20:13 sorear will have many qs later
20:13 jnthn :)
20:13 jnthn The relevant code-gen is in src/vm/jvm/QAST/Compiler.nqp
20:13 * colomon is still hung up getting Java 1.7 installed on OS X 10.6.8
20:14 * colomon is also lightly hacking on Niecza in another window
20:14 nwc10 colomon: I believe that it involves hacks. I haven't attempted it
20:14 sorear jnthn: https://gist.github.com/sorear/5694782, full list of fails FWIW
20:15 colomon nwc10: yes, I'm working on the hacks now.  :)
20:15 colomon http://stackoverflow.com/questions/13536667/how-to-install-java-jdk-7-on-snow-leopard
20:15 sorear nwc10: you here and/or coming?
20:16 nwc10 sorear: no, I'm not
20:18 xilo joined #perl6
20:19 colomon nwc10: :(
20:20 tadzik r: https://gist.github.com/tadzik/5694821 # bug?
20:20 camelia rakudo b2072f: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Two terms in a rowâ�¤at /tmp/NQkDK_ZQ2S:1â�¤------> [32mhttps[33mâ��[31m://gist.github.com/tadzik/5694821 # bug?[0mâ�¤    expecting any of:â�¤        argument listâ�¤        postfixâ�¤        infix stopperâ�¤        infix or meta-infixâ�¤        st…
20:20 tadzik eeh?
20:20 tadzik r: https://gist.github.com/tadzik/5694821
20:20 camelia rakudo b2072f: OUTPUT«rmdir is disallowed in restricted setting␤current instr.: 'throw' pc 347557 (src/gen/CORE.setting.pir:151689) (src/gen/CORE.setting:8887)␤called from Sub 'die' pc 36608 (src/gen/CORE.setting.pir:14320) (src/gen/CORE.setting:578)␤called from Sub 'die' pc 36506 (src/…
20:20 tadzik hrmng
20:20 tadzik ...can someone try it locally and tell me if it's a bug, or I just don't understand END?
20:22 kivutar joined #perl6
20:23 naids joined #perl6
20:23 avuserow r: sub foo { say "called foo"; END {say "in END";} }
20:23 camelia rakudo b2072f: OUTPUT«in END␤»
20:24 sorear jnthn: FWIW, the JVM use of PREFIX does not quite folllow the filesystem hierarchy standard.  can fix that and/or talk about it later
20:25 sorear colomon: any luck?
20:25 colomon not yet
20:25 colomon In theory I've got it on my machine now, but it's not finding it yet
20:25 sorear hash -r / ldconfig stuff?
20:26 colomon maybe?  hmmm
20:27 sorear > $*VM
20:27 sorear java.lang.NullPointerException
20:28 sorear I HAS A BUILT RAKUDO-JVM
20:28 arnsholt =D
20:28 colomon \o/
20:29 avuserow .oO( there should be a $*DEFAULT_CRASH_TYPE to tell you if it's a NullPMC, NullPointerException, segfault, etc)
20:30 dalek niecza: 67d20f1 | (Solomon Foster)++ | lib/CORE.setting:
20:30 dalek niecza: Try to get CUSTOM_LIB working.
20:30 dalek niecza:
20:30 dalek niecza: Very simple version, but hopefully enough to make panda happy.
20:30 dalek niecza: review: https://github.com/sorear/niecza/commit/67d20f1824
20:30 sorear 2+2 works, anyway
20:30 jnthn :)
20:31 mucker left #perl6
20:31 colomon sorear: jnthn might have hardcoded that.  Try 2+3 just to be sure.  ;)
20:32 nwc10 no, try 6*9 :-)
20:34 pmichaud 42.
20:35 lue r: say 6*9
20:35 camelia rakudo b2072f: OUTPUT«54␤»
20:35 lue .oO(darn, rakudo's broken)
20:43 ilogger2 joined #perl6
20:44 * colomon is giving up on Rakudo-JVM for the moment, in the interesting on not going insane trying to figure out how to get Java 1.7 working.  :\
20:45 colomon moritz++ # was just going to start working on fixing JSONPath, then realized you had already done it.  :)
20:45 woolfy joined #perl6
20:45 masak BUT...
20:45 masak r: say (6 * 9).base(13)
20:45 camelia rakudo b2072f: OUTPUT«42␤»
20:45 masak \o/
20:46 masak I'm not sure DNA was aware of that, at least at the time of writing the books... :)
20:47 colomon masak: I always just figured he ran out of letters before he could get to the rest of the formula.  :)
20:47 masak heh.
20:48 colomon tadzik: any notion how hard it would be to make the Rakudo version of Panda install modules for Niecza?
20:49 colomon tadzik: I realize you're involved in a big discussion at the whiteboard rather than checking IRC at the moment, I just wanted to float the idea before I forgot it.
20:50 lue masak: "I may be a sorry case, but I don't write jokes in base 13." -- Douglas Adams
20:51 masak lue: ah, yes. of course you'd know this, too :P
20:52 lue (hey, I had to look up the exact wording of the quote!)
20:52 masak yeah, yeah :P
20:52 * masak .oO( with a nick like yours... )
20:53 masak :P
20:54 lue r: say 42.chr; # only fitting the answer to LUE would mean this Perl 6
20:54 camelia rakudo b2072f: OUTPUT«*␤»
20:56 eiro hello everyone
20:57 sorear hello eiro
20:57 sorear you in austin?
20:57 eiro i wish i would be in austin. i can't afford it
20:57 tomyan joined #perl6
20:57 masak eiro! \o/
20:58 eiro just discovered MoarVM and submit patches to make it compile on linux
20:58 masak eiro++
20:58 eiro hey masak :) you're at austin ?
20:58 masak eiro: then you should probably check out #moarvm, too :)
20:58 masak eiro: I'm in Austin, hacking on my talks :D
20:59 colomon eiro: something to do with APR?
20:59 eiro colomon: yep. apr is just forgotten under linux (the string was empty)
21:00 eiro https://github.com/eiro/MoarVM/commit/74e2c80b0717d63a61758deaf964f67e97ba9749
21:00 colomon eiro: I built APR by hand to get it working on Linux.  If you've fixed that, that's a big eiro++
21:01 eiro thanks! now i can read a VM code :)
21:03 lue So, is MoarVM supposed to be another VM for NQP, like JVM and Parrot? (That's what it seems to be, but for a second I thought differently)
21:03 masak lue: yes.
21:04 masak lue: see http://strangelyconsistent.org/blog/a-vm-for-6model
21:04 eiro colomon: btw the patch is gross, just repeating a line in both if and else block. i don't know why it was empty (maybe there was a good reason?) but it fixed the problem to fill it
21:04 colomon eiro: atm I'm just focusing on verifying it works.
21:05 masak even a crude fix is better than a non-working state :)
21:06 sorear I guess I'm going to have to learn to read this JAST stuff
21:06 sorear jnthn: working on t/nqp/24-module.t atm.
21:07 dalek ecosystem: 424e795 | dagurval++ | META.list:
21:07 dalek ecosystem: Added Image::Resize
21:07 dalek ecosystem: review: https://github.com/perl6/ecosystem/commit/424e795ed7
21:08 masak sorear++!
21:09 jnthn sorear: That one is some closure related thingy, iirc...
21:09 jnthn sorear++ # hacking on nqp-jvm
21:15 colomon masak, jnthn: I can confirm that eiro++'s patch makes MoarVM build on my Linux box (without the workaround hack I used yesterday)
21:15 tadzik can someone give me a hand with https://gist.github.com/tadzik/5694821 ?
21:15 colomon tadzik: what's the problem?
21:16 tadzik colomon: try running it :)
21:16 sorear "fortunately, rakudo can work for quite a while with completely broken closures"
21:16 tadzik tl;dr: END is set up even if foo() is never ran
21:16 colomon big error there.
21:16 colomon ooh!
21:16 tadzik if that's correct, then it's at least confusing
21:17 sorear tadzik: That's how it's supposed to work.
21:17 jnthn tadzik: Yes, it hould be
21:17 jnthn tadzik: Did you want LEAVE?
21:17 sorear END blocks are called once, period.
21:17 sorear at process exit
21:17 colomon He wanted END but only if foo was called
21:17 sorear it does not work like atexit(3)
21:17 tadzik I do want it at process exit
21:18 sorear then you need to manage a list of directories-to-remove
21:18 tadzik if I have foo() creating a random tmpdir, but I want them all destroyed atexit
21:18 tadzik okay
21:18 jnthn OK. Have an our variable you set in the sub.
21:18 colomon how about END { rmdir $foo if $foo }
21:18 colomon ?
21:18 tadzik yeah, I could do that, but it's still broken if I have multiple foos
21:18 sorear (or use eval 'END { rmdir $s }', but ick)
21:18 tadzik I guess I have to go the way sorear++ says
21:19 colomon wouldn't an array of foos work?
21:20 tadzik bah
21:20 btyler joined #perl6
21:20 sorear colomon: 16:18 < sorear> then you need to manage a list of directories-to-remove
21:20 sorear btyler! o/
21:20 colomon sorear++
21:21 btyler hiya
21:23 * colomon 's version: https://gist.github.com/colomon/5695031
21:24 mst joined #perl6
21:25 sorear jnthn: got tips for reading jast?
21:25 jnthn Don't? ;)
21:25 masak ingy: https://github.com/perl6/nqp/tree/master/src/QRegex
21:26 masak and https://github.com/perl6/nqp/blob/master/src/QAST/Regex.nqp I guess.
21:30 Guest1337 joined #perl6
21:31 sorear 2081 lines of javap output \o/
21:32 * masak pictures the name "javap" as being the result of someone saying "java" and then quickly having his mouth covered by someone else
21:32 woolfy left #perl6
21:33 sorear qb_13, qb_14, ...
21:34 pmichaud it does somewhat feel as though we'd like to be able to dynamically add closures to be run at END time
21:34 masak r: my &closure = { say "OH HAI" }; END { &closure() }
21:34 camelia rakudo b2072f: OUTPUT«OH HAI␤»
21:35 masak pmichaud: worksforme ;)
21:35 colomon pmichaud: it is quite easily done by the end user already.
21:36 masak r: my $un-greeting; my &closure = { say $un-greeting }; $un-greeting = "bai bai!"; END { &closure() }
21:36 camelia rakudo b2072f: OUTPUT«bai bai!␤»
21:36 * colomon isn't trying to argue against the idea, mind you.
21:36 tadzik I would expect that to work :)
21:36 sorear I've always thought that atexit() would make a nice addition
21:36 sorear perhaps under that name
21:36 masak pmichaud: I'm also not arguing against the idea. I would really like a nice API for dynamically building regexes, even though most of that can be done with /<$rx>/
21:37 masak which is an analogous situation.
21:37 tadzik I think it should be a phaser, phasers are those things that mangle the regular control flow
21:37 pmichaud https://gist.github.com/pmichaud/5695072
21:37 tadzik as atexit would
21:37 pmichaud is the best I can come up with
21:37 lue .oO(ATEXIT { })
21:37 tadzik heh, yeah :)
21:38 dalek panda: d3042b8 | tadzik++ | lib/Panda.pm:
21:38 dalek panda: Fix GH-6
21:38 dalek panda: review: https://github.com/tadzik/panda/commit/d3042b8f68
21:41 colomon ATEXIT: https://gist.github.com/colomon/5695083
21:43 tadzik haha, awesome
21:43 tadzik now we only need a module
21:44 pmichaud and perhaps there's an argument to be made that the closures should be run LIFO  :)
21:44 colomon pmichaud: good point
21:44 masak yes.
21:44 FROGGS eiro: the empty string was intention
21:45 colomon should I do a module?  or should we just add it straight into the language?
21:45 masak pmichaud++ colomon++ tadzik++ # you just created a new phaser in, like, 10 minutes!
21:45 masak colomon: modules, IMO.
21:45 masak module*
21:45 tadzik haha! Beat that, spock!
21:45 pmichaud I'm thinking it's a potential language construct, but we don't have enough use cases to know it exactly yet
21:45 pmichaud so, module
21:45 colomon works for me.
21:45 pmichaud if it turns out to be incredibly common, then it can be added to language
21:45 colomon I've got it.
21:46 tadzik colomon++
21:46 colomon module name?
21:46 colomon Perl6::ATEXIT
21:46 colomon ?
21:47 tadzik Phaser::ATEXIT?
21:47 espadrine what's the difference with END if they're both LIFO?
21:47 sorear espadrine: END foo =~ BEGIN ATEXIT foo
21:48 sorear ATEXIT foo =~ eval 'END foo'
21:48 masak espadrine: this one is per-sub call. END never is.
21:49 espadrine ok
21:50 rindolf joined #perl6
21:51 lue is quoteharder a real thing? (Looking at S02:3831, "macro quote:<❰ ❱> ($text) { quasi { {{{$text}}}.quoteharder } }")
21:52 * sorear found enterFromMain
21:53 eiro FROGGS: why was it ?
21:55 masak lue: not in the sense that it's something more than just an example in S02.
21:56 aindilis joined #perl6
21:56 lue so it might as well be .foobar . OK then, it seemed like maybe it was an actual core language feature :)
21:57 ingy masak: ta
21:57 dalek ecosystem: d4da6a1 | colomon++ | META.list:
21:57 dalek ecosystem: Add Phaser::ATEXIT
21:57 dalek ecosystem: review: https://github.com/perl6/ecosystem/commit/d4da6a184a
21:58 colomon 12 minutes -- not bad.
21:58 colomon Guess I should make the README better.
22:01 colomon done.
22:02 * sorear wonders what hllName is about
22:03 arnsholt IIRC, it's the name of the HLL that owns a type
22:03 jnthn sorear: Are we in nqp, perl6, etc.
22:03 jnthn Used for type mapping, knowing when we cross a NQP/Rakudo boundary, etc.
22:03 arnsholt Since NQP objects leaking into Perl 6 is generally wrong
22:04 sorear Is a "static code ref" the same thing as a protosub in niecza?
22:06 avuserow does hyper or race work in any implementations?
22:07 sorear not afaik
22:08 jnthn sorear: I'm not sure I know what a protosub is in niecza.
22:08 jnthn sorear: Sounds like, though.
22:09 jnthn sorear: It's the original code ref before any closure cloning of it, etc.
22:09 FROGGS eiro: the thing in the string is a static lib, my patch from yesterday or so was meant to use the dynamic library instead, which get loaded by -lapr-1
22:10 FROGGS eiro: ... which fixed a build issue on my box
22:10 FROGGS eiro: so after your patch it will include the static library and load the dynamic one
22:11 FROGGS eiro: I'm curious that it works that way... what linux do you have? what architecture?
22:12 gdey joined #perl6
22:14 dalek roast: db118a6 | (Elizabeth Mattijsen)++ | S (2 files):
22:14 dalek roast: Unfudge tests that I made passing today
22:14 dalek roast: review: https://github.com/perl6/roast/commit/db118a6ca2
22:18 Guest1337 joined #perl6
22:19 thou joined #perl6
22:19 pmichaud lizmat++  # "tests that I made passing today"
22:20 pmichaud I always like hearing that.
22:28 labster joined #perl6
22:28 sorear jnthn: How does gather/take work?
22:29 tomyan joined #perl6
22:31 jnthn sorear: On Parrot uses coroutines. On the JVM uses the thing I didn't finish yet...
22:31 silug joined #perl6
22:32 jnthn sorear: It's the transform where at the point of a call you ensure the stack is empty.
22:32 jnthn And then uses exceptions as a stack iterator.
22:32 jnthn And save the stack
22:33 jnthn the "stack empty on call" invariant is already upheld.
22:33 sorear Neat
22:34 sorear How is invokedynamic being used here?
22:34 sorear I have no idea what's going on with this closure bug or how to instrument debugging for it
22:35 dalek v5: bb684f4 | (Tobias Leich)++ | lib/Perl5/ (3 files):
22:35 dalek v5: added support for: do EXPR
22:35 dalek v5: review: https://github.com/rakudo-p5/v5/commit/bb684f4a11
22:35 dalek v5: 4edddfd | (Tobias Leich)++ | / (2 files):
22:35 dalek v5: added "use vars" support, +128 passing tests
22:35 dalek v5: review: https://github.com/rakudo-p5/v5/commit/4edddfd850
22:36 colomon FROGGS++
22:38 sorear jnthn: how can I tell what part of the AST qb_13 corresponds to?  Is there anything else that's good to look at besides ast and classfile/javap?
22:39 jnthn sorear: Hmm...does javap not show annotations...
22:39 jnthn 'cus the actual name is in that.
22:44 havenwood joined #perl6
22:44 sorear RuntimeVisibleAnnotations:
22:44 sorear 0: #8(#9=s#109,#11=s#14,#13=s#110,#15=[s#111,s#112,s#113],#16=[],#17=[],#18=[],#19=[J#20])
22:44 sorear that?
22:44 FROGGS maybe you have to read that backwards
22:45 jnthn um. Yes, that. How...beautifully dumped :/
22:45 colomon http://71.10.146.50:3001/report # now with stats for for the first date, too.  +10 passing modules in the last two weeks (and I think maybe three more today, but not in the stats yet)
22:46 sorear Is it just me or is the sub name passed to every invokedynamic?
22:47 jnthn sorear: Yes; Perl 6 sub names don't comply with JVM method names
22:47 sorear jnthn: Is that just an ignored parameter for debugging?
22:47 sorear colomon: Nifto
22:47 jnthn sorear: Yeah
22:48 jnthn sorear: passing it as an arg isn't so bad, though; the JIT should be able to see that the bound callsite always drops it.
22:48 sorear niecza embeds a mangled version of the source name in the true method name
22:48 jnthn I pondered that, but was too lazy to mangle. :)
22:49 jnthn Don't mind if that happens.
22:50 * sorear wonders about wval_noa vs subcall_noa
22:52 jnthn sorear: wval_noa is for world values (lookups in an SC)
22:52 jnthn subcall_noa is a subroutine call, looking up the name lexically
22:52 jnthn So we do the resolution/lookup once.
23:07 gdey joined #perl6
23:07 timotimo i feel like someone has to build ACME::QuoteHarder
23:11 snearch joined #perl6
23:12 sorear jnthn: some qs for later: 1. is there a nice way to dump SixModelObjects?  2. what IDE if any do you use?
23:16 lue I don't suppose there's a way to do what I mean by "use Module::DebugStuff is export(:debug)" in a "main" module, is there?
23:17 lue (I could probably have the user set a global variable first and then "if ($*debug) { use Module::DebugStuff; }" in the main module, but I wonder if there's another way)
23:18 colomon joined #perl6
23:18 * colomon has retreated back to his hotel room
23:28 lue Seems like EXPORT might be my solution, but I can't get it to do anything :/
23:29 timotimo you mean sub EXPORT?
23:33 lue yes. Apparently I can't put in under C<module TheModule;>, and you must export it, in other words "sub EXPORT(*@a) is export {"
23:34 FROGGS joined #perl6
23:35 timotimo try our sub EXPORT
23:35 timotimo but yeah, that's what's currently implemented. it has a few holes
23:36 tadzik sorear: I'm quite sure jnthn uses Visual Studio
23:40 lue Also, use Module :named <positional> breaks (<positional> :named as well), and the use statement in the EXPORT sub doesn't work outside the sub (to be expected, but grr) :/
23:41 lue timotimo: I like sub EXPORT is export better. It feels more -Ofun :)
23:42 timotimo hehe
23:42 timotimo -Olol
23:42 japhb_ joined #perl6
23:49 lue r: macro foo(:$flag) { quasi { say 42 } if $flag }; foo(:flag);
23:49 camelia rakudo b2072f: OUTPUT«[31m===[0mSORRY![31m===[0m�Too many positional parameters passed; got 1 but expected 0�at /tmp/8bO2z1Miku:1�------> �»
23:50 * lue thought he had an answer for a second, but nope
23:56 lue Am I not allowed to use named parameters in macros, or is that a bug?

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

Perl 6 | Reference Documentation | Rakudo