»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by moritz on 3 May 2013.
eternaleye r: multi a( $b where * < 5 ) is pure { (state $c)++ }; multi a( $b where * >= 5 ) is pure { $b }; a( $_ ).say for 5..10; 00:00
camelia rakudo b2072f: OUTPUT«5␤6␤7␤8␤9␤10␤»
eternaleye r: multi a( $b where * < 5 ) is pure { (state $c)++ }; multi a( $b where * >= 5 ) is pure { (state $c)++ }; a( $_ ).say for 5..10;
camelia rakudo b2072f: OUTPUT«0␤1␤2␤3␤4␤5␤»
timotimo it's kinda not surprising that that can't be folded.
eternaleye The stub code one interests me, though. It looks like declaring is pure on a proto makes it fold the proto and ignore the multis. 00:01
lue Am I right in thinking the sentence about q:z<foo>// in S02:3877..3879 is useless, wrt S02:3752..3755 ?
eternaleye That might turn out to be a gotcha for people coming from Haskell. 00:02
timotimo oh
yeah, if you don't put a {*} somewhere in there, it won't work
eternaleye r: proto a( $b ) { ... }; multi a( Int $b ) is pure { $b }; multi a( Str $b ) is pure { $b }; a( $_ ).say for 5..10; 00:03
camelia rakudo b2072f: OUTPUT«Stub code executed␤ in method gist at src/gen/CORE.setting:10161␤ in method gist at src/gen/CORE.setting:893␤ in sub say at src/gen/CORE.setting:11047␤ in method say at src/gen/CORE.setting:890␤ in method say at src/gen/CORE.setting:889␤ in block at /tmp/Wd0…
eternaleye r: proto a( $b ) { * }; multi a( Int $b ) is pure { $b }; multi a( Str $b ) is pure { $b }; a( $_ ).say for 5..10;
camelia rakudo b2072f: OUTPUT«5␤6␤7␤8␤9␤10␤»
eternaleye r: proto a( $b ) { * }; multi a( Int $b ) is pure { $b+20 }; multi a( Str $b . 'zzz' ) is pure { $b }; a( $_ ).say for 5..10;
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Missing block␤at /tmp/xHNO3PbFJW:1␤------> $b ) is pure { $b+20 }; multi a( Str $b ⏏. 'zzz' ) is pure { $b }; a( $_ ).say fo␤ expecting any of:␤ postfix␤ statement end␤ statement modifier␤ …
eternaleye r: proto a( $b ) { * }; multi a( Int $b ) is pure { $b+20 }; multi a( Str $b ~ 'zzz' ) is pure { $b }; a( $_ ).say for 5..10;
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Missing block␤at /tmp/XeiI_4PcYX:1␤------> $b ) is pure { $b+20 }; multi a( Str $b ⏏~ 'zzz' ) is pure { $b }; a( $_ ).say fo␤ expecting any of:␤ postfix␤ statement end␤ statement modifier␤ …
eternaleye oldp5habits--
Gah
flussence lue: good catch, I think you're right 00:04
eternaleye r: proto a( $b ) { * }; multi a( Int $b ) is pure { $b+20 }; multi a( Str $b ) is pure { $b ~ 'zzz' }; a( $_ ).say for 5..10;
camelia rakudo b2072f: OUTPUT«25␤26␤27␤28␤29␤30␤»
eternaleye Mm
timotimo: Thanks
eternaleye timotimo: Haven't used proto since { ... } was the proper syntax. 00:04
timotimo oh! :)
i don't think i was born back then :P
eternaleye timotimo: Eh, it was less than 3yr back. 00:05
timotimo oh. yeah, i was indeed born before that
eternaleye r: proto a( $b ) is pure { * }; multi a( Int $b ) is pure { $b+20 }; multi a( Str $b ) is pure { $b ~ 'zzz' }; a( $_ ).say for 5..10;
camelia rakudo b2072f: OUTPUT«25␤26␤27␤28␤29␤30␤»
eternaleye Cool, that's more like I expected.
'is pure' + 'is cached' sounds like it'd be VERY nice for some mathematical work. 00:06
timotimo er. i dunno what i had expected to happen when i did pure on the 0 to 10 example with when * < 5 and when * >= 5 ...
it would seem like is cached should just blanket-apply is pure as well :P
eternaleye timotimo: Yeah, where clauses act differently IIRC. Don't they actually happen as a second stage in multi dispatch, and they just each get evaluated? 00:07
lue flussence: I think I'll remove it, unless someone who knows the design of Perl 6 like the back of their hand tells me that either it's the paren-only statement is wrong or that "no really, that's supposed to be like that"
timotimo eternaleye: i should have applied the same parater twice in a row to figure out if it'd get folded.
eternaleye timotimo: Mmm. for 5 xx 10? 00:08
timotimo more like that, yeah.
now i need a few seconds to think about how to generate a list 1, 1, 2, 2, 3, 3 in the most simple way.
r: say (1.0, 1.5 ... 10)>>.floor 00:09
camelia rakudo b2072f: OUTPUT«1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10␤»
timotimo almost.
r: say (1.0, 1.5 ... 10.5)>>.floor
camelia rakudo b2072f: OUTPUT«1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10␤»
timotimo r: say (^10 xx 2).flat.perl
camelia rakudo b2072f: OUTPUT«(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9).list␤»
timotimo oh, nope.
flussence std: q:z<foo>//
camelia std 2504c1a: OUTPUT«===SORRY!===␤Unrecognized quote modifier: z at /tmp/S8vTsyQXAY line 1:␤------> q⏏:z<foo>//␤Bogus term at /tmp/S8vTsyQXAY line 1 (EOF):␤------> q:z<foo>//⏏<EOL>␤Parse failed␤FAILED 00:00 42m␤»
eternaleye r: say (^10 xx 2).tree.zip.perl 00:10
camelia rakudo b2072f: OUTPUT«No such method 'zip' for invocant of type 'List'␤ in block at /tmp/5qLQU5oIOm:1␤␤»
eternaleye r: say ( [Z] (^10 xx 2).tree ).perl
camelia rakudo b2072f: OUTPUT«((0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9),).list␤»
timotimo oh, nope.cute idea.
eternaleye r: say ( [Z] (^10 xx 2) ).perl
camelia rakudo b2072f: OUTPUT«((0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9),).list␤»
timotimo r: say (^10 Z ^10).flat.perl
eternaleye r: say ( [Z] (^10) xx 2 ).perl
camelia rakudo b2072f: OUTPUT«(0, 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8, 9, 9).list␤» 00:11
rakudo b2072f: OUTPUT«((0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9),).list␤»
timotimo got it :)
eternaleye r: say ( [Z] (^10) xx 2 ).perl
camelia rakudo b2072f: OUTPUT«((0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9),).list␤»
timotimo no, look
eternaleye Yup, you did.
timotimo :)
eternaleye I couldn't tell who camelia was responding to, so I repeated mine.
timotimo ah. yeah, it would be something worth exploring in the evalbot source
if there were two requests very recently by different people, add a highlight to the front 00:12
eternaleye Mm, maybe make it "rakudo $revision (for $person):"
timotimo that sounds nice.
the evalbot is perl5, though, isn't it?
eternaleye timotimo: Even in general, not just close together. Simpler that way.
timotimo: Even if it is, that's find. P5 was my first programming languagd :D 00:13
flussence (while we're at it, how about playing tricks with sink context so code without a "say" returns the last value...)
timotimo nah, otherwise it's just more noise
eternaleye *language
timotimo you started programming that recently? ;)
eternaleye timotimo: I'm only 22, 23 on the 27th of this month.
timotimo my first experiences were when my father pointed me at qbasic and showed me nibbles and gorillas in qbasic ;)
eternaleye Started when I was 15.
timotimo but i was an incredibly priviledged child 00:14
eternaleye (Well, I used shell first, but I don't count that)
dalek ecs: 0cefa72 | lue++ | S02-bits.pod:
[S02] Removed useless statement concerning quoting forms.

Specifically, q:z<foo> will always be a quote form using angle brackets delimiters and an additional (imaginary in this case) :z adverb that didn't take any arguments, because according to the Adverbial Pair Forms section above, only parens give arguments to quote form adverbs.
00:15
ast: d0a1b09 | (Solomon Foster)++ | S32-io/dir.t:
Add some tests that Niecza can pass.
00:19
ecza: 5c63c23 | (Solomon Foster)++ | lib/CORE.setting:
IO::FileTestable doesn't have $!path.

Switch to ~self instead.
00:26
redditanalytics is there a way to see what release of Perl6 is in a certain ubuntu repository? 00:35
timotimo do you mean just by opening up the repository with a browser?
redditanalytics oh I was thinking from a command line 00:37
but I guess that would work
lol
I doubt the May 31 release is in the Repo
timotimo otherwise, apt should have some way to do it. i'm not sure either way
it's very unlikely, indeed.
redditanalytics what was the release before May 31?
timotimo i think earlier today i heard ubuntu has some 2011 version of rakudo star
march or february i think
rakudo.org/2013/02/24/rakudo-star-2...-released/ 00:38
redditanalytics let me go muck with it and find out. I'm running Mint 15 which I believe is using the 13.04 reps
timotimo if you mean a rakudo compiler release, it'd be a little bit more recent. hold on
tadzik FROGGS, FROGGS
FROGGS apt-cache showpkg rakudo
timotimo github.com/rakudo/rakudo/tree/nom/docs/announce - this page says we had compiler releases in 2013-04 and 2013-03, too
tadzik :D
FROGGS Versions:
0.1~2012.04.1-2build1
timotimo hah, were you sitting adjacent to each other? :P 00:39
tadzik FROGGS: where are you man?
redditanalytics haha
FROGGS tadzik: whats up?
in the hotel, finishing slides (hopefully)
tadzik FROGGS: we're about to have somefood, you're in?
FROGGS hmmm, no, really need to do this
redditanalytics what conference are you guys at?
timotimo FROGGS: i finished the last slide for my talk at the GPN literally 10 minutes before the talk.
YAPCNA in austin texas 00:40
FROGGS timotimo: well, I'm at least 30 slides from beeing done
tadzik FROGGS: ok, keep up the good work
FROGGS ó.ò
tadzik I'll eat for us both :)
FROGGS hehe
redditanalytics make sure you close your IRC window before presenting
timotimo i think i should try to popularize my incredibly niche presentation framework in the perl6 community :P
FROGGS tadzik++ # my man :o)
tadzik timotimo: please do! 00:41
timotimo it requires you to have perl6, use vim and also use i3 for a window manager.
FROGGS redditanalytics: my client is pretty silent
redditanalytics is Larry Walls as the conference?
FROGGS he is
timotimo but it's 10x better than using the perl6 repl that doesn't have line editing at all
redditanalytics you should go bar hopping later .. Austin is great 00:42
timotimo luckily it seems like many of the perl6 people like beer
FROGGS hehe, yes
timotimo (of course there's other nice stuff you can have at bars. like hot chocolate with rum :P ) 00:43
FROGGS but more important is that the perl6 ppl like ppl
tadzik timotimo: I'm fine with that :)
redditanalytics hahaha
what is the main site people should use to follow perl6? 00:44
FROGGS imagines a drunken snoring tadzik in his room later :P
redditanalytics perl6.org?
FROGGS redditanalytics: a site?
redditanalytics website
FROGGS what do you need a site for?
to stalk somebody?
timotimo i've put this: planetsix.perl.org/ into my feed reader
redditanalytics read blogs, follow releases, find ways to blow things up more spectacularly 00:45
FROGGS yeah, planetsix, masaks blog (strangelyconsistent) and jnthns blog (sixguts)
timotimo but those two blogs are in the planet, too, aren't they? 00:46
FROGGS ohh, are they?
I only read them when links are posted to this channel *gg*
timotimo yes, both are. 00:47
redditanalytics I am rewriting Reddit's entire search system in Perl ... this page has their real-time metrics ... www.redditanalytics.com/metrics.html
timotimo oh, a replacement for the notoriously bad search system? great!
redditanalytics perl doesn't even break a sweat
timotimo i have found it impossible to find a post that was on my front page after one or two days
timotimo even if i still know like three words out of the title 00:48
redditanalytics it is horrible. .. you'll be able to search by author, keywords, date range .. use filters (i.e. only show score > 1000)
any metric in their json
timotimo perl6 *would* break a sweat ;) 00:49
well, rakudo would anyway
redditanalytics hahaha
I'm keeping the last 500,000 comments in a hash
but I wanted to check out perl6 to see if it could do anything easier than perl 00:50
timotimo to my eyes, the perl6 code you'd end up with for any given problem is prettier, often simpler; but i'ven't done any perl5 whatsoever 00:51
FROGGS masak proved it today, reduce an >10 lines p5 script to a single line with about 25 chars, and it was of course more readable :o) 00:52
redditanalytics I really need to look at just how different the two languages are ....
timotimo i'm astounded by the comment-per-subreddit breakdown
redditanalytics the guys over on #perl say, "it's a different language completely" 00:53
and I said "but it's Perl"
timotimo indeed, it's perl. the difference isn't trivial, though
timotimo the fact that it has more type system in it than perl5 does makes a huge difference in my opinion 00:53
redditanalytics would a perl5 script run under perl6?
FROGGS like C and C++, there are at least as much differences as similarities
timotimo probably not, no.
FROGGS hey! 00:54
timotimo only if it's very simple (or if you use FROGGS++ v5 module)
FROGGS redditanalytics: some day, yes :o)
redditanalytics is there an Inline::Perl6 for perl 5?
just kidding
FROGGS no
but there are discussions about interaction of 5 and 6 that addresses this 00:55
timotimo indeed. i just heard about that concept when i watched the moarvm talk at yapcna 00:56
FROGGS right, but there is nothing to show right now 00:57
timotimo indeed. 00:57
flussence rn: my %foo = a=>1, b=>2; my %bar = a=>2, b=>3; say all(%foo{$_} < %bar{$_} for <a b>) 01:09
camelia niecza v24-65-g5c63c23: OUTPUT«===SORRY!===␤␤Unable to parse argument list at /tmp/QY0ZvUbI2W line 1:␤------> >1, b=>2; my %bar = a=>2, b=>3; say all(⏏%foo{$_} < %bar{$_} for <a b>)␤Couldn't find final ')'; gave up at /tmp/QY0ZvUbI2W line 1:␤------> …
..rakudo b2072f: OUTPUT«===SORRY!===␤Unable to parse expression in argument list; couldn't find final ')'␤at /tmp/uvzd9kNC7i:1␤------> a=>2, b=>3; say all(%foo{$_} < %bar{$_} ⏏for <a b>)␤ expecting any of:␤ postfix␤ statement end␤ …
flussence I don't understand what's wrong with that...
timotimo r: say all("a" ~ $_ lt "b" ~ $_ for <a b c>) 01:10
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Unable to parse expression in argument list; couldn't find final ')'␤at /tmp/w6BQAq2jpU:1␤------> say all("a" ~ $_ lt "b" ~ $_ ⏏for <a b c>)␤ expecting any of:␤ postfix␤»
timotimo r: say all(("a" ~ $_) lt ("b" ~ $_) for <a b c>)
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Unable to parse expression in argument list; couldn't find final ')'␤at /tmp/slZfVR8mor:1␤------> say all(("a" ~ $_) lt ("b" ~ $_) ⏏for <a b c>)␤ expecting any of:␤ postfix␤»
timotimo r: say all([("a" ~ $_) lt ("b" ~ $_) for <a b c>]) 01:11
camelia rakudo b2072f: OUTPUT«all(True True True)␤»
timotimo rn: my %foo = a=>1, b=>2; my %bar = a=>2, b=>3; say all(%foo{$_} < %bar{$_} for <a b>)
camelia niecza v24-65-g5c63c23: OUTPUT«===SORRY!===␤␤Unable to parse argument list at /tmp/xFhxlx7gNh line 1:␤------> >1, b=>2; my %bar = a=>2, b=>3; say all(⏏%foo{$_} < %bar{$_} for <a b>)␤Couldn't find final ')'; gave up at /tmp/xFhxlx7gNh line 1:␤------> …
..rakudo b2072f: OUTPUT«===SORRY!===␤Unable to parse expression in argument list; couldn't find final ')'␤at /tmp/NYNmjJFkYg:1␤------> a=>2, b=>3; say all(%foo{$_} < %bar{$_} ⏏for <a b>)␤ expecting any of:␤ postfix␤ statement end␤ …
timotimo rn: my %foo = a=>1, b=>2; my %bar = a=>2, b=>3; say all([%foo{$_} < %bar{$_} for <a b>])
camelia rakudo b2072f: OUTPUT«all(True True)␤»
..niecza v24-65-g5c63c23: OUTPUT«all([Bool::True, Bool::True], )␤»
timotimo rn: my %foo = a=>1, b=>2; my %bar = a=>2, b=>3; say all((%foo{$_} < %bar{$_} for <a b>))
camelia niecza v24-65-g5c63c23: OUTPUT«all(Bool::True, Bool::True)␤»
..rakudo b2072f: OUTPUT«all(True, True)␤»
timotimo rn: my %foo = a=>1, b=>2; my %bar = a=>2, b=>3; say all (%foo{$_} < %bar{$_} for <a b>)
camelia niecza v24-65-g5c63c23: OUTPUT«all(Bool::True, Bool::True)␤»
..rakudo b2072f: OUTPUT«all(True, True)␤»
timotimo i don't know why, but i fixed it
flussence I don't get it, but if one space fixes it I'm not gonna complain :) 01:12
timotimo rn: my %foo = a=>1, b=>2; my %bar = a=>2, b=>3; say all %foo{$_} < %bar{$_} for <a b>
camelia niecza v24-65-g5c63c23: OUTPUT«all(Bool::True, )␤all(Bool::True, )␤»
..rakudo b2072f: OUTPUT«all(True)␤all(True)␤»
timotimo now it's even shorter ;)
Nom- Curious... how is the JVM implementation of NQP coming along? I've been distracted lately with other $work related projects and haven't been able to keep up :) 01:13
I know it was mostly functional last I saw, but is the Rakudo implementation coming along? 01:14
timotimo nqp on jvm is already self-hosting
the rakudo implementation is coming along indeed, but a part of the work is being done on the MoarVM, too
redditanalytics apparently my xchat can't show all the unicode characters
FROGGS redditanalytics: you switched to utf8? 01:16
Nom- heh... building another VM :P
timotimo yeah, this time it's built from the start to work well with 6model
redditanalytics
ahhh there we go
timotimo also, it's got thoughts about threading put into it right from the start 01:17
Nom- I still question whether these guys would be better investing some time helping out on the Rubinius VM... there's already been some implementations of other languages on that VM and they're making pretty huge leaps on performance for Ruby code (at least)
timotimo do it ;)
parrot and jvm aren't going to be dropped just becaues of moarvm 01:18
so another vm will only mean the nqp code will improve
Nom- Yeah, I was starting to work on NQP on Rubinius a while back, but I was a bit out of my depth... went back to learning phase
FROGGS and having an vm that is made for 6model should actually be better in the end then, say, rubinius vm 01:19
timotimo yeah, that's also the conclusion fijal came to when i pulled him over here to discuss things
Nom- Maybe... but getting threading and a good concurrent GC is going to take a lot of hours
timotimo although he suggested reimplementing rakudo completely on rpython
(i think?)
Nom- Running on the JVM is probably a really good idea long term, but that startup time kills a lot of simple scripts 01:21
FROGGS true
Nom- I did come across RPython a while back ... interesting project if you want to learn stuff :)
timotimo yeah, moarvm is going to have killer startup times! (just my hopes and dreams)
FROGGS mine too 01:22
redditanalytics i know some of those words
timotimo :D 01:23
redditanalytics relevant for you guys later tonight ... xkcd.com/323/
timotimo already had a cocktail, but is very far away from texas 01:31
i may make a screencast tomorrowvabout and with my presentation framework 01:41
Nom- ls 01:46
oops wrong window :)
timotimo r: say dir 01:47
camelia rakudo b2072f: OUTPUT«IO::Path<star> IO::Path<src> IO::Path<.subversion> IO::Path<.bashrc> IO::Path<nom-inst1> IO::Path<toqast> IO::Path<test3.pl> IO::Path<.profile> IO::Path<t> IO::Path<nom-inst2> IO::Path<nom-inst> IO::Path<toqast-inst> IO::Path<toqast-inst2> IO::Path<examples> IO::Pa
timotimo r: say dir.>>Str 01:47
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Confused␤at /tmp/aNyzHlj3eO:1␤------> say dir.⏏>>Str␤ expecting any of:␤ dotty method or postfix␤»
timotimo say dir>>.Str 01:48
digh. darned phone keyboard 01:49
FROGGS r: say dir>>.Str
camelia rakudo b2072f: OUTPUT«star src .subversion .bashrc nom-inst1 toqast test3.pl .profile t nom-inst2 nom-inst toqast-inst toqast-inst2 examples Makefile .gitignore lib test2.pl .local bin rakudo p1 VERSION simple-tests .perlbrew std obj main.pl .lesshst nom niecza test.pl .cpanm log .bash_…
[Coke] gets a twitter password reset link IM and wonders who is trying to hack his account. 01:54
flussence rn: sub a { for ^5 { next if $_ %% 2; NEXT { say $_; START say 'first' } } }; a; a; 01:55
camelia rakudo b2072f: OUTPUT«0␤first␤1␤2␤3␤4␤0␤1␤2␤3␤4␤»
..niecza v24-65-g5c63c23: OUTPUT«===SORRY!===␤␤Action method statement_prefix:NEXT not yet implemented at /tmp/oWI6g8AoHc line 1:␤------> %% 2; NEXT { say $_; START say 'first' }⏏ } }; a; a;␤␤Unhandled exception: Unable to resolve method statement_leve…
timotimo wondering if there's any value to determining sequences (like 1, 2, 4 ... *) at compile-time 02:07
r: macro yoink(\x) { quasi { {{{ x }}}.Str }; yoink(123).perl.say 02:09
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Unable to parse expression in block; couldn't find final '}'␤at /tmp/6qYATakZwT:1␤------> i { {{{ x }}}.Str }; yoink(123).perl.say⏏<EOL>␤ expecting any of:␤ method arguments␤ statement end␤ s…
timotimo r: macro yoink(\x) { quasi { {{{ x }}}.Str } }; yoink(123).perl.say
camelia rakudo b2072f: OUTPUT«"123"␤»
timotimo looks to see what happens when sub WHAT is turned into a macro 02:11
flussence hm, one of these Text-Tabs-Wrap tests is taking an absurdly long time to run and I don't remember that being the case... 02:18
flussence ouch, profiler says regex matching is taking up 84% of the execution time. 02:31
timotimo "Code ref 'new' does not exist in serialization context" that wasn't helpful :(
Tene Hey, anyone here want my copy of The Art of the Metaobject Protocol? 04:02
I thought it would be a bit more general than it is; it's mostly a description of CLOS.
masak yeah, that's what it is, basically. 04:24
Teratogen meta-object 04:26
Teratogen is that like metaphysics? 04:26
PerlJam It's turtles all the way down 04:27
Teratogen damn straight 04:28
Teratogen PerlJam++ 04:28
colomon PerlJam++ # don't know why, must backlog 04:29
PerlJam heh
masak r: class A {}; say A.^name; say A.HOW.^name 04:31
camelia rakudo b2072f: OUTPUT«A␤Perl6::Metamodel::ClassHOW␤»
masak Teratogen: that second object there, is the object that knows a bunch of stuff about the class A. it's a class metaobject.
Teratogen ah, cooll 04:32
PerlJam Teratogen: I take it you haven't played with Moose much?
Teratogen BRAKE FOR MOOSE! (It Could Save Your Life) 04:33
no, haven't
eternaleye PerlJam: I thought it'd be ducks, really. 04:34
Teratogen there is a sci fi short story about a scientist who hated ducks, so he created a robot duck that slowly but surely killed all the ducks in the pond outside of his office 04:37
PerlJam I'm not sure why the platypus never made it as the Perl mascot
masak Teratogen: so for each class in a program, there's an instance of Perl6::Metamodel::ClassHOW, knowing the specifics of that class (attributes, methods, parent classes, etc) 04:38
PerlJam (more apt than ducks or camels or onions I think)
masak PerlJam: I'm guessing you've seen www.geekosystem.com/platypus-keytar...n-diagram/ ? :)
Teratogen this is how I imagine Perl 6 right now: 04:39
www.vagabondjourney.com/big-swiss-army-knife.jpg
sorear good * #perl6
PerlJam masak: actually, I hadn't (or at least hadn't remembered seeing it)
masak Teratogen: I generally find Perl 6 a bit more manageable than that -- but otherwise, yes ;)
PerlJam: the more I look at that Venn diagram, the more awesome it gets :) 04:40
PerlJam Teratogen: to me Perl 6 looks more like an ordinary swiss army knife that's bigger on the inside :) 04:41
It's like you can only see 4 or 5 tools at a time, but there are thousands of them to cycle through if you need them. 04:42
atroxaper r: say "Hello, #perl6!" 04:48
camelia rakudo b2072f: OUTPUT«Hello, #perl6!␤»
PerlJam atroxaper: greetings!
sorear o/ atroxaper, PerlJam 04:49
Jimmy__ Tene: ping 04:50
atroxaper o/ PerlJam, sorear
PerlJam sorear o/ 04:51
pmichaud 04:41 <PerlJam> Teratogen: to me Perl 6 looks more like an ordinary swiss army knife that's bigger on the inside :) 05:03
Perl 6: The Tardis of Programming Languages.
we even have daleks on our channel. 05:04
PerlJam fortunately they don't seem to be the kind that shout "EXTERMINATE!" all the time 05:05
sorear :) pmichaud 05:06
lue most git repos here consider the master authoritative :/
masak 'night, #perl6 05:09
PerlJam g'night masak 05:10
eternaleye lue: Yeah, but if something kills Davros our dalek goes too :P 05:11
lue
.oO(Suggestion: git repos use "doctor" instead of "master" for the main branch, and dalek is renamed to either Torchwood or UNIT :) )
05:13
eternaleye lue: The problem then would be that everyone would want to merge^Wdance with the 'harkness' branch. 05:35
trexy if perl5 is the swiss army knife of languages then perl6 is more like a sonic screwdriver with all the knobs "turned up to 11" ;-) 05:42
www.slideshare.net/trexy/nigel-hami...gameet2013
moritz good morning 06:47
daxim build.opensuse.org/project/monitor...ges:parrot # rakudo 2013.05 is all green 07:22
moritz \o/ 07:26
tadzik hello 12:02
daxim: have you seen panda with --prefix? 12:03
daxim bugs allewo!
github.com/tadzik/panda/issues/40 12:04
tadzik oh noes
duh
will look into it 12:05
tadzik I'm leaving austin soonish 12:05
nwc10 there is somewhere more exiting than Austin? 12:13
tadzik exams are pretty exciting, in a negative way 12:14
colomon colomon: huh, smoke test did not run overnight because Rakudo failed to build 12:16
[08:12am] colomon: gist.github.com/colomon/5705459
JimmyZ colomon: I got this failure sometimes, but after I rebuild parrot then no failure 12:27
colomon JimmyZ: I'm trying a full fresh build just to see what happens 12:27
redditanalytics good morning 12:28
colomon o/
redditanalytics I'm hoping to have some search components up later today. Will keep you posted (if you like Reddit)
colomon JimmyZ: yeah, looks like a clean build may work just fine. 12:32
JimmyZ colomon: :)
colomon Do you have ICU installed? 12:33
JimmyZ colomon: Nope
colomon I'm thinking I probably should for smoke testing. (Assuming I don't now... will check when the build is done.)
redditanalytics colomon, how long does your build process normally take and what kind of hardware do you use? 12:47
colomon redditanalytics: It's on the order of 5-10 minutes, I usually go do something else instead of watching it, if you know what I mean.
redditanalytics I am not a low level coder like you guys, but is C / Perl6 going to be able to take advantage of the new instructions in Haswell like FMA? 12:48
colomon This is on my Linux box. Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz, 32G RAM. I'm not trying to build in parallel. 12:51
redditanalytics that's a nice setup 12:52
I'm assuming you maxed out your ram with the 8gb SIMMS
unless you have a server motherboard
colomon my $work can take all the computing power I can throw at it. :) Having a nice machine for building p6 is just a bonus.
PerlJam redditanalytics: What's FMA? 12:54
redditanalytics Fused Multiply Add
en.wikipedia.org/wiki/FMA_instruction_set
colomon redditanalytics: I wouldn't expect Perl6 to take much advantage of FMA anytime soon.
redditanalytics I'm assuming they have to build it into the C compiler first
colomon redditanalytics: right, but even then it's a mismatch of levels. 12:55
redditanalytics I see
PerlJam looks like GCC supports FMA, so ... maybe? 12:56
colomon like I wouldn't expect $a + $b * $c in p6 to ever hit a level where C code would see a + b * c to optimize it.
PerlJam colomon: not even in the JIT?
moritz colomon: with natively typed ints and/or JIT it's thinkable, but a whole lot of work
colomon right 12:57
is FMA for ints or floating point?
moritz float 12:58
colomon but anyway, it seems like it would take special work to support it in p6, and even then you'd have to have just the right circumstances to make them work.
redditanalytics I believe integer
flaviusb You could perhaps have a quasi quoter or slang that took expressions like that and lowered them to FMA's, but it would be something that you would have to do around expressions you wanted to have that happen to. Right?
moritz "Both contain fused multiply–add (FMA) instructions for floating point scalar and SIMD operations" 12:59
redditanalytics what about Advanced Vector Extensions 2 (AVX2)?
moritz from that wiki page
redditanalytics that was the other big one
moritz fwiw I think that vectorization/SIMD is a better fit for Perl 6, at least when working with arrays of pre-declared type and shape 13:00
and then doing hyper/vector operations on them
but again, it's a topic far into the future
jnthn morning, #perl6
moritz for now we struggle with making our "normal" stuff fast
\o jnthn
PerlJam good morning jnthn 13:01
redditanalytics morning
colomon o/ 13:02
PerlJam (It's weird that we're all going through "morning" at the same time for a change :)
colomon \o/ morning has unbroken!
(except for moritz++)
nwc10 quick, write more tests for it
JimmyZ morning, jnthn 13:03
flussence (it's kind of odd just watching the join/quit messages scroll by all morning)
redditanalytics 90% of IRC traffic is probably just that 13:04
moritz generally ignores join/quit on #perl6
JimmyZ too
moritz redditanalytics: more about 33%, according to my logs 13:06
flussence I wish irssi was a bit smarter about pruning them; you generally don't care if $random_user reconnected 50 times if they've only said one line in the past month...
moritz (with a total of about 9 mio lines logged)
moritz flussence: irclog.perlgeek.de tries to be smarter, by collapsing multiple joins/quits into one 13:07
gfldex
.oO(if they finish Perl 6 as fast as they talk...)
moritz gfldex: it's not all #perl6
gfldex chatter is chatter even if you do it around the corner 13:08
moritz (only about 40% of logs is from #perl6)
moritz is data driven today 13:08
redditanalytics after downloading Perl6 last night (being a Perl5 guy) ... i.imgur.com/xVyoSl.jpg 13:09
JimmyZ :) 13:10
jnthn urgh, get it's gonna be hot outside 13:11
*bet
redditanalytics sunny and high of 75 here in Baltimore today 13:13
FROGGS sqirrel: whats up? 13:20
sqirrel: here: new.livestream.com/accounts/421914...ts/2150574 13:24
colomon is running an hour later this morning when the conference schedule is more or less an hour earlier.... 13:28
TimToady commuting & :) 13:32
timotimo oh wow, i got up right on time to see the first talk! that's amazing 13:36
timotimo am i the only one who doesn't hear anything on stream? 13:44
ah, now i do hear something (but there's humming) 13:45
(and a tiny bit of clipping)
mucker Oh boy, that "stranger than fact" talk was heavy. 13:52
mucker larry mentioned that you can go sigil-free in p6. can you ? 13:53
flaviusb Is there a non-flash version of new.livestream.com/accounts/4219145...ts/2150574 by any chance? 13:54
TimToady no, Larry pointed out that all Californians are liars
moritz r: my \x = 42; say x 13:55
camelia rakudo b2072f: OUTPUT«42␤»
timotimo flaviusb: you can grab the .smil file and it has a "http_base" and a "src" attribute, when you concatenate those you can directly feed mplayer or so 13:55
but whenever i tried, it would kick me off every few seconds :( 13:56
nwc10 mucker: IIRC that's using parcel syntax
flaviusb timotimo: Thanks, I'll give that a try. 13:57
mucker TimToady++ for non-linear talks 13:57
timotimo also, you need to watch out, it *may* change the contents of the file at specific intervals that i didn't try yet
(i just booted my windows box to watch the streams)
timotimo the humming is crazy; doesn't yapcna have enough money to hire a professional sound engineer for each lecture hall? 13:58
(but tbh, the one we had at gpn apparently wasn't competent enough to make sure the sound doesn't clip most of the time >_>)
btyler timotimo: as I understand, all the stream-tending people are volunteers 14:00
timotimo ah, mhm
d'aaw 14:01
TimToady mucker: rosettacode.org/wiki/First-class_fu...ons#Perl_6 is written partly in a sigilles style, except for $_ 14:02
*less
mucker TimToady: I don't know whether you have thought about it. But I have a lot of people use perly sytax in casual talk. Like $emphasis or $$greed. @many, `code goes here`. /yeah, a damn fine lie/ for regexes and snide(non-linear) comments. 14:04
I honestly think, the english needs to hacked with more syntax. 14:05
TimToady :)
mucker TimToady: if you are free, can we formalize perl as a markup language ? I'm a volunteer :) 14:06
pmichaud "Perglish"?
flaviusb timotimo: Cool. I'm getting cut off every few seconds too. I'll try monkeying around with mplayer streaming options.
timotimo flaviusb: the first thing you should try is -loop 0 14:07
:P
mucker: we could adopt Pod perhaps? :P
TimToady mucker: see perlcabal.org/syn/S26.html
mucker TimToady: I did read that, but I mean more along the lines of **acutual written** communation, for programming magazines. 14:08
raw.github.com/themucker/MarkPerl/...mo/test.mk 14:09
flaviusb timotimo: ... :)
mucker TimToady: new punctuation like ⸮, a hack on english alphabet or grammar for better communication (very silly idea :) 14:11
colomon had no idea FROGGS++ was that far along! 14:15
PerlJam FROGGS++
timotimo FROGGS++ :)
flussence If anyone else is still stuck, rtmpdump | mplayer is doing the job here :) 14:19
diakopter TimToady: maybe Mark wants to write an XS/C interpreter in Perl 6
pmichaud FROGGS++ 14:20
timotimo cool talk
pmichaud needs to be blogged somewhere :)
colomon +1
JimmyZ +1
timotimo once there's a reliable link to the video recording anyway 14:21
diakopter TimToady: or compiler
pmichaud I mean the work, not just the talk :)
timotimo ah, sure
lizmat FROGGS++ indeed! 14:22
colomon to me, it really feels like the last few years where it looked like p6 was treading water were regrouping to support all the awesome that is blossoming now...
atroxaper Was it FROGGS at stream? 14:22
timotimo yes
arnsholt Yeah
atroxaper Oh. I see :)
PerlJam colomon: indeed.
colomon (I mean, looked to the outside world like p6 was treading water)
timotimo oh yay pmichaud! i really enjoyed his presentations so far
colomon: "treading water" as in "stuck in the same place"? 14:23
gfldex timotimo: treading water is what you do when you swim in place. You just move your legs a little so you dont sink. 14:26
timotimo right, thanks
TimToady it was more like furious swimming to stay out of the whirlpool :)
timotimo hey FROGGS :) 14:27
FROGGS :o)
timotimo you missed all the "FROGGS++" messages from everyone :P
FROGGS is still alive, even when he not expected it 14:27
*g*
PerlJam FROGGS++ (here's one while you're here :)
FROGGS thank you all
sqirrel \o/ 14:28
masak FROGGS++ 14:29
FROGGS okay, now I can enjoy the yapc 14:29
TimToady FROGGS: I know the feeling :)
moritz you mean presenting was no fun? :-) 14:30
masak hugs TimToady
FROGGS moritz: well, it is better after than before
TimToady well, there's fun, and then there's "fun"
masak -O"fun"
FROGGS "hehe"
jnthn FROGGS++ # v5 talk
And v5 itself :) 14:31
labster FROGGS++
PerlJam FROGGS: A question no one asked you ... when will it be "done"? ;-)
FROGGS I even put my california-look-a-like shirt on *g*
masak TimToady: I just want to point out that I spotted the Epimenides in "all Californians are liars". ;)
timotimo that's why you were so colorful! :D
colomon n: say 1 eq "a" 14:32
camelia niecza v24-65-g5c63c23: OUTPUT«False␤»
colomon third try is the charm. # first two times I was offline, I think. stupid ATTWIFI.
FROGGS PerlJam: well, I'm at this for about three months, the first month was dedicated directly to get the grammar compiling and setting it up properly as a module
FROGGS PerlJam: so I expect much this year 14:32
masak n: sub letter-code { ord($^letter) - ord('a') + 1 }; say 1 eq letter-code "a"
camelia niecza v24-65-g5c63c23: OUTPUT«True␤»
colomon masak: I'm trying to sort out a problem where Niecza doesn't seem to be converting to Str before doing eq. Trying to sort out the boundaries of the problem... 14:33
PerlJam FROGGS: yeah, you've gotten much done (more than I expected anyway) Keep up the good work and have fun! :-)
FROGGS I'll do, it is quiet fun
pmichaud Indeed, it's impressive how far FROGGS++ has gotten.
FROGGS (it really is, Perl6 grammars are just awesome) 14:34
pmichaud I almost wish we could put up a "features matrix" like page for it.
colomon It's also a tribute to how nice Perl 6 grammars and NQP are.
PerlJam pmichaud: are you going to IRC during your talk? ;) 14:34
timotimo part of the amazingness is how great the tools already work to allow FROGGS to even start in the first place
FROGGS pmichaud: just email me the complete feature-list :P
timotimo: true
timotimo pmichaud: may i ask you to briefly repeat the questions of audience members during Q&A? 14:35
oh, he already disattended his laptop
masak n: sub letter-code { ((my %) = 'a'..'z' Z=> 1..*){$^letter} }; say 1 eq letter-code "a"
camelia niecza v24-65-g5c63c23: OUTPUT«True␤»
pmichaud timotimo: I will try to do so, but I often forget :) 14:35
masak n: sub letter-code { (hash 'a'..'z' Z=> 1..*){$^letter} }; say 1 eq letter-code "a"
camelia niecza v24-65-g5c63c23: OUTPUT«True␤»
timotimo thanks!
colomon masak: I suspect it knows how to do conversions of built-in types but not user-defined classes. 14:36
timotimo r: sub letter-code { (hash 'a'..'z' Z=> 1..*){$^letter} }; say "hello".comb>>.&letter-code;
camelia rakudo b2072f: OUTPUT«8 5 12 12 15␤»
masak colomon: I see.
colomon n: class Rasp { has $.berry; method Str { self.berry }; }; my $r = Rasp.new(:berry("blue")); say $r eq "blue" 14:37
camelia niecza v24-65-g5c63c23: OUTPUT«True␤»
timotimo i don't know who changed what, but stream audio is so much better right now
masak \o/
colomon masak: or, I am barking up the wrong tree
n: class Rasp { has $.berry; method Str { self.berry }; }; my $r = Rasp.new(:berry("blue")); say $r eq "straw" 14:38
camelia niecza v24-65-g5c63c23: OUTPUT«False␤»
masak n: class Tree { has $.type; method gist { "the $.type tree" }; class Dog { method bark-up(Tree $t) { say "woof: $t" } }; Dog.new.bark-up(Tree.new( :type<wrong> ) 14:39
camelia niecza v24-65-g5c63c23: OUTPUT«===SORRY!===␤␤Unable to parse argument list at /tmp/0SameNwoAj line 1:␤------> ) { say "woof: $t" } }; Dog.new.bark-up(⏏Tree.new( :type<wrong> )␤Couldn't find final ')'; gave up at /tmp/0SameNwoAj line 1 (EOF):␤------> …
masak n: class Tree { has $.type; method gist { "the $.type tree" }; class Dog { method bark-up(Tree $t) { say "woof: $t" } }; Dog.new.bark-up(Tree.new( :type<wrong> ))
camelia niecza v24-65-g5c63c23: OUTPUT«===SORRY!===␤␤Unable to parse block at /tmp/UeHTxmdQfp line 1:␤------> class Tree {⏏ has $.type; method gist { "the $.type t␤Couldn't find final '}'; gave up at /tmp/UeHTxmdQfp line 1 (EOF):␤------> og.new.bark-up(…
masak n: class Tree { has $.type; method gist { "the $.type tree" } }; class Dog { method bark-up(Tree $t) { say "woof: $t" } }; Dog.new.bark-up(Tree.new( :type<wrong> ))
camelia niecza v24-65-g5c63c23: OUTPUT«woof: Tree()<instance>␤»
masak hm.
n: class Tree { has $.type; method Str { "the $.type tree" } }; class Dog { method bark-up(Tree $t) { say "woof: $t" } }; Dog.new.bark-up(Tree.new( :type<wrong> ))
camelia niecza v24-65-g5c63c23: OUTPUT«woof: the wrong tree␤»
masak \o/ 14:40
btyler_ for the sake of stream watchers, please ask the speakers to repeat the question before they answer! 14:48
PerlJam btyler_: The question was "will JVM somehow show up in the feature matrix on perl6.org" give or take 14:49
btyler_: the other question that was asked while I was typing was "where does moarvm fall in the image on the screen" 14:50
btyler_ thanks. pmichaud++, this is a great talk
PerlJam I think he repeated the others. 14:50
masak pmichaud++'s talk is great. 14:51
colomon agreed. pmichaud++ 14:52
PerlJam masak++ (World :)
colomon n: my $p = IO::Path(:directory(""), :basename("Hello"), :volume("")); say $p eq "Hello" 14:54
camelia niecza v24-65-g5c63c23: OUTPUT«Unhandled exception: Unable to resolve method postcircumfix:<( )> in type Path␤ at /tmp/ZC7dpsnvEs line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4502 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4503 (module-CORE @ 58…
colomon :\
colomon n: my $p = IO::Path.new(:directory(""), :basename("Hello"), :volume("")); say $p eq "Hello" 14:54
camelia niecza v24-65-g5c63c23: OUTPUT«Unhandled exception: Cannot unbox a String from an object of repr P6opaque␤ at /tmp/c4o2vj35MP line 1 (mainline @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4502 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4503 (module-CORE @ 586) ␤…
colomon n: my $p = IO::Path.new(:directory(""), :basename("Hello"), :volume("")); say ~$p
camelia niecza v24-65-g5c63c23: OUTPUT«"Hello".IO␤» 14:55
moritz r: say IO::Path.can('Str')
camelia rakudo b2072f: OUTPUT«Str Str␤»
labster \o/
moritz so it doesn't look at .Str for comparisions?
PerlJam lots of jnthn++ love in the room
timotimo oh yeah
what's not to love
moritz don't cuddle him to death, we still need him! 14:56
TimToady the joke here was: just think how much he'd get done if he wasn't a slacker
moritz :-)
masak jnthn++ 14:57
timotimo hahaha
that would be mind blowing
JimmyZ jnthn++ indeed
masak PerlJam: "World" may be my best naming decision ever. 14:58
JimmyZ 麦高,明天就是你的演讲了? 14:59
PerlJam
.oO( "I don't speak wind" )
14:59
TimToady JimmyZ: it's still on the schedule 15:00
JimmyZ Yeah, so is 明天 15:01
TimToady 10:35 Texas time (CDT)
timotimo i'm interested in exploring deployment of perl6 on heroku. sounds like a nice blogpost?
JimmyZ agentzh‎ is here? 15:02
masak JimmyZ: 是。
PerlJam timotimo: why was that a question? :) 15:03
TimToady 是的
masak JimmyZ: 两个昨天,两个明天。^^ 15:03
timotimo ;)
JimmyZ 呵呵。
masak JimmyZ: 诚然,我值得的绰号“masakist” 15:04
:P
JimmyZ :)
TimToady Jam yesterday, jam tommorw, but never jam today...
*tommorow
masak TimToady: :D
for those who haven't seen that reference: en.wikipedia.org/wiki/Jam_tomorrow 15:05
PerlJam "at least the next few months" :-)
masak PerlJam: it's scary. 15:06
PerlJam: we're now moving so quickly that we're talking about *months*.
TimToady there has to be some advantage to thinking about this for 13 years in advance :) 15:08
PerlJam sounds like we're reaching critical mass. 15:09
(and we all know where that leads)
masak PerlJam: to the mushroom cloud? 15:10
PerlJam masak: a big explosion anywa. I'm not sure that a mushroom cloud is a given
masak "any sufficiently large blast will produce the same sort of effect" -- en.wikipedia.org/wiki/Mushroom_cloud 15:11
TimToady I want the World, I want the whole World, I want it now!!!
colomon just realized that even if he can't get Rakudo-JVM running on his Mac, he's still got a linux box. 15:13
PerlJam masak: if the explosion is big enough, atmospheric effects will be irrelevant 15:14
masak ok, you got me there.
points for thinking big enough.
colomon isn't seeing an obvious way to build the JVM version? Do you have to build nqp-jvm by hand (rather than by Configure.pl) ? 15:19
flussence cd nqp; perl ConfigureJVM.pl --stuff; ...;
(I'll let you know if I get it working :)
colomon flussence: there is no nqp directory at this point... 15:20
jnthn cd nqp; perl ConfigureJVM.pl --prefix=../install-jvm
jnthn make install 15:20
cd ..
jnthn perl ConfigureJVM.pl && make 15:20
colomon gist.github.com/colomon/5706754 # results of perl ConfigureJVM.pl for me 15:21
PerlJam chip++ 15:22
timotimo oh yes, you have to --with-nqp and ConfigureJVM.pl in nqp/
i think if you set the install path in ConfigureJVM to ../install-jvm it'll just work™
pmichaud++ # repeating questions 15:23
disco disco! 15:25
"i don't like that phase right now" or what was that?
avuserow yes
flussence phr
flussence "Stage parse : 78.584" 15:26
oh wow. I like this jvm stuff already...
timotimo pmichaud++ great talk, thanks! 15:27
avuserow and suddenly dozens of people start compiling rakudo-jvm :) pmichaud++ jnthn++
timotimo it's a good thing this java technology is cloud based, otherwise it would break under the strain of all the people trying to use it at once! 15:27
atroxaper 70? or 17? 15:28
flussence sounded like 70 to me
avuserow 70.
PerlJam 709
er 70
avuserow blog post: pmthium.com/2013/06/a-perl-6-develo...l-podcast/ 15:28
flussence well I managed to get a working java ./perl6 in the space of the last slide there :D 15:31
avuserow I only managed to get a broken nqp-jvm. What are the requirements of jvm? 15:32
flussence I dunno if there's anything special to do, I just did a `perl ConfigureJVM.pl --prefix=../installjvm && make && make install` in my rakudo/nqp/ and `perl ConfigureJVM.pl --with-nqp=installjvm/nqp --prefix=installjvm` in rakudo/ and it all seemed to build like normal 15:34
[Coke] rt.perl.org/rt3/Ticket/Display.html?id=118287 isn't stopping others from building rakudo-java? bother. 15:52
flussence is that on 32bit or 64? 15:52
[Coke] 64 bit. 15:53
[Coke] added to ticket. 15:54
flussence I've only had OOM problems like that with java when I go out of my way to mess with the system overcommit setting. 15:55
[Coke] RUN_PERL6 = $(JAVA) -Xms500m -Xmx2g -Xbootclasspath/a:.:$(NQP_JARS):rakudo-runtime.jar -cp $(NQP_PREFIX) perl6 15:57
[Coke] looks like the memory is only updated on the runtime, not on the build. 15:58
flussence how much RAM are you building with? mine's got about 4GB free.
[Coke] machine has 4gb, some swap. 15:59
flussence the one major difference I can see is I'm using IcedTea, not the standard java binary... 16:01
(but I can't imagine that's the cause, they're the same codebase IIRC)
[Coke] freed up some swap, trying again. 16:02
[Coke] no change. plenty of memory free, still dies. 16:04
flussence might be worth doing a sanity check: what does `cat /proc/sys/vm/overcommit_{memory,ratio}` show? (0 / 50 here) 16:07
PerlJam is probably going to wait for the "early majority" when it comes to the JVM 16:12
flussence isn't :) 16:17
colomon frankly, I think the faster we have JVM working, the better it will be for the community. So I'm going to do my best to get in early. 16:18
PerlJam colomon: oh, agreed on the importance of getting things working on the JVM 16:20
diakopter TimToady: combining diacritics should be optional forms of sigils 16:33
flussence flinches slightly 16:34
timotimo my ârray = 1, 2, 3 # like this?
timotimo slang it ;) 16:34
TimToady erm... 16:35
diakopter also 16:36
other diacritics could be declarators
anocelot diakopter: You're a diacritic.
colomon diakopter: was doubtful at first, but it would be pretty awesome to be about to put arrows over vector names... 16:37
diakopter anocelot: you're an anotation 16:37
TimToady you can already do all this with the sigilless style, at least by spec
grapheme classifiers are supposed to apply only to the base char 16:38
anocelot already uses sigils stylelessly. Oh, wait. You said... nvm
diakopter you can have a style in Perl?
flussence sure! just use the vim-to-html plugin... 16:39
TimToady std: my \Spin̈al_Tap = 42; say Spin̈al_Tap; 16:42
camelia std 2504c1a: OUTPUT«ok 00:00 42m␤»
TimToady nrp: my \Spin̈al_Tap = 42; say Spin̈al_Tap;
camelia pugs: OUTPUT«*** ␤ Unexpected "\\"␤ expecting "=", formal parameter, context, ":" or "("␤ at /tmp/_yVoDWeVXv line 1, column 4␤»
..rakudo b2072f: OUTPUT«===SORRY!===␤Method 'ast' not found for invocant of class 'NQPMu'␤»
..niecza v24-65-g5c63c23: OUTPUT«42␤»
colomon niecza++
FROGGS r: my \a = 42; say a 16:43
camelia rakudo b2072f: OUTPUT«42␤»
FROGGS r: my \ä = 42; say ä
camelia rakudo b2072f: OUTPUT«42␤»
TimToady rakudo's parser has not quite achieved STDhood yet... 16:44
FROGGS r: my \n̈ = 42; say n̈
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Method 'ast' not found for invocant of class 'NQPMu'␤»
FROGGS ahh
my irc client has problems with this character too
TimToady r: say 'n̈' 16:44
camelia rakudo b2072f: OUTPUT«n̈␤»
FROGGS when selecting text doesnt dispaly the dots, and the selection is one char off 16:45
TimToady r: sub n̈ { say '\o/' }; n̈
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Missing block␤at /tmp/ysqMX1Y9EC:1␤------> sub n⏏[31m̈ { say '\o/' }; n̈␤ expecting any of:␤ statement list␤ prefix or term␤ prefix or meta-prefix␤ new name to be defined␤»…
colomon sorear++
TimToady p: sub n̈ { say '\o/' }; n̈ 16:48
camelia pugs: OUTPUT«\o/␤»
TimToady au++ :)
FROGGS nqp: say 'n̈' ~~ / <identifier> / 16:49
camelia nqp: OUTPUT«Confused at line 2, near "say 'n\u0308' ~"␤current instr.: 'panic' pc 14721 (src/stage2/gen/NQPHLL.pir:5232) (src/stage2/gen/NQPHLL.nqp:279)␤»
FROGGS nrp: say 'n̈' ~~ / <identifier> /
camelia niecza v24-65-g5c63c23: OUTPUT«Unhandled exception: Unable to resolve method identifier in type Cursor␤ at /tmp/2Z1r4aDgHT line 1 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3026 (Regex.ACCEPTS @ 10) ␤ at /tmp/2Z1r4aDgHT line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/l…
..rakudo b2072f: OUTPUT«No such method 'identifier' for invocant of type 'Cursor'␤ in regex at /tmp/ZmBZmph2tm:1␤ in method ACCEPTS at src/gen/CORE.setting:10509␤ in method ACCEPTS at src/gen/CORE.setting:683␤ in block at /tmp/ZmBZmph2tm:1␤␤»
..pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.20120717/blib6/pugs/perl5/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** Can't locate P…
FROGGS nrp: say 'n̈' ~~ / <ident> /
camelia rakudo b2072f: OUTPUT«「n」␤ ident => 「n」␤␤»
..pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.20120717/blib6/pugs/perl5/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** Can't locate P…
..niecza v24-65-g5c63c23: OUTPUT«「n̈」␤ ident => 「n̈」␤␤»
FROGGS ha! 16:50
TimToady TMTOWTDI :)
FROGGS rn: say 'n̈' ~~ /^ <ident> $/
camelia rakudo b2072f: OUTPUT«Nil␤»
..niecza v24-65-g5c63c23: OUTPUT«「n̈」␤ ident => 「n̈」␤␤»
FROGGS there we go
colomon wait, why are rakudo and niecza different there?
lizmat because rakudo is wrong? 16:51
FROGGS rakudo doesnt treat it as an identifier, which is a bug
it doesnt treat is as a single character to be clear
lizmat but which should be easily fixable?
FROGGS .u n̈
yoleaux U+006E LATIN SMALL LETTER N [Ll] (n)
U+0308 COMBINING DIAERESIS [Mn] (◌̈)
colomon FROGGS: ah!
FROGGS so, what is it? is it one char or not? 16:52
colomon now I see, the n matches but not the diaeresis
flussence p5eval: use Unicode::UCD; Unicode::UCD::num("⅖")
p5eval flussence: ERROR: Can't locate Unicode/UCD.pm in @INC (you may need to install the Unicode::UCD module) (@INC contains: /lib) at (eval 7) line 1.BEGIN failed--compilation aborted at (eval 7) line 1.
TimToady what do you mean by "char"?
flussence :(
TimToady it's one grapheme
it's two codepoints
this is why we're moving toward NFG in p6
FROGGS what should <ident> match, a codepoint or a grapheme? 16:53
I guess the latter
TimToady p6 is supposed to default to grapheme semantics
no implementation does that completely yet
colomon rn: say 'n̈' ~~ /^ <ident>+ $/
camelia rakudo b2072f: OUTPUT«Nil␤»
..niecza v24-65-g5c63c23: OUTPUT«「n̈」␤ ident => 「n̈」␤␤»
FROGGS ohh
TimToady MoarVM is probably the closest, actually...
lizmat .u LATIN SMALL LETTER N 16:54
TimToady you don't need the +
yoleaux U+006E LATIN SMALL LETTER N [Ll] (n)
FROGGS hmm, looks like there is not much I can do about it right now for rakudo@parrot
TimToady rn: say "nnn" ~~ /^ <ident> $/
camelia rakudo b2072f, niecza v24-65-g5c63c23: OUTPUT«「nnn」␤ ident => 「nnn」␤␤»
TimToady FROGGS: NFG support on Parrot will probably be a retrofit from MoarVM 16:55
diakopter ... 16:56
FROGGS TimToady: would be cool...
k, off to lunch
benabik Well, that's fascinating: failed assertion 'Parrot_pcc_get_regs_used(interp, ctx, REGNO_STR) > idx' (compiling NQP master on Parrot master) 17:09
[Coke] flussence: this is OS X. there is no /proc 17:12
Tene no procfs?! abomination! 17:55
What's next? No sysfs?
diederich ouch 17:59
flussence [Coke]: whoops, no idea then. sorry 18:14
.oO( maybe I should keep a darwin VM around just in case... )
flussence
.oO( wow, freenode's being quite violent today... )
18:17
d4l3k_ : 206e529 | (Tobias Leich)++ | lib/Perl5/ (3 files):
let `say` be a term, rather then a sub

If it is a sub it will pollute nested v6 blocks.
18:18
: 37adf8f | (Tobias Leich)++ | lib/Perl5/Grammar.nqp:
allow subroutine forward declarations
: 53e3a12 | (Tobias Leich)++ | lib/Perl5/Actions.nqp:
handle missing <block> in routine_def
RedditAnalytics did they build this IRC network on a Fisher Price "My first topological IRC" kit? 18:20
timotimo have i missed the grammars talk already? 18:21
flussence they did build it on a "My first chat protocol"...
FROGGS timotimo: what grammar talk?
timotimo what if perl6 grammars could generate 18:22
is the title i think
flussence that's in 100 mins from now 18:23
timotimo great?
!
FROGGS right! 18:23
:o) 18:24
tadzik hello folks
timotimo hey 18:24
tadzik any idea which terminal on jfk has both power sockets and wifi? :) 18:25
FROGGS tadzik: no idea :/ 18:25
tadzik dang
timotimo i ended up locking myself out of my VM rather than making my screencat
flussence
.oO( the vt100 in the broom cupboard might )
tadzik what's up in austin?
FROGGS .oO( nice purly screencat ) 18:26
tadzik: reini urban about p2/potion/p5/p6 in a few minutes
tadzik nice
tadzik I'll continue my hunt for a place of power 18:27
tadzik ha! I look like a first-world bum now 18:30
colomon jnthn++ 18:50
colomon has a working Rakudo-JVM! 19:01
tadzik yay 19:01
colomon tadzik: your trip home going okay? 19:01
tadzik colomon: yeah 19:02
colomon tadzik: it was great to meet you.
tadzik colomon: I went to terminal 5 (I'm supposed to be at 1), sat next to the cold wall becuase there's a power socket and a wifi here
tadzik colomon: yeah, it was terrific to meet you too! 19:02
tadzik you're not done with me yet, I'll be back :) 19:03
I like the US
jnthn ;)
colomon :)
tadzik everyone speaks the language I understand 19:03
nwc10 and I'm sure the US loves you, and your money 19:03
tadzik and people are generally merry and chatty
of course it does, it's a fair deal I guess
colomon tadzik: that varies from place to place, but it's certainly true where I'm from, too. 19:04
(Merry and chatty, I mean.)
colomon tadzik: If you're trying to get around the country, you're certainly welcome to crash in our guest room for a few nights. 19:05
tadzik colomon: thanks, I appreciate that 19:05
tadzik diakopter scared me though with the "renting a car is expensive if you're under 25" 19:06
but I guess I could as well train around
colomon tadzik: that's actually of limited use in the US -- limited coverage and frequently pretty expensive. 19:07
tadzik hm 19:07
colomon but I bet if you put your mind to it, you can figure it out. (or we can -- bet you can get us locals to give you a hand, too.) 19:08
tadzik yeah, I guess so 19:10
tadzik bah 19:14
I can drive manual, are you not entertained? ;) 19:15
stupid regulations
colomon actually, taking the bus between cities might be the best solution. But I haven't done that since I was very young.
tadzik so I can basically rent a car at a sane price 7 years after I get my driving license? :)
nwc10 IIRC insisting on a manual and a couple of other obscure features was a way to game hire places to get a better car for the same price
as all the features weren't often available in a low-end manual 19:16
tadzik hmm
nwc10 I don't try this
I don't like hiring cars. Or driving, hugely
nwc10 Other people can drive. I'll drink 19:16
benabik There are some very surprising bus deals. I think there's a $15 bus between DC and NYC. 19:17
dalek kudo/nom: 2975f0b | pmichaud++ | README:
Add another note to README.
tadzik that's reasonable
colomon tadzik: I guess properly speaking, I'd rather encourage you to come visit the states! Just warning that the planing may be tricky if you're intent on seeing a lot of different places in one trip. 19:18
tadzik sure :) 19:18
colomon n: my $p = "hello".path; say ~$p; 19:27
camelia niecza v24-65-g5c63c23: OUTPUT«"hello".IO␤»
colomon thats... weird, actually
moritz it is 19:28
why would the stringification contain .IOI
moritz s/I$/?/ 19:29
quietfanatic rn: say "hello".path.gist; say "hello".path.perl 19:30
camelia rakudo b2072f: OUTPUT«IO::Path<hello>␤IO::Path.new(basename => "hello", directory => ".", volume => "")␤»
..niecza v24-65-g5c63c23: OUTPUT«Path<hello>␤IO::Path.new(:basename("hello"), :directory(""), :volume(""))␤»
quietfanatic rn: say ~("hello".path) 19:31
camelia niecza v24-65-g5c63c23: OUTPUT«"hello".IO␤»
..rakudo b2072f: OUTPUT«hello␤»
colomon moritz: yes, I'm considering that a bug. Just getting distracted here...
quietfanatic I thought say was supposed to use .gist
quietfanatic or wait 19:31
that's strinigification
nm
moritz n: print 'hello'.path 19:32
camelia niecza v24-65-g5c63c23: OUTPUT«Unhandled exception: Cannot unbox a String from an object of repr P6opaque␤ at /home/p6eval/niecza/lib/CORE.setting line 1734 (List.join @ 6) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1378 (print @ 6) ␤ at /tmp/QN8vu7XurQ line 1 (mainline @ 4) ␤ a…
moritz n: print 'hello'.path.Str
camelia niecza v24-65-g5c63c23: OUTPUT«hello»
moritz n: print ~'hello'.path 19:33
camelia niecza v24-65-g5c63c23: OUTPUT«hello»
moritz wwwwat?
oh
n: say (~'hello'.path).WHAT
camelia niecza v24-65-g5c63c23: OUTPUT«(IO)␤»
moritz n: say IO ~~ Str
camelia niecza v24-65-g5c63c23: OUTPUT«False␤»
moritz so the stringification with ~ doesn't return a Str 19:34
that is indeed "curious"
colomon moritz: yes, I think that's the source of my problem. or maybe a related problem. 19:37
PerlJam stevan++
colomon PerlJam: Moe talk? 19:38
PerlJam colomon: aye 19:39
colomon PerlJam: good stuff, I take it?
PerlJam yep
colomon PerlJam: I'll have to watch the video sometime 19:39
PerlJam moe == Perl 5½ :-) 19:40
colomon on scala.
PerlJam he borrows more from Perl 6 than I thought he did.
colomon how much is working now?
colomon moritz: oh, I see what you're saying. this is seriously odd. 19:44
timotimo oooh, i like karen pauleys accent a lot 19:46
colomon moritz: oooo, I'm an idiot.
timotimo: +1
colomon moritz: bug fixed locally. 19:47
dalek ecza: aee6525 | (Solomon Foster)++ | lib/CORE.setting:
Path.Str should NOT return an IO object.
19:48
colomon and pushed 19:50
PerlJam colomon: you'll have to watch the talk or grab the slides from stevan :)
colomon PerlJam: what, he swore you to secrecy?! ;) 19:51
PerlJam colomon: nah, I'm actually too busy helping a student worker back home on another channel, sorryt. 19:52
s/yt/y/
colomon PerlJam: that's a very good excuse. PerlJam++
dalek rl6-roast-data: 9cab527 | coke++ | / (4 files):
today (automated commit)
19:53
kudo-star-daily: fb63bcf | coke++ | log/ (5 files):
today (automated commit)
[Coke]_ looks the daily rakudo-star build exploded. wtf. 19:55
[Coke] (complained about being in a .git directory for the first time ever.) 19:56
[Coke] jnthn: any pointers on a workaround for rt.perl.org/rt3/Ticket/Display.html?id=118287 ? 19:57
.ask Juerd if he can install some version of java 7 on feather1 ? 19:58
yoleaux [Coke]: I'll pass your message to Juerd.
anocelot [Coke]: jnthn may be setting up a presentation just now.
timotimo seems so from the stream, he's plugging in VGA right now 19:59
[Coke] yes, I'm not expecting an immediate reply.
Juerd [Coke]: feather already has a severe RAM shortage, so no, I'm not going to install any java vm. 20:00
yoleaux 19:58Z <[Coke]> Juerd: if he can install some version of java 7 on feather1 ?
[Coke] Juerd: ... lovely. ok. guess I'll have to spring for a box somewhere. 20:03
colomon [Coke]: I haven't managed to get the smoke test first last night to run yet either. Not sure why, many things distracting me.
[Coke] colomon: your smoke test is all modules? 20:04
colomon [Coke]: yes
Its first attempt to build rakudo failed, I fixed that but now panda is not happy.
[Coke] colomon: is that different from sergot's thing that does that? 20:05
colomon you're just wanting to run make spectest for JVM?
[Coke]: I have no idea what sergot has.
[Coke] just? I'm wanting to integrate it with the existing one. which is currently (but not for long I guess) running on feather. 20:05
timotimo wow, holy hell, those pbc files, like CORE.setting.pbc could *really* benefit from a bit of compression 20:06
[Coke] colomon: feather.perl6.nl/~sergot/modules/, which I think was recently killed for running on feather.
timotimo as in: a huge part of it is just lots and lots of nullbytes
004a780: 0c00 0000 0000 0000 7a00 0000 0000 0000 ........z.......
colomon [Coke]: I've just got a straightforward script which updates rakudo, rebootstraps panda, and then runs emmentaler and stores the results.
timotimo almost every line seems to look like that
[Coke] it used to have a list of all modules in the ecosystem, and list whether or not they passed their tests.
colomon [Coke]: well, I do that: 71.10.146.50:3001/report 20:07
but that's running on my home box and isn't up all the time.
FROGGS_ moritz aka sinkmaster, are you there? 20:08
okay, already solved it I guess
[Coke] colomon: I'm just trying to make sure we're avoiding duplicate of effrot. 20:14
we already had sergot's list which was a rework of modules.perl6.org that never got backfilled.
colomon lexical classes... woah.
anocelot Is it a coincidence that ==> looks a bit like a shovel? 20:16
quietfanatic anocelot: Er, probably
pmichaud at jnthn's talk
btw, I've been thinking that i^Hperhaps n^Hinfinities should be transitive throuh .map, .grep, etc.
colomon [Coke]: my goal is mainly to highlight regressions in module performance
timotimo does somebody know if bailador will work on niecza?
anocelot Because when jnthn said "shove it into" I immediately thought, "Hey! It's a shovel!" 20:17
colomon timotimo: tend to doubt it, but I haven't tried.
timotimo does it sound sensible to want to try to put niecza onto heroku and plug it into the web framework using psgi?
colomon timotimo: no clue. I don't really know what heroku is, to be honest. :( 20:18
timotimo heroku is a pretty awesome hosting platform thingie 20:19
platform as a service
colomon timotimo: remind me to take a look sometime later? jnthn++ is being insanely cool at the moment. :) 20:21
timotimo i'm watching that right now, too 20:21
and yeah, he's so cool!
[Coke] Does anyone have a place where we can do a daily build for testing that isn't feather? 20:22
benabik What's jnthn talking about? 20:22
colomon www.yapcna.org/yn2013/talk/4657 20:23
timotimo mhhh, delicious yeti 20:24
timotimo why is the ( backslashed? 20:25
benabik misses YAPC. 20:25
timotimo ah, a capture 20:25
pmichaud I've been thinking that list infinity should be transitive through .map and .grep 20:26
(having seen jnthn++'s example at the beginning of the talk)
timotimo i wonder if there's some enhancement to be had from making "make" refer to the $/ more directly at compile time and spare some lookups? maybe that already happens
timotimo what was the good example? 20:28
pmichaud my @a = 1..Inf; my @b = @a.grep( *.is-prime ); # not jnthn's exact example, but it gets the point across 20:29
the first assignment knows to stop. the second one is eager and goes.... forever.
timotimo wait, is jnthn about to present a fully working generative grammar? 20:34
colomon_ timotimo: he's certainly sketching one out...
timotimo it would seem like jnthn to have a talk entitled "what if we could $really_cool_thing" and then presents "well, here's a fully working implementation." 20:35
colomon_ oh, absolutely
TimToady btw, I think that close to none of those <?> were actually necessary, given the invoking rules always eat whitespace first 20:36
[Coke] Unhandled exception: Method 'backtrace' not found (rakudo-java on any test file) 20:38
timotimo :(
[Coke] it's the "use v6"; 20:40
timotimo huh! 20:41
weird.
[Coke] if I comment those out, a lot of the sanity tests pass.
colomon_ Backtions++ 20:41
TimToady Bactions?
timotimo haha 20:42
what jnthn calls reverse action classes
colomon_ timotimo: that was a spelling question, TimToady++ is sitting in front of me. 20:43
timotimo ah
TimToady is always looking to save a character of typing :) 20:44
colomon_ github.com/jnthn/grammar-generative 20:45
[Coke] r: "is always looking to save a character of typing :)".chars.say 20:46
camelia rakudo b2072f: OUTPUT«50␤»
timotimo wow. 20:47
flussence *applause*
TimToady "vicariously lazy"
flussence a few years ago I wondered out loud if there was a way to turn the sprintf grammar in the spec into a list of valid patterns for autogenerating tests... but I'd written it off as impossible. 20:49
TimToady back to LBJ... 20:50
FROGGS_ flussence: if the goal is to support exactly the same like perl5, there are tests
colomon_ flussence: "impossible" means it takes jnthn++ 401 lines of code
flussence :D
timotimo jnthn++ # excellent stuff!
FROGGS_ hehe
flussence jnthn++ indeed 20:50
flussence (actually I think it was also suggested back then GGE could maybe be hacked up to do it, but I can't imagine that being fun to use :) 20:53
timotimo i'm currently discussing in #parrot that the CORE.setting.pbc has basically hundreds of screenfuls of one column of data, three columns of nullbytes in its hexdump output 20:55
0000190: 725f 414e 4e00 0000 a21f 2000 0000 0000 r_ANN..... .....
004b800: 0800 0100 0000 0000 e1ff ffff ffff ffff ................
every line between those two has '0000 0000 0000' in it.
flussence looks like lots of 32bit ints used to store tiny values? 20:56
timotimo 64 bits, no? 20:58
anyway, gzipping the core setting pbc turns it into 1.7 mb down from 17 mb. i wonder if startup time would be improved that way. 20:59
flussence why settle for gzip? xz -9 gets it down to 850kB :) 20:59
benabik Because integrating xz into Parrot is more effort than using the GzipHandle dynpmc that's already there? :-D 21:00
flussence oh fine :)
timotimo does xz decompress similarly fast?
flussence IIRC it's somewhere between gz and bz2, toward the gzip end of the scale 21:01
.oO( hmm, I have a zpaq compressor here... )
21:02
flussence well that's a bit rubbish, it actually came out larger than xz. 21:03
timotimo huh? 21:04
flussence lrzip -z
timotimo apparently the core setting also contains a boatload of binary data expressed in some kind of "don't have any non-printable characters" format
flussence (-z uses PAQ, which is ridiculously impractical and only meant to win benchmarks)
timotimo endless blocks of data that looks a lot like "AAAAAN0LAABUAAAA" 21:05
(perhaps the same kind of 2 bytes data + 6 bytes padding rubbish from the earlier parts of the pbc) 21:06
flussence I'm looking at the hex dump and wow. That's a ton of repeated values, not just null bytes
benabik Compiling for Parrot requires a step through a text format. Avoiding non-printables probably makes life easier. 21:07
timotimo >_< 21:08
benabik I was trying to fix that in last year's GSoC, but ran out of time. Having a baby tends to ruin schedules. 21:09
timotimo is the code severely bitrotted? 21:09
benabik Only got up to disassembly. IIRC, I was mucking around with making the Key PMC easier to deal with. It's fairly magical. 21:10
timotimo well, it's not really very pressing, is it?
benabik The disassembler still seems to work, for some value of work, although Rosella seems to be bit rotted. 21:11
timotimo i was just kind of astounded when i saw it.
flussence you know what works better than xxd for showing this stuff? 21:14
import it in gimp as raw data, indexed, ~1000 wide
jnthn survived his first talk :) 21:20
TimToady hopes to survive his last talk :) 21:21
timotimo he's calling perl dead! 21:24
could the serializing format be bettered? it's mostly As and \es 21:25
eternaleye flussence: zpaq is less a compression algorithm and more a container, which has a.) a compression algo written in a DSL and b.) a bytestream to decompress it.
flussence: So it can be faster than gzip or smaller than XZ, depending on the compressor. 21:26
flussence oh jeez, that's scary. Sounds like RAR...
eternaleye flussence: No, no
flussence: It was meant to fix the mutual imcompatibility problem in the PAQ family
flussence are you sure? I've seen an article that demonstrated rar is turing-complete... :)
eternaleye flussence: By encoding 'how to decode' inside the file.
flussence: Intriguing 21:27
flussence: Anyway, mattmahoney.net/dc/zpaq.html
^^^ Original author's page
flussence ok, based on my extremely crude .pbc viewer, the first third of the setting is poorly packed integers, the middle is base64 data, and the end is also poorly packed integers, but *sequential* too. 21:30
timotimo m) 21:30
could improving this yield better startup time?
i mean ... base64? seriously? :(
flussence it certainly couldn't hurt. 21:31
benabik The "first third" is the bytecode. The end is debug annotations with line numbers.
jnthn The JVM one is smaller, fwiw
timotimo jnthn: and the moarvm one is probably also smaller, no? 21:32
jnthn don't have it yet :)
benabik Well, Line numbers and bytecode indexes.
timotimo completely unreadable slide :(
oooh, i'm hungrin' for some noodles now :( 21:33
masak rn: say "Just Another Perl 6 Hacker".comb[22, 13, 5, 6].join.uc 21:35
camelia rakudo b2072f, niecza v24-66-gaee6525: OUTPUT«CPAN␤»
masak \o/
timotimo one thousand lines of push $P5463, 0 or push $P5463, 1
yaaaay
jnthn :P
masak r: say "CPAN".comb (<) "Just Another Perl 6 Hacker".uc.comb 21:36
camelia rakudo b2072f: OUTPUT«True␤»
timotimo ... (<) is syntax for what? 21:37
masak timotimo: set membership.
er.
subsethood, sorry.
timotimo mhm, ok 21:38
masak strict subsethood, even.
masak r: say set() (<) set() 21:38
camelia rakudo b2072f: OUTPUT«False␤»
timotimo oh, it's the texas subsetness thing 21:38
masak yeah.
flussence
.oO( the setting is quite literally painful to look at when rendered as 8bit grayscale )
21:39
timotimo flussence: what's the button to get that? 21:40
flussence open in gimp, filetype raw data, set to 1536w×20000h, indexed colour 21:40
timotimo cute
i missed the slide about marketing that made people applaud :( 21:41
anocelot (Mark Keating)
flussence stream stopped working entirely for me on this one :(
timotimo oh, ok 21:42
anocelot It looks like it's still sendind data. Reset the stream client?
flussence no good, it doesn't even get past the connecting stage for me :/ 21:43
(worked fine before...)
oh well, they'll all end up on youtube eventually... 21:45
anocelot Hmm... I'd check it on my lappy, but I'm trying very hard not to install flash on this machine. >.>
flussence I've been getting by using rtmpdump all day 21:46
timotimo what's the @{$x} for? o_O 21:47
this person is not making me interested in learning perl 5 after perl 6 :P 21:48
anocelot timotimo: I think it's basically @$x. Just unnecessicarily disambiguated. 21:49
timotimo oh, @$x == 0 would be "list inside $x has 0 elements"? 21:50
anocelot That's my understanding. It's certainly possible I'm wrong. ;) 21:51
huf the *array* 21:52
the array referred to by $x has 0 elements
timotimo sorry. okay.
dalek rl6-bench: 8a6d7d7 | (Geoffrey Broadwell)++ | timeall:
Fix a regression in timeall that broke the ability to save partial timing results if the user ctrl-c's in the middle of a chain of tests
21:57
rl6-bench: 911bc4d | (Geoffrey Broadwell)++ | components.json:
Add a moarvm component
rl6-bench: a24fce5 | (Geoffrey Broadwell)++ | compilers.pl:
Add an nqp-moarvm compiler entry
rl6-bench: cb84c62 | (Geoffrey Broadwell)++ | mi (2 files):
Add skips for tests that make nqp-moarvm blow up
huf oh jesus fuck timotimo dont listen to me
i'm mouthing off in the wrong channel. damn.
sorry.
OH, you *were* talking about p5? oh. 21:58
timotimo haha :) 21:59
huf yeah, p5 works like i said. :D
pmichaud "Easier to write a new language than patch p5." Hmmmm, where have I heard that before? ;-)
anocelot pmichaud: FORTRAN. 22:00
flussence PH... *ducks tomatoes*
timotimo nietzsche perl6? :) 22:06
anocelot "Quick language design is the opiate of the masses." -- Nietzsche perl6 22:08
timotimo d'aaw, pixel art doggie 22:12
timotimo haha, what's going on i don't even ... 22:17
anocelot Someone handed a very excitable person the microphone and he had everyone go Yeeeehaw and do the wave. 22:18
timotimo yeah, i kind of saw that 22:18
anocelot XD I'm good at answering rhetorical questions. 22:19
timotimo :)
dalek rl6-bench: 2c55b57 | (Geoffrey Broadwell)++ | compilers.pl:
nqp-jvm is no longer its own specialized compiler, it's just part of nqp, so fix the compiler def
22:26
timotimo oh so much text on a slide 22:30
two thousand packages o_O 22:31
timotimo i think he forgot to click the "does this setup look okay?" "yes" button 22:33
anocelot timotimo: Yerp. 22:35
dalek : 85b720e | (Tobias Leich)++ | / (4 files):
make `scalar` a token
22:37
lue hello world o/ 22:39
timotimo hey lue 22:40
masak .oO( o/ Hey, Lue o/ ) 22:41
timotimo i'd love to be able to use o/ ... o/ for regexes %)
masak r: sub circumfix:<o/ o/>($music) { say $music }; o/ Hey, Lue o/ 22:42
camelia rakudo b2072f: OUTPUT«===SORRY!===␤Unable to parse expression in circumfix:sym<o/ o/>; couldn't find final $stopper␤at /tmp/diwTHMN2B7:1␤------> >($music) { say $music }; o/ Hey, Lue o/⏏<EOL>␤ expecting any of:␤ argument list␤ prefix…
sorear o/ folks
masak :/
timotimo sorear: do you have experience with niecza and psgi? 22:43
FROGGS_ *schock*, I missed jnthn's talk >.<
-c
sorear timotimo: i got it working once, long ago
don't remember much 22:44
timotimo unfortunately i have no idea how psgi and friends work ... hm.
masak FROGGS_: I'm sure it'll be online.
FROGGS_ masak: whut? 22:54
damnit
do he'll give me a privat lesson if I buy beer? 22:55
masak: ahh, I read a 'not' in your sentence
sorear which jnthn talk?
FROGGS_ not it is gone
'if grammars could generate' 22:56
sorear it was cute
FROGGS_ 'cute' ?
jnthn or "crazy"
FROGGS_ "aha" 22:57
timotimo that's ingy, isn't it? 22:58
jnthn yes :) 22:59
timotimo perl6 to snusp compiler incoming? 23:00
sorear pretty sure shachaf(?) was working on a brainfuck backend for gcc 23:03
shachaf Me? No.
timotimo pretty sweet. 23:03
shachaf sorear: I think you're thinking of ais523.
sorear probably. 23:04
timotimo wait, where did i hear this before
wasn't the same talk in lightning talks yesterday too? 23:05
sorear not quite
timotimo oh, right, yesterday was a rage-filled cgi.pm must die talk
sorear Saywer X was doing his CGI MUST DIE talk
which incidentally was given, exactly the same, in Frankfurt 23:06
FROGGS_ incidentally?
sorear by the way
timotimo Frankfurt am Main?
sorear yes
yapc::eu 2012
I was there 23:07
FROGGS_ ya, seen it
timotimo mhm
TimToady chow & 23:09
sorear bye TimToady 23:10
timotimo oops, the psgi module fails with current rakudo :( 23:13
timotimo investigates
FROGGS_ masak: are you going to the game night or is there a hackathon planned? 23:14
timotimo gaaah, it fails only when using blib/lib 23:16
rakudo seems to be trying to go for $orig in a Stash 23:17
which doesn't seem to be the way you do it. 23:18
timotimo it seems like it doesn't supply an iteration thingie at all, so do i have to nqp::getattr? 23:19
timotimo /home/timo/build/rakudo/perl6.c:921: undefined reference to `Parrot_set_config_hash' 23:31
how did ... i do that?
timotimo oh i know 23:35
i stripped files that i shouldn't have stripped