Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2013-08-12

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:04 ggoebel joined #perl6
00:05 lizmat joined #perl6
00:06 crab2313 joined #perl6
00:16 stevan_ joined #perl6
00:18 woolfy joined #perl6
00:19 FROGGS joined #perl6
00:28 li joined #perl6
00:30 li left #perl6
00:32 raiph joined #perl6
00:40 FROGGS joined #perl6
00:48 FROGGS joined #perl6
00:49 mikemol joined #perl6
01:04 [Coke] f
01:07 dalek perl6-roast-data: eb1ee6f | coke++ | / (4 files):
01:07 dalek perl6-roast-data: today (automated commit)
01:07 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/eb1ee6f692
01:07 dalek perl6-roast-data: 8c6e5e7 | coke++ | / (4 files):
01:07 dalek perl6-roast-data: today (automated commit)
01:07 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/8c6e5e7e6d
01:09 [Coke] 98.9 and falling.
01:19 FROGGS joined #perl6
01:19 dayangkun joined #perl6
01:32 Mouq joined #perl6
01:42 diakopter [Coke]: are you still able to make use of host06?
01:43 xinming joined #perl6
01:43 benabik joined #perl6
01:53 [Coke] Using it every day for the rakudo.jvm run, yes
01:53 [Coke] still haven't gotten pugs working.
01:53 [Coke] so haven't done the cutover to run all the tests there.
01:56 colomon joined #perl6
01:57 diakopter cool; how long does the rakudo.jvm run take btw
01:59 grondilu joined #perl6
02:10 FROGGS joined #perl6
02:13 benabik joined #perl6
02:18 [Coke] 80min
02:18 [Coke] give or take
02:19 [Coke] and that's -with- the server. :(
02:19 [Coke] er, the eval-server
02:26 diakopter wat.
02:33 flussence joined #perl6
02:36 sorear sounds about right
02:36 sorear diakopter: [Coke] is running tests on a single core, I think
02:37 prevost joined #perl6
02:44 [Coke] yes.
02:47 diakopter sorear: oh
02:47 diakopter [Coke]: oh
03:00 Mouq Are trees in perl6 just (usually) Pairs?
03:01 Mouq Are pears grow trees, and trees grow pears, ∴ trees are pairs
03:01 colomon …. you could do it that way, I guess
03:02 Mouq s/are//
03:02 Mouq Yeah, I didn't really know if there was a better way to do it
03:04 colomon I know trees  are mentioned several times in the specs, but I don't recall ever actually running into them in practice.
03:05 colomon http://perlcabal.org/syn/S02.html#list%2C_flat%2C_item%2C_and_.tree
03:05 FROGGS joined #perl6
03:06 colomon I've got to say I don't understand the tree bits of that section at all.
03:08 Mouq I'll look into Takadonet's Tree::Simple [https://github.com/Takadonet/Tree--Simple/]
03:09 Mouq I think it .tree just turns Parcels of Parcels into LoLs?
03:09 Mouq Yeah
03:10 Mouq r: my $a = (4,5,(6,7)); say $a.tree.perl
03:10 camelia rakudo 2566e7: OUTPUT«LoL.new(4, 5, $(6, 7))␤»
03:10 sorear those trees are constructed from Lists, not pairs.
03:11 wtw joined #perl6
03:12 Mouq Yeah
03:12 Mouq Honestly Pairs are dumb way to do it :L
03:12 Mouq In general
03:15 Mouq Maybe. Yeah, like I said, I still have to look at it. It's weird since Pairs can have anything as a 'key'
03:22 colomon you can certainly build a binary tree structure with pairs.
03:23 colomon dunno that anything in p6 would automatically know how to make sense of it, though.
03:23 colomon whereas making it a LoL is at least something p6 has a notion of.
03:40 crab2313 joined #perl6
03:51 FROGGS joined #perl6
04:01 [Coke] is there a magic class method that is called immediately post-new if I wish to do some setup? (BUILD is too early, it seems). should I just override new?
04:04 sorear probably
04:04 sorear I mean that's what I'd do
04:05 sorear not to be confused with 'there probably is'
04:09 [Coke] Roger.
04:28 [Coke] star: use Rakudo::Debugger;
04:28 camelia star 2013.02: OUTPUT«===SORRY!===␤Could not find Rakudo::Debugger in any of: /home/p6eval/star/lib/parrot/4.10.0/languages/perl6/site/lib, /home/p6eval/star/lib/parrot/4.10.0/languages/perl6/vendor/lib, /home/p6eval/star/lib/parrot/4.10.0/languages/perl6/lib, /home/p6eval/.perl6/2013.02…
04:28 FROGGS joined #perl6
04:43 birdwindupbird joined #perl6
04:58 logie joined #perl6
05:07 timotimo o/
05:09 FROGGS joined #perl6
05:11 timotimo [Coke]: are you sure you're still running the rakudo.jvm tests? it seems to me it really should have changed some of the numbers in the recent runs
05:11 timotimo at least the amount of specd'd tests etc
05:15 timotimo i just want to make sure those numbers are right. and if they aren't, which is my gut feeling, they shouldn't be used during the yapc that's currently happening
05:26 aindilis joined #perl6
05:26 crab2313 joined #perl6
05:27 mdn joined #perl6
05:30 denisboyun joined #perl6
05:36 lizmat timotimo: they're the counts of number of tests in the block
05:37 lizmat what is #?DOES ?
05:39 timotimo #?does is sometimes needed to tell the fudger how many tests to skip or mark todo
05:41 mdn joined #perl6
05:41 locsmif joined #perl6
05:43 xenoterracide joined #perl6
05:44 vovka joined #perl6
05:45 FROGGS joined #perl6
05:46 xinming joined #perl6
06:01 sorear lizmat: if you have a test in a loop, the fudger needs to be told how many tests it's replacing so that the plan isn't invalidated
06:08 dmol joined #perl6
06:09 berekuk joined #perl6
06:15 moritz there's another use for #?DOES
06:15 timotimo good morning moritz :)
06:15 moritz you can put it before a subroutine declaration
06:15 moritz and then it tells the fudger how many tests that routine runs
06:15 moritz good morning timotimo
06:15 moritz and good morning #perl6
06:16 moritz going to the venue now&
06:17 logie joined #perl6
06:22 dayangkun joined #perl6
06:32 moritz at the venue, and online again :-)
06:34 timotimo if you have an epiphany about how to get at the signature of the Block node i'm currently in at any time of the day, just drop me a line :P
06:34 moritz sure
06:38 timotimo at first i was thinking i must be able to get the Code from the lexpad, but for little things that have no names, i most probably can't
06:39 kaleem joined #perl6
06:42 woolfy left #perl6
06:45 moritz
06:46 FROGGS joined #perl6
06:47 FROGGS o/
06:47 tadzik hai
06:48 moritz it's tadzik!
06:48 tadzik it's moritz! and FROGGS!
06:48 tadzik what a day
06:48 FROGGS :o)
06:48 * tadzik is warming up a place in Main Room
06:49 moritz timotimo: I wonder if we should just mark lexicals coming from signatures in some way in the QAST tree
06:50 timotimo it would be handy for my optimisation at least :)
06:50 moritz decl<sigvar> or so
06:50 moritz or maybe a totally new flag
06:50 moritz I'll try to talk to jnthn later
06:50 timotimo the nodes already have a hash interface for all kinds of annotations
06:51 timotimo i use that for storing poison bits at the moment
06:51 moritz that's not very nice in terms of memory usage, afaict
06:53 timotimo our memory usage is already atrocious; this is just during the compilation phase, though. the QAST nodes are thrown away after compiling, are they not?
06:54 timotimo i need to associate name-of-variable with bits anyway, so i probably won't get around a hash anyway
06:55 moritz never mind for now
06:56 Vico_Wu joined #perl6
07:00 mdn joined #perl6
07:03 ivanoff joined #perl6
07:06 berekuk joined #perl6
07:08 kaleem joined #perl6
07:10 dalek roast: 480b93c | ivanoff++ | S02-lexical-conventions/sub-block-parsing.t:
07:10 dalek roast: movie sub/hash syntax
07:10 dalek roast: review: https://github.com/perl6/roast/commit/480b93cb01
07:10 dalek roast: 89021cf | ivanoff++ | S13-syntax/sub-hash.t:
07:10 dalek roast: movie sub/hash syntax
07:10 dalek roast: review: https://github.com/perl6/roast/commit/89021cf8e0
07:25 sqirrel joined #perl6
07:31 SamuraiJack joined #perl6
07:33 leont joined #perl6
07:42 leont joined #perl6
07:48 berekuk joined #perl6
07:50 FROGGS .tell lizmat the Jugendherberge in frankfurt has no vacancies left ó.ò
07:50 yoleaux FROGGS: I'll pass your message to lizmat.
08:05 timotimo oh no
08:05 timotimo way too many perlers coming to the workshop :(
08:06 FROGGS ?
08:06 FROGGS ahh, to the one in frankfurt you mean
08:06 timotimo yes
08:07 leont joined #perl6
08:07 fhelmberger joined #perl6
08:08 timotimo running all the benchmarks through my optimisation with verbosity cranked up to max, it seems like only one of the benchmarks so far actually got anything out of the optimisation; unless others use something in the setting that benefitted from it, which i don't know
08:09 moritz r: sub f { say $CALLER::x }; my $x; f()
08:09 camelia rakudo 2566e7: OUTPUT«Cannot access '$x' through CALLER, because it is not declared as dynamic␤  in method at_key at src/gen/CORE.setting:8754␤  in method postcircumfix:<{ }> at src/gen/CORE.setting:1855␤  in method postcircumfix:<{ }> at src/gen/CORE.setting:1843␤  in sub f at /tmp/Idr…
08:10 moritz I wonder why/how I messed it up locally
08:11 timotimo messed up how?
08:12 FROGGS r: sub f { say CALLER::<$x> }; my $x; f()
08:12 camelia rakudo 2566e7: OUTPUT«Cannot access '$x' through CALLER, because it is not declared as dynamic␤  in method at_key at src/gen/CORE.setting:8754␤  in method postcircumfix:<{ }> at src/gen/CORE.setting:1855␤  in method postcircumfix:<{ }> at src/gen/CORE.setting:1843␤  in sub f at /tmp/lQl…
08:12 timotimo given that cancer is the uncontrolled multiplication of cells in ones body, "TimToady++" seems scarily appropriate ...
08:12 moritz timotimo: doesn't die anymore
08:13 moritz timotimo: I'm trying to make it a typed exception
08:14 timotimo oh? i had thought i've already seen the typed exception get committed to rakudo's repository
08:14 timotimo r: try { sub f { say CALLER::<$x> }; my $x; f() }; say $!.WHAT
08:14 camelia rakudo 2566e7: OUTPUT«(X::Caller::NotDynamic)␤»
08:14 moritz oh, I pushed that?
08:14 timotimo yes :)
08:14 timotimo and it seems to work, too
08:15 moritz \o/
08:15 timotimo now i'm wondering why i can't run rakudo-jvm for benchmarks any more ...
08:15 timotimo ah. it's because it exploded while compiling
08:15 timotimo "bad type on operand stack", probably my optimisation exploding earlier for the jvm because of the bytecode verifier
08:18 timotimo huh. sometimes faster, sometimes slower.
08:18 timotimo http://t.h8.lv/p6bench/locals-optimisation.html
08:19 timotimo well, since it's "wrong" anyway, better not pay too close attention to performance just yet. the benchmark doesn't verify that what it's doing is actually correct.
08:27 danaj joined #perl6
08:31 berekuk joined #perl6
08:35 SmokeMachine joined #perl6
08:35 crab2313 joined #perl6
08:45 SamuraiJack_ joined #perl6
08:54 BigBear joined #perl6
08:57 PuercoPop joined #perl6
09:08 leont joined #perl6
09:12 berekuk joined #perl6
09:17 FROGGS $ perl6 -e 'use QAST:from<NQP>; say QAST::WVal.HOW.name(QAST::WVal)'
09:17 FROGGS QAST::WVal
09:17 FROGGS \o/
09:17 FROGGS jnthn: ping
09:18 grondilu joined #perl6
09:19 tadzik woot
09:21 cognominal joined #perl6
09:23 leont joined #perl6
09:23 jnthn yays, I got on the wifi...
09:23 jnthn FROGGS: pong
09:23 lizmat joined #perl6
09:23 dmol joined #perl6
09:24 FROGGS jnthn: I can now load modules :from<NQP>, but this required updating stage0 for vm/parrot...
09:24 jnthn FROGGS: Oh?
09:24 jnthn FROGGS: OK, that is sometimes needed
09:24 jnthn Can I see the patch?
09:24 FROGGS jnthn: I'm going to push this to a branch, but vm/jvm needs an updated stage0 too I guess
09:24 FROGGS jnthn: I just push to a branch, okay?
09:25 denis_boyun joined #perl6
09:25 jnthn ok
09:26 jnthn will review
09:26 nugged joined #perl6
09:26 masak o/
09:26 jnthn o/ masak
09:26 jnthn which talk are you in? :)
09:26 FROGGS awesome!
09:26 FROGGS hi masak
09:26 cosimo joined #perl6
09:26 moritz jnthn: masak and me are in the redis takl
09:26 dalek nqp/from_nqp: ab98aad | (Tobias Leich)++ | src/ (12 files):
09:26 dalek nqp/from_nqp: give load_module the same signature as rakudo's
09:26 dalek nqp/from_nqp:
09:26 dalek nqp/from_nqp: This makes it pluggable, so we will be able to use modules :from<NQP>
09:26 dalek nqp/from_nqp: review: https://github.com/perl6/nqp/commit/ab98aad041
09:27 jnthn moritz: same!
09:27 jnthn masak: I dare you to ask at question time at the end, "Is it web scale?" :D
09:28 masak jnthn: :P
09:28 masak jnthn: you do it :P
09:29 nwc10 masak: you are on the same row. Look left
09:29 dalek rakudo/from_nqp: f63899e | (Tobias Leich)++ | src/Perl6/ (2 files):
09:29 dalek rakudo/from_nqp: wire NQP's ModuleLoader to use ...:from<NQP>
09:29 dalek rakudo/from_nqp: review: https://github.com/rakudo/rakudo/commit/f63899e31d
09:30 masak nwc10: yes, I know.
09:30 masak I spotted jnthn before the talk. we looked directly at each other, but I was assuming that jnthn didn't see me at that distance.
09:31 jnthn I didn't see you :)
09:31 FROGGS >.<
09:32 masak how would you do a sorted set in Perl 6?
09:33 moritz masak: with an RB tree?
09:33 lizmat fwiw, I think a sorted list primitive is missing from the spec, or did I miss that?
09:33 yoleaux 07:50Z <FROGGS> lizmat: the Jugendherberge in frankfurt has no vacancies left ó.ò
09:33 lizmat sh*t
09:33 FROGGS jnthn: btw, since I'm now able to use QAST, does it makes sense to expose this as AST (or rename it?)
09:33 jnthn No
09:34 FROGGS NO! :P
09:34 jnthn QAST isn't a Perl 6 standard thing, far from. We should require an explicit use of it.
09:34 FROGGS jnthn: don't we need it for slangs?
09:35 * jnthn waves hands
09:35 FROGGS hehe
09:35 jnthn Anyway, no, for now we should not pull it in by default.
09:35 jnthn But it's nice we can
09:35 jnthn Is it just those two last patches?
09:35 FROGGS true
09:35 arnsholt I think macros and slangs are going to need a standardised AST API of some kind
09:36 arnsholt Basing that off QAST may or may not be a good idea =)
09:36 FROGGS jnthn: yes, but nqp@jvm might be broken within these branches as I mentioned
09:36 jnthn wait, what...
09:37 jnthn What's that FLATTENABLE_HASH call doing in the NQP module loader...that's a Perl 6 specific thing...
09:37 FROGGS arnsholt: the good thing would be that we dont have not Yet Anotehr Translation Layer(tm), the bad thing is that we break code when changing QAST
09:37 FROGGS arnsholt: btw, do you have time to look at my zavolaj pull request?
09:38 dmol joined #perl6
09:38 arnsholt Oh, right. That thing
09:38 FROGGS jnthn: ohh, then this might need to happen in Perl6::ModuleLoader
09:38 arnsholt It looks mergeable, but I'm not sure if Inline::C should be using APIs that internal
09:38 FROGGS jnthn: otherwise is exploded because it can't iterate over a Stash ($*GLOBALish from Perl6)
09:38 jnthn ah...
09:38 FROGGS arnsholt: well, I can duplicate the code, but I wanted to avoid that for Inline::C
09:39 jnthn I'd rather we do that unpacking in Perl 6 module loader before passing it along, though.
09:39 arnsholt Yeah, but I think we might want to expose different internals =)
09:39 FROGGS jnthn: alright
09:39 arnsholt Like role Native
09:39 FROGGS arnsholt: ahh, okay...
09:39 FROGGS arnsholt: I must dig into this this evening, my HEAD is somewhere else atm :o)
09:40 arnsholt 's ok
09:40 arnsholt If that approach doesn't work out, we can expose the other stuff
09:40 arnsholt But I think if we export role Native, your code won't have to do very much plumbing
09:41 FROGGS that would be great
09:45 patspam joined #perl6
09:46 dalek roast: fe9f77c | moritz++ | S32-exceptions/misc.t:
09:46 dalek roast: Test for X::Caller::NotDynamic
09:46 dalek roast: review: https://github.com/perl6/roast/commit/fe9f77cb36
09:47 berekuk joined #perl6
09:47 moritz can anybody please test if that test passes for you?
09:48 moritz if not, I neet to push a local commit; if yes, I can drop it
09:50 TimToady joined #perl6
09:51 * moritz increases the rate limits for the irc logs, because we seem to be all NATted to the same IP
09:51 tadzik hah
09:52 nwc10 TimToady has landed in a talk? Gloria is at far left of 3rd row in the main room
09:54 mathw everyone having fun I hope?
09:55 * moritz has
09:56 patspam joined #perl6
09:56 tadzik yeah, fun :)
10:01 mathw :)
10:01 * mathw deleted loads of code this morning, that makes him very happy
10:01 FROGGS ?
10:02 FROGGS mathw: finished a project?
10:03 mathw removed a stupid feature that should never have been there in the first place
10:04 patspam joined #perl6
10:05 FROGGS ahh
10:10 nugged joined #perl6
10:14 nugged joined #perl6
10:19 fridim_ joined #perl6
10:23 leont joined #perl6
10:24 dalek rakudo/nom: 9bbec9c | moritz++ | src/ (2 files):
10:24 dalek rakudo/nom: typed exceptions for if()
10:24 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/9bbec9cb01
10:24 dalek roast: 52b7dda | moritz++ | S32-exceptions/misc.t:
10:24 dalek roast: RT #73806: if() as function call should warn if it leads to a syntax error
10:24 dalek roast: review: https://github.com/perl6/roast/commit/52b7dda0dc
10:25 jnthn moritz: Is "IfAsFunction" maybe better as "KeywordAsFunction" if it can be for non-if things too?
10:25 jnthn :)
10:25 moritz oh yes
10:27 dalek rakudo/nom: 2a26f1a | moritz++ | src/ (2 files):
10:27 dalek rakudo/nom: Rename exception class, jnthn++
10:27 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/2a26f1ad1e
10:28 dalek roast: 6b2c663 | moritz++ | S32-exceptions/misc.t:
10:28 dalek roast: Track exception class rename
10:28 dalek roast: review: https://github.com/perl6/roast/commit/6b2c663233
10:28 jnthn moritz++
10:31 moritz anyway, typed exceptions are great for testing
10:31 jnthn aye
10:31 * jnthn should track down the leftover typed exception test failures on jvm at some point :)
10:32 moritz what I'd really like is to have the NoSuchMethod error actually contain the invocant
10:33 jnthn Hm
10:33 moritz parrot only gives us the name of the class
10:33 moritz which is LLTA, actually :-)
10:33 jnthn I can probably work out how to do that on the JVM :)
10:33 moritz well yes, the code has the invocant right there when it throws the exception
10:33 jnthn aye
10:33 moritz or it least it did when I last looked at it
10:34 * jnthn ponders how to convey it
10:35 FROGGS jnthn: and I thought it would be easy :o( https://gist.github.com/FROGGS/7373412a4e57f42af5fa
10:36 dmol joined #perl6
10:37 jnthn FROGGS: hmm..
10:40 FROGGS exactly
10:40 FROGGS I'm going to put things in my stomach now... bbiab
11:11 berekuk joined #perl6
11:16 dmol joined #perl6
11:19 berekuk joined #perl6
11:22 denis_boyun joined #perl6
11:22 denis_boyun moritz: Hi moritz. Can you ever work with nqp?
11:28 moritz denis_boyun: yes
11:28 denis_boyun moritz: can you help. Do you know how to run a shell command with nqp? Is it possible?
11:29 denis_boyun moritz: such like system or `` on perl5
11:30 berekuk joined #perl6
11:30 lizmat joined #perl6
11:30 moritz denis_boyun: look at rakudo's src/core/control.pm, sub run
11:31 masak Just Another Perl Heckler
11:31 moritz denis_boyun: it uses pir::spawnw__IP($RPA), but that only works on parrot
11:31 moritz masak: ... & Koch
11:31 FROGGS denis_boyun: I guess you are interested in parrot specific code?
11:33 masak moritz: that some German brand that I don't know about? :)
11:33 denis_boyun FROGGS: yeap, at this year i'm working on gsoc with parrot foundation. It's needing for my project
11:33 * masak accidentally googles "Heckler and Kock"
11:33 masak oh, defense manufacturing.
11:34 FROGGS denis_boyun: yeah, I've seen you in #parrot
11:34 [Sno] joined #perl6
11:34 denis_boyun moritz: thanks a lot
11:35 moritz denis_boyun: and you can build an RPA in NQP just with []
11:35 FROGGS nqp: pir::spawnw__IP(["ls"])
11:35 camelia nqp: OUTPUT«bin␤boot␤dalek␤dalek-queue␤docs␤evalbot␤evalbot.log␤examples␤FETCH_URL␤foo␤harmless_testfile.txt␤js␤lib␤LICENSE␤log␤main.pl␤Makefile␤niecza␤nom␤nom-inst␤nom-inst1␤nom-inst2␤nqp-jvm␤obj␤p␤p1␤p2␤p5test␤perf␤perl5␤Pugs.hs␤rakudo␤rakudo-star-2012.02.tar.gz␤README.pod␤run␤simple-…
11:35 moritz nqp: pir::spawnw__IP(['echo', 'hellow', 'world'])
11:35 camelia nqp: OUTPUT«hellow world␤»
11:35 * moritz too slow on that unusable keyboard
11:36 moritz FROGGS++
11:36 FROGGS I'm sitting on a comfortable keyboard at $work.... :o(
11:37 FROGGS well, in front of* actually
11:37 nwc10 has anyone run the range code on riba's t-shirt yet?
11:38 berekuk joined #perl6
11:38 FROGGS denis_boyun: and you can pull out pure pir by running: nqp --target=pir -e 'pir::spawnw__IP(["ls"])'
11:39 FROGGS search for "spawnw" and you find the four lines you need
11:39 masak nwc10: what range code is that?
11:40 lizmat joined #perl6
11:40 nwc10 I don't have enough unicode in this terminal (etc) and it's screen
11:40 nwc10 camel .. camel onion
11:40 nwc10 gmmm, the onion isn't in Unicode
11:40 nwc10 it cheats!
11:42 moritz .u onion
11:42 yoleaux No characters found
11:42 moritz indeed :/
11:43 denis_boyun joined #perl6
11:43 BabsSeed What's the best way to get involved in compiler development? I'd like to learn and get involved.
11:43 nugged joined #perl6
11:43 masak rn: .say for '☃' .. '☄'
11:43 masak BabsSeed: where do you live?
11:43 camelia rakudo 2a26f1: OUTPUT«☃␤☄␤»
11:43 camelia ..niecza v24-88-g1f87209: OUTPUT«(timeout)☃␤WTF␤WTG␤WTH␤WTI␤WTJ␤WTK␤WTL␤WTM␤WTN␤WTO␤WTP␤WTQ␤WTR␤WTS␤WTT␤WTU␤WTV␤WTW␤WTX␤WTY␤WTZ␤WUA␤WUB␤WUC␤WUD␤WUE␤WUF␤WUG␤WUH␤WUI␤WUJ␤WUK␤WUL␤WUM␤WUN␤WUO␤WUP␤WUQ␤WUR␤WUS␤WUT␤WUU␤WUV␤WUW␤WUX␤WUY␤WUZ␤WVA␤WVB␤WVC␤WVD␤WVE␤WVF␤WVG␤WVH␤WVI␤WVJ␤WVK␤WVL␤WVM␤WVN␤W…
11:44 BabsSeed masak: Southern Europe
11:44 FROGGS BabsSeed: maybe start reading here, and then choose a project you would like to contribute? http://www.josetteorama.com/all-about-perl-6-interview-of-jonathan-worthington-part-1-of-3/
11:44 masak BabsSeed: it's funny you should ask tight now.
11:44 masak BabsSeed: there's a workshop being planned in September in Frankfurt.
11:44 BabsSeed masak: Oh, Frankfurt isn't far at all
11:45 FROGGS BabsSeed: where exactly is "Southern Europe"?
11:45 * masak .oO( Niecza goes "what the fuck?", "what the... Godwin?", "what the hell?"... )
11:46 BabsSeed FROGGS: I'm averse to saying too much in a probably logged public channel :P
11:46 FROGGS *g*
11:46 FROGGS it is logged, yes
11:47 FROGGS btw, many devs here are from europe
11:47 * TimToady always says too much, except when he doesn't...
11:47 berekuk joined #perl6
11:47 FROGGS hehe
11:47 BabsSeed FROGGS: I'm from England, but I live in southern Europe :P
11:47 FROGGS statement_mod:<TimToady>
11:48 FROGGS ahh, jnthn used to flee from England too
11:48 BabsSeed Flee those taxes :D
11:49 tadzik haha
11:49 moritz I guess jnthn's problem with England was that "England" doesn't start with "S"
11:49 tadzik isn't taxes somehow related to "you get what you pay for"? :P
11:49 BabsSeed tadzik: I didnt get what I paid for :P
11:49 tadzik heh
11:49 TimToady has he got to Switzerland yet?
11:50 BabsSeed Without delving too much into detail, I paid a fair punch of social security 2005-2008 then got hit by the recession and had no real safety net
11:51 colomon_ joined #perl6
11:51 BabsSeed So, yeah I wouldn't mind if I got what I paid for :D But unfortunately other people got what I paid for.
11:59 leont joined #perl6
11:59 BabsSeed What areas are most desired for contribution? I'm very comfortable in C
11:59 denisboyun joined #perl6
11:59 TimToady you'd probably be more comfortable working on MoarVM than on JVM, I imagine
11:59 tadzik you may want to look into MoarVM
12:00 nwc10 you might be more answers once jnthn finishes this talk: http://act.yapc.eu/ye2013/talk/4912
12:00 nugged joined #perl6
12:00 TimToady there's a #moarvm irc channel
12:00 TimToady but we're all kinda busy conferencing at the moment
12:01 BabsSeed Thanks
12:01 TimToady I think #moarvm is logged by the bot, if you want to backlog
12:01 nwc10 it is, http://irclog.perlgeek.de/moarvm/2013-08-12
12:01 BabsSeed Will take a look, definitely want to get involved
12:01 TimToady http://irclog.perlgeek.de/moarvm/2013-08-12 for today's log
12:04 BabsSeed I need to get a github, have been meaning to for ages
12:04 [Coke] timotimo: I'm fairly certain I'm updating it properly yes (it involves scp'ing the output file back to feather and rerunning the bit that generates the CSV file, but yah. I'll double check today.
12:04 nugged joined #perl6
12:05 BabsSeed [Coke]: Someone said you've implemented a Twitter bot in p6, did you make a Twitter module for that?
12:05 BabsSeed I'm considering writing a Twitter module but don't want to reinvent the wheel
12:10 lizmat joined #perl6
12:11 [Coke] BabsSeed: http://feather.perl6.nl/~coke/cj.pl
12:11 [Coke] no module, it was pretty simplistic.
12:13 [Coke] issues: that used to take forever to startup the connection to IRC, and MuEvent predates actual threads; there is probably a better solution these days in rakudo.jvm (but rakudo.jvm has no sockets yet)
12:14 dalek Perlito: 261d4e2 | (Flavio S. Glock)++ | / (3 files):
12:14 dalek Perlito: Perlito5 - grammar - subroutine predeclaration
12:14 dalek Perlito: review: https://github.com/fglock/Perlito/commit/261d4e27e3
12:16 FROGGS [Coke]: I think it has sockets in a development branch # tadzik++
12:17 nwc10 translucent is the new shiny
12:19 denis_boyun joined #perl6
12:19 JimmyZ I see mortiy‎ and moritz there...
12:20 tadzik [Coke]: actually, I was planning to port MuEvent to rakud-jvm on the hackathon yesterday
12:20 tadzik but then I noticed the lack of sockets, so I started that first :)
12:20 tadzik and of course Java makes it bizzare
12:20 * FROGGS .oO( and then you noticed the lack of $a, and started that frist )
12:20 tadzik because you can't have juts a Socket. You also have a SocketServer, and SocketImple, which is not useful for anything, really
12:21 tadzik and then InetAddress, and 5 subclasses of it
12:21 tadzik </rant>
12:22 dalek Perlito: d6ada0a | (Flavio S. Glock)++ | src5/lib/Perlito5/Grammar/Map.pm:
12:22 dalek Perlito: Perlito5 - grammar - TODO cleanup
12:22 dalek Perlito: review: https://github.com/fglock/Perlito/commit/d6ada0aeb8
12:23 moritz JimmyZ: we also have Ulti and Util :-)
12:23 mathw tadzik: Java's choices in the class library are one of the reasons why I never liked it much
12:24 tadzik the way Java treats sockets is a lot different that what we're used to
12:24 moritz tadzik: as long as there is no InetAddressFactory :-)
12:24 tadzik moritz: there is SocketFactory, don't worry
12:24 dakkar joined #perl6
12:25 moritz tadzik: oh great!
12:25 mathw I think the only reason Java doesn't have StringFactory is because they cheated and built functionality into String that's unachievable in a user-defined class to make it efficient
12:26 mathw (still bitter about how String gets an overridden infix:<+>)
12:26 JimmyZ moritz: yeah, I saw too :P
12:27 nugged joined #perl6
12:30 [Sno] left #perl6
12:34 kbaker joined #perl6
12:36 BabsSeed [Coke]: Thanks
12:38 BabsSeed [Coke]: Oh OK, I'm looking to implement the full oauth/streaming API
12:38 BabsSeed For use with clients/bots for Twitter
12:39 nine_ joined #perl6
12:45 JimmyZ nine_: wb
12:53 BenGoldberg joined #perl6
12:57 tadzik apparently, I'm Bad Luck Brian ;) http://imgur.com/ufHyjzi
13:00 moritz tadzik: :-)
13:01 leont joined #perl6
13:04 FROGGS *g*
13:06 nugged tadzik: non-LOL like in python channel :)
13:07 tadzik haha
13:07 tadzik This is #perl6, I can say what I want: LOL
13:07 FROGGS *lol*
13:07 tadzik FROGGS: watch your mouth!
13:07 FROGGS SCHMETTERLING!
13:07 tadzik :D
13:07 FROGGS ohh, okay ;o)
13:07 tadzik I mean, lol
13:08 tadzik compiling src/Perl6/ConstantFolder.nqp, all I saw is src/Perl6/CannonFodder.nqp
13:10 FROGGS what would be CannonFodder in Perl 6 world?
13:10 moritz do we even need the ConstantFolder.nqp thingy?
13:10 tadzik let's rewrite it in Perl6
13:10 tadzik the game, I mean
13:10 FROGGS there is a game called cannon fodder?
13:11 moritz iirc it's very limited, and these days we have the optimizer doing a much better job
13:11 tadzik sure
13:13 tadzik aww, yiss
13:14 dalek nqp/jvm-sockets: 66ee630 | tadzik++ | src/vm/jvm/ (3 files):
13:14 dalek nqp/jvm-sockets: Add opcodes for bind, listen, accept
13:14 dalek nqp/jvm-sockets: review: https://github.com/perl6/nqp/commit/66ee630378
13:19 nugged joined #perl6
13:20 dmol joined #perl6
13:23 TimToady finally got on again; room 2 is kind of dead...
13:23 masak o/
13:23 GlitchMr grep(&is-prime, 1..*)[25000] is slow compared to alternatives in C++ and JavaScript - http://v8-io12.appspot.com/#16
13:24 GlitchMr But it's way shorter, so perhaps there is nothing wrong with that.
13:25 PerlJam GlitchMr: When do you need the first 25,000 primes quickly?
13:25 PerlJam :-)
13:25 GlitchMr For benchmarks.
13:25 GlitchMr I gave up after waiting for two minutes.
13:26 masak GlitchMr: you can probably speed it up by writing it without ranges and grep and the call to &is-prime.
13:26 dmol joined #perl6
13:26 TimToady say grep(&is-prime, 2,3,5,7 ... *)[^20]
13:26 TimToady nr: say grep(&is-prime, 2,3,5,7 ... *)[^20]
13:26 camelia niecza v24-88-g1f87209: OUTPUT«(Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any) (Any)␤»
13:26 camelia ..rakudo 2a26f1: OUTPUT«␤»
13:26 TimToady nr: say grep(&is-prime, 2,(3,5,7 ... *))[^20]
13:26 masak heh.
13:26 camelia rakudo 2a26f1, niecza v24-88-g1f87209: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71␤»
13:27 TimToady nr: say grep(&is-prime, 2,(3,5,7 ... *))[1000]
13:27 camelia rakudo 2a26f1, niecza v24-88-g1f87209: OUTPUT«7927␤»
13:27 TimToady nr: say grep(&is-prime, 2,(3,5,7 ... *))[10000]
13:28 * FROGGS .oO( say grep(&is-prime, 2,(3,5,7 .. *))[^20] # excluding Inf )
13:28 camelia rakudo 2a26f1, niecza v24-88-g1f87209: OUTPUT«(timeout)»
13:28 jnthn yay, survived my talk :)
13:28 moritz and very well at that :-)
13:28 TimToady did you mean ...^ ?
13:29 TimToady using .. will only count by 1's again
13:29 FROGGS TimToady: yeah, looks like :/
13:30 TimToady I suspect 3,5,7...* will already exclude Inf somehow :)
13:30 FROGGS yeah, it was supposed to be a joke
13:30 TimToady though it might take it a while to do so :)
13:30 lizmat joined #perl6
13:30 ajr joined #perl6
13:30 FROGGS nr: say Inf.is-prime
13:31 camelia rakudo 2a26f1: OUTPUT«No such method 'is-prime' for invocant of type 'Num'␤  in block  at /tmp/eLhUkXMquP:1␤␤»
13:31 camelia ..niecza v24-88-g1f87209: OUTPUT«Unhandled exception: Unable to resolve method is-prime in type Num␤  at /tmp/1zYPK7CGgC line 1 (mainline @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4579 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4580 (module-CORE @ 576) ␤  at /ho…
13:31 berekuk joined #perl6
13:31 moritz I hate that "Unable to resolve method" error message
13:31 moritz *WHY* is it not able to resolve the method?
13:31 FROGGS No such method...
13:31 FROGGS is way better
13:33 FROGGS I like to quote M$ here: "Unexpected exception..." - Ohh, so it wasn't expected to fail? How exceptional -.-
13:34 TimToady it would be nice to list all the classes that have that method name, at least if there aren't too many
13:34 TimToady maybe figure out relate classes somehow
13:34 TimToady related*
13:34 kaare_ joined #perl6
13:35 FROGGS ohh, that is nice
13:35 FROGGS that saves me to grep for a method name from time to time
13:35 notjack joined #perl6
13:35 nugged_ joined #perl6
13:36 moritz well, it's not easy to compile a list of classes that have a certain method without leaking anonymous classes
13:38 lizmat wouldn't that just be a matter of adding an attribute in a metaclass indicating whether it should be included in such a list ?
13:39 jnthn Potentially
13:39 PerlJam Perhaps that it's anonymous already accomplishes this.
13:39 jnthn Just need to be careful not to make stuff never-collectable
13:42 mdn joined #perl6
13:44 BenGoldberg Maybe it could be done as a debugging option, turned on by an environment variable?  If it's not done by default, it might not matter if it makes some classes not-garbage-collectible
13:46 rindolf joined #perl6
13:50 PuercoPop left #perl6
13:52 pmichaud good morning, #perl6
13:52 masak pmichaud! \o/
13:52 PerlJam pmichaud!  o/
13:52 pmichaud just passing through for a bit
13:53 * masak .oO( you can never pass through the same #perl6 twice )
13:53 PerlJam pmichaud: Then I'll tell you Happy Birthday!  now in case I don't see you tomorrow  :-)
13:53 pmichaud PerlJam: thanks :-)
13:54 FROGGS pmichaud: \o/
13:54 lizmat happy bday, pmichaud!
13:54 masak haooy birthday minus one, pmichaud!
13:54 masak happy*
13:54 pmichaud I finally reached the big 3-0
13:54 * FROGGS is going to wear a funny heat for his b-day
13:55 FROGGS ?
13:55 FROGGS is there something missing in the middle?
13:55 lizmat I remember the big 3-0
13:55 pmichaud no, at the beginning.
13:55 masak happy 0x30, pmichaud :)
13:55 FROGGS ahh
13:55 pmichaud well, I haven't reached it yet........
13:55 FROGGS I was thinking about that but I am too tired to calc
13:56 pmichaud still have 15h04 to go here :)
13:56 masak ♪ Hann á afmæli í dag, hann á afmæli í dag, hann á afmæli hann pmichaud, hann á afmæli í dag ♪
13:56 masak (well, except it's not i dag...)
13:56 [Coke] pmichaud: happy pre-birthday.
13:56 pmichaud how goes yapc::eu, for those who are there?
13:57 masak pmichaud: it's very interesting.
13:57 FROGGS jnthn held his first talk...
13:57 FROGGS masak: how are the ant-talks?
13:57 FROGGS anti
13:57 lizmat and TimToady held a very inspiring keynote  :-)
13:57 FROGGS coll
13:57 FROGGS cool*
13:57 masak I almost feel a bit sorry for the p6-trolls. even if they were coherent (which they weren't) or didn't have talks which weren't as bad as their talk titles...
13:58 FROGGS I wish I was there :/
13:58 masak ...TimToady completely cut them off by saying in his keynote that what they're proposing isn't gonna fly.
13:58 PerlJam Is YAPC::EU being recorded as well?
13:58 lizmat I've seen cameras
13:58 masak basically, "Perl 5 + Moose isn't Perl 7".
13:58 lizmat not sure whether they we're one
13:58 lizmat on
13:59 pmichaud wow (keynote cutoff).  TimToady++
13:59 pmichaud I'm glad to hear TimToady++ come out directly at the conference on the topic
13:59 masak pmichaud: I wrote a prepared speech-let for the Wednesday morning discussion... but I don't feel as compelled to deliver it now.
14:00 masak it's basically more of the same. Perl 6 is now at the point it's doing things Perl 5 doesn't and probably never will. talking about Perl 7 *now* is... silly, and self-defeating.
14:00 pmurias joined #perl6
14:00 masak but basically, all the proponents of that seem very out of touch.
14:00 * [Coke] wonders what is up with the 2 separate "movie" commits.
14:00 notjack what's the general consensus in this forum for Perl5's versioning issue?
14:00 masak "movie" commits?
14:01 masak notjack: oh hai.
14:01 notjack hi masak
14:01 masak notjack: could you be more specific? what issue?
14:01 masak (there are several) :)
14:01 notjack the issue that is motivating people to try to name something "perl 7"
14:01 masak right. what issue?
14:01 [Coke] 480b93cb01 & 89021cf8e0
14:01 notjack in short, the cieling put in place by perl6, which ultimately turned out to be a completely different language than perl5
14:01 [Coke] (roast)
14:02 masak notjack: "completely different language" is a bit excessive, I think.
14:02 notjack I'm not trying to troll, in case it smells like that
14:02 masak notjack: no, I know.
14:02 masak and I'm trying to be precise in replying, after I know what the discussion is.
14:02 notjack from my outsider perspective, it looks and feels completely different
14:02 FROGGS well, Perl 6 is in a position to break backwards compat with Perl 5 code, that is all
14:03 notjack FROGGS: true, but then I could write a Perl5 interpreter in any new language I create
14:03 masak notjack: I think one of the opinions here is that it's not *completely* different. it's still "Perlish".
14:03 masak notjack: that said, Perl 5 and Perl 6 are on different paths, so to speak, and have been for many years.
14:03 FROGGS notjack: what do you want a new perl 5 interpreter for?
14:03 masak notjack: though there is work being done to bring them closer.
14:04 notjack FROGGS: only to (arbitrarily) claim the "perl5 backwards compatible" badge
14:04 notjack masak: what's  motivating that work?
14:04 notjack masak: IMO, I like Perl6's path, and don't see a reason to converge on perl5 (though I can see reasons for p5 to borrow ideas from 6)
14:04 masak notjack: the view that combining existing code bases and Perl 6 would be useful and interesting.
14:04 masak it would be a transition path, if nothing else.
14:05 notjack hmm, I forgot about CPAN
14:05 masak :)
14:05 notjack yeah, I can see that as a huge reason to maintain compat and compatible paths
14:05 masak FROGGS++ is helping us with another bit of that. you can write Perl 5 code inside a block in Perl 6.
14:05 notjack what's the syntax?
14:05 notjack {{{ use v5; blahh blah blah }}} ?
14:06 masak one set of braces is enough :)
14:06 masak but yes. that.
14:06 notjack cool
14:06 FROGGS no XS though, just to be clear
14:06 notjack can you make me a {{{ use APL; iota rho etc }}}?
14:06 notjack kthx
14:06 FROGGS *g*
14:07 mdn joined #perl6
14:07 FROGGS slangs are possible, yes
14:07 masak notjack: :D
14:07 masak notjack: that's a goal, yes.
14:07 masak one schtepp at a time.
14:07 notjack heh
14:07 notjack I've been writing a small script in v5 this week
14:08 masak FROGGS: yes, "no XS" would be why we also want diakopter++'s grant.
14:08 notjack I'd forgotten how quirky (though loveable) it is
14:08 kst`` joined #perl6
14:08 masak Perl 6 is less quirky and more lovable, IMO ;)
14:08 FROGGS masak: I was about to typing that, yes
14:08 notjack yeah, but everyone thinks that about their own weird-looking baby
14:08 notjack :)
14:08 bruges_ joined #perl6
14:08 logie joined #perl6
14:09 masak notjack: granted, but I hear many outsiders saying it, too.
14:09 tsmith joined #perl6
14:09 masak notjack: not all of them, mind. but many.
14:09 bluescreen10 joined #perl6
14:09 FROGGS just explain a perl n00b perl 5's changing sigils >.<
14:10 masak FROGGS: I have, on occasion. delivering our Perl course.
14:10 FROGGS yeah, my co-worker was on a 1-week perl course last week
14:11 timotimo joined #perl6
14:11 pmichaud I wonder if I can find TimToady's keynote somewhere...
14:12 FROGGS so he knows a bit about scalars now :o)
14:13 lizmat pmichaud: we'll let you know as soon as we know it
14:14 lizmat but I don't expect it to be before the end of the conference, though
14:14 notjack I'm not sure how I feel about making everything a metachar in a regex
14:14 pmichaud lizmat: email me a link, please, because I don't know how often I can check in on #perl6
14:14 lizmat will do
14:14 notjack is there an equivalent to qq inside a regex?
14:14 notjack or q I mean
14:14 arnsholt Quotes
14:14 notjack legit?
14:14 arnsholt Simple as that =)
14:15 notjack as in /'hi, I am not a metachar'/?
14:15 arnsholt And making all the non-alnums metacharacters cuts down all the memorisation of what's what
14:15 berekuk joined #perl6
14:16 notjack that's fair
14:16 BenGoldberg notjack: In perl5 regexes, it is of course /\Qhi, I am not a metachar\E/
14:17 notjack BenGoldberg: yeah, but that was rarely neccesary
14:17 tadzik heh, ported IO::Socket::INET to jvm sockets. Server dies with java.lang.NullPointerException, client with java.lang.IndexOutOfBoundsException
14:17 tadzik progress! :)
14:18 tadzik that means they connect to each other, since they die at the same time :)
14:20 FROGGS tadzik: yay! :o)
14:20 denisboyun joined #perl6
14:23 [Coke] tadzik++
14:23 [Coke] tadzik: gentle reminder, you have a failing Star module.
14:24 tadzik [Coke]: oh! MEvent?
14:25 [Coke] tadzik: https://github.com/coke/rakudo-star-daily/blob/master/log/test-modules.log#L216
14:25 [Coke] If I counted right, that's: https://github.com/tadzik/perl6-File-Tools/blob/master/t/02-shell-command.t#L26
14:26 tadzik [Coke]: oh,m I think I fixed that yesterday
14:26 tadzik oh, no I didn't
14:26 [Coke] failed in the latest daily star run. (about 2pm eastern)
14:26 tadzik I should just remove File::Tools, it's now splitted between File::Find and Shell::Command
14:26 tadzik I fixed File::Find, but not File::Tools
14:26 tadzik I'll remove it from the ecosystem
14:27 [Coke] urk. ok. be sure to (have someone) update star to include the new modules instead of the old ones? (and maybe add a changelog to that effect?)
14:27 tadzik oh!
14:27 tadzik I will do that
14:27 masak tadzik++
14:27 dmol joined #perl6
14:29 dalek ecosystem: b8b90c7 | tadzik++ | META.list:
14:29 dalek ecosystem: Remove File::Tools
14:29 dalek ecosystem: review: https://github.com/perl6/ecosystem/commit/b8b90c75c4
14:30 [Coke] tadzik++
14:31 dalek rakudo-js: 2d2fef3 | (Pawel Murias)++ | run_tests:
14:31 dalek rakudo-js: Add test 46 to run_tests (which passes if icu is installed).
14:31 dalek rakudo-js: review: https://github.com/pmurias/rakudo-js/commit/2d2fef3420
14:31 dalek rakudo-js: e1face9 | (Pawel Murias)++ | src/QAST/Compiler/JavaScript.nqp:
14:31 dalek rakudo-js: Implement the eos anchor.
14:31 dalek rakudo-js: review: https://github.com/pmurias/rakudo-js/commit/e1face9fa3
14:31 dalek rakudo-js: 9605581 | (Pawel Murias)++ | src/QAST/Compiler/JavaScript.nqp:
14:31 dalek rakudo-js: Emit an extra newline for better code readablity.
14:31 dalek rakudo-js: review: https://github.com/pmurias/rakudo-js/commit/96055817fd
14:31 dalek rakudo-js: 8056f61 | (Pawel Murias)++ | / (2 files):
14:31 dalek rakudo-js: Refactor run_tests so that it's more visible which tests are missing.
14:31 dalek rakudo-js: review: https://github.com/pmurias/rakudo-js/commit/8056f61a74
14:31 dalek rakudo-js: 8f90bad | (Pawel Murias)++ | / (2 files):
14:31 dalek rakudo-js: Update the setting, subst now works, pass test 62
14:31 dalek rakudo-js: review: https://github.com/pmurias/rakudo-js/commit/8f90bad804
14:32 pmurias 4 tests from nqp/t/nqp are missing
14:32 crab2313 joined #perl6
14:32 tadzik oh shit no
14:32 [Coke] pmurias++
14:32 tadzik raukdostar uses submodules
14:32 tadzik I ned help then :|
14:33 pmurias [Coke]: actually it's 5 will be 4 very soon ;)
14:34 lizmat joined #perl6
14:35 FROGGS pmurias: and then I can run Perl 6 code in my browser?
14:35 [Coke] so are you at cross-compile, nearly passing all tests?
14:38 Psyche^_ joined #perl6
14:41 Kiralexis joined #perl6
14:43 pmurias [Coke]: cross-compile, most of the tests in nqp/t/nqp, there also tests in nqp/t/serialization and nqp/t/qregex
14:44 colomon what's going on at YAPC::EU today?
14:44 nwc10 lightning talks
14:44 nwc10 right now ribasushi is telling us about what's new in DBIx::Class
14:44 nwc10 everyone is hiring in Ukraine, it seems
14:45 nwc10 leont and I have jnthn surrounded and are going to kidnap him and force him to work on Perl 5
14:45 nwc10 :-)
14:45 nwc10 oh, wait, he'll commit things to make it slower. D'oh!
14:46 lizmat .oO(all part of a cunning plan to recruit leont and nwc10)
14:52 telex joined #perl6
14:59 kaleem joined #perl6
15:01 kaleem joined #perl6
15:04 nwc10 lizmat: Perl 6 is hiring? :-)
15:06 moritz sure. Our pay is competitive with most other open source projects!
15:06 FROGGS *g*
15:07 * [Coke] started working on a kenken solver last night, after being stuck for hours on a particularly difficult kenken.
15:07 * [Coke] sees that there is already a python one that has been translated to JS, and wonders if this sort of constraint based thing would be fun/easy/nifty/whatever
15:08 btyler joined #perl6
15:15 colomon joined #perl6
15:16 leont joined #perl6
15:18 _jaldhar joined #perl6
15:21 timotimo [Coke]: it's just so weird to me that all numbers from the jvm test run havae stayed the same when really there were changes to the test suite that added/untodo'd/... tests
15:24 mathw [Coke]: Last time I tried a constraint-based thing it was not fun or nifty or easy at all :(
15:26 TimToady it's one of those things that we're speccing to be easy to implement...someday...
15:26 TimToady andthen and orelse are part of that, which together with lazy lists should allow implementation of a normal backtracker
15:26 TimToady in fact, that's basically how STD works inside
15:27 PerlJam 6/win 15
15:28 [Coke] timotimo: untodoing it doesn't mean it passes on .jvm
15:29 [Coke] plus someone broke several dozen tests - it's possible your passes are hidden behind that.
15:29 ajr joined #perl6
15:29 dalek nqp: 376098c | (Pawel Murias)++ | t/nqp/63-slurp.t:
15:29 dalek nqp: Check if we actually slurp something with a correct length.
15:29 dalek nqp: review: https://github.com/perl6/nqp/commit/376098c62a
15:29 pmurias a good puzzle is how to write a test for slurps that checks if it slurps itself correctly ;)
15:30 timotimo [Coke]: it wouldn't make them pass, it would make them fail, which i didn't see either (but i didn't actually go through with a calculator)
15:31 timotimo it was suspicious to me that the rakudo.jvm-summary file had no change whatsoever
15:32 timotimo but yeah, you're right. is_default blew up completely now
15:37 timotimo and substr for some reason
15:37 dalek rakudo-js: 491ab5e | (Pawel Murias)++ | / (5 files):
15:37 dalek rakudo-js: Implement the ops used by nqp::slurp.
15:37 dalek rakudo-js:
15:37 dalek rakudo-js: Run tests inside the nqp directory so that relative paths inside of them work correctly.
15:37 dalek rakudo-js: Pass test 63.
15:37 dalek rakudo-js: review: https://github.com/pmurias/rakudo-js/commit/491ab5e125
15:37 pmurias [Coke]: 4 test left now ;)
15:37 pmurias * tests
15:38 kbaker joined #perl6
15:38 [Coke] pmurias: woooo
15:38 * [Coke] is handed a box of cider apple croissant donuts
15:44 timotimo wow that souds amazing
15:45 pmurias jnthn: why do we call decont on self?
15:45 pmurias jnthn: in nqp code
15:52 grondilu Is today TimToady's keynote online yet?
15:52 rindolf joined #perl6
15:56 colomon What all has happened today?  p6y, I mean.
16:02 Ayiko joined #perl6
16:06 timotimo it kind of seems like fudging in the substr tests is b0rked
16:08 timotimo fudge and run seems to fudge it for .rakudo and then runs the .t instead
16:10 [Coke] are you fudging for "rakudo" or "rakudo.jvm" ?
16:10 [Coke] "rakudo" is no longer a valid thing to pass in, I think.
16:10 timotimo i'm just running fudgeandrun, should i have to change that? i thought it figured out the implementation o nits own?
16:11 [Coke] that -should- work, yes.
16:11 [Coke] this on jvm?
16:11 timotimo yes
16:11 [Coke] what does your ./perl6 --version say ?
16:11 [Coke] (that's what t/fudgeandrun is using)
16:11 timotimo +S32-str/substr.rakudo.jvm aborted 63 test(s)
16:11 timotimo looking at why this is b0rked
16:12 timotimo if you fudgeandrun it correctly, you will indeed get only one failure
16:12 timotimo This is perl6 version 2013.07-206-g2a26f1a built on JVM
16:12 [Coke] that should fudge for rakudo.jvm, then.
16:13 [Coke] ah, that file has no vm specific fudges. testing
16:14 timotimo it does have rakudo.jvm in the imlp variable
16:15 [Coke] https://gist.github.com/coke/6212404
16:15 timotimo oh wait, yeah, that could still be it
16:15 timotimo locally i get the "too many parameters" problem from very early on
16:16 [Coke] my rakudo-jvm on this box is probably 2 days old.
16:16 ajr joined #perl6
16:17 grondilu cool, someone got rid of the _MVM in MoarVM.
16:32 lizmat joined #perl6
16:39 spider-mario joined #perl6
16:42 BenGoldberg r: say grep(&is-prime, 2,(3,5..*))[1000]
16:42 camelia rakudo 2a26f1: OUTPUT«7927␤»
16:43 BenGoldberg r: .say for grep(&is-prime, 2,(3,5..*))[1000..*]
16:43 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:43 colomon 3,5..* is 3, 5, 6, 7, 8, 9 … etc
16:43 BenGoldberg r: .say for grep(&is-prime, 2,(3,5...*))[100..*]
16:43 colomon you probably want 3, 5… *  -- right
16:44 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:44 BenGoldberg r: .say for grep(&is-prime, 2,(3,5...*))
16:44 colomon why are you asking for an infinite list of primes?
16:44 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:44 BenGoldberg r: .say for grep(True, 2,(3,5...*))
16:45 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:45 colomon or inifinitely many odd numbers?
16:45 BenGoldberg r: .say for grep(True, ^Inf)
16:45 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:46 BenGoldberg r: .say for ^Inf
16:46 camelia rakudo 2a26f1: OUTPUT«(timeout)0␤1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤11␤12␤13␤14␤15␤16␤17␤18␤19␤20␤21␤22␤23␤24␤25␤26␤27␤28␤29␤30␤31␤32␤33␤34␤35␤36␤37␤38␤39␤40␤41␤42␤43␤44␤45␤46␤47␤48␤49␤50␤51␤52␤53␤54␤55␤56␤57␤58␤59␤60␤61␤62␤63␤64␤65␤66␤67␤68␤69␤70␤71␤72␤73␤74␤75␤76␤77␤78␤79␤80␤
16:48 BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[^1000]; say @p[$_] for ^Inf
16:48 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:49 colomon that one may just be slow… at least, I don't see a reason for an infinite loop
16:49 BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[1000..Inf]; say @p[$_] for 0,1
16:49 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:49 BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[100..Inf]; say @p[$_] for 0,1
16:50 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:50 BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[0..Inf]; say @p[$_] for 0,1
16:50 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:50 BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[0..Inf];
16:50 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:50 BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[0...Inf];
16:51 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:51 BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*));
16:51 camelia rakudo 2a26f1: OUTPUT«(timeout)»
16:52 colomon n: my @p := grep(&is-prime, 2,(3,5...*));
16:52 camelia niecza v24-88-g1f87209: OUTPUT«Potential difficulties:â�¤  @p is declared but not used at /tmp/KJ6eO8lJsj line 1:â�¤------> [32mmy [33mâ��[31m@p := grep(&is-prime, 2,(3,5...*));[0mâ�¤â�¤Â»
16:52 colomon n: my @p := grep(&is-prime, 2,(3,5...*)); say "Hi!"
16:52 camelia niecza v24-88-g1f87209: OUTPUT«Potential difficulties:â�¤  @p is declared but not used at /tmp/YWHP_K1pCt line 1:â�¤------> [32mmy [33mâ��[31m@p := grep(&is-prime, 2,(3,5...*)); say [0mâ�¤â�¤Hi!â�¤Â»
16:53 BenGoldberg rn: my @p := grep(True, ^Inf)
16:53 camelia niecza v24-88-g1f87209: OUTPUT«Potential difficulties:â�¤  @p is declared but not used at /tmp/HVV1ayCSp2 line 1:â�¤------> [32mmy [33mâ��[31m@p := grep(True, ^Inf)[0mâ�¤â�¤Â»
16:53 camelia ..rakudo 2a26f1: OUTPUT«(timeout)»
16:53 BenGoldberg rn: my @p; say 'hi'
16:53 camelia niecza v24-88-g1f87209: OUTPUT«Potential difficulties:â�¤  @p is declared but not used at /tmp/yXj28nFt2b line 1:â�¤------> [32mmy [33mâ��[31m@p; say 'hi'[0mâ�¤â�¤hiâ�¤Â»
16:53 camelia ..rakudo 2a26f1: OUTPUT«hi␤»
16:54 BenGoldberg rn: my @p; @p := grep(True, [1])
16:54 camelia rakudo 2a26f1, niecza v24-88-g1f87209:  ( no output )
16:54 BenGoldberg rn: my @p; @p := grep(True, [^Inf])
16:54 camelia niecza v24-88-g1f87209: OUTPUT«(timeout)»
16:54 camelia ..rakudo 2a26f1:  ( no output )
16:55 ssutch joined #perl6
16:58 moritz n: [^Inf]; say "alive"
16:58 camelia niecza v24-88-g1f87209: OUTPUT«(timeout)»
17:03 btyler joined #perl6
17:13 crab2313 joined #perl6
17:17 benabik joined #perl6
17:19 slava joined #perl6
17:20 vovka667 joined #perl6
17:23 fhelmberger joined #perl6
17:25 MDn joined #perl6
17:26 dalek Perlito: d435a99 | (Flavio S. Glock)++ | / (3 files):
17:26 dalek Perlito: Perlito5 - grammar - more 'variable' rules
17:26 dalek Perlito: review: https://github.com/fglock/Perlito/commit/d435a9910b
17:31 Mouq joined #perl6
17:34 Vico_Wu1 joined #perl6
17:35 nugged joined #perl6
17:37 nugged_ joined #perl6
17:40 ssutch joined #perl6
17:40 vovka667 joined #perl6
17:41 crab2313 joined #perl6
17:44 Rotwang joined #perl6
17:44 ajr joined #perl6
17:47 notjack has anyone done any work on a p6 plugin for Eclipse?
17:47 notjack in particular, syntax higlighting
17:48 notjack padre seems like a dead project
17:48 [Coke] no, but I'd be vaguely interested in that project. (I want to write a p6 plugin for eclipse for syntax highlighting for other languages (written in p6)
17:48 Vico_Wu joined #perl6
17:50 notjack take a p6 grammar and generate an Eclipse syntax-highlighting file, you mean?
17:50 FROGGS joined #perl6
17:51 notjack I guess you could do that for any editor or IDE (generate a correctly-formatted syntax higlighting file for that editor, given a p6 grammar)
17:54 dmol joined #perl6
17:55 moritz well, if the syntax hilighting file for that IDE is Turing complete, yes
17:59 berekuk joined #perl6
17:59 notjack speaking of Turing completeness
17:59 notjack in order to learn p6 better, I'd like to implement a little J parser in it
18:00 notjack probably is, I know less about parsing than I do about p6, as a practical matter
18:00 notjack problem is*
18:00 moritz notjack: have you read the "Using Perl 6" ebook?
18:00 moritz it has a chapter about p6 regexes, and one about grammars
18:00 notjack I've grepped it, but not read it
18:01 notjack p6 regexes I think I can get hold of pretty quick
18:01 moritz grepping isn't enough for learning -)
18:01 notjack grammars are harder ... because I don't understand grammars as a general topic
18:01 denis_boyun joined #perl6
18:01 notjack unless the book offers enough of an education on the topic as well?
18:01 timotimo for your use case, grammars will be easier, though :P
18:02 mdn joined #perl6
18:02 notjack Let me ask: the J language has 9 production rules in its grammar, and a 4-token fixed lookahead (http://www.jsoftware.com/help/dictionary/dicte.htm)
18:03 notjack in lines of p6 code ... my gut says it should take about a page of code
18:03 notjack anyone think differently?
18:03 moritz notjack: the book has a complete JSON parser as an example, and explains it
18:03 notjack ooh
18:03 notjack "and explains it"++
18:03 PerlJam moritz++
18:03 moritz (sorry, I have to finish my slides for tomorrow before looking at the J grammar)
18:04 PerlJam notjack: a "page" for just the parser?  sure.
18:04 PerlJam notjack: add another for the semantics.
18:04 notjack forget the semantics
18:04 notjack that would take a lot of work
18:04 notjack I mean just building an AST
18:04 moritz that's explained for the JSON stuff too
18:05 notjack J has ~120ish primitives and things like implicit composition, etc.
18:05 notjack I can write that (SMOP), but I need a place to start, i.e. an AST
18:10 _jaldhar joined #perl6
18:13 moritz masak: http://elixir-lang.org/blog/2013/08/08/elixir-design-goals/ has some macro examples, including some we considered for Perl 6
18:14 moritz (peeking into the components of a boolean expression to explain why a test/assert failed)
18:21 japhb_ joined #perl6
18:23 moritz ok, talks if finally "prepared"
18:24 btyler joined #perl6
18:29 moritz notjack: I can't actually deduce the J grammar rules from the page you gave us
18:30 timotimo jnthn: do you have an idea how i can get the Code object that belongs to a QAST::Block? Or another way to figure out what the signature of a QAST::Block is? or another way to figure out which variables are to be set by the signature?
18:31 moritz timotimo: I talkded to jnthn about that earlier today, and off-hand he didn't know any way
18:31 timotimo ah damn
18:31 moritz timotimo: but he made the interesting observation that nqp could greatly benefit from that optimization
18:31 timotimo yes, i thought so, too
18:32 timotimo but there's no optimizer for nqp and i'm not sure how to hook it up properly
18:32 moritz timotimo: and since nqp doesn't have a custom binder, the parameters doen't need to be lexical
18:32 moritz yes, might not be trivial to get started
18:32 timotimo also, the obvious way to prevent the signature problem is to bail out as soon as the "bind a signature" op is encountered, which is probably ... everywhere :(
18:33 jaffa4 joined #perl6
18:39 fhelmberger joined #perl6
18:39 slava [op] 1,2,3,4 ... this is awesome, it looks like a folding from haskell
18:41 timotimo and indeed it is :)
18:43 dmol joined #perl6
18:43 jaffa4 anybody tied jvm?
18:44 * moritz tied the JVM into a knot
18:44 jaffa4 tried
18:45 moritz last time I tried to build rakudo on JVM, the build failed
18:45 moritz complained about a missing op
18:46 moritz but before it worked great
18:46 timotimo i compiled a rakudo.jvm today from newest nqp and rakudo master/nom branches
18:48 jaffa4 99% complete I heard
18:49 FROGGS jaffa4: true
18:50 TimToady more accurately, 99% as complete as rakudo/parrot is
18:50 flussence that missing 1% still includes things like "running rakudo from anywhere but the current directory"
18:51 TimToady if you set the CLASSPATH it seems to help with that
18:51 skids joined #perl6
18:51 TimToady was doing commands like: CLASSPATH=~/nom perl6 rad.p6
18:52 geekosaur although using ~ there is kinda dubious
18:52 TimToady where perl6 is a symlink /usr/local/bin/perl6 -> /home/larry/nom/install-jvm/bin/perl6
18:53 TimToady obviously, it'd be nice not to have to set the CLASSPATH explicitly like that
18:53 bloonix joined #perl6
18:53 sidus joined #perl6
18:58 jaffa4 it is not clear how to compile for jvm
19:00 TimToady did you read the README?
19:00 TimToady has some directions there that seem relatively adequate
19:01 TimToady as in, worked for me :)
19:02 * TimToady is scheduled to sleep like a log for the 120° or so of earthly rotation
19:02 jaffa4 120?
19:03 jaffa4 is that 8 hours?
19:03 timotimo r: (120 / 360) * 24
19:03 camelia rakudo 2a26f1:  ( no output )
19:03 timotimo r: say (120 / 360) * 24
19:03 camelia rakudo 2a26f1: OUTPUT«8␤»
19:03 jaffa4 TimToady: thank you
19:04 notjack common misconception. the logs, they're not sleeping. they're merely biding their time.
19:04 TimToady has $.timotimo handles 'thank-you';
19:04 timotimo :D
19:04 TimToady zzz &
19:06 PacoAir joined #perl6
19:06 Targen joined #perl6
19:07 jaffa4 Are you sleeping and wawke at the same time?
19:07 jaffa4 awake<<
19:08 zwut00 joined #perl6
19:15 lue I don't suppose there's a way to search for anything *but* a sequence of characters in a regex, like /-'::'/, is there?
19:15 btyler joined #perl6
19:16 timotimo there's ugly hacks like /..<!after '::'>/
19:16 benabik I wonder if /<!before '::'>../ would be faster.
19:16 timotimo may be, yes
19:16 flussence something like /<?{ $0 eq '::' }>/?
19:16 dalek rakudo/nom: a379871 | moritz++ | / (4 files):
19:16 dalek rakudo/nom: remove Perl6::ConstantFolder
19:16 dalek rakudo/nom:
19:16 dalek rakudo/nom: it was not really needed, and the optimizer does
19:16 dalek rakudo/nom: constant folding much more generally
19:17 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/a379871ee1
19:17 colomon something with split, maybe?  depends on what you are trying to do.
19:18 moritz negating a regex usually only makes sense if you anchor it somewhere
19:20 timotimo that's a good point
19:20 timotimo flussence: i think using a lookahead/lookbehind instead of a code assertion is better because it doesn't have to turn the matched part into a string and it can reason about matching
19:22 flussence yeah, you're right
19:25 zby_home joined #perl6
19:27 mdn joined #perl6
19:28 * moritz blug: http://perlgeek.de/blog-en/perl-6/2013-yapc-day-1.html
19:30 timotimo almost as good as being there in person!
19:31 colomon \o/
19:31 moritz there were some hard choices to make
19:32 timotimo "but in the didn't."
19:32 moritz for example I would have loved to attend http://ye2013.yapc-surveys.org/talk/32 but it clashed with another one I wanted to see
19:32 timotimo "invalid code"
19:34 btyler joined #perl6
19:35 moritz timotimo: fixed, thanks
19:37 lue here's what I'm trying to do --> https://gist.github.com/lue/6214323  it currently prints the key (debug statement) and then fails the match.
19:39 Rotwang joined #perl6
19:42 btyler_ joined #perl6
19:44 [Coke] moritz: I've been building the jvm daily, no problems.
19:44 dalek rakudo-star-daily: 25328ae | coke++ | log/ (5 files):
19:44 dalek rakudo-star-daily: today (automated commit)
19:44 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/25328aefa0
19:44 moritz ok, then it's probably my fault somehow :-)
19:47 [Coke] timotimo: can you open a rakudo star ticket about dropping tools and adding the 2 other modules?
19:47 [Coke] that way it doesn't get lost.
19:47 timotimo i guess i can, even though it's tadzik's changes?
19:47 [Coke] (er, timotimo? tadzik? my t-space is pretty full...)
19:47 timotimo i forgot which modules those are, though
19:47 timotimo yeah, let tadzik do it :)
19:48 [Coke] timotimo++, I meant tadzik++, apologies.
19:48 * [Coke] wonders what secrete project jnthn++ is revealing this week
19:49 [Coke] *secret
19:49 jaffa4 what makes you think that?\
19:49 [Coke] ... it's a perl conference where jnthn++ is presenting.
19:49 colomon jaffa4: it's kind of a tradition
19:50 jaffa4 again something?
19:50 moritz [Coke]: not really secret, but jnthn has been doing some impressive hacking on concurrency stuff on JVM
19:50 [Coke] it's how we got moarvm, rakudo::debugger...
19:50 colomon grammar::debugger...
19:51 colomon the grammar generator stuff..
19:51 jaffa4 is this communism?
19:52 dmol joined #perl6
19:52 [Coke] I don't think it's an ism.
19:53 celesta joined #perl6
19:53 celesta ahoi #perl6
20:00 [Coke] aloha
20:00 Vico_Wu1 joined #perl6
20:07 BigBear joined #perl6
20:10 celesta r: sub foo( @x, @y ){ for @x >>~~<< @y { return False if !$_; }; return True; }; say foo( 1..5, 5..9);
20:10 camelia rakudo 2a26f1: OUTPUT«False␤»
20:10 celesta how do i "catch" the error if lists are not of same length?
20:11 celesta r: sub validate( @x, @y ){ for @x >>~~<< @y { return False if !$_; }; return True; }; say validate( 1..6, 5..9);
20:11 camelia rakudo 2a26f1: OUTPUT«Lists on both side of non-dwimmy hyperop of infix:<~~> are not of the same length␤left: 6 elements, right: 5 elements␤  in sub hyper at src/gen/CORE.setting:15492␤  in sub hyper at src/gen/CORE.setting:15480␤  in block  at src/gen/CORE.setting:15471␤  in sub valida…
20:12 FROGGS well, you can use a try { CATCH {} } block or you just check the array lengths yourself
20:14 celesta FROGGS: thank you
20:21 Mouq joined #perl6
20:23 jaffa4 it lool=ks fot nmake
20:24 jaffa4 rakudo: say "hello
20:24 camelia rakudo a37987: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/sbUmHsHVVtâ�¤Unable to parse expression in double quotes; couldn't find final '"'â�¤at /tmp/sbUmHsHVVt:1â�¤------> [32msay "hello[33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        argument listâ�¤        prefix or …
20:24 jaffa4 rakudo: say "hello"
20:24 camelia rakudo a37987: OUTPUT«hello␤»
20:30 Guest2529 I know this code is wrong, and why; is the error message right?
20:30 Guest2529 perl6 -e 'say "say "Hi!"; say "Perl 6 here";'
20:30 Guest2529 ===SORRY!===
20:30 Guest2529 Two terms in a row
20:30 Guest2529 at -e:1
20:30 Guest2529 ------> say "say "⏏Hi!"; say "Perl 6 here";
20:30 Guest2529 expecting any of:
20:30 Guest2529 postfix
20:30 Guest2529 infix stopper
20:30 Guest2529 infix or meta-infix
20:30 Guest2529 statement end
20:30 Guest2529 statement modifier
20:30 Guest2529 statement modifier loop
20:31 benabik ajr_: Too many double quotes.
20:32 ajr_ As I said, I understand the error, I just wonder about the message.
20:32 colomon joined #perl6
20:33 benabik ajr_: It's parsing Hi! as a term, immediately after the string term.  I think most unknown words are assumed to be terms.
20:36 [Coke] r: say what the
20:36 camelia rakudo a37987: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/BsigbqGP4pâ�¤Undeclared routines:â�¤    the used at line 1â�¤    what used at line 1â�¤â�¤Â»
20:36 [Coke] r: say "what" "the"
20:36 camelia rakudo a37987: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/6j4io3qQjSâ�¤Two terms in a rowâ�¤at /tmp/6j4io3qQjS:1â�¤------> [32msay "what" [33mâ��[31m"the"[0mâ�¤    expecting any of:â�¤        postfixâ�¤        infix stopperâ�¤        infix or meta-infixâ�¤        statement endâ�¤ …
20:36 [Coke] r: say "what" the
20:36 camelia rakudo a37987: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/C6BIL3JEiyâ�¤Two terms in a rowâ�¤at /tmp/C6BIL3JEiy:1â�¤------> [32msay "what" [33mâ��[31mthe[0mâ�¤    expecting any of:â�¤        postfixâ�¤        infix stopperâ�¤        infix or meta-infixâ�¤        statement endâ�¤   …
20:37 [Coke] ^^
20:46 notjack r: say 'thefuck'.WHAT;
20:46 camelia rakudo a37987: OUTPUT«(Str)␤»
20:46 notjack r: say <tee eff>.WHAT;
20:46 camelia rakudo a37987: OUTPUT«(Parcel)␤»
20:46 notjack r: say <tee eff>.WHAT.WHAT;
20:46 camelia rakudo a37987: OUTPUT«(Parcel)␤»
20:52 [Coke] r: say.WHAT.say, now?
20:52 camelia rakudo a37987: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/UaVu6ZOYBTâ�¤Two terms in a rowâ�¤at /tmp/UaVu6ZOYBT:1â�¤------> [32msay.WHAT.say, now[33mâ��[31m?[0mâ�¤    expecting any of:â�¤        postfixâ�¤        infix stopperâ�¤        infix or meta-infixâ�¤        statement end…
20:52 [Coke] aw.
20:55 masak o/
20:55 [Coke] o/~
20:58 masak moritz++ # blug post
21:00 lue hello masak o/
21:07 masak hellue, lo. \o
21:08 celesta heja masak
21:09 dalek perl6-roast-data: 8c0f742 | coke++ | / (5 files):
21:09 dalek perl6-roast-data: today (automated commit)
21:09 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/8c0f742380
21:09 [Coke] rakudo.parrot failing more tests today. rakudo.jvm dropped another .01%
21:10 [Coke] looks like it's "is dynamic" related fallout. also rakudo.parrot has passing todos.
21:14 masak aloha celesta.
21:14 lue masak: Any clues as to why https://gist.github.com/lue/6214323 fails on the first :: ?
21:17 [Coke] lue: have you tried the regex debugger?
21:18 lue panda's refusing to work :/ (I may have to update rakudo, it's been a little while)
21:20 masak lue: I think it's failing in the <?before
21:20 rjbs I wrote some p6 code that might be fun to turn into a shared distribution.  What do I want to read to do that?
21:20 masak lue: that's just an hypothesis, but it seems reasonable to me.
21:20 masak rjbs: http://wiki.perl6.org/Create%20and%20Distribute%20Modules
21:20 rjbs Thanks!
21:21 masak uvtc++ siddhant++ # creating that article
21:21 * [Coke] wonders how many times rjbs will have to do that before we get a distzilla.
21:23 lue Yep, it gets the colon (added debug msg), so it fails on the before.
21:24 masak lue: that's because $<closer> does not contain the whitespace after the '::'
21:25 augustus joined #perl6
21:26 lue I'm going to go update rakudo, and then see if panda will work for me.
21:34 augustus so is anyone around here today, or is everybody off at YAPC?
21:34 masak augustus: we're all off at YAPC ;)
21:35 lue not all of us :(
21:35 augustus ok.  That's all week long, or like 3 days?  Or really, when might I expect to find people hanging out in the chat room again?
21:37 masak augustus: oh, people are still hanging out here.
21:37 masak augustus: and you're likely to get replies pretty quickly.
21:37 jercos I'm trapped in a sphere made of my own self-loathing, so I'm usually watching, rather than talking.
21:37 masak through advanced 20th-century technologies like "IRC backlog" and "screen".
21:37 augustus hehe nice :)
21:38 masak jercos: that's not altogether uncommon, IIUC.
21:38 jercos U?
21:38 masak jercos: fsvo "self-loathing". sometimes it's simply shyness or lack of sense of being able to contribute (rightly or wrongly)
21:38 dduncan joined #perl6
21:38 masak I know I was a lurker for many years.
21:39 masak dduncan! \o/
21:39 dduncan hello
21:39 augustus so I got the latest code from git, and I want to play around with the latest java interop, since the latest Perl6 changes mentioned some improvements.  anyone here familiar with the latest and greatest in this area?
21:39 masak dduncan: long time no Muldis! :D
21:39 * lue should do something more significant for Perl 6. Oh wait, there's the secret project.
21:40 dduncan masak, well I've been busy with work that's paying me
21:40 masak dduncan: haven't seen you around for a while.
21:40 dduncan If I ever get a couple months to dedicate to it, I could have an executable, I"m that ready otherwise.
21:40 masak dduncan: nor on the p6l list.
21:41 dduncan To be honest I rarely sign on to IRC because its so distractinv.
21:41 jercos augustus: personally I'm at the point of "Woah, use :java actually works"
21:42 augustus jercos: yes it is impressive, and much easier than before :)
21:42 masak "before"?
21:42 dduncan My sign-on here was actually an auto-join bookmark ... I had opened freenode to use another channel actually.
21:42 Vico_Wu joined #perl6
21:43 dduncan People don't usually greet me so quicky when that happens.
21:43 dduncan that is all
21:43 masak dduncan: we've missed you too! :P
21:43 dduncan My day job is all Perl / databases though, thankfully.
21:44 PerlJam dduncan: you are a wise man to avoid IRC at much as possible  :)
21:44 augustus masak: previously you had to load classes manually with typeForName etc.  use is much more direct and meaningful
21:44 masak augustus: was that Perl 5 or Rakudo?
21:45 augustus masak: to the best of my knowledge Perl 5 doesn't run on JVM?  Anyway, Rakudo.
21:46 jnthn Yes, you could do it that way before.
21:46 PerlJam augustus: how long ago was "previously"?
21:46 jnthn If you knew how :)
21:46 jnthn It's the way Threading.pm is written.
21:46 masak oh!
21:48 Vico_Wu1 joined #perl6
21:51 labster So how is YAPC going, folks?
21:52 labster Is there going to be a way to see these talks online?
21:53 augustus PerlJam: Perviously was like 10-15 days ago? I forget exactly
21:54 augustus jnthn: I was wondering if the latest and greatest includes an easier way to call a java constructor? Or is that still the Threading.pm way?
21:54 colomon joined #perl6
21:54 aindilis` joined #perl6
21:54 jnthn augustus: Still the Treading.pm way if there are overloads, otherwise it's just .new
21:55 jnthn augustus: Mentioned in my talk today that I'm working on getting a kinda multidispatch - overload mapping in place, but it's a little tricky.
21:55 augustus jnthn: ahh, ok that's why it didn't work, because of the overloads.  Also, is it possible to access static properties?
21:55 augustus jnthn: very exciting!
21:55 augustus jnthn: too bad I couldn't be there
21:56 * timotimo is waiting eagerly for the recording
21:56 timotimo i won't get anything else done :(
21:56 jnthn augustus: yeah, I think the SWT example I showed today did it, it's something like field/get_name/typedesc
21:57 colomon jnthn! \o/
21:57 augustus jnthn: does it follow the java byte code format?  I tried that a couple of weeks ago and couldn't get that to work
21:57 timotimo jnthn: since you said there's no obvious way to get the signature from a QAST::Block, should perhaps the actions attach a signature to the QAST::Block in some way?
22:00 jnthn timotimo: No obvious way? Hm
22:00 jnthn There's an unobvious way if you can get the code object though ;)
22:03 timotimo that's the problem, i see no way to get the Code object :(
22:03 timotimo the problem i'm really trying to solve is find out which vars are signature variables and not change those into locals
22:03 colomon joined #perl6
22:04 augustus jnthn: I'm trying   my \Out := System.'field/get_out/Ljava/io/PrintStream'(); to no avail.  Any thoughts on what it should be?
22:04 zacts joined #perl6
22:05 Targen joined #perl6
22:05 jaldhar_ joined #perl6
22:06 jeff_s1 joined #perl6
22:06 perigrin joined #perl6
22:06 segomos joined #perl6
22:06 BigBear_ joined #perl6
22:10 jnthn r: loop say 42;
22:10 camelia rakudo a37987: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/FVoNwiggviâ�¤Missing blockâ�¤at /tmp/FVoNwiggvi:1â�¤------> [32mloop [33mâ��[31msay 42;[0mâ�¤    expecting any of:â�¤        statement listâ�¤        scoped blockâ�¤Â»
22:13 masak r: say 42 while True
22:13 timotimo jnthn: is "turn lexicals into locals" the only really worthwhile optimisation for nqp that you can think of?
22:13 camelia rakudo a37987: OUTPUT«(timeout)42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤42␤
22:14 timotimo how big would the difference be between the nqp optimizer and the rakudo optimizer? both operate on qast, both have a world object afaict, both have Blocks, Routines, lexicals?
22:15 silug joined #perl6
22:22 jnthn timotimo: NQP is a simpler language and can assume more. For one, parameters are simpler and can be reduced to locals much more easily :)
22:22 jnthn timotimo: The very worthwhile thing is spotting when we have native int variables and turning add_n on them into add_i to avoid coercions.
22:23 timotimo so add_n is for both integers and nums?
22:23 jnthn no, the code-gen coerces!
22:24 jnthn augustus: SWT.'field/get_NONE/I'() is my example from my SWT code
22:24 timotimo i don't think i understand how that works
22:24 jnthn timotimo: add_n needs two nums, so if you feed it to ints then it will do a int -> num on each arg, then the add_n
22:24 timotimo so we currently have native ints that are fed to add_n and we get a int->num inside and an num->int outside?
22:25 jnthn Something like
22:25 jnthn yes
22:25 jnthn Which is sucky :)
22:25 jnthn And a good target for the optimizer.
22:25 jnthn Rakudo lacks this problem
22:26 jnthn As its numeric type system is much stronger
22:26 timotimo OK
22:27 timotimo any big differences in the way the optimizer would be set up (simply a class with an optimize method that turns QAST into QAST?) and hooked to the compiler?
22:27 timotimo maybe tomorrow i'll hack up something very simple
22:29 masak r: my %h = 'look' => { say "I see things!" }; sub run_command($cmd) { (%h{$cmd} // {;})() }; run_command 'look'
22:29 camelia rakudo a37987: OUTPUT«I see things!␤»
22:31 segomos joined #perl6
22:32 jnthn timotimo: Should be about the same
22:33 timotimo neato.
22:33 lizmat joined #perl6
22:34 jnthn ok, enough slide hacking...think I'm gonna sleep :)
22:34 jnthn 'night, #perl6
22:34 timotimo gnite :)
22:35 celesta good night everybody
22:35 Vico_Wu2 joined #perl6
22:35 lizmat joined #perl6
22:38 skids joined #perl6
22:45 dduncan left #perl6
22:52 ssutch joined #perl6
22:53 augustus Well, I think everybody has signed off, but I think these chat logs are saved for posterity, so I'll answer my own question.  I was missing the semicolon.  my $out := System.'field/get_out/Ljava/io/PrintStream'(); needs to be my $out := System.'field/get_out/Ljava/io/PrintStream;'(); with a semicolon terminating the classname inside the quotes.
22:56 masak no, we're still here.
22:57 masak most people are either sleeping already, or still out being social.
22:57 fridim_ joined #perl6
22:58 FROGGS joined #perl6
23:00 masak 'night, #perl6
23:00 timotimo good night masak :)
23:19 ssutch joined #perl6
23:22 nebuchadnezzar joined #perl6
23:25 lue can anyone here get Rakudo::Debugger to work?
23:29 timotimo i can't get panda to work actually
23:29 lue wfm; have you updated your PATH?
23:30 timotimo i can't get shell::command to work from the ext/ dir
23:30 timotimo gimme a minute
23:32 timotimo ufo is so slow :(
23:52 cognominal joined #perl6
23:56 * lue wishes there was something that explained <before> and <after> so he would do better than bumble around with placement in the regex and the right prefix.

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

Perl 6 | Reference Documentation | Rakudo