»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 sorear on 25 June 2013.
00:02 woolfy left 00:03 raiph left 00:04 ggoebel joined, lizmat left 00:05 lizmat joined 00:06 crab2313 joined 00:16 stevan_ joined 00:18 woolfy joined 00:19 FROGGS joined 00:28 li joined 00:30 li left 00:31 FROGGS left 00:32 raiph joined 00:39 benabik left 00:40 FROGGS joined 00:45 crab2313 left, FROGGS left 00:48 FROGGS joined 00:49 mikemol joined 00:52 stevan_ left 01:01 aindilis left
[Coke] f 01:04
01:06 FROGGS left
dalek rl6-roast-data: eb1ee6f | coke++ | / (4 files):
today (automated commit)
 01:07
rl6-roast-data: 8c6e5e7 | coke++ | / (4 files):
today (automated commit)
[Coke] 98.9 and falling. 01:09
01:19 FROGGS joined, dayangkun joined 01:21 dayangkun left 01:32 Mouq joined 01:34 FROGGS left 01:41 xinming left
diakopter [Coke]: are you still able to make use of host06? 01:42
01:43 xinming joined, benabik joined
[Coke] Using it every day for the rakudo.jvm run, yes 01:53
still haven't gotten pugs working.
so haven't done the cutover to run all the tests there.
01:56 colomon joined
diakopter cool; how long does the rakudo.jvm run take btw 01:57
01:58 raiph left 01:59 grondilu joined 02:08 benabik left 02:10 FROGGS joined 02:13 benabik joined
[Coke] 80min 02:18
give or take
and that's -with- the server. :( 02:19
er, the eval-server
02:20 tsmith left
diakopter wat. 02:26
02:33 flussence joined
sorear sounds about right 02:36
diakopter: [Coke] is running tests on a single core, I think
02:37 prevost joined 02:43 prevost left
[Coke] yes. 02:44
diakopter sorear: oh 02:47
[Coke]: oh
02:59 FROGGS left
Mouq Are trees in perl6 just (usually) Pairs? 03:00
Are pears grow trees, and trees grow pears, ∴ trees are pairs 03:01
colomon …. you could do it that way, I guess
Mouq s/are// 03:02
Yeah, I didn't really know if there was a better way to do it
colomon I know trees are mentioned several times in the specs, but I don't recall ever actually running into them in practice. 03:04
perlcabal.org/syn/S02.html#list%2C_..._and_.tree 03:05
03:05 FROGGS joined
colomon I've got to say I don't understand the tree bits of that section at all. 03:06
Mouq I'll look into Takadonet's Tree::Simple [github.com/Takadonet/Tree--Simple/] 03:08
I think it .tree just turns Parcels of Parcels into LoLs? 03:09
Yeah
r: my $a = (4,5,(6,7)); say $a.tree.perl 03:10
camelia rakudo 2566e7: OUTPUT«LoL.new(4, 5, $(6, 7))␤»
sorear those trees are constructed from Lists, not pairs.
03:11 wtw joined
Mouq Yeah 03:12
Honestly Pairs are dumb way to do it :L
In general
03:14 wtw_ left
Mouq Maybe. Yeah, like I said, I still have to look at it. It's weird since Pairs can have anything as a 'key' 03:15
colomon you can certainly build a binary tree structure with pairs. 03:22
dunno that anything in p6 would automatically know how to make sense of it, though. 03:23
whereas making it a LoL is at least something p6 has a notion of.
03:34 FROGGS left 03:39 fridim_ left 03:40 crab2313 joined 03:52 FROGGS joined
[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? 04:01
sorear probably 04:04
I mean that's what I'd do
not to be confused with 'there probably is' 04:05
[Coke] Roger. 04:09
04:23 xenoterracide left
[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 left, FROGGS joined 04:31 crab2313 left 04:43 birdwindupbird joined 04:55 FROGGS left 04:58 logie joined
timotimo o/ 05:07
05:09 FROGGS joined 05:10 logie left
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
at least the amount of specd'd tests etc
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 05:15
05:23 Mouq left 05:26 aindilis joined, crab2313 joined 05:27 mdn joined 05:30 denisboyun joined 05:32 FROGGS left
lizmat timotimo: they're the counts of number of tests in the block 05:36
what is #?DOES ? 05:37
timotimo #?does is sometimes needed to tell the fudger how many tests to skip or mark todo 05:39
05:41 mdn left, mdn joined, locsmif joined 05:43 xenoterracide joined 05:44 xinming left, vovka joined 05:45 FROGGS joined 05:46 xinming joined 06:00 mdn left, denisboyun left
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 06:01
06:03 vovka left 06:05 FROGGS left 06:08 dmol joined 06:09 berekuk joined
moritz there's another use for #?DOES 06:15
timotimo good morning moritz :)
moritz you can put it before a subroutine declaration
and then it tells the fudger how many tests that routine runs
good morning timotimo
and good morning #perl6
going to the venue now& 06:16
06:17 logie joined 06:22 dayangkun joined 06:23 logie left 06:31 crab2313 left
moritz at the venue, and online again :-) 06:32
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 dmol left, berekuk left
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 06:38
06:39 kaleem joined 06:41 FROGGS[mobile] left 06:42 woolfy left 06:44 dayangkun left 06:45 lizmat left
moritz 06:45
06:46 FROGGS joined
FROGGS o/ 06:47
tadzik hai
moritz it's tadzik! 06:48
tadzik it's moritz! and FROGGS!
what a day
FROGGS :o)
tadzik is warming up a place in Main Room
moritz timotimo: I wonder if we should just mark lexicals coming from signatures in some way in the QAST tree 06:49
timotimo it would be handy for my optimisation at least :) 06:50
moritz decl<sigvar> or so
or maybe a totally new flag
I'll try to talk to jnthn later
timotimo the nodes already have a hash interface for all kinds of annotations
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 BenGoldberg left
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? 06:53
i need to associate name-of-variable with bits anyway, so i probably won't get around a hash anyway 06:54
moritz never mind for now 06:55
06:56 Vico_Wu joined 07:00 mdn joined 07:03 ivanoff joined 07:05 mdn left 07:06 berekuk joined 07:08 kaleem left, kaleem joined
dalek ast: 480b93c | ivanoff++ | S02-lexical-conventions/sub-block-parsing.t:
movie sub/hash syntax
 07:10
ast: 89021cf | ivanoff++ | S13-syntax/sub-hash.t:
movie sub/hash syntax
07:11 ivanoff left 07:25 sqirrel joined 07:31 SamuraiJack joined 07:33 leont joined 07:41 leont left 07:42 leont joined 07:44 leont left 07:47 berekuk left 07:48 berekuk joined
FROGGS .tell lizmat the Jugendherberge in frankfurt has no vacancies left ó.ò 07:50
yoleaux FROGGS: I'll pass your message to lizmat.
timotimo oh no 08:05
way too many perlers coming to the workshop :(
FROGGS ? 08:06
ahh, to the one in frankfurt you mean
timotimo yes
08:07 leont joined, fhelmberger joined
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 08:08
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…
moritz I wonder why/how I messed it up locally 08:10
timotimo messed up how? 08:11
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…
timotimo given that cancer is the uncontrolled multiplication of cells in ones body, "TimToady++" seems scarily appropriate ...
moritz timotimo: doesn't die anymore
timotimo: I'm trying to make it a typed exception 08:13
timotimo oh? i had thought i've already seen the typed exception get committed to rakudo's repository 08:14
r: try { sub f { say CALLER::<$x> }; my $x; f() }; say $!.WHAT
camelia rakudo 2566e7: OUTPUT«(X::Caller::NotDynamic)␤»
moritz oh, I pushed that?
timotimo yes :)
and it seems to work, too
moritz \o/ 08:15
timotimo now i'm wondering why i can't run rakudo-jvm for benchmarks any more ...
08:15 berekuk left
timotimo ah. it's because it exploded while compiling 08:15
"bad type on operand stack", probably my optimisation exploding earlier for the jvm because of the bytecode verifier
huh. sometimes faster, sometimes slower. 08:18
t.h8.lv/p6bench/locals-optimisation.html
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. 08:19
08:20 leont left 08:27 danaj joined 08:31 berekuk joined 08:35 SmokeMachine joined, crab2313 joined 08:45 SamuraiJack_ joined 08:49 SamuraiJack left 08:54 BigBear joined 08:57 PuercoPop joined 08:58 berekuk left 09:00 ssutch left 09:08 leont joined 09:12 berekuk joined
FROGGS $ perl6 -e 'use QAST:from<NQP>; say QAST::WVal.HOW.name(QAST::WVal)' 09:17
QAST::WVal
\o/
09:17 grondilu left
FROGGS jnthn: ping 09:17
09:18 grondilu joined
tadzik woot 09:19
09:21 cognominal left, cognominal joined 09:22 berekuk left, leont left 09:23 leont joined
jnthn yays, I got on the wifi... 09:23
FROGGS: pong
09:23 lizmat joined, dmol joined
FROGGS jnthn: I can now load modules :from<NQP>, but this required updating stage0 for vm/parrot... 09:24
jnthn FROGGS: Oh?
FROGGS: OK, that is sometimes needed
Can I see the patch?
FROGGS jnthn: I'm going to push this to a branch, but vm/jvm needs an updated stage0 too I guess
jnthn: I just push to a branch, okay?
09:25 denis_boyun joined
jnthn ok 09:26
will review
09:26 nugged joined
masak o/ 09:26
jnthn o/ masak
which talk are you in? :)
FROGGS awesome!
hi masak
09:26 cosimo joined
moritz jnthn: masak and me are in the redis takl 09:26
dalek p/from_nqp: ab98aad | (Tobias Leich)++ | src/ (12 files):
give load_module the same signature as rakudo's

This makes it pluggable, so we will be able to use modules :from<NQP>
jnthn moritz: same! 09:27
masak: I dare you to ask at question time at the end, "Is it web scale?" :D
masak jnthn: :P 09:28
jnthn: you do it :P
nwc10 masak: you are on the same row. Look left 09:29
dalek kudo/from_nqp: f63899e | (Tobias Leich)++ | src/Perl6/ (2 files):
wire NQP's ModuleLoader to use ...:from<NQP>
masak nwc10: yes, I know. 09:30
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.
jnthn I didn't see you :) 09:31
FROGGS >.<
masak how would you do a sorted set in Perl 6? 09:32
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?
yoleaux 07:50Z <FROGGS> lizmat: the Jugendherberge in frankfurt has no vacancies left ó.ò
lizmat sh*t
FROGGS jnthn: btw, since I'm now able to use QAST, does it makes sense to expose this as AST (or rename it?)
jnthn No
FROGGS NO! :P 09:34
jnthn QAST isn't a Perl 6 standard thing, far from. We should require an explicit use of it.
FROGGS jnthn: don't we need it for slangs?
jnthn waves hands 09:35
FROGGS hehe
jnthn Anyway, no, for now we should not pull it in by default.
But it's nice we can
Is it just those two last patches?
FROGGS true
arnsholt I think macros and slangs are going to need a standardised AST API of some kind
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
jnthn wait, what...
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 dmol left
FROGGS arnsholt: btw, do you have time to look at my zavolaj pull request? 09:37
09:38 dmol joined
arnsholt Oh, right. That thing 09:38
FROGGS jnthn: ohh, then this might need to happen in Perl6::ModuleLoader
arnsholt It looks mergeable, but I'm not sure if Inline::C should be using APIs that internal
FROGGS jnthn: otherwise is exploded because it can't iterate over a Stash ($*GLOBALish from Perl6)
jnthn ah...
FROGGS arnsholt: well, I can duplicate the code, but I wanted to avoid that for Inline::C
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 =)
FROGGS jnthn: alright
arnsholt Like role Native
FROGGS arnsholt: ahh, okay...
arnsholt: I must dig into this this evening, my HEAD is somewhere else atm :o)
arnsholt 's ok 09:40
If that approach doesn't work out, we can expose the other stuff
But I think if we export role Native, your code won't have to do very much plumbing
FROGGS that would be great 09:41
09:42 nugged left, lizmat left 09:45 denis_boyun left, patspam joined
dalek ast: fe9f77c | moritz++ | S32-exceptions/misc.t:
Test for X::Caller::NotDynamic
 09:46
09:47 berekuk joined
moritz can anybody please test if that test passes for you? 09:47
if not, I neet to push a local commit; if yes, I can drop it 09:48
09:50 leont left, TimToady joined
moritz increases the rate limits for the irc logs, because we seem to be all NATted to the same IP 09:51
tadzik hah
nwc10 TimToady has landed in a talk? Gloria is at far left of 3rd row in the main room 09:52
mathw everyone having fun I hope? 09:54
moritz has 09:55
09:55 patspam left 09:56 patspam joined
tadzik yeah, fun :) 09:56
mathw :) 10:01
mathw deleted loads of code this morning, that makes him very happy
FROGGS ?
10:02 patspam left
FROGGS mathw: finished a project? 10:02
mathw removed a stupid feature that should never have been there in the first place 10:03
10:04 patspam joined
FROGGS ahh 10:05
10:10 nugged joined 10:14 nugged left, nugged joined 10:19 fridim_ joined 10:23 leont joined
dalek kudo/nom: 9bbec9c | moritz++ | src/ (2 files):
typed exceptions for if()
 10:24
ast: 52b7dda | moritz++ | S32-exceptions/misc.t:
RT #73806: if() as function call should warn if it leads to a syntax error
jnthn moritz: Is "IfAsFunction" maybe better as "KeywordAsFunction" if it can be for non-if things too? 10:25
:)
moritz oh yes
10:27 nugged left
dalek kudo/nom: 2a26f1a | moritz++ | src/ (2 files):
Rename exception class, jnthn++
 10:27
ast: 6b2c663 | moritz++ | S32-exceptions/misc.t:
Track exception class rename
 10:28
jnthn moritz++
moritz anyway, typed exceptions are great for testing 10:31
jnthn aye
jnthn should track down the leftover typed exception test failures on jvm at some point :)
moritz what I'd really like is to have the NoSuchMethod error actually contain the invocant 10:32
jnthn Hm 10:33
moritz parrot only gives us the name of the class
which is LLTA, actually :-)
jnthn I can probably work out how to do that on the JVM :)
moritz well yes, the code has the invocant right there when it throws the exception
jnthn aye
moritz or it least it did when I last looked at it
jnthn ponders how to convey it 10:34
FROGGS jnthn: and I thought it would be easy :o( gist.github.com/FROGGS/7373412a4e57f42af5fa 10:35
10:36 dmol left, dmol joined 10:37 berekuk left
jnthn FROGGS: hmm.. 10:37
FROGGS exactly 10:40
I'm going to put things in my stomach now... bbiab
10:41 dmol left 10:42 leont left 11:11 fridim_ left, berekuk joined 11:16 dmol joined 11:18 berekuk left 11:19 berekuk joined 11:22 denis_boyun joined
denis_boyun moritz: Hi moritz. Can you ever work with nqp? 11:22
moritz denis_boyun: yes 11:28
11:28 berekuk left
denis_boyun moritz: can you help. Do you know how to run a shell command with nqp? Is it possible? 11:28
moritz: such like system or `` on perl5 11:29
11:30 berekuk joined, lizmat joined
moritz denis_boyun: look at rakudo's src/core/control.pm, sub run 11:30
masak Just Another Perl Heckler 11:31
moritz denis_boyun: it uses pir::spawnw__IP($RPA), but that only works on parrot
masak: ... & Koch
FROGGS denis_boyun: I guess you are interested in parrot specific code?
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
masak accidentally googles "Heckler and Kock"
oh, defense manufacturing.
FROGGS denis_boyun: yeah, I've seen you in #parrot 11:34
11:34 [Sno] joined
denis_boyun moritz: thanks a lot 11:34
moritz denis_boyun: and you can build an RPA in NQP just with [] 11:35
FROGGS nqp: pir::spawnw__IP(["ls"])
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-…
moritz nqp: pir::spawnw__IP(['echo', 'hellow', 'world'])
camelia nqp: OUTPUT«hellow world␤»
moritz too slow on that unusable keyboard 11:36
FROGGS++
11:36 lizmat left
FROGGS I'm sitting on a comfortable keyboard at $work.... :o( 11:36
well, in front of* actually 11:37
11:37 berekuk left
nwc10 has anyone run the range code on riba's t-shirt yet? 11:37
11:38 berekuk joined
FROGGS denis_boyun: and you can pull out pure pir by running: nqp --target=pir -e 'pir::spawnw__IP(["ls"])' 11:38
search for "spawnw" and you find the four lines you need 11:39
masak nwc10: what range code is that?
11:40 lizmat joined, lizmat left
nwc10 I don't have enough unicode in this terminal (etc) and it's screen 11:40
camel .. camel onion
gmmm, the onion isn't in Unicode
it cheats!
11:40 denis_boyun left
moritz .u onion 11:42
yoleaux No characters found
moritz indeed :/
11:43 denis_boyun joined
BabsSeed What's the best way to get involved in compiler development? I'd like to learn and get involved. 11:43
11:43 nugged joined
masak rn: .say for '☃' .. '☄' 11:43
BabsSeed: where do you live?
camelia rakudo 2a26f1: OUTPUT«☃␤☄␤»
..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…
BabsSeed masak: Southern Europe 11:44
FROGGS BabsSeed: maybe start reading here, and then choose a project you would like to contribute? www.josetteorama.com/all-about-perl...rt-1-of-3/
masak BabsSeed: it's funny you should ask tight now.
BabsSeed: there's a workshop being planned in September in Frankfurt.
BabsSeed masak: Oh, Frankfurt isn't far at all
FROGGS BabsSeed: where exactly is "Southern Europe"? 11:45
masak .oO( Niecza goes "what the fuck?", "what the... Godwin?", "what the hell?"... )
11:45 berekuk left
BabsSeed FROGGS: I'm averse to saying too much in a probably logged public channel :P 11:46
FROGGS *g*
it is logged, yes
btw, many devs here are from europe 11:47
TimToady always says too much, except when he doesn't...
11:47 berekuk joined
FROGGS hehe 11:47
BabsSeed FROGGS: I'm from England, but I live in southern Europe :P
FROGGS statement_mod:<TimToady>
ahh, jnthn used to flee from England too 11:48
BabsSeed Flee those taxes :D
tadzik haha 11:49
moritz I guess jnthn's problem with England was that "England" doesn't start with "S"
tadzik isn't taxes somehow related to "you get what you pay for"? :P
BabsSeed tadzik: I didnt get what I paid for :P
tadzik heh
TimToady has he got to Switzerland yet?
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 11:50
11:51 colomon_ joined
BabsSeed So, yeah I wouldn't mind if I got what I paid for :D But unfortunately other people got what I paid for. 11:51
11:53 colomon left, colomon_ is now known as colomon 11:55 denis_boyun left 11:58 nugged left 11:59 leont joined
BabsSeed What areas are most desired for contribution? I'm very comfortable in C 11:59
11:59 denisboyun joined
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
nwc10 you might be more answers once jnthn finishes this talk: act.yapc.eu/ye2013/talk/4912 12:00
12:00 nugged joined
TimToady there's a #moarvm irc channel 12:00
but we're all kinda busy conferencing at the moment
BabsSeed Thanks 12:01
TimToady I think #moarvm is logged by the bot, if you want to backlog
nwc10 it is, irclog.perlgeek.de/moarvm/2013-08-12
BabsSeed Will take a look, definitely want to get involved
12:01 sqirrel left
TimToady irclog.perlgeek.de/moarvm/2013-08-12 for today's log 12:01
12:03 nugged left
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
BabsSeed [Coke]: Someone said you've implemented a Twitter bot in p6, did you make a Twitter module for that? 12:05
I'm considering writing a Twitter module but don't want to reinvent the wheel
12:10 lizmat joined, lizmat left
[Coke] BabsSeed: feather.perl6.nl/~coke/cj.pl 12:11
no module, it was pretty simplistic.
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) 12:13
dalek rlito: 261d4e2 | (Flavio S. Glock)++ | / (3 files):
Perlito5 - grammar - subroutine predeclaration
 12:14
FROGGS [Coke]: I think it has sockets in a development branch # tadzik++ 12:16
nwc10 translucent is the new shiny 12:17
12:19 denis_boyun joined, denisboyun left
JimmyZ I see mortiy‎ and moritz there... 12:19
tadzik [Coke]: actually, I was planning to port MuEvent to rakud-jvm on the hackathon yesterday 12:20
but then I noticed the lack of sockets, so I started that first :)
12:20 leont left
tadzik and of course Java makes it bizzare 12:20
FROGGS .oO( and then you noticed the lack of $a, and started that frist )
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
and then InetAddress, and 5 subclasses of it
</rant>
dalek rlito: d6ada0a | (Flavio S. Glock)++ | src5/lib/Perlito5/Grammar/Map.pm:
Perlito5 - grammar - TODO cleanup
 12:22
12:22 nugged left
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
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 :-)
tadzik moritz: there is SocketFactory, don't worry
12:24 dakkar joined
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
(still bitter about how String gets an overridden infix:<+>) 12:26
12:26 denis_boyun left
JimmyZ moritz: yeah, I saw too :P 12:26
12:27 berekuk left, nugged joined 12:30 [Sno] left 12:34 kbaker joined
BabsSeed [Coke]: Thanks 12:36
[Coke]: Oh OK, I'm looking to implement the full oauth/streaming API 12:38
For use with clients/bots for Twitter
12:39 nine_ joined
JimmyZ nine_: wb 12:45
12:46 crab2313 left 12:53 BenGoldberg joined, fhelmberger left 12:54 dakkar left
tadzik apparently, I'm Bad Luck Brian ;) imgur.com/ufHyjzi 12:57
moritz tadzik: :-) 13:00
13:01 leont joined
FROGGS *g* 13:04
nugged tadzik: non-LOL like in python channel :) 13:06
tadzik haha 13:07
This is #perl6, I can say what I want: LOL
FROGGS *lol*
tadzik FROGGS: watch your mouth!
FROGGS SCHMETTERLING!
tadzik :D
FROGGS ohh, okay ;o)
tadzik I mean, lol
compiling src/Perl6/ConstantFolder.nqp, all I saw is src/Perl6/CannonFodder.nqp 13:08
FROGGS what would be CannonFodder in Perl 6 world? 13:10
moritz do we even need the ConstantFolder.nqp thingy?
tadzik let's rewrite it in Perl6
the game, I mean
FROGGS there is a game called cannon fodder?
moritz iirc it's very limited, and these days we have the optimizer doing a much better job 13:11
tadzik sure
aww, yiss 13:13
dalek p/jvm-sockets: 66ee630 | tadzik++ | src/vm/jvm/ (3 files):
Add opcodes for bind, listen, accept
 13:14
13:15 nugged left, dmol left 13:19 nugged joined 13:20 dmol joined
TimToady finally got on again; room 2 is kind of dead... 13:23
masak o/
GlitchMr grep(&is-prime, 1..*)[25000] is slow compared to alternatives in C++ and JavaScript - v8-io12.appspot.com/#16 13:24
But it's way shorter, so perhaps there is nothing wrong with that.
13:24 dmol left
PerlJam GlitchMr: When do you need the first 25,000 primes quickly? 13:25
:-)
GlitchMr For benchmarks.
I gave up after waiting for two minutes.
masak GlitchMr: you can probably speed it up by writing it without ranges and grep and the call to &is-prime. 13:26
13:26 dmol joined
TimToady say grep(&is-prime, 2,3,5,7 ... *)[^20] 13:26
nr: say grep(&is-prime, 2,3,5,7 ... *)[^20]
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)␤»
..rakudo 2a26f1: OUTPUT«␤»
TimToady nr: say grep(&is-prime, 2,(3,5,7 ... *))[^20]
masak heh.
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␤»
TimToady nr: say grep(&is-prime, 2,(3,5,7 ... *))[1000] 13:27
camelia rakudo 2a26f1, niecza v24-88-g1f87209: OUTPUT«7927␤»
TimToady nr: say grep(&is-prime, 2,(3,5,7 ... *))[10000]
FROGGS .oO( say grep(&is-prime, 2,(3,5,7 .. *))[^20] # excluding Inf ) 13:28
camelia rakudo 2a26f1, niecza v24-88-g1f87209: OUTPUT«(timeout)»
jnthn yay, survived my talk :)
moritz and very well at that :-)
TimToady did you mean ...^ ?
using .. will only count by 1's again 13:29
FROGGS TimToady: yeah, looks like :/
TimToady I suspect 3,5,7...* will already exclude Inf somehow :) 13:30
FROGGS yeah, it was supposed to be a joke
TimToady though it might take it a while to do so :)
13:30 lizmat joined, ajr joined
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␤␤»
..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, ajr is now known as Guest95365
moritz I hate that "Unable to resolve method" error message 13:31
*WHY* is it not able to resolve the method?
13:31 Guest95365 is now known as ajr_
FROGGS No such method... 13:31
is way better
I like to quote M$ here: "Unexpected exception..." - Ohh, so it wasn't expected to fail? How exceptional -.- 13:33
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
maybe figure out relate classes somehow
related*
13:34 kaare_ joined
FROGGS ohh, that is nice 13:35
that saves me to grep for a method name from time to time
13:35 notjack joined, nugged_ joined
moritz well, it's not easy to compile a list of classes that have a certain method without leaking anonymous classes 13:36
13:37 nugged left
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 ? 13:38
jnthn Potentially 13:39
PerlJam Perhaps that it's anonymous already accomplishes this.
jnthn Just need to be careful not to make stuff never-collectable
13:40 berekuk left 13:41 dmol left 13:42 mdn joined
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 13:44
13:45 nugged_ left 13:46 rindolf joined 13:47 mdn left 13:50 PuercoPop left
pmichaud good morning, #perl6 13:52
masak pmichaud! \o/
PerlJam pmichaud! o/
pmichaud just passing through for a bit
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 :-)
pmichaud PerlJam: thanks :-)
FROGGS pmichaud: \o/ 13:54
lizmat happy bday, pmichaud!
masak haooy birthday minus one, pmichaud!
happy*
pmichaud I finally reached the big 3-0
FROGGS is going to wear a funny heat for his b-day
? 13:55
is there something missing in the middle?
lizmat I remember the big 3-0
pmichaud no, at the beginning.
masak happy 0x30, pmichaud :)
FROGGS ahh
pmichaud well, I haven't reached it yet........
FROGGS I was thinking about that but I am too tired to calc
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 ♪
(well, except it's not i dag...)
[Coke] pmichaud: happy pre-birthday.
pmichaud how goes yapc::eu, for those who are there?
masak pmichaud: it's very interesting. 13:57
FROGGS jnthn held his first talk...
masak: how are the ant-talks?
anti
lizmat and TimToady held a very inspiring keynote :-)
FROGGS coll
cool*
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...
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.
PerlJam Is YAPC::EU being recorded as well?
lizmat I've seen cameras
masak basically, "Perl 5 + Moose isn't Perl 7".
lizmat not sure whether they we're one
on
pmichaud wow (keynote cutoff). TimToady++ 14:00
I'm glad to hear TimToady++ come out directly at the conference on the topic
masak pmichaud: I wrote a prepared speech-let for the Wednesday morning discussion... but I don't feel as compelled to deliver it now.
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
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.
notjack what's the general consensus in this forum for Perl5's versioning issue?
masak "movie" commits? 14:01
notjack: oh hai.
notjack hi masak
masak notjack: could you be more specific? what issue?
(there are several) :)
notjack the issue that is motivating people to try to name something "perl 7"
masak right. what issue?
[Coke] 480b93cb01 & 89021cf8e0
(roast)
notjack in short, the cieling put in place by perl6, which ultimately turned out to be a completely different language than perl5
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
masak notjack: no, I know.
and I'm trying to be precise in replying, after I know what the discussion is.
notjack from my outsider perspective, it looks and feels completely different
FROGGS well, Perl 6 is in a position to break backwards compat with Perl 5 code, that is all
14:02 rindolf left
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".
notjack: that said, Perl 5 and Perl 6 are on different paths, so to speak, and have been for many years.
FROGGS notjack: what do you want a new perl 5 interpreter for?
masak notjack: though there is work being done to bring them closer.
notjack FROGGS: only to (arbitrarily) claim the "perl5 backwards compatible" badge 14:04
masak: what's motivating that work?
14:08 bruges left
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) 14:09
masak notjack: the view that combining existing code bases and Perl 6 would be useful and interesting.
it would be a transition path, if nothing else.
notjack hmm, I forgot about CPAN
masak :)
notjack yeah, I can see that as a huge reason to maintain compat and compatible paths
masak FROGGS++ is helping us with another bit of that. you can write Perl 5 code inside a block in Perl 6.
notjack what's the syntax?
{{{ use v5; blahh blah blah }}} ?
masak one set of braces is enough :)
but yes. that.
notjack cool
FROGGS no XS though, just to be clear
notjack can you make me a {{{ use APL; iota rho etc }}}?
kthx
FROGGS *g*
14:09 mdn joined
FROGGS slangs are possible, yes 14:09
masak notjack: :D
notjack: that's a goal, yes.
one schtepp at a time.
notjack heh
I've been writing a small script in v5 this week
masak FROGGS: yes, "no XS" would be why we also want diakopter++'s grant.
notjack I'd forgotten how quirky (though loveable) it is
14:09 kst`` joined
masak Perl 6 is less quirky and more lovable, IMO ;) 14:09
FROGGS masak: I was about to typing that, yes
notjack yeah, but everyone thinks that about their own weird-looking baby
:)
14:09 bruges_ joined, logie joined
masak notjack: granted, but I hear many outsiders saying it, too. 14:09
14:09 tsmith joined
masak notjack: not all of them, mind. but many. 14:09
14:09 bluescreen10 joined
FROGGS just explain a perl n00b perl 5's changing sigils >.< 14:09
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, kst` left
pmichaud I wonder if I can find TimToady's keynote somewhere... 14:11
FROGGS so he knows a bit about scalars now :o) 14:12
lizmat pmichaud: we'll let you know as soon as we know it 14:13
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 benabik left
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
notjack is there an equivalent to qq inside a regex?
or q I mean
arnsholt Quotes
notjack legit?
arnsholt Simple as that =)
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
notjack that's fair 14:16
BenGoldberg notjack: In perl5 regexes, it is of course /\Qhi, I am not a metachar\E/
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
progress! :)
that means they connect to each other, since they die at the same time :) 14:18
14:18 berekuk left
FROGGS tadzik: yay! :o) 14:20
14:20 denisboyun joined 14:22 mdn left
[Coke] tadzik++ 14:23
tadzik: gentle reminder, you have a failing Star module.
tadzik [Coke]: oh! MEvent? 14:24
[Coke] tadzik: github.com/coke/rakudo-star-daily/...s.log#L216 14:25
If I counted right, that's: github.com/tadzik/perl6-File-Tools...mand.t#L26
tadzik [Coke]: oh,m I think I fixed that yesterday 14:26
oh, no I didn't
[Coke] failed in the latest daily star run. (about 2pm eastern)
tadzik I should just remove File::Tools, it's now splitted between File::Find and Shell::Command
I fixed File::Find, but not File::Tools
I'll remove it from the ecosystem
[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!
I will do that
masak tadzik++
14:27 dmol joined
dalek osystem: b8b90c7 | tadzik++ | META.list:
Remove File::Tools
 14:29
[Coke] tadzik++ 14:30
dalek kudo-js: 2d2fef3 | (Pawel Murias)++ | run_tests:
Add test 46 to run_tests (which passes if icu is installed).
 14:31
kudo-js: e1face9 | (Pawel Murias)++ | src/QAST/Compiler/JavaScript.nqp:
Implement the eos anchor.
kudo-js: 9605581 | (Pawel Murias)++ | src/QAST/Compiler/JavaScript.nqp:
Emit an extra newline for better code readablity.
kudo-js: 8056f61 | (Pawel Murias)++ | / (2 files):
Refactor run_tests so that it's more visible which tests are missing.
kudo-js: 8f90bad | (Pawel Murias)++ | / (2 files):
Update the setting, subst now works, pass test 62
pmurias 4 tests from nqp/t/nqp are missing 14:32
14:32 crab2313 joined
[Coke] pmurias++ 14:32
tadzik oh shit no
raukdostar uses submodules
I ned help then :|
14:33 lizmat left
pmurias [Coke]: actually it's 5 will be 4 very soon ;) 14:33
14:34 lizmat joined
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 14:41 Kiralexis joined 14:42 Psyche^ left
pmurias [Coke]: cross-compile, most of the tests in nqp/t/nqp, there also tests in nqp/t/serialization and nqp/t/qregex 14:43
14:43 kaleem left
colomon what's going on at YAPC::EU today? 14:44
nwc10 lightning talks
14:44 leont left
nwc10 right now ribasushi is telling us about what's new in DBIx::Class 14:44
everyone is hiring in Ukraine, it seems
leont and I have jnthn surrounded and are going to kidnap him and force him to work on Perl 5 14:45
:-)
oh, wait, he'll commit things to make it slower. D'oh!
lizmat
.oO(all part of a cunning plan to recruit leont and nwc10)
 14:46
14:47 crab2313 left 14:50 telex left 14:51 lizmat left 14:52 telex joined 14:59 kaleem joined 15:01 kaleem left, kaleem joined 15:04 colomon left
nwc10 lizmat: Perl 6 is hiring? :-) 15:04
moritz sure. Our pay is competitive with most other open source projects! 15:06
FROGGS *g*
[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 15:15 colomon joined 15:16 leont joined 15:18 _jaldhar joined
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 15:21
mathw [Coke]: Last time I tried a constraint-based thing it was not fun or nifty or easy at all :( 15:24
15:25 denisboyun left
TimToady it's one of those things that we're speccing to be easy to implement...someday... 15:26
andthen and orelse are part of that, which together with lazy lists should allow implementation of a normal backtracker
in fact, that's basically how STD works inside
15:27 birdwindupbird left
PerlJam 6/win 15 15:27
15:27 FROGGS left 15:28 dmol left, ajr_ left
[Coke] timotimo: untodoing it doesn't mean it passes on .jvm 15:28
15:29 leont left
[Coke] plus someone broke several dozen tests - it's possible your passes are hidden behind that. 15:29
15:29 ajr joined
dalek p: 376098c | (Pawel Murias)++ | t/nqp/63-slurp.t:
Check if we actually slurp something with a correct length.
 15:29
pmurias a good puzzle is how to write a test for slurps that checks if it slurps itself correctly ;)
15:29 ajr is now known as Guest24616 15:30 Guest24616 is now known as ajr_
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) 15:30
it was suspicious to me that the rakudo.jvm-summary file had no change whatsoever 15:31
but yeah, you're right. is_default blew up completely now 15:32
15:35 kaleem left
timotimo and substr for some reason 15:37
dalek kudo-js: 491ab5e | (Pawel Murias)++ | / (5 files):
Implement the ops used by nqp::slurp.

Run tests inside the nqp directory so that relative paths inside of them work correctly. Pass test 63.
pmurias [Coke]: 4 test left now ;)
* tests
15:37 kbaker left 15:38 kbaker joined
[Coke] pmurias: woooo 15:38
[Coke] is handed a box of cider apple croissant donuts
timotimo wow that souds amazing 15:44
pmurias jnthn: why do we call decont on self? 15:45
jnthn: in nqp code
grondilu Is today TimToady's keynote online yet? 15:52
15:52 rindolf joined
colomon What all has happened today? p6y, I mean. 15:56
16:02 Ayiko joined
timotimo it kind of seems like fudging in the substr tests is b0rked 16:06
fudge and run seems to fudge it for .rakudo and then runs the .t instead 16:08
16:10 pmurias left
[Coke] are you fudging for "rakudo" or "rakudo.jvm" ? 16:10
"rakudo" is no longer a valid thing to pass in, I think.
timotimo i'm just running fudgeandrun, should i have to change that? i thought it figured out the implementation o nits own?
[Coke] that -should- work, yes. 16:11
this on jvm?
timotimo yes
[Coke] what does your ./perl6 --version say ?
(that's what t/fudgeandrun is using)
timotimo +S32-str/substr.rakudo.jvm aborted 63 test(s)
looking at why this is b0rked
if you fudgeandrun it correctly, you will indeed get only one failure 16:12
This is perl6 version 2013.07-206-g2a26f1a built on JVM
[Coke] that should fudge for rakudo.jvm, then.
ah, that file has no vm specific fudges. testing 16:14
timotimo it does have rakudo.jvm in the imlp variable
[Coke] gist.github.com/coke/6212404 16:15
16:15 ajr_ left
timotimo oh wait, yeah, that could still be it 16:15
locally i get the "too many parameters" problem from very early on
[Coke] my rakudo-jvm on this box is probably 2 days old. 16:16
16:16 ajr joined 16:17 ajr is now known as Guest79469
grondilu cool, someone got rid of the _MVM in MoarVM. 16:17
16:24 Guest79469 is now known as ajr_ 16:32 lizmat joined 16:39 spider-mario joined
BenGoldberg r: say grep(&is-prime, 2,(3,5..*))[1000] 16:42
camelia rakudo 2a26f1: OUTPUT«7927␤»
BenGoldberg r: .say for grep(&is-prime, 2,(3,5..*))[1000..*] 16:43
camelia rakudo 2a26f1: OUTPUT«(timeout)»
colomon 3,5..* is 3, 5, 6, 7, 8, 9 … etc
BenGoldberg r: .say for grep(&is-prime, 2,(3,5...*))[100..*]
colomon you probably want 3, 5… * -- right
camelia rakudo 2a26f1: OUTPUT«(timeout)» 16:44
BenGoldberg r: .say for grep(&is-prime, 2,(3,5...*))
colomon why are you asking for an infinite list of primes?
camelia rakudo 2a26f1: OUTPUT«(timeout)»
BenGoldberg r: .say for grep(True, 2,(3,5...*))
camelia rakudo 2a26f1: OUTPUT«(timeout)» 16:45
colomon or inifinitely many odd numbers?
BenGoldberg r: .say for grep(True, ^Inf)
camelia rakudo 2a26f1: OUTPUT«(timeout)»
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␤
BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[^1000]; say @p[$_] for ^Inf 16:48
camelia rakudo 2a26f1: OUTPUT«(timeout)»
colomon that one may just be slow… at least, I don't see a reason for an infinite loop 16:49
BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[1000..Inf]; say @p[$_] for 0,1
camelia rakudo 2a26f1: OUTPUT«(timeout)»
BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[100..Inf]; say @p[$_] for 0,1
camelia rakudo 2a26f1: OUTPUT«(timeout)» 16:50
BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[0..Inf]; say @p[$_] for 0,1
camelia rakudo 2a26f1: OUTPUT«(timeout)»
BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[0..Inf];
camelia rakudo 2a26f1: OUTPUT«(timeout)»
BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*))[0...Inf];
camelia rakudo 2a26f1: OUTPUT«(timeout)» 16:51
BenGoldberg r: my @p := grep(&is-prime, 2,(3,5...*));
camelia rakudo 2a26f1: OUTPUT«(timeout)»
colomon n: my @p := grep(&is-prime, 2,(3,5...*)); 16:52
camelia niecza v24-88-g1f87209: OUTPUT«Potential difficulties:␤ @p is declared but not used at /tmp/KJ6eO8lJsj line 1:␤------> my ⏏@p := grep(&is-prime, 2,(3,5...*));␤␤»
colomon n: my @p := grep(&is-prime, 2,(3,5...*)); say "Hi!"
camelia niecza v24-88-g1f87209: OUTPUT«Potential difficulties:␤ @p is declared but not used at /tmp/YWHP_K1pCt line 1:␤------> my ⏏@p := grep(&is-prime, 2,(3,5...*)); say ␤␤Hi!␤»
BenGoldberg rn: my @p := grep(True, ^Inf) 16:53
camelia niecza v24-88-g1f87209: OUTPUT«Potential difficulties:␤ @p is declared but not used at /tmp/HVV1ayCSp2 line 1:␤------> my ⏏@p := grep(True, ^Inf)␤␤»
..rakudo 2a26f1: OUTPUT«(timeout)»
BenGoldberg rn: my @p; say 'hi'
camelia niecza v24-88-g1f87209: OUTPUT«Potential difficulties:␤ @p is declared but not used at /tmp/yXj28nFt2b line 1:␤------> my ⏏@p; say 'hi'␤␤hi␤»
..rakudo 2a26f1: OUTPUT«hi␤»
BenGoldberg rn: my @p; @p := grep(True, [1]) 16:54
camelia rakudo 2a26f1, niecza v24-88-g1f87209: ( no output )
BenGoldberg rn: my @p; @p := grep(True, [^Inf])
camelia niecza v24-88-g1f87209: OUTPUT«(timeout)»
..rakudo 2a26f1: ( no output )
16:55 ssutch joined
moritz n: [^Inf]; say "alive" 16:58
camelia niecza v24-88-g1f87209: OUTPUT«(timeout)»
17:00 vk___ left, btyler left 17:03 btyler joined 17:12 ajr_ left 17:13 crab2313 joined 17:15 BigBear left 17:17 benabik joined 17:19 slava joined 17:20 vovka667 joined 17:23 fhelmberger joined 17:25 MDn joined
dalek rlito: d435a99 | (Flavio S. Glock)++ | / (3 files):
Perlito5 - grammar - more 'variable' rules
 17:26
17:28 MDn left 17:29 fhelmberger left 17:31 Mouq joined 17:34 Vico_Wu1 joined 17:35 nugged joined 17:37 nugged_ joined, nugged left, crab2313 left 17:38 Vico_Wu left 17:39 ssutch left 17:40 vovka667 left, ssutch joined, vovka667 joined 17:41 crab2313 joined 17:44 nugged_ left, Rotwang joined, ajr joined 17:45 ajr is now known as Guest2529
notjack has anyone done any work on a p6 plugin for Eclipse? 17:47
in particular, syntax higlighting
17:48 ssutch left
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 17:49 Vico_Wu1 left
notjack take a p6 grammar and generate an Eclipse syntax-highlighting file, you mean? 17:50
17:50 FROGGS joined
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) 17:51
17:52 _jaldhar left 17:54 dmol joined
moritz well, if the syntax hilighting file for that IDE is Turing complete, yes 17:55
17:59 crab2313 left, berekuk joined
notjack speaking of Turing completeness 17:59
in order to learn p6 better, I'd like to implement a little J parser in it
probably is, I know less about parsing than I do about p6, as a practical matter 18:00
problem is*
moritz notjack: have you read the "Using Perl 6" ebook?
it has a chapter about p6 regexes, and one about grammars
notjack I've grepped it, but not read it 18:01
p6 regexes I think I can get hold of pretty quick
moritz grepping isn't enough for learning -)
notjack grammars are harder ... because I don't understand grammars as a general topic
18:01 denis_boyun joined
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
notjack Let me ask: the J language has 9 production rules in its grammar, and a 4-token fixed lookahead (www.jsoftware.com/help/dictionary/dicte.htm) 18:02
in lines of p6 code ... my gut says it should take about a page of code 18:03
anyone think differently?
moritz notjack: the book has a complete JSON parser as an example, and explains it
notjack ooh
18:03 Mouq left
notjack "and explains it"++ 18:03
PerlJam moritz++
moritz (sorry, I have to finish my slides for tomorrow before looking at the J grammar)
PerlJam notjack: a "page" for just the parser? sure. 18:04
notjack: add another for the semantics.
notjack forget the semantics
that would take a lot of work
I mean just building an AST
moritz that's explained for the JSON stuff too 18:05
notjack J has ~120ish primitives and things like implicit composition, etc.
I can write that (SMOP), but I need a place to start, i.e. an AST
18:10 _jaldhar joined 18:12 dmol left
moritz masak: elixir-lang.org/blog/2013/08/08/eli...ign-goals/ has some macro examples, including some we considered for Perl 6 18:13
18:14 btyler left
moritz (peeking into the components of a boolean expression to explain why a test/assert failed) 18:14
18:14 kst`` is now known as kst, mdn left 18:15 vovka667 left 18:18 rindolf left 18:21 japhb_ joined 18:22 SamuraiJack_ left
moritz ok, talks if finally "prepared" 18:23
18:24 btyler joined
moritz notjack: I can't actually deduce the J grammar rules from the page you gave us 18:29
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? 18:30
moritz timotimo: I talkded to jnthn about that earlier today, and off-hand he didn't know any way 18:31
timotimo ah damn
moritz timotimo: but he made the interesting observation that nqp could greatly benefit from that optimization
timotimo yes, i thought so, too
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
yes, might not be trivial to get started
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 18:39 fhelmberger joined
slava [op] 1,2,3,4 ... this is awesome, it looks like a folding from haskell 18:39
timotimo and indeed it is :) 18:41
18:43 dmol joined
jaffa4 anybody tied jvm? 18:43
moritz tied the JVM into a knot 18:44
jaffa4 tried
moritz last time I tried to build rakudo on JVM, the build failed 18:45
complained about a missing op
but before it worked great 18:46
timotimo i compiled a rakudo.jvm today from newest nqp and rakudo master/nom branches
18:46 fhelmberger left
jaffa4 99% complete I heard 18:48
FROGGS jaffa4: true 18:49
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"
TimToady if you set the CLASSPATH it seems to help with that 18:51
18:51 skids joined
TimToady was doing commands like: CLASSPATH=~/nom perl6 rad.p6 18:51
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
obviously, it'd be nice not to have to set the CLASSPATH explicitly like that 18:53
18:53 bloonix joined, sidus joined 18:55 btyler left
jaffa4 it is not clear how to compile for jvm 18:58
TimToady did you read the README? 19:00
has some directions there that seem relatively adequate
as in, worked for me :) 19:01
TimToady is scheduled to sleep like a log for the 120° or so of earthly rotation 19:02
jaffa4 120?
is that 8 hours? 19:03
timotimo r: (120 / 360) * 24
camelia rakudo 2a26f1: ( no output )
timotimo r: say (120 / 360) * 24
camelia rakudo 2a26f1: OUTPUT«8␤»
jaffa4 TimToady: thank you
notjack common misconception. the logs, they're not sleeping. they're merely biding their time. 19:04
TimToady has $.timotimo handles 'thank-you';
timotimo :D
TimToady zzz &
19:06 PacoAir joined, Targen joined
jaffa4 Are you sleeping and wawke at the same time? 19:07
awake<<
19:08 zwut00 joined 19:12 dmol left
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
19:15 btyler joined
timotimo there's ugly hacks like /..<!after '::'>/ 19:16
benabik I wonder if /<!before '::'>../ would be faster.
timotimo may be, yes
flussence something like /<?{ $0 eq '::' }>/?
dalek kudo/nom: a379871 | moritz++ | / (4 files):
remove Perl6::ConstantFolder

it was not really needed, and the optimizer does constant folding much more generally		 19:17
colomon something with split, maybe? depends on what you are trying to do.
moritz negating a regex usually only makes sense if you anchor it somewhere 19:18
timotimo that's a good point 19:20
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
flussence yeah, you're right 19:22
19:25 zby_home joined, btyler left 19:26 Rotwang left, zby_home left 19:27 mdn joined
moritz blug: perlgeek.de/blog-en/perl-6/2013-yapc-day-1.html 19:28
timotimo almost as good as being there in person! 19:30
colomon \o/ 19:31
moritz there were some hard choices to make
timotimo "but in the didn't." 19:32
moritz for example I would have loved to attend ye2013.yapc-surveys.org/talk/32 but it clashed with another one I wanted to see
timotimo "invalid code"
19:34 btyler joined
moritz timotimo: fixed, thanks 19:35
lue here's what I'm trying to do --> gist.github.com/lue/6214323 it currently prints the key (debug statement) and then fails the match. 19:37
19:39 Rotwang joined 19:42 btyler_ joined
[Coke] moritz: I've been building the jvm daily, no problems. 19:44
dalek kudo-star-daily: 25328ae | coke++ | log/ (5 files):
today (automated commit)
19:44 btyler left
moritz ok, then it's probably my fault somehow :-) 19:44
[Coke] timotimo: can you open a rakudo star ticket about dropping tools and adding the 2 other modules? 19:47
that way it doesn't get lost.
timotimo i guess i can, even though it's tadzik's changes?
[Coke] (er, timotimo? tadzik? my t-space is pretty full...)
timotimo i forgot which modules those are, though
yeah, let tadzik do it :)
[Coke] timotimo++, I meant tadzik++, apologies. 19:48
[Coke] wonders what secrete project jnthn++ is revealing this week
*secret 19:49
jaffa4 what makes you think that?\
[Coke] ... it's a perl conference where jnthn++ is presenting.
colomon jaffa4: it's kind of a tradition
jaffa4 again something? 19:50
moritz [Coke]: not really secret, but jnthn has been doing some impressive hacking on concurrency stuff on JVM
[Coke] it's how we got moarvm, rakudo::debugger...
colomon grammar::debugger... 19:51
the grammar generator stuff..
jaffa4 is this communism?
19:52 dmol joined
[Coke] I don't think it's an ism. 19:52
19:53 celesta joined
celesta ahoi #perl6 19:53
[Coke] aloha 20:00
20:00 Vico_Wu1 joined 20:02 mdn left 20:03 Vico_Wu left 20:07 BigBear joined 20:08 dmol left
celesta r: sub foo( @x, @y ){ for @x >>~~<< @y { return False if !$_; }; return True; }; say foo( 1..5, 5..9); 20:10
camelia rakudo 2a26f1: OUTPUT«False␤»
celesta how do i "catch" the error if lists are not of same length?
20:10 kaare_ left
celesta r: sub validate( @x, @y ){ for @x >>~~<< @y { return False if !$_; }; 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…
FROGGS well, you can use a try { CATCH {} } block or you just check the array lengths yourself 20:12
celesta FROGGS: thank you 20:14
20:17 colomon left 20:21 Mouq joined
jaffa4 it lool=ks fot nmake 20:24
rakudo: say "hello
camelia rakudo a37987: OUTPUT«===SORRY!=== Error while compiling /tmp/sbUmHsHVVt␤Unable to parse expression in double quotes; couldn't find final '"'␤at /tmp/sbUmHsHVVt:1␤------> say "hello⏏<EOL>␤ expecting any of:␤ argument list␤ prefix or …
jaffa4 rakudo: say "hello"
camelia rakudo a37987: OUTPUT«hello␤»
Guest2529 I know this code is wrong, and why; is the error message right? 20:30
perl6 -e 'say "say "Hi!"; say "Perl 6 here";'
===SORRY!===
Two terms in a row
at -e:1
------> say "say "⏏Hi!"; say "Perl 6 here";
expecting any of:
postfix
infix stopper
infix or meta-infix
statement end
statement modifier
statement modifier loop
20:30 Guest2529 is now known as ajr_
benabik ajr_: Too many double quotes. 20:31
ajr_ As I said, I understand the error, I just wonder about the message. 20:32
20:32 colomon joined
benabik ajr_: It's parsing Hi! as a term, immediately after the string term. I think most unknown words are assumed to be terms. 20:33
[Coke] r: say what the 20:36
camelia rakudo a37987: OUTPUT«===SORRY!=== Error while compiling /tmp/BsigbqGP4p␤Undeclared routines:␤ the used at line 1␤ what used at line 1␤␤»
[Coke] r: say "what" "the"
camelia rakudo a37987: OUTPUT«===SORRY!=== Error while compiling /tmp/6j4io3qQjS␤Two terms in a row␤at /tmp/6j4io3qQjS:1␤------> say "what" ⏏"the"␤ expecting any of:␤ postfix␤ infix stopper␤ infix or meta-infix␤ statement end␤ …
[Coke] r: say "what" the
camelia rakudo a37987: OUTPUT«===SORRY!=== Error while compiling /tmp/C6BIL3JEiy␤Two terms in a row␤at /tmp/C6BIL3JEiy:1␤------> say "what" ⏏the␤ expecting any of:␤ postfix␤ infix stopper␤ infix or meta-infix␤ statement end␤ … 20:37
[Coke] ^^
notjack r: say 'thefuck'.WHAT; 20:46
camelia rakudo a37987: OUTPUT«(Str)␤»
notjack r: say <tee eff>.WHAT;
camelia rakudo a37987: OUTPUT«(Parcel)␤»
notjack r: say <tee eff>.WHAT.WHAT;
camelia rakudo a37987: OUTPUT«(Parcel)␤»
20:47 jaffa4 left 20:48 kbaker left
[Coke] r: say.WHAT.say, now? 20:52
camelia rakudo a37987: OUTPUT«===SORRY!=== Error while compiling /tmp/UaVu6ZOYBT␤Two terms in a row␤at /tmp/UaVu6ZOYBT:1␤------> say.WHAT.say, now⏏?␤ expecting any of:␤ postfix␤ infix stopper␤ infix or meta-infix␤ statement end…
[Coke] aw.
masak o/ 20:55
[Coke] o/~
masak moritz++ # blug post 20:58
lue hello masak o/ 21:00
masak hellue, lo. \o 21:07
celesta heja masak 21:08
dalek rl6-roast-data: 8c0f742 | coke++ | / (5 files):
today (automated commit)
 21:09
[Coke] rakudo.parrot failing more tests today. rakudo.jvm dropped another .01%
21:09 skids left
[Coke] looks like it's "is dynamic" related fallout. also rakudo.parrot has passing todos. 21:10
masak aloha celesta. 21:14
lue masak: Any clues as to why gist.github.com/lue/6214323 fails on the first :: ?
21:14 benabik left
[Coke] lue: have you tried the regex debugger? 21:17
lue panda's refusing to work :/ (I may have to update rakudo, it's been a little while) 21:18
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?
masak lue: that's just an hypothesis, but it seems reasonable to me.
rjbs: wiki.perl6.org/Create%20and%20Distr...%20Modules
rjbs Thanks! 21:21
masak uvtc++ siddhant++ # creating that article
[Coke] wonders how many times rjbs will have to do that before we get a distzilla.
lue Yep, it gets the colon (added debug msg), so it fails on the before. 21:23
masak lue: that's because $<closer> does not contain the whitespace after the '::' 21:24
21:25 augustus joined
lue I'm going to go update rakudo, and then see if panda will work for me. 21:26
augustus so is anyone around here today, or is everybody off at YAPC? 21:34
masak augustus: we're all off at YAPC ;)
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?
masak augustus: oh, people are still hanging out here. 21:37
augustus: and you're likely to get replies pretty quickly.
jercos I'm trapped in a sphere made of my own self-loathing, so I'm usually watching, rather than talking.
masak through advanced 20th-century technologies like "IRC backlog" and "screen".
augustus hehe nice :)
masak jercos: that's not altogether uncommon, IIUC. 21:38
jercos U?
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
masak I know I was a lurker for many years. 21:39
dduncan! \o/
21:39 spider-mario left
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?
masak dduncan: long time no Muldis! :D
lue should do something more significant for Perl 6. Oh wait, there's the secret project.
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.
dduncan If I ever get a couple months to dedicate to it, I could have an executable, I"m that ready otherwise.
masak dduncan: nor on the p6l list.
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"
augustus jercos: yes it is impressive, and much easier than before :) 21:42
masak "before"?
dduncan My sign-on here was actually an auto-join bookmark ... I had opened freenode to use another channel actually.
21:42 Vico_Wu1 left 21:43 Vico_Wu joined
dduncan People don't usually greet me so quicky when that happens. 21:43
that is all
masak dduncan: we've missed you too! :P
dduncan My day job is all Perl / databases though, thankfully.
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
masak augustus: was that Perl 5 or Rakudo?
augustus masak: to the best of my knowledge Perl 5 doesn't run on JVM? Anyway, Rakudo. 21:45
jnthn Yes, you could do it that way before. 21:46
PerlJam augustus: how long ago was "previously"?
jnthn If you knew how :)
It's the way Threading.pm is written.
masak oh!
21:48 Vico_Wu1 joined 21:49 Vico_Wu left
labster So how is YAPC going, folks? 21:51
21:51 Rotwang left 21:52 FROGGS left
labster Is there going to be a way to see these talks online? 21:52
augustus PerlJam: Perviously was like 10-15 days ago? I forget exactly 21:53
21:53 colomon left 21:54 notjack left
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
21:54 colomon joined, aindilis` joined
jnthn augustus: Still the Treading.pm way if there are overloads, otherwise it's just .new 21:54
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?
jnthn: very exciting!
jnthn: too bad I couldn't be there
21:56 aindilis left
timotimo is waiting eagerly for the recording 21:56
i won't get anything else done :(
jnthn augustus: yeah, I think the SWT example I showed today did it, it's something like field/get_name/typedesc
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
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?
jnthn timotimo: No obvious way? Hm 22:00
There's an unobvious way if you can get the code object though ;)
22:02 jeff_s1 left
timotimo that's the problem, i see no way to get the Code object :( 22:03
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 left, colomon joined
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
22:04 zacts left, Targen left, zacts joined 22:05 Targen joined, jaldhar_ joined 22:06 jeff_s1 joined, perigrin joined, denis_boyun left, _jaldhar left, segomos joined, BigBear_ joined 22:08 BigBear left
jnthn r: loop say 42; 22:10
camelia rakudo a37987: OUTPUT«===SORRY!=== Error while compiling /tmp/FVoNwiggvi␤Missing block␤at /tmp/FVoNwiggvi:1␤------> loop ⏏say 42;␤ expecting any of:␤ statement list␤ scoped block␤»
22:11 BigBear_ left
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?
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␤
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? 22:14
22:16 silug joined 22:18 bluescreen10 left
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
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.
timotimo so add_n is for both integers and nums? 22:23
jnthn no, the code-gen coerces!
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
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
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
yes
Which is sucky :)
And a good target for the optimizer.
Rakudo lacks this problem
As its numeric type system is much stronger 22:26
timotimo OK
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
maybe tomorrow i'll hack up something very simple
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 left, segomos joined
jnthn timotimo: Should be about the same 22:32
22:33 lizmat left
timotimo neato. 22:33
22:33 lizmat joined, lizmat left
jnthn ok, enough slide hacking...think I'm gonna sleep :) 22:34
'night, #perl6
timotimo gnite :)
celesta good night everybody 22:35
22:35 Vico_Wu2 joined, celesta left, lizmat joined 22:38 Vico_Wu1 left, skids joined 22:45 dduncan left 22:47 Mouq left 22:52 ssutch joined
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. 22:53
masak no, we're still here. 22:56
most people are either sleeping already, or still out being social. 22:57
22:57 fridim_ joined 22:58 FROGGS joined
masak 'night, #perl6 23:00
timotimo good night masak :)
23:03 skids left 23:08 sidus left 23:11 augustus left 23:18 ssutch left 23:19 ssutch joined 23:20 FROGGS left 23:22 nebuchadnezzar joined
lue can anyone here get Rakudo::Debugger to work? 23:25
timotimo i can't get panda to work actually 23:29
lue wfm; have you updated your PATH?
timotimo i can't get shell::command to work from the ext/ dir 23:30
gimme a minute
ufo is so slow :( 23:32
23:45 ajr_ left 23:52 cognominal left, cognominal joined
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. 23:56

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!