»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 dansamo joined
japhb__ jnthn: excellent. 00:06
00:09 dmol left 00:10 dansamo left 00:12 Psyche^ left, risou_awy is now known as risou, Psyche^ joined 00:15 ajr_ left 00:25 benabik joined 00:30 risou is now known as risou_awy 00:42 rurban joined 00:44 benabik left
dalek p: 183b8af | dwarring++ | examples/rubyish/ (2 files):
rubyish =begin .. =end multiline comments. simplify slurpy hashes
00:45
00:46 rurban left
dalek kudo/nom: cf0b0ca | jonathan++ | src/vm/jvm/core/Threading.pm:
Better enforce Exception object coercion.
00:56
kudo/nom: 3208450 | jonathan++ | src/vm/jvm/core/Threading.pm:
Use Promise::Keeper in Channel.

To be more robust.
kudo/nom: 9d54481 | jonathan++ | src/vm/jvm/core/Threading.pm:
Add Subscribable, and initial publish bits.

This is the first piece towards enabling reactive programming with push-collections, a la Rx. Many combinators to come.
ast: b9d4b05 | jonathan++ | S17-concurrency/subscribable.t:
Add basic Subscribable tests.
jnthn 'night 00:59
01:02 rurban joined 01:11 benabik joined
dalek -pb: 5f9f01b | (Geoffrey Broadwell)++ | / (2 files):
Add PB::Binary::WireTypes with tests

Put WireType enumeration in a common place, with tests to make sure the values don't accidentally change (they're part of the PB standard).
01:12
-pb: ad86091 | (Geoffrey Broadwell)++ | / (4 files):
Switch to using PB::Binary::WireTypes

Move code and tests using either WIRE_TYPE_* or hard-coded numbers to the WireType::* enum instead.
01:17 PacoAir left 01:20 Rotwang left 01:21 PacoAir joined 01:33 Vendethi_ joined 01:38 Vendethi_ left 01:39 dwarring joined 01:40 colomon left
dwarring has the module smoke server gone out? I've been using 97.95.33.142:3001/report 01:41
01:41 colomon joined
dwarring no results since 25th Oct 01:41
colomon panda has been broken 01:46
r: say "TH̘E􏿽xCD􏿽x84􏿽xCC􏿽x89􏿽xCD􏿽x96 ͠P̯͍̭O̚​N̐Y̡ H􏿽xCD􏿽xA8􏿽xCD􏿽x8A􏿽xCC􏿽xBD􏿽xCC􏿽x85􏿽xCC􏿽xBE􏿽xCC􏿽x8E􏿽xCC􏿽xA1􏿽xCC􏿽xB8􏿽xCC􏿽xAA􏿽xCC􏿽xAFE􏿽xCC􏿽xBE􏿽xCD􏿽x9B􏿽xCD􏿽xAA􏿽xCD􏿽x84􏿽xCC􏿽x80􏿽xCC􏿽x81􏿽xCC􏿽xA7􏿽xCD􏿽x98􏿽xCC􏿽xAC􏿽xCC􏿽xA9 􏿽xCD􏿽xA7􏿽xCC􏿽xBE􏿽xCD􏿽xAC􏿽xCC􏿽xA7􏿽xCC􏿽xB6􏿽xCC􏿽xA8􏿽xCC􏿽xB1􏿽xCC􏿽xB9􏿽xCC􏿽xAD􏿽xCC􏿽xAFC􏿽xCD􏿽xAD􏿽xCC􏿽x8F􏿽xCD􏿽xA5􏿽xCD􏿽xAE􏿽xCD􏿽x9F􏿽xCC􏿽xB7􏿽xCC􏿽x99􏿽xCC􏿽xB2􏿽xCC􏿽x9D􏿽xCD􏿽x96Oͮ͏̮̪̝͍M􏿽xCD􏿽x8A􏿽xCC􏿽x92􏿽xCC􏿽x9A􏿽xCD􏿽xAA􏿽xCD􏿽xA9􏿽xCD􏿽xAC􏿽xCC􏿽x9A􏿽xCD􏿽x9C􏿽xCC􏿽xB2􏿽xCC􏿽x96E􏿽xCC􏿽x91􏿽xCD􏿽xA9􏿽xCD􏿽x8C􏿽xCD􏿽x9D􏿽xCC􏿽xB4􏿽xCC􏿽x9F􏿽xCC􏿽x9F􏿽xCD􏿽x99􏿽xCC􏿽x9ES􏿽xCD􏿽xAF􏿽xCC􏿽xBF􏿽xCC􏿽x94􏿽xCC􏿽xA8􏿽xCD􏿽x80􏿽xCC􏿽xA5􏿽xCD􏿽x85􏿽xCC􏿽xAB􏿽xCD􏿽x8E􏿽xCC􏿽xAD".ords 02:02
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«84 72 792 69 836 777 854 32 864 80 815 845 813 79 794 8203 78 784 89 801 32 72 872 842 829 773 830 782 801 824 810 815 69 830 859 874 836 768 769 807 856 812 809 32 871 830 876 807 822 808 817 825 813 815 67 877 783 869 878 863 823 793…»
dwarring heh
japhb__ What is with the pony, anyway? I mean, I remember the "you can't parse HTML with regexes" rant on SO that ended with "Tony the pony he comes", but where did *that* come from? 02:04
colomon has no clue, but was looking for a wicked Unicode string to use as a test case for work.
japhb__ That qualifies.
I tested PB serialization of strings with just two: an ASCII-only one, and one that contained one-, two-, and three-byte Unicode codepoint values. 02:07
I figure if those both roundtrip, there's half the battle right there.
colomon That's essentially what I'm looking to do here, too.
japhb__ Oh interesting! When I switched screens (ctrl-A ctrl-A) to another screen and back, the rendering of your test string *changed*. 02:08
colomon serialization via the STEP file format.
japhb__ is trying to think of a properly sixish and yet still performant and relatively space-efficient way to represent messages in decoded form in memory 02:10
colomon insidescoopsf.sfgate.com/blog/2013/...ween-2013/ 02:12
02:13 risou_awy is now known as risou
colomon BTW, that pony string in STEP encoding (according to my code I"m testing): TH\X2\0318\X0\E\X2\0344\X0\\X2\0309\X0\\X2\0356\X0\ \X2\0360\X0\P\X2\032F\X0\\X2\034D\X0\\X2\032D\X0\O\X2\031A\X0\\X2\200B\X0\N\X2\0310\X0\Y\X2\0321\X0\ 02:20
H\X2\0368\X0\\X2\034A\X0\\X2\033D\X0\\X2\0305\X0\\X2\033E\X0\\X2\030E\X0\\X2\0321\X0\\X2\0338\X0\\X2\032A\X0\\X2\032F\X0\E\X2\033E\X0\\X2\035B\X0\\X2\036A\X0\\X2\0344\X0\\X2\0300\X0\\X2\0301\X0\\X2\0327\X0\\X2\0358\X0\\X2\032C\X0\\X2\0329\X0\
\X2\0367\X0\\X2\033E\X0\\X2\036C\X0\\X2\0327\X0\\X2\0336\X0\\X2\0328\X0\\X2\0331\X0\\X2\0339\X0\\X2\032D\X0\\X2\032F\X0\C\X2\036D\X0\\X2\030F\X0\\X2\0365\X0\\X2\036E\X0\\X2\035F\X0\\X2\0337\X0\\X2\0319\X0\\X2\0332\X0\\X2\031D\X0\\X2\0356\X0\O\X2\036E\X0\\X2\034F\X0\\X2\032E\X0\\X2\032A\X0\\X2\031D\X0\\X2\034D\X0\M\X2\034A\X0\\X2\0312\X0\\X2\031A\X0\\X2\036A\X0\\X2\0369\X0\\X2\036C\X0\\X2\031A\X0\\X2\035C\X0\\X2\0332\X0\\X2\0316\X0\E\X
11\X0\\X2\0369\X0\\X2\034C\X0\\X2\035D\X0\\X2\0334\X0\\X2\031F\X0\\X2\031F\X0\\X2\0359\X0\\X2\031E\X0\S\X2\036F\X0\\X2\033F\X0\\X2\0314\X0\\X2\0328\X0\\X2\0340\X0\\X2\0325\X0\\X2\0345\X0\\X2\032B\X0\\X2\034E\X0\\X2\032D\X0\
kind of ugly, I guess 02:22
japhb__ uh, yeah.
colomon The good news is, it seems to work. 02:23
assuming my test isn't broken. ;)
japhb__ Well, there is that. 02:24
r: say Nil.^mro 02:26
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«Nil (Iterator) (Iterable) (Cool) (Any) (Mu)␤»
02:30 risou is now known as risou_awy 02:35 rurban left
colomon okay, scary bug. One version of the STEP translator handled the unicode well. The other version, instead of converting to utf32 and writing out the corresponding codes (like \X2\033D\X0\), just wrote out the utf8 bytes. The test passed because it round-tripped perfectly, but the actual format of the STEP file it work was quite illegal. :\ 02:44
02:45 fridim_ joined 03:04 tobyink left 03:14 SamuraiJack joined
japhb__ OK, I need to carry around a lot of meta information about all the attributes of each message class. I'm not sure of the best way to do this. One thought is to subclass ClassHOW so that I can build message classes using a subclass of Attribute that has extra storage, and then adding some trait_mod:<is> multis for the new info. 03:17
I don't know if that "Just Works", however.
If it does work, it feels like the most sixish way of doing it. 03:18
Anyone have any better ideas, before I run off and hack some crazy prototype? 03:19
03:22 johnny5_ left
timotimo i'm not sure if it helps, but i've done some class generation at compile time in my ADT module 03:25
03:28 rurban joined 03:29 johnny5_ joined
dwarring Tony the pony: msmvps.com/blogs/jon_skeet/archive/...-fail.aspx 03:37
04:05 cognominal__ left 04:06 cognominal__ joined 04:14 risou_awy is now known as risou
BenGoldberg goodnight 04:20
04:24 BenGoldberg left 04:33 risou is now known as risou_awy 04:52 rurban left 04:53 preflex_ joined, ChanServ sets mode: +v preflex_, preflex left 04:54 preflex_ is now known as preflex 05:02 odoacre joined 05:05 grondilu left
dwarring rn: say "Les Misérables".split("").reverse.reverse.join 05:15
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33, niecza v24-98-g473bd20: OUTPUT«Les Misérables␤»
dwarring :)
rn: say 0.3 == 3/10 05:16
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33, niecza v24-98-g473bd20: OUTPUT«True␤» 05:17
dwarring :) :)
05:23 rurban joined
retupmoca rn: say "Les Misérables".comb.reverse.join 05:28
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33, niecza v24-98-g473bd20: OUTPUT«selbarésiM seL␤»
05:32 rurban left 05:36 djanatyn left 05:37 Tene joined, Tene left, Tene joined 05:38 djanatyn joined 05:41 xalbo joined
dwarring rn: my $s = "Mise" ~ chr(0x301) ~ "rables"; say $s; say $s.comb.reverse.join("") 05:42
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33, niecza v24-98-g473bd20: OUTPUT«Mise􏿽xCC􏿽x81rables␤selbar􏿽xCC􏿽x81esiM␤»
dwarring needs normalization I guess 05:43
05:47 logie joined 05:54 lue left
TimToady needs NFG 05:59
06:09 logie left 06:15 risou_awy is now known as risou 06:25 kaleem joined 06:32 rurban joined 06:33 risou is now known as risou_awy 06:36 rurban left 06:57 darutoko joined 06:59 denis_boyun joined 07:14 nnunley left 07:15 araujo left 07:31 tobyink joined 07:33 rurban joined 07:37 rurban left, tobyink left
masak mooring, #perl6 07:40
FROGGS morning 07:49
07:52 xinming left 07:53 denis_boyun_ joined 07:54 denis_boyun left 08:02 tobyink joined 08:05 muixirt joined
muixirt good morning 08:05
tried to build the recent R* package but it failed because of icu issues 08:07
how do I fix that?
moritz start by nopasting the error messages 08:08
muixirt: do you happen to use a newish Ubuntu?
muixirt moritz: yes 08:09
moritz muixirt: irclog.perlgeek.de/perl6/2013-10-25#i_7763459 describes a workaround
08:10 nnunley joined 08:14 nnunley left 08:16 risou_awy is now known as risou
hoelzro доброе утро, #perl6! 08:16
muixirt thanks moritz that helped 08:17
08:19 zakharyas joined
masak www.aaronbeppu.com/blog/2013-10-30-...pdate.html -- this reminded me of hugme. :) 08:20
masak hugs hugme :)
muixirt first I thought parrot was the guilty part
08:26 sqirrel joined 08:27 lizmat joined 08:32 risou is now known as risou_awy 08:33 rurban joined 08:35 jaffa4 joined 08:39 PacoAir left 08:40 PacoAir joined 08:42 lizmat_ joined 08:43 denis_boyun_ left 08:44 lizmat left 08:47 PacoAir left 08:51 BooK_ joined 08:52 pecastro_ joined, ingyfoo joined 08:53 cxreg2 joined 08:54 robins joined, FOAD_ joined, dmol joined 08:56 szbalint_ joined 08:58 FOAD left, pecastro left, DarthGandalf left, robinsmidsrod left, ggoebel left, tadzik left, BooK left, cxreg left, szbalint left, ingy left, FOAD_ is now known as FOAD 08:59 rurban left, DarthGandalf joined 09:00 xinming joined 09:01 robins is now known as robinsmidsrod 09:05 tadzik joined 09:16 isBEKaml joined
FROGGS moritz / muixirt: I filed a bug report at bugs.launchpad.net/ubuntu/+source/...ug/1247024 09:19
isBEKaml hey guys, I ran into an issue with error reporting today when I was building nqp jvm: gist.github.com/anonymous/7262942
It's funny because I'm still building stage1, where did stage2 come from? The directory doesn't even exist! 09:20
FROGGS stage2 should be in nqp/src 09:22
nvm, the path has changed
isBEKaml Should it be coming from those bootstrapped jars in src/vm/jvm/stage0? If that's the case, directory reported are all wrong. 09:25
FROGGS the reported dir could be reported wrongly, because the shipped stage0 was a stage2 of somebody else 09:27
isBEKaml yes, that's what I think. 09:28
FROGGS still we need to cope with the error
trying now to build an nqp@jvm 09:29
isBEKaml that'll build fine if you're using the shipped Configure.pl and makefiles. I'm using ant since I was not able to install nmake last I tried. 09:30
09:30 lizmat_ is now known as lizmat
isBEKaml I'm building the ant script as I poke around in nqp. :-) 09:30
ant -> ant.apache.org 09:31
the only way to reproduce this would be to make some trivial change in $(J_STAGE1)/$(CORE_SETTING_JAR) so that it fails. Then the stage0 jar will throw this error. 09:32
09:38 woolfy joined
FROGGS ahh 09:40
yes, it built fine 09:41
maybe moritz an explain the changes he made to the build system better
I must go now and will be here in an hour or so 09:42
09:44 lizmat_ joined
isBEKaml Sure, thank you. 09:44
09:46 lizmat left, woolfy left, Rotwang joined 09:51 tobyink left 09:57 xinming_ joined
jnthn -> salzburg for APW 09:57
10:01 xinming left
isBEKaml jnthn: salty fortress? :-) 10:02
10:05 tobyink joined
lizmat_ is off to see more Salzburg sights 10:10
10:17 risou_awy is now known as risou 10:21 woosley left 10:27 isBEKaml left
masak isBEKaml: isn't de:"burg" in this sense en:"city" rather than en:"fortress"? 10:28
I mean, en:"citizen" is de:"burger", so...
10:28 moukeddar joined
masak moukeddar! \o/ 10:30
10:32 preflex left 10:33 risou is now known as risou_awy 10:34 preflex joined, ChanServ sets mode: +v preflex 10:37 tobyink left 10:45 spider-mario joined 10:46 lizmat_ left 10:47 amkrankr1leuen is now known as amkrankruleuen, amkrankruleuen left, amkrankruleuen joined 10:53 kivutar joined 11:16 darutoko left
hoelzro class attributes are implicitly ro, right? 11:17
oh, wait 11:18
I think I misunderstood
'is rw' only applies to the generated methods
hoelzro thought he found a bug
masak hoelzro: right -- the attribute slots are 'rw' by default. accessors are 'readonly' by default. 11:22
hoelzro why is the ro trait named 'readonly'? why not 'ro' to be more consistent with 'rw'? 11:23
hmm
I presume that return-rw does something with containers to implements its semantics? 11:24
hoelzro muses on how to implement triggers when using return-rw for attribute accessors 11:25
masak hoelzro: the official party line is that 'readonly' is more Huffmanly correct than 'ro'. 11:31
moritz readonly is the default for attributes and parameters, so there's not much point shortening it 11:36
when would you even write it explicitly?
masak I, personally, never use it. 11:37
11:47 benabik left
mathw it makes sense. read-write is the exceptional circumstance, read-only is the one you almost never need to explicitly say 11:48
masak the only thing that nags me about it is that people keep asking "why isn't is written 'ro'?", and so it becomes a FAQ. 11:49
and the only reason it's a FAQ rather than being spelled 'ro' is some stubborn notion of Huffmanliness.
huf you cant have it ro! it'll upset your stomach! 11:50
cook it first
mathw personally I think it's easier to figure out what 'rw' means than 'ro'
'is ro', what's ro? my brain goes
probably because I've watched too much Star Trek
11:51 awwaiid joined
masak 'is ro' also has prior art in Moose. 11:53
11:55 daniel-s__ is now known as daniel-s 11:59 rindolf joined 12:13 Rotwang left 12:17 risou_awy is now known as risou 12:18 isBEKaml joined, jaffa4 left, kaleem left
isBEKaml masak: maybe. I just picked it directly off wikipedia. Also, de:salzig means en:salty. :-) 12:19
12:20 pippo joined
pippo nr: my @array; for ^20_000 {@array[$_]=$_}; say @array 12:22
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33, niecza v24-98-g473bd20: OUTPUT«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…»
pippo nr: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',')
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33, niecza v24-98-g473bd20: OUTPUT«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…»
pippo nr: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(',')
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«(timeout)» 12:23
..niecza v24-98-g473bd20: OUTPUT«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 81 82 83 84 85 86 8…»
pippo There is something wrong with that "split" on rakudo jvm. It eats up too much memory and takes to much time 12:26
12:34 risou is now known as risou_awy 12:38 moukeddar left
masak isBEKaml: also, de:salz means en:salt :) 12:38
isBEKaml: hence my pun about Säureburg and Baseburg. 12:39
isBEKaml masak: yessir :-)
sauerburg :) 12:40
masak autopun-ish: twitter.com/schuyler/status/396167385868021761
isBEKaml ehh, who is cmasak and masak on twitter? 12:41
arnsholt masak: That's awesome! =D
isBEKaml ehh, I kinda vaguely get it - given that their entire vocabulary is 23 words or so. :P 12:43
masak isBEKaml: no, that's not it. 12:45
isBEKaml: maybe en.wikipedia.org/wiki/Snowclone will help elucidate the matter.
isBEKaml masak: then I read the wikipedia page on it and the first sentence gave it away. :P 12:46
masak the tweet is more of a self-referential snowclone than an autopun, actually.
isBEKaml yeah, they have something like 50 words for snow. each adding to the other.
so it's difficult to think objectively about any one thing in Inuit/Eskimo language. 12:47
I was looking at this and then it fell in: en.wikipedia.org/wiki/Eskimo_words_for_snow 12:48
masak "Snowclones are memechés, if you will: meme-ified clichés with the operative words removed, leaving spaces for you or the masses to Mad Lib their own versions."
isBEKaml: I don't find it very well-defined to talk about "N words for <anything>" in a strongly agglutinative language. 12:49
it's about as useful as saying "C has 93,741 ways to index the first element of an array". 12:50
12:50 denis_boyun joined
isBEKaml yeah, that's what I meant when I said "each adding to the other" about their words for snow. :-) 12:51
Okay, now back to lurking mode - fiddling with nqp. :D 12:52
12:55 jaffa4 joined
arnsholt IIRC serious people have actually looked into how many distinct roots various Inuit languages have for snow 12:57
It varies between languages, obviously. But apparently it can be framed in a way that makes sense for the languages
masak roots would be more informative. :) 12:58
arnsholt Yeah. As you said, word isn't a very useful entity in polysynthetic languages =) 13:00
At least not in this context
If we're discussing the boundaries between morphology and syntax, they're very interesting indeed =D
13:01 SmokeMachine joined 13:03 logie joined
pippo should'nt this be linear in time with the number of cycles ? "my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(',')" 13:04
masak pippo: yes, it should. 13:06
pippo: maybe there is a Schlemiel in there somewhere.
pippo in rakudo jvm it is not. the above takes far more time than this: "my @array; for ^41 {for ^501 {@array[$_]=$_}; say @array.join(',').split(',')"} 13:07
13:07 rurban joined
pippo masak: what is a "Schlemiel" 13:10
isBEKaml bungler 13:11
13:11 rurban left
masak pippo: "Who is Shlemiel? He's the guy in this joke [...]" www.joelonsoftware.com/articles/fog...00319.html 13:13
[Coke] lizmat++
(mixy)
13:14 ajr joined, ajr is now known as Guest18596
pippo masak: are you referring to me? 13:15
13:15 Guest18596 is now known as ajr_
isBEKaml ayttm? 13:15
masak pippo: no, I was referring to the very specific software mistake of introducing an O(x**2) algorithm inside of a process that need only take O(x) time.
pippo: sorry, I was chunking that whole concept down into that word. 13:16
but the article makes it very clear what I meant. :)
isBEKaml pippo: Schlemiel means "a bungler" - a slang word 13:17
dalek rl6-roast-data: 55d18c6 | coke++ | / (5 files):
today (automated commit)
13:18
rl6-roast-data: 688234a | coke++ | doit:
avoid git failures when extra .out files exist
pippo masak: I am a newbie. I just like perl6 very much and I would like to use it for practical problems. But rakudo JVM cannot even read a cvs file (the rason for the split above) without crashing or taking too much time
[Coke] rakudo-jvm has large startup costs relative to rakudo-parrot, btw. 13:19
pippo It is not the startup that makes me unhappy. 13:20
13:22 denis_boyun left
jaffa4 pippo: life? 13:23
pippo: it has not crashed for me yet
pippo jaffa4: what is the time of execution on your machine for this: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(',') 13:24
jaffa4 quite fast 13:26
pippo on rakudo jvm?
jaffa4 rakudo 10.2013
parrot based
pippo I was referring to jvm? 13:27
I was referring to jvm!
jaffa4 pippo:ok
jvm: print 'hello' 13:28
13:28 tobyink joined
[Coke] rj: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(','); 13:28
camelia rakudo-jvm 882e33: OUTPUT«(timeout)» 13:29
jaffa4 what is rj?
FROGGS rj: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',')
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
FROGGS rj: my @array; for ^20_000 {@array[$_]=$_}; say 1
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
FROGGS :/
[Coke] jaffa4: rakudo-jvm
13:29 salv0 joined
jaffa4 rj: my @array; for 1..20000 {@array[$_]=$_}; say 1 13:30
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
jaffa4 rakudo: my @array; for 1..20000 {@array[$_]=$_}; say 1
camelia rakudo-jvm 882e33: OUTPUT«(timeout)» 13:31
..rakudo-parrot 9d5448: OUTPUT«1␤»
masak jaffa4: you can see it from the camelia output -- when one types 'rj', it replies 'rakudo-jvm'.
pippo rj: say "Hello"
masak (timeout)
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
masak rj: print "(timeout)"
jaffa4 that is cheating 13:32
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
masak rj: sleep 10; print "(timeout)"
moritz it seems that the eval soerver isn't stable enough yet
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
masak :)
moritz j: say 'alive again' 13:33
camelia rakudo-jvm 882e33: OUTPUT«alive again␤»
FROGGS C:\rakudo>perl6 -e "use Shell::Command; mkpath('a\\b')"
Failed to create directory 'a\b/' with mode '0o777': mkdir failed: Das System kann den angegebenen Pfad nicht finden.
something there does not like backslashes 13:34
moritz ouch
isBEKaml eh, what? chmod o777?
moritz windows
no umask set
or something
pippo I think rj has crashed like when executing on my machine when I do "my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(',');"
isBEKaml but it doesn't make sense to use it on windows or even display that, right? 13:35
moritz shrugs
pippo rj: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(','); # this is safe 13:36
camelia rakudo-jvm 882e33: OUTPUT«0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,…»
pippo and not safe when you add the split at the end... 13:37
13:37 logie left
[Coke] (eval server not stable) - wonder if someone else is finally seeing the issue I had with test_summary.pl where it would eventually just stop responding. 13:37
isBEKaml pippo: assign and split? 13:38
pippo this: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(','); <== this split
moritz [Coke]: I've had that too, with spectest 13:39
[Coke] moritz: yay, it's not just me? :|
[Coke] commutes
isBEKaml rj: my @array; for ^20_000 {@array[$_]=$_}; my @temp=@array.join(',');say @temp.split(','); # why not use this? 13:40
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
moritz pippo: yes, the split consumes way too much memory on my machine too
pippo moritz: this is annoying! 13:41
isBEKaml moritz: do you happen to know why this nqp build malfunctions? gist.github.com/anonymous/7262942 13:42
colomon is it just jvm? It seemed like the parrot version also timed out in the backlog...
pippo colomon: I only have jvm and niecza on my machine. 13:43
FROGGS tadzik / jaffa4: see github.com/tadzik/Shell-Command/pull/1 13:45
masak rj: say (^20_000).join(',')
camelia rakudo-jvm 882e33: OUTPUT«0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,…»
colomon hmmm, your 9:38 code seems to run on parrot rakudo in about 3-4 seconds
masak rj: say (^20_000).join(',').split(',')
camelia rakudo-jvm 882e33: OUTPUT«(timeout)» 13:46
13:46 bluescreen10 joined
masak yeah, there's something rotten in the state of .split 13:46
13:46 kivutar left 13:47 kivutar joined
isBEKaml gives up on getting nqp to build 13:48
jaffa4 FROGGS: more about it? 13:49
moritz isBEKaml: where does this invocation come from?
isBEKaml moritz: I don't know - it happens when I try to build core.setting jar in stage1. 13:50
FROGGS jaffa4: this should fix the panda problem on windows, where it was unable to create a directory when installing a module
moritz isBEKaml: this does not look like typical output from make 13:51
isBEKaml moritz: no, that's not make. it's ant.
moritz isBEKaml: well, that's not how you build NPQ :-)
isBEKaml If you look at the comments there, I have tried reproducing that from the cmd prompt. using relative and absolute paths. 13:52
Both failed.
13:52 pernatiy joined
isBEKaml moritz: I know. I can 13:52
moritz isBEKaml: thing is, the NQP build is a bit tricky
using stage0 to build stage1, stage1 to build stage2
FROGGS jaffa4: currently I am creating a new msi, and then try to install modules
moritz and all that without mixing up the stages
isBEKaml moritz: I know. Last I tried, I couldn't install nmake and I'm not looking forward to downloading VS just for nmake. :-) 13:53
13:53 araujo_ joined
isBEKaml moritz: yes, that's exactly how I set it up. What's confusing is, it throws up saying it's from stage2! 13:53
moritz isBEKaml: that's probably because stage0 is a copy from a previous stage2
13:54 nnunley joined
moritz isBEKaml: and the source file statements inside JVM bytecode don't change after compile time 13:54
isBEKaml moritz: stage0 is from the repo. Is that why it's broken?
moritz no
isBEKaml: what I can offer is to give you a log of a successful nqp build with make
isBEKaml: and then you can compare that with the commands you executed, and see where they differ 13:55
isBEKaml moritz: Let me tell you what I'm doing here: I use ant, hand coding the build script.
moritz: and then I compare output from every stage as it happens as I build with ant. (Obviously, makefile build would fail with some stupid error over path name conventions). I then go looking at the makefile to reproduce the same steps with ant. 13:56
tadzik FROGGS: merged
FROGGS jaffa4: progress, at least: ==> Successfully installed Math::RungeKutta
Failed to remove the directory '.work\1383314078_1\lib\Math'
tadzik: thanks! :o)
isBEKaml moritz: so far, I've been successful until I came to hll in stage1. Then somewhere down the path, core.setting got f'd up. :-)
13:57 araujo_ left
moritz isBEKaml: that's why I'm offering you a log from a reference build 13:57
isBEKaml moritz: once I get this building, you can expect a PR with build.xml :-)
moritz so that you can see where you instructions differ from the "offical" build 13:58
isBEKaml moritz: okay, please paste the "official" log. :-)
timotimo rj: say (^20_000).join(",").split(",")[20000] 13:59
moritz makes it
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
timotimo (so not the stringification of th list...)
moritz rakudo-jvm is probably borked again 14:00
14:00 araujo_ joined, araujo_ left
moritz using timeouts of camelia's rakudo-jvm target is NOT a reliable way to gather information 14:00
timotimo rj: say("is this working?")
you seem to be right 14:01
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
pippo timotimo: I think this " say (^20_000).join(",").split(",")[20000]" killed rj 14:02
moritz since the evalserver isn't spawned off of camelia, its resource limits don't apply
14:03 pmurias joined
pmurias what could cause "Missing or wrong version of dependency 'src/stage2/QRegex.nqp'" after updating nqp? 14:03
moritz pmurias: interference with an installed NQP
pmurias with leftover parts of installed NQP perhaps? 14:04
moritz isBEKaml: moritz.faui2k3.org/tmp/j-build.log
pmurias: yes
pmurias: or a missing dependency in the Makefile
isBEKaml moritz: danke! 14:05
pippo rj: say "I am back"
camelia rakudo-jvm 882e33: OUTPUT«I am back␤»
14:06 araujo_ joined
pippo timotimo: if you think it is aproblem of stingification then why this works fine... 14:07
rj: say (^20_000).join(",")
moritz does anybody have an idea for how to create a more robust rakudo-jvm setup for camelia
camelia rakudo-jvm 882e33: OUTPUT«0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,…»
moritz pippo: timotimo merely tested a hypothesis
timotimo because the list version puts a [ and ] at the end? :P
and beginning*
pmurias ahh, the nqp executable has been renamed to nqp-p? 14:14
was renaming (or aliasing) rakudo's executable to jakudo,parakudo,mokudo considered? 14:17
14:18 risou_awy is now known as risou
pippo bie #perl6 14:23
14:23 pippo left, xinming_ left 14:24 araujo_ is now known as araujo 14:25 araujo is now known as Guest54701, Guest54701 is now known as araujo, araujo left, araujo joined, xinming joined
isBEKaml mokudo? 14:29
14:32 SamuraiJack left
pmurias isBEKaml: rakudo on moarvm 14:32
isBEKaml ah 14:34
pmurias seems much easier to type then perl6-moarvm ;)
14:35 risou is now known as risou_awy
isBEKaml that it is. :-) 14:35
isBEKaml afk - be back later. 14:36
14:36 isBEKaml left 14:38 kaare_ joined
colomon moritz++ # building parrakudo and jakudo together now EASY! 14:43
hmmm, the 20,000 split thing takes 9 seconds on perl6-p, 25 on perl6-j. That's a definite weak point for perl6-j, but it doesn't seem like a catastrophe? 14:45
oh, wait!
perl6-j didn't finish it, it took 25 seconds to die with java.lang.StackOverflowError 14:46
okay, catastrophe.
masak :/
did I submit this as a rakudobug?
I don't remember.
oh, I submat rt.perl.org/Ticket/Display.html?id=120380 14:47
14:54 iSlug joined
diakopter show me the split one 15:00
?
jaffa4 I have problem with current compiler 15:01
It gives me too short error message
Could not locate compile-time value for symbol P6
no line number, or column number
Is that normal?
timotimo masak: i seem tor ecall that the performance improved ridiculously when you use [\d+] instead of (\d+), or if you didn't say the match object
masak: if you only say the last submatch of $/, you get the answer almost instantaneously 15:02
rj: my $file = "";
camelia ( no output )
timotimo oops
15:02 kivutar_ joined
timotimo rj: my $file = ""; for ^20000 {$file ~= "$_;"}; say ($file ~~ /(\d+) +% ';'/)[0][20000]; 15:02
camelia rakudo-jvm 882e33: OUTPUT«Nil␤» 15:03
timotimo rj: my $file = ""; for ^20000 {$file ~= "$_;"}; say ($file ~~ /(\d+) +% ';'/)[0][19999];
camelia rakudo-jvm 882e33: OUTPUT«「19999」␤␤»
jaffa4 masak: is it a bug?
timotimo there we go.
rj: my $file = ""; for ^20000 {$file ~= "$_;"}; say $file ~~ /[\d+] +% ';'/);
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/D56RfDAV_F␤Unexpected closing bracket␤at /tmp/D56RfDAV_F:1␤------> e ~= "$_;"}; say $file ~~ /[\d+] +% ';'/⏏);␤»
timotimo rj: my $file = ""; for ^20000 {$file ~= "$_;"}; say $file ~~ /[\d+] +% ';'/;
camelia rakudo-jvm 882e33: OUTPUT«「0;1;2;3;4;5;6;7;8;9;10;11;12;13;14;15;16;17;18;19;20;21;22;23;24;25;26;27;28;29;30;31;32;33;34;35;36;37;38;39;40;41;42;43;44;45;46;47;48;49;50;51;52;53;54;55;56;57;58;59;60;61;62;63;64;65;66;67;68;69;70;71;72;73;74;75;76;77;78;79;80;81;82;83;84;85;86;87;…»
masak jaffa4: I don't know what you did to get that error, but sounds like a possible bug, eys.
yes* 15:04
timotimo "could not locate compile-time value for symbol Foo" usually comes from the optimizer, try --optimize=0
colomon diakopter: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(',');
timotimo or rather --optimize=off
jaffa4 it is easy, clone std.... then execute something like that erl6 -I D:\m\p6\perl6grammer\std\lib -I D:\m\p6\perl6grammer\std D:\m\p6\perl6grammer\std\STD.pm6
15:04 kaleem joined
masak jaffa4: ah. I tried to tell you yesterday, but maybe I didn't get all the way: people don't use STD like that. 15:05
nor are you really expected to.
15:05 kivutar left
masak what you have to understand about STD is that it's all one big glorious hack. 15:05
jaffa4 masakL still what I got, is not ok
masak it is in my book.
STD is kind of using Perl 5 as a trampoline to explore Perl 6 parsing. 15:06
15:06 benabik joined
masak it was never meant to be included from a Rakudo script like you're trying to. 15:06
now, granted, you may have stumbled on an interesting rakudobug nevertheless.
but I predict it will be a big pain to golf :) 15:07
15:07 kivutar joined
jaffa4 I would like to parse perl 6 from perl 6 15:07
Is this such a big dream>
pmurias jaffa4: use the rakudo grammar 15:08
15:08 kaleem left
jaffa4 What rakudo grammar? 15:08
diakopter colomon: I'm curious how nqp-m handles an nqp version of that program 15:09
15:09 risou_awy is now known as risou
diakopter (I think there's a not small possibility it handles it even worse..) 15:10
15:13 ajr joined, ajr_ left
masak jaffa4: you'll find it in the Rakudo repository, in src/Perl6/Grammar.nqp 15:13
15:14 ajr is now known as Guest8772
colomon diakopter: it's pippo++'s code, and I'm sure he wouldn't mind you if translated it to nqp. ;) 15:14
masak jaffa4: but I'm not 100% clear on what pmurias++ means by "use the rakudo grammar". 15:15
jaffa4: I haven't seen anyone doing that, from start to end.
jaffa4: it's a big dream indeed, but a rather attractive one. please keep dreaming in that way :)
(by which I mean, I would like to see it work, too)
15:17 kivutar_ left
pmurias masak: I meant that to parse perl6 he should try to use Perl6::Grammar 15:18
masak pmurias: believe me, I'm with you so far. 15:21
pmurias: do you have a Perl 6 script that uses Perl6::Grammar to parse Perl 6 code?
pmurias so the problem is the Perl6/nqp distinction? 15:22
timotimo isn't there some :from<nqp> adverb for use already?
masak pmurias: the problem is parsing Perl 6 from (Rakudo) Perl 6. 15:23
15:24 jnap joined, denis_boyun joined
diakopter colomon: I'm having serious trouble translating it. 15:27
15:27 Guest8772 is now known as ajr_
FROGGS tadzik: should the calls to rmdir shouldn't be rm_rf instead? github.com/tadzik/Shell-Command/bl...and.pm#L29 15:28
kivutar where is try.rakudo.org/ ? is it dead? 15:33
is it something like tour.golang.org/#1 ? 15:34
moritz kivutar: it's dead, yes :( 15:35
(I know a dead rakudo when I see one) 15:36
15:39 sftp left
tadzik FROGGS: obtw, the pull request is a bit wrong I think 15:39
FROGGS (This is a stiff rakudo)
tadzik: really?
tadzik it created the Shell directpyu
directory
this can't be right
or maybe it was there before
FROGGS tadzik: I don't understand 15:40
tadzik FROGGS: there?s a toplevel Shell directory in the repo noew
* now
timotimo pmurias: when do you have the time to continue working on rakudo-js? 15:41
tadzik I'll take a closer look when I'm on a laptop
FROGGS tadzik: weird, I just used githubs magic, clicked "edit" on the Command.pm, which did a fork and a PR 15:42
tadzik: it was there before
tadzik maybe I screwed something up before then 15:43
FROGGS tadzik: d0442646c482b6b23f0871ea874022d6435b5a35 is the bad one
tadzik dang 15:44
ok= I'll fix it up later today
the local today, at least 15:45
tadzik in california
masak tadzik: how's CA? :)
tadzik veru US-y :) 15:46
FROGGS *g*
nice :o)
tadzik which is just what I wanted :)
masak .oO( we phase-shifted tadzik about 45 degrees west, so his local day is a bit offset right now )
FROGGS have fun in the U, S and A
that is not right 15:47
that is left
15:47 muixirt left
tadzik "west -- it can't be right" 15:50
FROGGS exactly :o)
but: "life is peaceful there" 15:51
15:51 denis_boyun left 15:53 nebuchadnezzar left
japhb__ tadzik: Where in CA are you? 15:56
masak .oO( slightly below the crossbar of the A ) 15:57
15:57 nebuchadnezzar joined
japhb__ FROGGS: The Shell::Command line you pointed to is correct; the line *above* it is wrong, and I already fixed it. Maybe my fix got reverted? 15:58
15:58 iSlug left, kivutar left
FROGGS japhb: what is wrong with the line above? 15:59
japhb__ The result of the map should be sorted before being reversed, so that all directory contents will be gone before the directory is rmdir'ed. 16:00
FROGGS japhb__: that is in the Command.pm under lib/
japhb__ Gah, I wonder if my fix is not on a branch -- visible to panda, but not on Shell-Command's master. 16:01
FROGGS japhb__: there are two Command.pm files, you've seem to edited the correct one
16:03 p5eval left
japhb__ Ah, OK, at least the fix got in there. 16:03
Would you mind applying my fix to the other copy (or figuring out why there even *is* another copy)? I'm going to be AFK for much of the day. 16:04
tadzik japhb__: San Mateo 16:05
FROGGS japhb__: the other one will be removed... currently I'm hunting for a bug on windows
tadzik staying at diakopter++'s 16:06
japhb__: where are you?
dalek kudo-js: a1679de | (Pawel Murias)++ | / (7 files):
Unbitrot to work with the newest nqp.

Adapt to nqp-p, handle QAST::Blocks with declaration_static, add bool_I op.
16:07
pmurias timotimo: you mean when will I have the time?
japhb__ Idea for a useful tool for hunting down performance weirdness in Rakudo ... run a micro-benchmark over and over with sub profiling on, using increasing iteration counts, and look at the collected profiles searching for either A) counts that don't increase linearly with benchmark iterations, or B) subs that take more time *per iteration* as iterations increase.
I saw something similar once for another language, but I forget where.
tadzik: I'm right now in Mountain View (not too far south of you). 16:08
How long are you hear?
FROGGS pmurias: what is your plan to get it into nqp/ ?
japhb__ *here
tadzik about 3 weeks, I'm getting back on like 18th
japhb__ W00t! We should get the gang together!
pmurias timotimo: I plan to resume work on it from now on, it's mostly a matter of managing my time correctly 16:09
japhb__ is almost to bus stop, but will backlog to continue meet-up discussion soon thereafter. :-)
16:09 p5eval joined
retupmoca r: "asdf".index('a').Bool.say; say "asdf".index('x').Bool.say; 16:10
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«False␤False␤True␤»
retupmoca wait, what? 16:11
FROGGS *g*
retupmoca r: "asdf".index('s').Bool.say
pmurias FROGGS: once it fully bootstraps and I figure out how to boostrap things correctly, I think it would make sense to have my own releases on the npm repository so it's something I'll done onse ready
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«True␤»
FROGGS r: "asdf".index('a').Bool.say; "asdf".index('x').Bool.say;
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«False␤False␤»
retupmoca ah, extra say
FROGGS retupmoca: index("a") is zero, so it it false-ish
retupmoca ok
is there a good way to tell the difference between 'substring at index 0' and 'substring not found' ? 16:12
FROGGS check for >= 0 to know success
r: say "asdf".index('x')
retupmoca r: say "asdf".index('x') >= 0
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«(Int)␤»
rakudo-jvm 882e33: OUTPUT«Cannot look up attributes in a type object␤ in sub infix:<>=> at gen/jvm/CORE.setting:4128␤ in block at /tmp/Salw33Ls9O:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1086␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1292␤ in any command_eval…»
..rakudo-parrot 9d5448: OUTPUT«Invocant requires an instance, but a type object was passed␤ in method Bridge at gen/parrot/CORE.setting:4214␤ in sub infix:<>=> at gen/parrot/CORE.setting:4139␤ in sub infix:<>=> at gen/parrot/CORE.setting:3991␤ in block at /tmp/PX0tcoTOvd:…»
FROGGS r: say "asdf".index('a')
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«0␤»
FROGGS ahh 16:13
it is "undef" for the not-found case
r: say "asdf".index('x').defined
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«False␤»
FROGGS r: say "asdf".index('a').defined
camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«True␤»
retupmoca ah, there we go
.defined is what I was looking for 16:14
S32 was a little unclear on how failure is represented 16:16
japhb__ So tadzik, diakopter, TimToady, anyone else in the SF Bay Area next week: Care for lunch at the Googleplex? 16:23
16:25 jeffreykegler joined 16:27 sftp joined 16:28 lizmat joined
FROGGS retupmoca: improve it :o) 16:30
16:32 woolfy joined, jaffa4 left 16:42 tobyink left 16:45 ajr_ left
tadzik japhb__: I don't see why not :) 16:50
16:58 colomon left
japhb tadzik et al., How does Monday sound? 17:04
tadzik I don't have plans for monday :) 17:07
so, good
diakopter you don't? :) 17:10
oh wait.
tadzik I do? :)
diakopter backlogged
japhb__: sure
japhb Excellent, anyone else interested? 17:11
17:19 kaare_ left, rurban joined 17:26 colomon joined 17:29 colomon left 17:30 ajr joined, ajr is now known as Guest14420 17:31 Guest14420 is now known as ajr_, woolfy left 17:32 lizmat left
FROGGS japhb: I am, but I am still in ol' Europe 17:34
tadzik quick, come here! 17:35
japhb Dangit FROGGS, jump on a plane! ;-) 17:36
17:36 colomon joined
FROGGS I would like to, problem is, I have to care about my kids next week (besides not having enough money) 17:36
japhb yoleaux, where are you when I need you? 17:37
17:37 BitPuffin left
colomon tadzik: you're in California? 17:37
japhb FROGGS, totally understood. If you do end up coming this way sometime, ping me. :-) 17:38
FROGGS I will :o)
tadzik / jaffa4: btw, panda finally works now (again) on windows 17:39
tadzik colomon: yes :) 17:40
FROGGS tadzik: problem was, "you" opened .pm files to look up foreign module usages, but you didn't close the FH
colomon tadzik: \o/ so glad you made it out to the states!
cxreg2 "While the JVM backend mostly implements the same features as the Parrot backend, many bits are still missing, most prominently the native call interface" # wait, is that still true?
17:40 cxreg2 is now known as cxreg
FROGGS tadzik: so it was unable to delete the file, and died later when trying to delete the dir 17:40
tadzik ah
colomon: yeah, me too :)
I dramatically needed some time off
cxreg i love the "Flip flops can be negated." change in the spec. "AUTUMN casts COLD WEATHER. Your FLIP FLOPS are NEGATED" 17:44
FROGGS japhb++ # you already fixed what I recognized on windows! github.com/tadzik/panda/commit/d8b...055d3be5fb 17:45
tadzik: do we need to update panda in star? 17:46
tadzik FROGGS: I don't know, do we? :P 17:47
FROGGS hmmm, that is weird
tadzik: the submodule commit-id is new enough, but the star release contains an older version 17:48
colomon tadzik: panda hasn't worked (on Linux) for the smoke tester in a week. 17:49
17:50 rindolf left
tadzik colomon: I just noticed that rakudobrew didn't build it for me, checking why 17:50
17:52 rurban left, rurban joined
colomon huh. 17:53
I've been getting
===SORRY!===
Could not find Panda in any of: /home/smoker/rakudo/install/lib/parrot/5.9.0-devel/languages/perl6/site/lib, /home/smoker/rakudo/install/lib/parrot/5.9.0-devel/languages/perl6/vendor/lib, /home/smoker/rakudo/install/lib/parrot/5.9.0-devel/languages/perl6/lib, /home/smoker/.perl6/2013.10-120-g9d54481/lib
but just now I got
build stage failed for File::Tools: Failed building lib/File/Find.pm
==> Please make sure that /home/smoker/rakudo/install/lib/parrot/5.9.0-devel/languages/perl6/site/bin is in your PATH
(and yes, that is in my path) 17:54
FROGGS colomon: maybe you are missing this commit? github.com/tadzik/panda/commit/9f2...b8480dd1c2
colomon that does seem like a significant commit!
FROGGS it does 17:55
colomon I'm actually trying pandabrew rebootstrap when I get these messages.
which I think should mean I've got the latest panda?
FROGGS I dunno how that magic works 17:56
colomon oh, maybe not! 17:57
FROGGS damn, the star release really contains old modules :/
17:58 zakharyas left
colomon crap, rebootstrap doesn't try to update panda at all 17:59
17:59 rurban left
FROGGS example: moritz++ did a Hash.delete fix a month ago, which is not included in the star tgz file... even when the submodule hash-ids are up-to-date 17:59
colomon my pandabrew panda dated back to YAPC::NA :\ 18:00
FROGGS O.o 18:01
18:03 benabik left
colomon so, tadzik: shouldn't pandabrew rebootstrap get the latest panda? 18:07
18:09 lue joined
lue hello world o/ 18:09
japhb o/ 18:10
18:13 darutoko joined
colomon after updating pandabrew's panda by hand, smoke tester appears to be working again 18:14
18:20 bluescreen10 left
FROGGS ool 18:22
cool
18:23 bluescreen10 joined
dalek p: 149a377 | (Pawel Murias)++ | Configure.pl:
Make Configure.pl --help use the real list of backends.
18:24
18:28 dansamo joined
japhb OK, so far it looks like we've got 5-6 people coming for Monday lunch at Google. Anyone else interested? 18:36
18:37 jeffreykegler left
moritz interested, but not anywhere near :-) 18:47
18:47 eternaleye left
timotimo pmurias: oh cool! i thought you had other stuff in your life from now on 18:47
18:52 SamuraiJack joined 18:56 bluescreen10 left 18:57 bluescreen10 joined, bluescreen10 left, bluescreen10 joined 19:05 sqirrel left
moritz any p6 or nqp LHFs I could hack on tonight? 19:10
19:11 SamuraiJack left
diakopter moritz: might be nice to have the remaining required moar ops stubbed. would certainly be helpful. 19:12
FROGGS masak++ # very entertaining: www.joelonsoftware.com/articles/fog...00319.html 19:13
diakopter moritz: jnthn did most of them for rakudo in the moar-support branch.. but most of them will need op entries in moarvm proper (oplist), OR in the qast->mast compiler in nqp 19:14
FROGGS .oO( nqp$ prove -v -e nqp-p t/docs/opcodes.t | grep "not ok" )
diakopter moritz: er, the 2nd "them" there is the remaining ones for moar
FROGGS moritz: and there is still RT, if you don't feel op-ish :o) 19:15
diakopter moritz: these definitely need native moar ops gist.github.com/diakopter/7270438 19:16
er, skip the nativecall ones though
editing gist
19:17 slava-work joined
diakopter moritz: (native moar ops as opposed to ones with non-core-mappings in the qast->mast compiler) 19:17
slava-work I hate silly people 19:22
FROGGS good day to you too slava-work :o) 19:23
slava-work :(
it would be a good day if I had rakudo-moarvm to play with, *hint* *hint* *wink* *wink*, jnthn :) 19:24
FROGGS hehe
slava-work does p6 have the whole before/after method keywords that moose has? (wondering if they lifted it out or invented it) 19:25
because that is seriously some awesome stuff
19:26 risou is now known as risou_awy, risou_awy is now known as risou
FROGGS slava-work: if it not in "core", you could implement it (as a module) with, like, 3 lines each for before and after 19:26
because you can create a trait, that wraps the thing (sub) it is applied to, and does something before or after for example 19:27
19:28 darutoko left
PerlJam those would read weird as traits 19:29
slava-work FROGGS: ty 19:30
FROGGS PerlJam: maybe
19:31 bluescreen10 left, sqirrel joined
slava-work I kind of realised how much that is used in lots of places of "I need this from base class, but to pre-treat the data" or something of that sort, so in java, you'd have to use composition, where inheritance makes more sense otherwise 19:31
19:31 tobyink joined, bluescreen10 joined 19:32 colomon left
FROGGS or make a role that you mix into the sub-object... that should work too 19:32
slava-work yes 19:33
FROGGS in the end you wanna do this? before &sub-name, { <code goes here> }
slava-work I knew I was right for liking perl and hating java :P
FROGGS *g* 19:34
slava-work pretty much
PerlJam might as well just call .wrap
slava-work I am also not a big fan of java annotations, which allow spring to automagically link things together
19:36 colomon joined 19:39 colomon left
Ulti grrr OSX Mavericks has lost its OpenGL headers so my build is failing 19:39
FROGGS rp: sub test() { say 42 }; &test.wrap: -> { say "before"; callsame }; test
camelia rakudo-parrot 9d5448: OUTPUT«before␤42␤»
FROGGS PerlJam++
rp: sub test() { say 42 }; &test.wrap: -> { callsame; say "after" }; test 19:40
camelia rakudo-parrot 9d5448: OUTPUT«42␤after␤»
19:40 colomon joined
FROGGS rp: sub test() { say 42 }; &test.wrap: -> { say "before"; callsame }; &test.wrap: -> { say "even beforer"; callsame }; test 19:40
camelia rakudo-parrot 9d5448: OUTPUT«even beforer␤before␤42␤»
FROGGS jaffa4: please test froggs.de/perl6/rakudo-star-2013.10-RC1.msi (other testers welcome) 19:44
19:44 risou is now known as risou_awy 19:50 sqirrel left 19:59 Rotwang joined
Ulti why does parrot even have a binding to OpenGL? 20:00
moritz because that way, all the dozens of HLLs based on parrot can use OpenGL 20:01
Ulti :/ 20:03
lue Ulti: don't you know about the OpenGL object in Perl 6? It's the best thing ever. :P 20:05
Ulti it is kind of lame Apple randomly moved all framework headers into the XCode.app directory
lue no, but I'd prefer it was an OpenCL Promise object :P 20:06
rakudo-fpga why go virtual machine when you can have hardware 20:07
lue
.oO(rakudo-2A03)
Ulti arent 2A03 incredibly hard to acquire thanks to chiptune mayhem? 20:08
lue The 2A03 specifically? Maybe, but the genericer 6502 should be incredibly easy. 20:10
s/genericer/genericker/ # pronunciation rules
Ulti :S
uh oh its not just paths its the whole API has changed and the bindings are wrong win win win
moritz after some more thinking and tinkering, I'm pretty sure I don't know how I would patch rakudo for "no strict" 20:13
20:15 PacoAir joined
lue moritz: ooc, is it just making the phrase "no strict" work, or do you have to implement its effects too? 20:16
moritz lue: just the latter 20:17
lue Doesn't -e already do (some of) that?
moritz no
it should, but doesn't
FROGGS shall I give it a whirl? 20:18
moritz yes please
there's one more thing I'm trying to do, but don't wait for it 20:19
20:23 dansamo left
diakopter wouldn't you just auto-declare a lexical in just the same place it currently errors 20:26
arnsholt My goal for the evening: Implement CStr on JVM 20:28
FROGGS diakopter: that is what v5 does
arnsholt And if there's time after that, CArrays of value types
moritz diakopter: TimToady explictly said that's not what happens
rather it's a direct lookup into the symbol table
jnthn slava-work: Some day in the not too distant future, you will...as for today, I need to write my talk for Austrian Perl Workshop tomorrow! ;) 20:29
moritz: Anything I can give a hint on?
FROGGS moritz: well, you have to declare it in order looking it up, no?
moritz p5eval: $blerg = 42; { use strict; say $blerg }
p5eval moritz: ERROR: Global symbol "$blerg" requires explicit package name at (eval 7) line 1.
FROGGS well, this would error out this way too 20:30
hmm, no, maybe not
moritz not if I'd autodeclare $blerg
jnthn: how would the generated code look like? 20:32
jnthn: an anonymous variable or container bound to the slot of the symbol table?
'cause just binding into the symbol table doesn't seem to be enough (segfaults rakudo :-) 20:33
FROGGS moritz: gimme 10min to try my v5 patch
jnthn moritz: Hang on... :) What are the semantics you're after?
our-like? 20:34
moritz jnthn: an undeclared variable under 'no strict;' should not declare a user-visible variable, but point straight into the symbol table
jnthn moritz: Of the current package? 20:36
moritz jnthn: yes
basically OUR::<$x>
jnthn QAST::Op.new( :op('callmethod'), :name('at_key'), QAST::WVal.new( :value($*PACKAGE) ), QAST::SVal.new( :value($varname) ) ) # and rely on auto-viv...maybe... 20:37
oh, oops
That WVal wantsa QAST::Op.new( :op('who'), ... ) around it
moritz oh
good idea :-)
and I used :op<atkey>, which doesn't autovivify 20:38
diakopter moritz: oh; I missed that 20:41
well you have to declare at it least somewhere, even if every lookup/store of it is through the symbol table, right? 20:42
moritz why?
jnthn not afaik 20:45
20:45 lowpro30 joined
moritz I've managed to stop the declaration from happening, but the lookup is still for a lexical variable 20:47
and that segfaults, because there's no such var 20:48
diakopter
.oO( segfault? )
20:50
20:50 denis_boyun joined
diakopter should transform every occurrence in that package to Foo::<foo> or whatever? 20:50
moritz yes, that's how I understand it 20:51
20:51 PacoAir left 20:54 lowpro30 left 20:57 colomon left
moritz calls it a day 21:00
dalek kudo/nom: 3f92477 | jonathan++ | src/vm/jvm/core/Threading.pm:
Various combinators for Subscribables.
kudo/nom: 2a0f6c9 | jonathan++ | src/vm/jvm/core/Threading.pm:
Naming cleanup in Publish.for implementation.
ast: dc0b47e | jonathan++ | S17-concurrency/subscribable.t:
Tests for some Subscribable combinators.
21:01
jnthn L'hacking de tren...
FROGGS froggs@TL02L-R8RXTCW-linux:~/dev/rakudo$ perl6-p -e 'use 6; $foo = 42; say $foo'
42
froggs@TL02L-R8RXTCW-linux:~/dev/rakudo$ perl6-p -e 'use v6; $foo = 42; say $foo'
===SORRY!=== Error while compiling -e
I guess this is a problem then: 21:02
perl6-p -e 'use 6; $foo = 42; { use v6; say $foo }'
42
(and yes, it should be "6" rather than "use 6") 21:03
that is what I have: gist.github.com/FROGGS/6f8989611659551dae1f 21:09
dunno what would be needed to make 'use 6; $foo = 42; { use v6; say $foo }' throw 21:10
21:18 bluescreen10 left, bluescreen10 joined, bluescreen10 left
tipdbmp If Array extends List, how come when I: use MONKEY_TYPING; augment class List { method sum { [+] self; } }; the expression [1 .. 10].sum.say dies with 'no such method' error? 21:22
FROGGS jnthn: is there already a way to check in the gramar's comp_unit if we are running as -e? 21:23
tipdbmp: that is because the method cache of Array does not get updated 21:24
pmurias jnthn: if I change nqp-js to use the multi-backend-build-system should I keep it in a nqp branch or a github fork?
FROGGS tipdbmp: you can work-around it by doing "augment class Array { };" after the List...
tipdbmp FROGGS, so I should augment both? 21:25
FROGGS tipdbmp: you have to to renew the method cache
tipdbmp Oh, I see.
FROGGS tipdbmp: but for Array you don't actually have to change a method
21:25 colomon joined
FROGGS an empty augment will do 21:25
tipdbmp Yes, it worked. Thanks. 21:26
jnthn FROGGS: I think (yes, ugh) that $?FILE or whatever ends up as -e :)
FROGGS jnthn: ahh, yes!
rp: say "$?FILE"
camelia rakudo-parrot 9d5448: OUTPUT«/tmp/XVplQXYf2S␤»
jnthn pmurias: Branch is fine, that's how JVM and MoarVM happened before landing in master
FROGGS err, okay... so I have to wait until my local one compiles :o) 21:27
21:27 risou_awy is now known as risou
arnsholt Rakudo's make j-install doesn't seem to work on my machine 21:30
Is that a known thing?
FROGGS unknown to me at least 21:31
arnsholt Creating '/nqp/install/bin/perl6-j' 21:33
open: No such file or directory at tools/build/create-jvm-runner.pl line 35.
moritz: You seen that happen?
My config.default is: 21:34
--backends=jvm,parrot
--prefix=../nqp/install
FROGGS perl6 -e '$foo = 42; { use v6; say $foo }' # I don't see why this should throw in the strict-enabled block 21:38
21:45 risou is now known as risou_awy 22:03 lowpro30 joined 22:05 PacoAir joined 22:07 PacoAir left 22:08 PacoAir joined 22:12 tobyink left 22:14 lowpro30 left 22:27 rurban joined 22:30 BenGoldberg joined 22:31 colomon left 22:38 xenoterracide left, colomon joined
japhb Ulti: Parrot has an OpenGL binding because I wrote it when I was still hoping that Parrot would build momentum and actually *have* the multiple languages interoperating that it was supposed to be for, and that moritz is teasing about. 22:41
But Tene and I discovered that language interop didn't get quite the support we'd both hoped. 22:42
Tene Yes, unfortunately.
japhb Yes! I got another person up for lunch on Monday. 22:43
22:45 denis_boyun left
BenGoldberg I would hate to sleep past lunch, on a Monday ;) 22:46
japhb heh 22:48
22:57 rurban left
tadzik colomon: oh, I don't think pandabrew does that, no 23:00
colomon tadzik: I don't know that rebootstrap should be the one to do that, but there definitely needs to be some way to update pandabrew's version of panda from a pandabrew command. 23:03
tadzik colomon: fwiw, I'm not really maintaining pandabrew anymore 23:04
the reason for it was for panda to self-update after rakudo update, and now rakudobrew autoupdates panda after each build
colomon is rakudobrew ready for prime time?
tadzik I thik so
I use it without any problems
like, I don't build rakudos outside rakudobrew anymore 23:05
23:07 moartl joined
japhb That's certainly a positive endorsement 23:07
23:07 moartl left
jnthn heh, if Moar's startup time sucked then imoartal would be a good name for an eval server.. 23:08
diakopter :D 23:11
moritz arnsholt: what --prefix did you configure? 23:12
arnsholt: /nqp/install/... looks slightly wrong to me 23:13
23:16 johnny5_ left 23:17 spider-mario left, johnny5_ joined 23:20 jnap left 23:22 johnny5_ left 23:23 johnny5_ joined 23:24 johnny5_ left
lue jnthn: a sucky startup time would warrant "moaratorium", surely :) 23:25
23:25 johnny5_ joined 23:27 risou_awy is now known as risou 23:28 labster left 23:29 lizmat joined 23:30 woolfy joined
moritz a slow startup would moartify the authors 23:31
japhb moritz: Yes, precisely: the authors would die of boredom. 23:33
Ulti japhb which planned target languages have graphics tightly integrated? I've only used a couple like Icon and I guess Logo 23:34
japhb There are a number of those now: PostScript, Processing, JavaScript (Canvas/WebGL), and so on. But that wasn't actually why I did it. 23:36
I wanted all languages to have a good OpenGL binding available "for free" so that users of more classically textual languages could start thinking graphically. 23:37
FROGGS moritz: I think that works gist.github.com/FROGGS/6f8989611659551dae1f
japhb Just like SDL + OpenGL + OpenAL = suddenly people go "Hey, I'm going to write a game!"
FROGGS moritz: at least it survived basic tests
23:38 woolfy left
FROGGS gnight all 23:38
Ulti japhb so parrot provides an abstraction of OpenGL?
moritz FROGGS: doesn't $*W.is_lexical suffice? 23:39
FROGGS no, I need to stuff the <lax> somewhere
japhb Ulti: The idea was that any language written on top of parrot could use any library that was bound for Parrot just once.
Library/API binding becomes an M+N instead of M*N problem. 23:40
FROGGS moritz: and I didn't want it to loop over the blocks twice
japhb That was the whole reason I worked on Parrot.
23:41 ssutch_ joined
ssutch_ chello 23:41
japhb I'm sure ingy would tell me it was some form of acmeism. :-)
utf8hello
23:43 risou is now known as risou_awy
dalek rl6-roast-data: a8fb2b7 | coke++ | / (5 files):
today (automated commit)
23:44
23:57 dmol left 23:59 Rotwang left