»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
00:46 rurban left
dalek kudo/nom: cf0b0ca | jonathan++ | src/vm/jvm/core/Threading.pm:
Better enforce Exception object coercion.
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).
-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 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
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
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++
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)
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)»
[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