Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2013-11-01

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:02 dansamo joined #perl6
00:06 japhb__ jnthn: excellent.
00:12 Psyche^ joined #perl6
00:25 benabik joined #perl6
00:42 rurban joined #perl6
00:45 dalek nqp: 183b8af | dwarring++ | examples/rubyish/ (2 files):
00:45 dalek nqp: rubyish =begin .. =end multiline comments. simplify slurpy hashes
00:45 dalek nqp: review: https://github.com/perl6/nqp/commit/183b8af658
00:56 dalek rakudo/nom: cf0b0ca | jonathan++ | src/vm/jvm/core/Threading.pm:
00:56 dalek rakudo/nom: Better enforce Exception object coercion.
00:56 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/cf0b0ca163
00:56 dalek rakudo/nom: 3208450 | jonathan++ | src/vm/jvm/core/Threading.pm:
00:56 dalek rakudo/nom: Use Promise::Keeper in Channel.
00:56 dalek rakudo/nom:
00:56 dalek rakudo/nom: To be more robust.
00:56 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/3208450d36
00:56 dalek rakudo/nom: 9d54481 | jonathan++ | src/vm/jvm/core/Threading.pm:
00:56 dalek rakudo/nom: Add Subscribable, and initial publish bits.
00:56 dalek rakudo/nom:
00:56 dalek rakudo/nom: This is the first piece towards enabling reactive programming with
00:56 dalek rakudo/nom: push-collections, a la Rx. Many combinators to come.
00:56 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/9d544811ea
00:56 dalek roast: b9d4b05 | jonathan++ | S17-concurrency/subscribable.t:
00:56 dalek roast: Add basic Subscribable tests.
00:56 dalek roast: review: https://github.com/perl6/roast/commit/b9d4b056eb
00:59 jnthn 'night
01:02 rurban joined #perl6
01:11 benabik joined #perl6
01:12 dalek p6-pb: 5f9f01b | (Geoffrey Broadwell)++ | / (2 files):
01:12 dalek p6-pb: Add PB::Binary::WireTypes with tests
01:12 dalek p6-pb:
01:12 dalek p6-pb: Put WireType enumeration in a common place, with tests to make sure the
01:12 dalek p6-pb: values don't accidentally change (they're part of the PB standard).
01:12 dalek p6-pb: review: https://github.com/samurais​am/p6-pb/commit/5f9f01bac4
01:12 dalek p6-pb: ad86091 | (Geoffrey Broadwell)++ | / (4 files):
01:12 dalek p6-pb: Switch to using PB::Binary::WireTypes
01:12 dalek p6-pb:
01:12 dalek p6-pb: Move code and tests using either WIRE_TYPE_* or hard-coded numbers to
01:12 dalek p6-pb: the WireType::* enum instead.
01:12 dalek p6-pb: review: https://github.com/samurais​am/p6-pb/commit/ad860913c3
01:21 PacoAir joined #perl6
01:33 Vendethi_ joined #perl6
01:39 dwarring joined #perl6
01:41 dwarring has the module smoke server gone out? I've been using http://97.95.33.142:3001/report
01:41 colomon joined #perl6
01:41 dwarring no results since 25th Oct
01:46 colomon panda has been broken
02:02 colomon r: say "TH̘Ë͖́̉ ͠P̯͍̭O̚​N̐Y̡ H̸̡̪̯ͨ͊̽̅̾̎Ȩ̬̩̾͛ͪ̈́̀́͘ ̶̧̨̱̹̭̯ͧ̾ͬC̷̙̲̝͖ͭ̏ͥͮ͟Oͮ͏̮̪̝͍M͊̒ͪ̚​̲̖ͩͬ̚͜Ȇ̴̟̟͙̞ͩ͌͝S̨̥̫͎̭ͯ̿̔̀ͅ".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…»
02:02 dwarring heh
02:04 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.
02:04 japhb__ That qualifies.
02:07 japhb__ 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 japhb__ I figure if those both roundtrip, there's half the battle right there.
02:07 colomon That's essentially what I'm looking to do here, too.
02:08 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.
02:10 * 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
02:12 colomon http://insidescoopsf.sfgate.com/blog​/2013/10/31/chicago-restaurant-dress​es-up-as-alinea-wins-halloween-2013/
02:20 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 colomon 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\\X​2\0327\X0\\X2\0358\X0\\X2\032C\X0\\X2\0329\X0\
02:20 colomon \X2\0367\X0\\X2\033E\X0\\X2\036C\X0\\X2\0327\X0\\​X2\0336\X0\\X2\0328\X0\\X2\0331\X0\\X2\0339\X0\\X​2\032D\X0\\X2\032F\X0\C\X2\036D\X0\\X2\030F\X0\\X​2\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\0​31A\X0\\X2\035C\X0\\X2\0332\X0\\X2\0316\X0\E\
02:20 colomon 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\0​345\X0\\X2\032B\X0\\X2\034E\X0\\X2\032D\X0\
02:22 colomon kind of ugly, I guess
02:22 japhb__ uh, yeah.
02:23 colomon The good news is, it seems to work.
02:23 colomon assuming my test isn't broken.  ;)
02:24 japhb__ Well, there is that.
02:26 japhb__ r: say Nil.^mro
02:26 camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«Nil (Iterator) (Iterable) (Cool) (Any) (Mu)␤»
02:45 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.  :\
02:45 fridim_ joined #perl6
03:14 SamuraiJack joined #perl6
03:17 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 japhb__ I don't know if that "Just Works", however.
03:18 japhb__ If it does work, it feels like the most sixish way of doing it.
03:19 japhb__ Anyone have any better ideas, before I run off and hack some crazy prototype?
03:25 timotimo i'm not sure if it helps, but i've done some class generation at compile time in my ADT module
03:28 rurban joined #perl6
03:29 johnny5_ joined #perl6
03:37 dwarring Tony the pony: http://msmvps.com/blogs/jon_skeet/archive/2009​/11/02/omg-ponies-aka-humanity-epic-fail.aspx
04:06 cognominal__ joined #perl6
04:20 BenGoldberg goodnight
04:53 preflex_ joined #perl6
05:02 odoacre joined #perl6
05:15 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␤»
05:15 dwarring :)
05:16 dwarring rn: say 0.3 == 3/10
05:17 camelia rakudo-parrot 9d5448, rakudo-jvm 882e33, niecza v24-98-g473bd20: OUTPUT«True␤»
05:17 dwarring :) :)
05:23 rurban joined #perl6
05:28 retupmoca rn: say "Les Misérables".comb.reverse.join
05:28 camelia rakudo-parrot 9d5448, rakudo-jvm 882e33, niecza v24-98-g473bd20: OUTPUT«selbarésiM seL␤»
05:37 Tene joined #perl6
05:37 Tene joined #perl6
05:38 djanatyn joined #perl6
05:41 xalbo joined #perl6
05:42 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«Misérables␤selbaŕesiM␤»
05:43 * dwarring needs normalization I guess
05:47 logie joined #perl6
05:59 TimToady needs NFG
06:25 kaleem joined #perl6
06:32 rurban joined #perl6
06:57 darutoko joined #perl6
06:59 denis_boyun joined #perl6
07:31 tobyink joined #perl6
07:33 rurban joined #perl6
07:40 masak mooring, #perl6
07:49 FROGGS morning
07:53 denis_boyun_ joined #perl6
08:02 tobyink joined #perl6
08:05 muixirt joined #perl6
08:05 muixirt good morning
08:07 muixirt tried to build the recent R* package but it failed because of icu issues
08:07 muixirt how do I fix that?
08:08 moritz start by nopasting the error messages
08:08 moritz muixirt: do you happen to use a newish Ubuntu?
08:09 muixirt moritz: yes
08:09 moritz muixirt: http://irclog.perlgeek.de/​perl6/2013-10-25#i_7763459 describes a workaround
08:10 nnunley joined #perl6
08:16 hoelzro доброе утро, #perl6!
08:17 muixirt thanks moritz that helped
08:19 zakharyas joined #perl6
08:20 masak http://www.aaronbeppu.com/blog/2013​-10-30-feel-better-bot-update.html -- this reminded me of hugme. :)
08:20 * masak hugs hugme :)
08:20 muixirt first I thought parrot was the guilty part
08:26 sqirrel joined #perl6
08:27 lizmat joined #perl6
08:33 rurban joined #perl6
08:35 jaffa4 joined #perl6
08:40 PacoAir joined #perl6
08:42 lizmat_ joined #perl6
08:51 BooK_ joined #perl6
08:52 pecastro_ joined #perl6
08:52 ingyfoo joined #perl6
08:53 cxreg2 joined #perl6
08:54 robins joined #perl6
08:54 FOAD_ joined #perl6
08:54 dmol joined #perl6
08:56 szbalint_ joined #perl6
08:59 DarthGandalf joined #perl6
09:00 xinming joined #perl6
09:05 tadzik joined #perl6
09:16 isBEKaml joined #perl6
09:19 FROGGS moritz / muixirt: I filed a bug report at https://bugs.launchpad.net/ub​untu/+source/icu/+bug/1247024
09:19 isBEKaml hey guys, I ran into an issue with error reporting today when I was building nqp jvm: https://gist.github.com/anonymous/7262942
09:20 isBEKaml It's funny because I'm still building stage1, where did stage2 come from? The directory doesn't even exist!
09:22 FROGGS stage2 should be in nqp/src
09:22 FROGGS nvm, the path has changed
09:25 isBEKaml Should it be coming from those bootstrapped jars in src/vm/jvm/stage0? If that's the case, directory reported are all wrong.
09:27 FROGGS the reported dir could be reported wrongly, because the shipped stage0 was a stage2 of somebody else
09:28 isBEKaml yes, that's what I think.
09:28 FROGGS still we need to cope with the error
09:29 FROGGS trying now to build an nqp@jvm
09:30 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 isBEKaml I'm building the ant script as I poke around in nqp. :-)
09:31 isBEKaml ant -> http://ant.apache.org
09:32 isBEKaml 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.
09:38 woolfy joined #perl6
09:40 FROGGS ahh
09:40 FROGGS yes, it built fine
09:41 FROGGS maybe moritz an explain the changes he made to the build system better
09:42 FROGGS I must go now and will be here in an hour or so
09:44 lizmat_ joined #perl6
09:44 isBEKaml Sure, thank you.
09:46 Rotwang joined #perl6
09:57 xinming_ joined #perl6
09:57 * jnthn -> salzburg for APW
10:02 isBEKaml jnthn: salty fortress? :-)
10:05 tobyink joined #perl6
10:10 * lizmat_ is off to see more Salzburg sights
10:21 woosley left #perl6
10:28 masak isBEKaml: isn't de:"burg" in this sense en:"city" rather than en:"fortress"?
10:28 masak I mean, en:"citizen" is de:"burger", so...
10:28 moukeddar joined #perl6
10:30 masak moukeddar! \o/
10:34 preflex joined #perl6
10:44 spider-mario joined #perl6
10:47 amkrankruleuen joined #perl6
10:53 kivutar joined #perl6
11:17 hoelzro class attributes are implicitly ro, right?
11:18 hoelzro oh, wait
11:18 hoelzro I think I misunderstood
11:18 hoelzro 'is rw' only applies to the generated methods
11:18 * hoelzro thought he found a bug
11:22 masak hoelzro: right -- the attribute slots are 'rw' by default. accessors are 'readonly' by default.
11:23 hoelzro why is the ro trait named 'readonly'? why not 'ro' to be more consistent with 'rw'?
11:23 hoelzro hmm
11:24 hoelzro I presume that return-rw does something with containers to implements its semantics?
11:25 * hoelzro muses on how to implement triggers when using return-rw for attribute accessors
11:31 masak hoelzro: the official party line is that 'readonly' is more Huffmanly correct than 'ro'.
11:36 moritz readonly is the default for attributes and parameters, so there's not much point shortening it
11:36 moritz when would you even write it explicitly?
11:37 masak I, personally, never use it.
11:48 mathw it makes sense. read-write is the exceptional circumstance, read-only is the one you almost never need to explicitly say
11:49 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 masak and the only reason it's a FAQ rather than being spelled 'ro' is some stubborn notion of Huffmanliness.
11:50 huf you cant have it ro! it'll upset your stomach!
11:50 huf cook it first
11:50 mathw personally I think it's easier to figure out what 'rw' means than 'ro'
11:50 mathw 'is ro', what's ro? my brain goes
11:50 mathw probably because I've watched too much Star Trek
11:51 awwaiid joined #perl6
11:53 masak 'is ro' also has prior art in Moose.
11:59 rindolf joined #perl6
12:18 isBEKaml joined #perl6
12:19 isBEKaml masak: maybe. I just picked it directly off wikipedia. Also, de:salzig means en:salty. :-)
12:20 pippo joined #perl6
12:22 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…»
12:22 pippo nr: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',')
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​,16,17,18,19,20,21,22,23,24,25,26,27,28,29,3​0,31,32,33,34,35,36,37,38,39,40,41,42,43,44,​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…»
12:22 pippo nr: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(',')
12:23 camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«(timeout)»
12:23 camelia ..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…»
12:26 pippo There is something wrong with that "split" on rakudo jvm. It eats up too much memory and takes to much time
12:38 masak isBEKaml: also, de:salz means en:salt :)
12:39 masak isBEKaml: hence my pun about Säureburg and Baseburg.
12:39 isBEKaml masak: yessir :-)
12:40 isBEKaml sauerburg :)
12:40 masak autopun-ish: https://twitter.com/schuyle​r/status/396167385868021761
12:41 isBEKaml ehh, who is cmasak and masak on twitter?
12:41 arnsholt masak: That's awesome! =D
12:43 isBEKaml ehh, I kinda vaguely get it - given that their entire vocabulary is 23 words or so. :P
12:45 masak isBEKaml: no, that's not it.
12:45 masak isBEKaml: maybe https://en.wikipedia.org/wiki/Snowclone will help elucidate the matter.
12:46 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.
12:46 isBEKaml yeah, they have something like 50 words for snow. each adding to the other.
12:47 isBEKaml so it's difficult to think objectively about any one thing in Inuit/Eskimo language.
12:48 isBEKaml I was looking at this and then it fell in: https://en.wikipedia.org/w​iki/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."
12:49 masak isBEKaml: I don't find it very well-defined to talk about "N words for <anything>" in a strongly agglutinative language.
12:50 denis_boyun joined #perl6
12:50 masak it's about as useful as saying "C has 93,741 ways to index the first element of an array".
12:51 isBEKaml yeah, that's what I meant when I said "each adding to the other" about their words for snow. :-)
12:52 isBEKaml Okay, now back to lurking mode - fiddling with nqp. :D
12:55 jaffa4 joined #perl6
12:57 arnsholt IIRC serious people have actually looked into how many distinct roots various Inuit languages have for snow
12:57 arnsholt It varies between languages, obviously. But apparently it can be framed in a way that makes sense for the languages
12:58 masak roots would be more informative. :)
13:00 arnsholt Yeah. As you said, word isn't a very useful entity in polysynthetic languages =)
13:00 arnsholt At least not in this context
13:00 arnsholt If we're discussing the boundaries between morphology and syntax, they're very interesting indeed =D
13:01 SmokeMachine joined #perl6
13:03 logie joined #perl6
13:04 pippo should'nt this be linear in time with the number of cycles ? "my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(',')"
13:06 masak pippo: yes, it should.
13:06 masak pippo: maybe there is a Schlemiel in there somewhere.
13:07 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 rurban joined #perl6
13:10 pippo masak: what is a "Schlemiel"
13:11 isBEKaml bungler
13:13 masak pippo: "Who is Shlemiel? He's the guy in this joke [...]" http://www.joelonsoftware.com​/articles/fog0000000319.html
13:13 [Coke] lizmat++
13:13 [Coke] (mixy)
13:14 ajr joined #perl6
13:15 pippo masak: are you referring to me?
13:15 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.
13:16 masak pippo: sorry, I was chunking that whole concept down into that word.
13:16 masak but the article makes it very clear what I meant. :)
13:17 isBEKaml pippo: Schlemiel means "a bungler" - a slang word
13:18 dalek perl6-roast-data: 55d18c6 | coke++ | / (5 files):
13:18 dalek perl6-roast-data: today (automated commit)
13:18 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/55d18c68c9
13:18 dalek perl6-roast-data: 688234a | coke++ | doit:
13:18 dalek perl6-roast-data: avoid git failures when extra .out files exist
13:18 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/688234a872
13:18 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
13:19 [Coke] rakudo-jvm has large startup costs relative to rakudo-parrot, btw.
13:20 pippo It is not the startup that makes me unhappy.
13:23 jaffa4 pippo: life?
13:23 jaffa4 pippo: it has not crashed for me yet
13:24 pippo jaffa4: what is the time of execution on your machine for this: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(',')
13:26 jaffa4 quite fast
13:26 pippo on rakudo jvm?
13:26 jaffa4 rakudo 10.2013
13:26 jaffa4 parrot based
13:27 pippo I was referring to jvm?
13:27 pippo I was referring to jvm!
13:27 jaffa4 pippo:ok
13:28 jaffa4 jvm: print 'hello'
13:28 tobyink joined #perl6
13:28 [Coke] rj: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(',');
13:29 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
13:29 jaffa4 what is rj?
13:29 FROGGS rj: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',')
13:29 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
13:29 FROGGS rj: my @array; for ^20_000 {@array[$_]=$_}; say 1
13:29 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
13:29 FROGGS :/
13:29 [Coke] jaffa4: rakudo-jvm
13:29 salv0 joined #perl6
13:30 jaffa4 rj: my @array; for 1..20000 {@array[$_]=$_}; say 1
13:30 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
13:30 jaffa4 rakudo: my @array; for 1..20000 {@array[$_]=$_}; say 1
13:31 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
13:31 camelia ..rakudo-parrot 9d5448: OUTPUT«1␤»
13:31 masak jaffa4: you can see it from the camelia output -- when one types 'rj', it replies 'rakudo-jvm'.
13:31 pippo rj: say "Hello"
13:31 masak (timeout)
13:31 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
13:31 masak rj: print "(timeout)"
13:32 jaffa4 that is cheating
13:32 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
13:32 masak rj: sleep 10; print "(timeout)"
13:32 moritz it seems that the eval soerver isn't stable enough yet
13:32 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
13:32 masak :)
13:33 moritz j: say 'alive again'
13:33 camelia rakudo-jvm 882e33: OUTPUT«alive again␤»
13:33 FROGGS C:\rakudo>perl6 -e "use Shell::Command; mkpath('a\\b')"
13:33 FROGGS Failed to create directory 'a\b/' with mode '0o777': mkdir failed: Das System kann den angegebenen Pfad nicht finden.
13:33 FROGGS something there does not like backslashes
13:34 moritz ouch
13:34 isBEKaml eh, what? chmod o777?
13:34 moritz windows
13:34 moritz no umask set
13:34 moritz or something
13:34 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(',');"
13:35 isBEKaml but it doesn't make sense to use it on windows or even display that, right?
13:35 * moritz shrugs
13:36 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,…»
13:37 pippo and not safe when you add the split at the end...
13:37 [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.
13:38 isBEKaml pippo: assign and split?
13:38 pippo this: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(','); <== this split
13:38 moritz [Coke]: I've had that too, with spectest
13:39 [Coke] moritz: yay, it's not just me? :|
13:39 * [Coke] commutes
13:40 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)»
13:40 moritz pippo: yes, the split consumes way too much memory on my machine too
13:41 pippo moritz: this is annoying!
13:42 isBEKaml moritz: do you happen to know why this nqp build malfunctions? https://gist.github.com/anonymous/7262942
13:42 colomon is it just jvm?  It seemed like the parrot version also timed out in the backlog...
13:43 pippo colomon: I only have jvm and niecza on my machine.
13:45 FROGGS tadzik / jaffa4: see https://github.com/tadzik/Shell-Command/pull/1
13:45 masak rj: say (^20_000).join(',')
13:45 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,…»
13:45 colomon hmmm, your 9:38 code seems to run on parrot rakudo in about 3-4 seconds
13:45 masak rj: say (^20_000).join(',').split(',')
13:46 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
13:46 bluescreen10 joined #perl6
13:46 masak yeah, there's something rotten in the state of .split
13:47 kivutar joined #perl6
13:48 * isBEKaml gives up on getting nqp to build
13:49 jaffa4 FROGGS: more about it?
13:49 moritz isBEKaml: where does this invocation come from?
13:50 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
13:51 moritz isBEKaml: this does not look like typical output from make
13:51 isBEKaml moritz: no, that's not make. it's ant.
13:51 moritz isBEKaml: well, that's not how you build NPQ :-)
13:51 isBEKaml If you look at the comments there, I have tried reproducing that from the cmd prompt. using relative and absolute paths.
13:52 isBEKaml Both failed.
13:52 pernatiy joined #perl6
13:52 isBEKaml moritz: I know. I can
13:52 moritz isBEKaml: thing is, the NQP build is a bit tricky
13:52 moritz using stage0 to build stage1, stage1 to build stage2
13:52 FROGGS jaffa4: currently I am creating a new msi, and then try to install modules
13:52 moritz and all that without mixing up the stages
13:52 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 araujo_ joined #perl6
13:53 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 #perl6
13:54 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?
13:54 moritz no
13:54 moritz isBEKaml: what I can offer is to give you a log of a successful nqp build with make
13:55 moritz 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.
13:56 isBEKaml 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
13:56 FROGGS jaffa4: progress, at least: ==> Successfully installed Math::RungeKutta
13:56 FROGGS Failed to remove the directory '.work\1383314078_1\lib\Math'
13:56 FROGGS tadzik: thanks! :o)
13:56 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 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 :-)
13:57 moritz so that you can see where you instructions differ from the "offical" build
13:58 isBEKaml moritz: okay, please paste the "official" log. :-)
13:59 timotimo rj: say (^20_000).join(",").split(",")[20000]
13:59 * moritz makes it
13:59 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
13:59 timotimo (so not the stringification of th list...)
14:00 moritz rakudo-jvm is probably borked again
14:00 araujo_ joined #perl6
14:00 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?")
14:01 timotimo you seem to be right
14:01 camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
14:02 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 #perl6
14:03 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
14:04 pmurias with leftover parts of installed NQP perhaps?
14:04 moritz isBEKaml: http://moritz.faui2k3.org/tmp/j-build.log
14:04 moritz pmurias: yes
14:04 moritz pmurias: or a missing dependency in the Makefile
14:05 isBEKaml moritz: danke!
14:05 pippo rj: say "I am back"
14:05 camelia rakudo-jvm 882e33: OUTPUT«I am back␤»
14:06 araujo_ joined #perl6
14:07 pippo timotimo: if you think it is aproblem of stingification then why this works fine...
14:07 pippo rj: say (^20_000).join(",")
14:07 moritz does anybody have an idea for how to create a more robust rakudo-jvm setup for camelia
14:07 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,…»
14:07 moritz pippo: timotimo merely tested a hypothesis
14:07 timotimo because the list version puts a [ and ] at the end? :P
14:07 timotimo and beginning*
14:14 pmurias ahh, the nqp executable has been renamed to nqp-p?
14:17 pmurias was renaming (or aliasing) rakudo's executable to jakudo,parakudo,mokudo considered?
14:23 pippo bie #perl6
14:23 pippo left #perl6
14:25 araujo joined #perl6
14:25 xinming joined #perl6
14:28 isBEKaml mokudo?
14:32 pmurias isBEKaml: rakudo on moarvm
14:34 isBEKaml ah
14:34 pmurias seems much easier to type then perl6-moarvm ;)
14:35 isBEKaml that it is. :-)
14:36 * isBEKaml afk - be back later.
14:38 kaare_ joined #perl6
14:43 colomon moritz++ # building parrakudo and jakudo together now EASY!
14:45 colomon 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 colomon oh, wait!
14:46 colomon perl6-j didn't finish it, it took 25 seconds to die with java.lang.StackOverflowError
14:46 colomon okay, catastrophe.
14:46 masak :/
14:46 masak did I submit this as a rakudobug?
14:46 masak I don't remember.
14:47 masak oh, I submat https://rt.perl.org/Ticket/Display.html?id=120380
14:54 iSlug joined #perl6
15:00 diakopter show me the split one
15:00 diakopter ?
15:01 jaffa4 I have problem with current compiler
15:01 jaffa4 It gives me too short error message
15:01 jaffa4 Could not locate compile-time value for symbol P6
15:01 jaffa4 no line number, or column number
15:01 jaffa4 Is that normal?
15:01 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
15:02 timotimo masak: if you only say the last submatch of $/, you get the answer almost instantaneously
15:02 timotimo rj: my $file = "";
15:02 camelia rakudo-jvm 882e33: ( no output )
15:02 timotimo oops
15:02 kivutar_ joined #perl6
15:02 timotimo rj: my $file = ""; for ^20000 {$file ~= "$_;"}; say ($file ~~ /(\d+) +% ';'/)[0][20000];
15:03 camelia rakudo-jvm 882e33: OUTPUT«Nil␤»
15:03 timotimo rj: my $file = ""; for ^20000 {$file ~= "$_;"}; say ($file ~~ /(\d+) +% ';'/)[0][19999];
15:03 camelia rakudo-jvm 882e33: OUTPUT«「19999」␤␤»
15:03 jaffa4 masak: is it a bug?
15:03 timotimo there we go.
15:03 timotimo rj: my $file = ""; for ^20000 {$file ~= "$_;"}; say $file ~~ /[\d+] +% ';'/);
15:03 camelia rakudo-jvm 882e33: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/D56RfDAV_F�Unexpected closing bracket�at /tmp/D56RfDAV_F:1�------> [32me ~= "$_;"}; say $file ~~ /[\d+] +% ';'/[33m�[31m);[0m�»
15:03 timotimo rj: my $file = ""; for ^20000 {$file ~= "$_;"}; say $file ~~ /[\d+] +% ';'/;
15:03 camelia rakudo-jvm 882e33: OUTPUT«「0;1;2;3;4;5;6;7;8;9;10;11;12;13;14;1​5;16;17;18;19;20;21;22;23;24;25;26;27;28;29;​30;31;32;33;34;35;36;37;38;39;40;41;42;43;44​;45;46;47;48;49;50;51;52;53;54;55;56;57;58;5​9;60;61;62;63;64;65;66;67;68;69;70;71;72;73;​74;75;76;77;78;79;80;81;82;83;84;85;86;87;…»
15:03 masak jaffa4: I don't know what you did to get that error, but sounds like a possible bug, eys.
15:03 masak yes*
15:04 timotimo "could not locate compile-time value for symbol Foo" usually comes from the optimizer, try --optimize=0
15:04 colomon diakopter: my @array; for ^20_000 {@array[$_]=$_}; say @array.join(',').split(',');
15:04 timotimo or rather --optimize=off
15:04 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 #perl6
15:05 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 masak nor are you really expected to.
15:05 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
15:05 masak it is in my book.
15:06 masak STD is kind of using Perl 5 as a trampoline to explore Perl 6 parsing.
15:06 benabik joined #perl6
15:06 masak it was never meant to be included from a Rakudo script like you're trying to.
15:06 masak now, granted, you may have stumbled on an interesting rakudobug nevertheless.
15:07 masak but I predict it will be a big pain to golf :)
15:07 kivutar joined #perl6
15:07 jaffa4 I would like to parse perl 6 from perl 6
15:07 jaffa4 Is this such a big dream>
15:08 pmurias jaffa4: use the rakudo grammar
15:08 jaffa4 What rakudo grammar?
15:09 diakopter colomon: I'm curious how nqp-m handles an nqp version of that program
15:10 diakopter (I think there's a not small possibility it handles it even worse..)
15:13 ajr joined #perl6
15:13 masak jaffa4: you'll find it in the Rakudo repository, in src/Perl6/Grammar.nqp
15:14 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 masak jaffa4: I haven't seen anyone doing that, from start to end.
15:15 masak jaffa4: it's a big dream indeed, but a rather attractive one. please keep dreaming in that way :)
15:15 masak (by which I mean, I would like to see it work, too)
15:18 pmurias masak: I meant that to parse perl6 he should try to use Perl6::Grammar
15:21 masak pmurias: believe me, I'm with you so far.
15:21 masak pmurias: do you have a Perl 6 script that uses Perl6::Grammar to parse Perl 6 code?
15:22 pmurias so the problem is the Perl6/nqp distinction?
15:22 timotimo isn't there some :from<nqp> adverb for use already?
15:23 masak pmurias: the problem is parsing Perl 6 from (Rakudo) Perl 6.
15:24 jnap joined #perl6
15:24 denis_boyun joined #perl6
15:27 diakopter colomon: I'm having serious trouble translating it.
15:28 FROGGS tadzik: should the calls to rmdir shouldn't be rm_rf instead? https://github.com/tadzik/Shell-Comm​and/blob/master/Shell/Command.pm#L29
15:33 kivutar where is http://try.rakudo.org/ ? is it dead?
15:33 kivutar is it something like http://tour.golang.org/#1 ?
15:35 moritz kivutar: it's dead, yes :(
15:36 moritz (I know a dead rakudo when I see one)
15:39 tadzik FROGGS: obtw, the pull request is a bit wrong I think
15:39 FROGGS (This is a stiff rakudo)
15:39 FROGGS tadzik: really?
15:39 tadzik it created the Shell directpyu
15:39 tadzik directory
15:39 tadzik this can't be right
15:39 tadzik or maybe it was there before
15:40 FROGGS tadzik:  I don't understand
15:40 tadzik FROGGS: there?s a toplevel Shell directory in the repo noew
15:40 tadzik * now
15:41 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
15:42 FROGGS tadzik: weird, I just used githubs magic, clicked "edit" on the Command.pm, which did a fork and a PR
15:42 FROGGS tadzik: it was there before
15:43 tadzik maybe I screwed something up before then
15:43 FROGGS tadzik: d0442646c482b6b23f0871ea874022d6435b5a35 is the bad one
15:44 tadzik dang
15:44 tadzik ok= I'll fix it up later today
15:45 tadzik the local today, at least
15:45 * tadzik in california
15:45 masak tadzik: how's CA? :)
15:46 tadzik veru US-y :)
15:46 FROGGS *g*
15:46 FROGGS nice :o)
15:46 tadzik which is just what I wanted :)
15:46 * masak .oO( we phase-shifted tadzik about 45 degrees west, so his local day is a bit offset right now )
15:46 FROGGS have fun in the U, S and A
15:47 FROGGS that is not right
15:47 FROGGS that is left
15:50 tadzik "west -- it can't be right"
15:50 FROGGS exactly :o)
15:51 FROGGS but: "life is peaceful there"
15:56 japhb__ tadzik: Where in CA are you?
15:57 * masak .oO( slightly below the crossbar of the A )
15:57 nebuchadnezzar joined #perl6
15:58 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?
15:59 FROGGS japhb: what is wrong with the line above?
16:00 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/
16:01 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 japhb__ Ah, OK, at least the fix got in there.
16:04 japhb__ 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.
16:05 tadzik japhb__: San Mateo
16:05 FROGGS japhb__: the other one will be removed... currently I'm hunting for a bug on windows
16:06 tadzik staying at diakopter++'s
16:06 tadzik japhb__: where are you?
16:07 dalek rakudo-js: a1679de | (Pawel Murias)++ | / (7 files):
16:07 dalek rakudo-js: Unbitrot to work with the newest nqp.
16:07 dalek rakudo-js:
16:07 dalek rakudo-js: Adapt to nqp-p, handle QAST::Blocks with declaration_static, add bool_I op.
16:07 dalek rakudo-js: review: https://github.com/pmurias/​rakudo-js/commit/a1679defa6
16:07 pmurias timotimo: you mean when will I have the time?
16:07 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.
16:07 japhb__ I saw something similar once for another language, but I forget where.
16:08 japhb__ tadzik: I'm right now in Mountain View (not too far south of you).
16:08 japhb__ How long are you hear?
16:08 FROGGS pmurias: what is your plan to get it into nqp/ ?
16:08 japhb__ *here
16:08 tadzik about 3 weeks, I'm getting back on like 18th
16:08 japhb__ W00t!  We should get the gang together!
16:09 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 #perl6
16:10 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␤»
16:11 retupmoca wait, what?
16:11 FROGGS *g*
16:11 retupmoca r: "asdf".index('s').Bool.say
16:11 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
16:11 camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«True␤»
16:11 FROGGS r: "asdf".index('a').Bool.say; "asdf".index('x').Bool.say;
16:11 camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«False␤False␤»
16:11 retupmoca ah, extra say
16:11 FROGGS retupmoca: index("a") is zero, so it it false-ish
16:11 retupmoca ok
16:12 retupmoca 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
16:12 FROGGS r: say "asdf".index('x')
16:12 retupmoca r: say "asdf".index('x') >= 0
16:12 camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«(Int)␤»
16:12 camelia 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…»
16:12 camelia ..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:…»
16:12 FROGGS r: say "asdf".index('a')
16:12 camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«0␤»
16:13 FROGGS ahh
16:13 FROGGS it is "undef" for the not-found case
16:13 FROGGS r: say "asdf".index('x').defined
16:13 camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«False␤»
16:13 FROGGS r: say "asdf".index('a').defined
16:13 camelia rakudo-parrot 9d5448, rakudo-jvm 882e33: OUTPUT«True␤»
16:13 retupmoca ah, there we go
16:14 retupmoca .defined is what I was looking for
16:16 retupmoca S32 was a little unclear on how failure is represented
16:23 japhb__ So tadzik, diakopter, TimToady, anyone else in the SF Bay Area next week: Care for lunch at the Googleplex?
16:25 jeffreykegler joined #perl6
16:27 sftp joined #perl6
16:28 lizmat joined #perl6
16:30 FROGGS retupmoca: improve it :o)
16:32 woolfy joined #perl6
16:50 tadzik japhb__: I don't see why not :)
17:04 japhb tadzik et al., How does Monday sound?
17:07 tadzik I don't have plans for monday :)
17:07 tadzik so, good
17:10 diakopter you don't? :)
17:10 diakopter oh wait.
17:10 tadzik I do? :)
17:10 * diakopter backlogged
17:10 diakopter japhb__: sure
17:11 japhb Excellent, anyone else interested?
17:19 rurban joined #perl6
17:26 colomon joined #perl6
17:30 ajr joined #perl6
17:34 FROGGS japhb: I am, but I am still in ol' Europe
17:35 tadzik quick, come here!
17:36 japhb Dangit FROGGS, jump on a plane!  ;-)
17:36 colomon joined #perl6
17:36 FROGGS I would like to, problem is, I have to care about my kids next week (besides not having enough money)
17:37 japhb yoleaux, where are you when I need you?
17:37 colomon tadzik: you're in California?
17:38 japhb FROGGS, totally understood.  If you do end up coming this way sometime, ping me.  :-)
17:38 FROGGS I will :o)
17:39 FROGGS tadzik / jaffa4: btw, panda finally works now (again) on windows
17:40 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
17:40 colomon tadzik: \o/   so glad you made it out to the states!
17:40 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 FROGGS tadzik: so it was unable to delete the file, and died later when trying to delete the dir
17:40 tadzik ah
17:40 tadzik colomon: yeah, me too :)
17:40 tadzik I dramatically needed some time off
17:44 cxreg i love the "Flip flops can be negated." change in the spec.  "AUTUMN casts COLD WEATHER.  Your FLIP FLOPS are NEGATED"
17:45 FROGGS japhb++ # you already fixed what I recognized on windows! https://github.com/tadzik/panda/commit/d​8b4f4f12dafae1ecd1bf84773a317055d3be5fb
17:46 FROGGS tadzik: do we need to update panda in star?
17:47 tadzik FROGGS: I don't know, do we? :P
17:47 FROGGS hmmm, that is weird
17:48 FROGGS tadzik: the submodule commit-id is new enough, but the star release contains an older version
17:49 colomon tadzik: panda hasn't worked (on Linux) for the smoke tester in a week.
17:50 tadzik colomon: I just noticed that rakudobrew didn't build it for me, checking why
17:52 rurban joined #perl6
17:52 colomon huh.
17:53 colomon I've been getting
17:53 colomon ===SORRY!===
17:53 colomon 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/parr​ot/5.9.0-devel/languages/perl6/lib, /home/smoker/.perl6/2013.10-120-g9d54481/lib
17:53 colomon but just now I got
17:53 colomon build stage failed for File::Tools: Failed building lib/File/Find.pm
17:53 colomon ==> Please make sure that /home/smoker/rakudo/install/lib/parrot​/5.9.0-devel/languages/perl6/site/bin is in your PATH
17:54 colomon (and yes, that is in my path)
17:54 FROGGS colomon: maybe you are missing this commit? https://github.com/tadzik/panda/commit/9​f2591b3b0445b294db631e2d5f317b8480dd1c2
17:54 colomon that does seem like a significant commit!
17:55 FROGGS it does
17:55 colomon I'm actually trying pandabrew rebootstrap when I get these messages.
17:55 colomon which I think should mean I've got the latest panda?
17:56 FROGGS I dunno how that magic works
17:57 colomon oh, maybe not!
17:57 FROGGS damn, the star release really contains old modules :/
17:59 colomon crap, rebootstrap doesn't try to update panda at all
17:59 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
18:00 colomon my pandabrew panda dated back to YAPC::NA :\
18:01 FROGGS O.o
18:07 colomon so, tadzik: shouldn't pandabrew rebootstrap get the latest panda?
18:09 lue joined #perl6
18:09 lue hello world o/
18:10 japhb o/
18:13 darutoko joined #perl6
18:14 colomon after updating pandabrew's panda by hand, smoke tester appears to be working again
18:22 FROGGS ool
18:22 FROGGS cool
18:23 bluescreen10 joined #perl6
18:24 dalek nqp: 149a377 | (Pawel Murias)++ | Configure.pl:
18:24 dalek nqp: Make Configure.pl --help use the real list of backends.
18:24 dalek nqp: review: https://github.com/perl6/nqp/commit/149a377a55
18:28 dansamo joined #perl6
18:36 japhb OK, so far it looks like we've got 5-6 people coming for Monday lunch at Google.  Anyone else interested?
18:47 * moritz interested, but not anywhere near :-)
18:47 timotimo pmurias: oh cool! i thought you had other stuff in your life from now on
18:52 SamuraiJack joined #perl6
18:56 bluescreen10 joined #perl6
18:57 bluescreen10 joined #perl6
19:10 moritz any p6 or nqp LHFs I could hack on tonight?
19:12 diakopter moritz: might be nice to have the remaining required moar ops stubbed. would certainly be helpful.
19:13 FROGGS masak++ # very entertaining: http://www.joelonsoftware.com​/articles/fog0000000319.html
19:14 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" )
19:14 diakopter moritz: er, the 2nd "them" there is the remaining ones for moar
19:15 FROGGS moritz: and there is still RT, if you don't feel op-ish :o)
19:16 diakopter moritz: these definitely need native moar ops https://gist.github.com/diakopter/7270438
19:16 diakopter er, skip the nativecall ones though
19:16 diakopter editing gist
19:17 slava-work joined #perl6
19:17 diakopter moritz: (native moar ops as opposed to ones with non-core-mappings in the qast->mast compiler)
19:22 slava-work I hate silly people
19:23 FROGGS good day to you too slava-work :o)
19:23 slava-work :(
19:24 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
19:25 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 slava-work because that is seriously some awesome stuff
19:26 FROGGS slava-work: if it not in "core", you could implement it (as a module) with, like, 3 lines each for before and after
19:27 FROGGS because you can create a trait, that wraps the thing (sub) it is applied to, and does something before or after for example
19:29 PerlJam those would read weird as traits
19:30 slava-work FROGGS: ty
19:30 FROGGS PerlJam: maybe
19:31 sqirrel joined #perl6
19:31 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 tobyink joined #perl6
19:31 bluescreen10 joined #perl6
19:32 FROGGS or make a role that you mix into the sub-object... that should work too
19:33 slava-work yes
19:33 FROGGS in the end you wanna do this? before &sub-name, { <code goes here> }
19:33 slava-work I knew I was right for liking perl and hating java :P
19:33 FROGGS *g*
19:33 slava-work pretty much
19:34 PerlJam might as well just call .wrap
19:34 slava-work I am also not a big fan of java annotations, which allow spring to automagically link things together
19:36 colomon joined #perl6
19:39 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
19:39 camelia rakudo-parrot 9d5448: OUTPUT«before␤42␤»
19:39 FROGGS PerlJam++
19:40 FROGGS rp: sub test() { say 42 }; &test.wrap: -> { callsame; say "after" }; test
19:40 camelia rakudo-parrot 9d5448: OUTPUT«42␤after␤»
19:40 colomon joined #perl6
19:40 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␤»
19:44 FROGGS jaffa4: please test http://froggs.de/perl6/rakudo-star-2013.10-RC1.msi (other testers welcome)
19:59 Rotwang joined #perl6
20:00 Ulti why does parrot even have a binding to OpenGL?
20:01 moritz because that way, all the dozens of HLLs based on parrot can use OpenGL
20:03 Ulti :/
20:05 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
20:06 Ulti lue no, but I'd prefer it was an OpenCL Promise object :P
20:07 Ulti rakudo-fpga why go virtual machine when you can have hardware
20:07 lue .oO(rakudo-2A03)
20:08 Ulti arent 2A03 incredibly hard to acquire thanks to chiptune mayhem?
20:10 lue The 2A03 specifically? Maybe, but the genericer 6502 should be incredibly easy.
20:10 lue s/genericer/genericker/ # pronunciation rules
20:10 Ulti :S
20:10 Ulti uh oh its not just paths its the whole API has changed and the bindings are wrong win win win
20:13 moritz after some more thinking and tinkering, I'm pretty sure I don't know how I would patch rakudo for "no strict"
20:15 PacoAir joined #perl6
20:16 lue moritz: ooc, is it just making the phrase "no strict" work, or do you have to implement its effects too?
20:17 moritz lue: just the latter
20:17 lue Doesn't -e already do (some of) that?
20:17 moritz no
20:17 moritz it should, but doesn't
20:18 FROGGS shall I give it a whirl?
20:18 moritz yes please
20:19 moritz there's one more thing I'm trying to do, but don't wait for it
20:26 diakopter wouldn't you just auto-declare a lexical in just the same place it currently errors
20:28 arnsholt My goal for the evening: Implement CStr on JVM
20:28 FROGGS diakopter: that is what v5 does
20:28 arnsholt And if there's time after that, CArrays of value types
20:28 moritz diakopter: TimToady explictly said that's not what happens
20:28 moritz rather it's a direct lookup into the symbol table
20:29 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 jnthn moritz: Anything I can give a hint on?
20:29 FROGGS moritz: well, you have to declare it in order looking it up, no?
20:29 moritz p5eval: $blerg = 42; { use strict; say $blerg }
20:29 p5eval moritz: ERROR: Global symbol "$blerg" requires explicit package name at (eval 7) line 1.
20:30 FROGGS well, this would error out this way too
20:30 FROGGS hmm, no, maybe not
20:30 moritz not if I'd autodeclare $blerg
20:32 moritz jnthn: how would the generated code look like?
20:32 moritz jnthn: an anonymous variable or container bound to the slot of the symbol table?
20:33 moritz 'cause just binding into the symbol table doesn't seem to be enough (segfaults rakudo :-)
20:33 FROGGS moritz: gimme 10min to try my v5 patch
20:33 jnthn moritz: Hang on... :) What are the semantics you're after?
20:34 jnthn our-like?
20:34 moritz jnthn: an undeclared variable under 'no strict;' should not declare a user-visible variable, but point straight into the symbol table
20:36 jnthn moritz: Of the current package?
20:36 moritz jnthn: yes
20:36 moritz basically OUR::<$x>
20:37 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 jnthn oh, oops
20:37 jnthn That WVal wantsa QAST::Op.new( :op('who'), ... ) around it
20:37 moritz oh
20:37 moritz good idea :-)
20:38 moritz and I used :op<atkey>, which doesn't autovivify
20:41 diakopter moritz: oh; I missed that
20:42 diakopter 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?
20:45 jnthn not afaik
20:45 lowpro30 joined #perl6
20:47 moritz I've managed to stop the declaration from happening, but the lookup is still for a lexical variable
20:48 moritz and that segfaults, because there's no such var
20:50 diakopter .oO( segfault? )
20:50 denis_boyun joined #perl6
20:50 diakopter should transform every occurrence in that package to Foo::<foo> or whatever?
20:51 moritz yes, that's how I understand it
21:00 * moritz calls it a day
21:00 dalek rakudo/nom: 3f92477 | jonathan++ | src/vm/jvm/core/Threading.pm:
21:00 dalek rakudo/nom: Various combinators for Subscribables.
21:00 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/3f924771bb
21:00 dalek rakudo/nom: 2a0f6c9 | jonathan++ | src/vm/jvm/core/Threading.pm:
21:00 dalek rakudo/nom: Naming cleanup in Publish.for implementation.
21:00 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/2a0f6c9a79
21:01 dalek roast: dc0b47e | jonathan++ | S17-concurrency/subscribable.t:
21:01 dalek roast: Tests for some Subscribable combinators.
21:01 dalek roast: review: https://github.com/perl6/roast/commit/dc0b47ed17
21:01 jnthn L'hacking de tren...
21:01 FROGGS froggs@TL02L-R8RXTCW-linux:~/dev/rakudo$ perl6-p -e 'use 6; $foo = 42; say $foo'
21:01 FROGGS 42
21:01 FROGGS froggs@TL02L-R8RXTCW-linux:~/dev/rakudo$ perl6-p -e 'use v6; $foo = 42; say $foo'
21:01 FROGGS ===SORRY!=== Error while compiling -e
21:02 FROGGS I guess this is a problem then:
21:02 FROGGS perl6-p -e 'use 6; $foo = 42; { use v6; say $foo }'
21:02 FROGGS 42
21:03 FROGGS (and yes, it should be "6" rather than "use 6")
21:09 FROGGS that is what I have: https://gist.github.com/FR​OGGS/6f8989611659551dae1f
21:10 FROGGS dunno what would be needed to make 'use 6; $foo = 42; { use v6; say $foo }' throw
21:18 bluescreen10 joined #perl6
21:22 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?
21:23 FROGGS jnthn: is there already a way to check in the gramar's comp_unit if we are running as -e?
21:24 FROGGS 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?
21:24 FROGGS tipdbmp: you can work-around it by doing "augment class Array { };" after the List...
21:25 tipdbmp FROGGS, so I should augment both?
21:25 FROGGS tipdbmp: you have to to renew the method cache
21:25 tipdbmp Oh, I see.
21:25 FROGGS tipdbmp: but for Array you don't actually have to change a method
21:25 colomon joined #perl6
21:25 FROGGS an empty augment will do
21:26 tipdbmp Yes, it worked. Thanks.
21:26 jnthn FROGGS: I think (yes, ugh) that $?FILE or whatever ends up as -e :)
21:26 FROGGS jnthn: ahh, yes!
21:26 FROGGS rp: say "$?FILE"
21:26 camelia rakudo-parrot 9d5448: OUTPUT«/tmp/XVplQXYf2S␤»
21:26 jnthn pmurias: Branch is fine, that's how JVM and MoarVM happened before landing in master
21:27 FROGGS err, okay... so I have to wait until my local one compiles :o)
21:30 arnsholt Rakudo's make j-install doesn't seem to work on my machine
21:30 arnsholt Is that a known thing?
21:31 FROGGS unknown to me at least
21:33 arnsholt Creating '/nqp/install/bin/perl6-j'
21:33 arnsholt open: No such file or directory at tools/build/create-jvm-runner.pl line 35.
21:33 arnsholt moritz: You seen that happen?
21:34 arnsholt My config.default is:
21:34 arnsholt --backends=jvm,parrot
21:34 arnsholt --prefix=../nqp/install
21:38 FROGGS perl6 -e '$foo = 42; { use v6; say $foo }' # I don't see why this should throw in the strict-enabled block
22:03 lowpro30 joined #perl6
22:05 PacoAir joined #perl6
22:08 PacoAir joined #perl6
22:27 rurban joined #perl6
22:30 BenGoldberg joined #perl6
22:38 colomon joined #perl6
22:41 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.
22:42 japhb But Tene and I discovered that language interop didn't get quite the support we'd both hoped.
22:42 Tene Yes, unfortunately.
22:43 japhb Yes! I got another person up for lunch on Monday.
22:46 BenGoldberg I would hate to sleep past lunch, on a Monday ;)
22:48 japhb heh
23:00 tadzik colomon: oh, I don't think pandabrew does that, no
23:03 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.
23:04 tadzik colomon: fwiw, I'm not really maintaining pandabrew anymore
23:04 tadzik the reason for it was for panda to self-update after rakudo update, and now rakudobrew autoupdates panda after each build
23:04 colomon is rakudobrew ready for prime time?
23:04 tadzik I thik so
23:04 tadzik I use it without any problems
23:05 tadzik like, I don't build rakudos outside rakudobrew anymore
23:06 moartl joined #perl6
23:07 japhb That's certainly a positive endorsement
23:08 jnthn heh, if Moar's startup time sucked then imoartal would be a good name for an eval server..
23:11 diakopter :D
23:12 moritz arnsholt: what --prefix did you configure?
23:13 moritz arnsholt: /nqp/install/... looks slightly wrong to me
23:17 johnny5_ joined #perl6
23:23 johnny5_ joined #perl6
23:25 lue jnthn: a sucky startup time would warrant "moaratorium", surely :)
23:25 johnny5_ joined #perl6
23:29 lizmat joined #perl6
23:30 woolfy joined #perl6
23:31 moritz a slow startup would moartify the authors
23:33 japhb moritz: Yes, precisely: the authors would die of boredom.
23:34 Ulti japhb which planned target languages have graphics tightly integrated? I've only used a couple like Icon and I guess Logo
23:36 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.
23:37 japhb 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 https://gist.github.com/FR​OGGS/6f8989611659551dae1f
23:37 japhb Just like SDL + OpenGL + OpenAL = suddenly people go "Hey, I'm going to write a game!"
23:37 FROGGS moritz: at least it survived basic tests
23:38 woolfy left #perl6
23:38 FROGGS gnight all
23:38 Ulti japhb so parrot provides an abstraction of OpenGL?
23:39 moritz FROGGS: doesn't $*W.is_lexical suffice?
23:39 FROGGS no, I need to stuff the <lax> somewhere
23:39 japhb Ulti: The idea was that any language written on top of parrot could use any library that was bound for Parrot just once.
23:40 japhb 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
23:40 japhb That was the whole reason I worked on Parrot.
23:41 ssutch_ joined #perl6
23:41 ssutch_ chello
23:41 japhb I'm sure ingy would tell me it was some form of acmeism.  :-)
23:41 japhb utf8hello
23:44 dalek perl6-roast-data: a8fb2b7 | coke++ | / (5 files):
23:44 dalek perl6-roast-data: today (automated commit)
23:44 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/a8fb2b7695

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

Perl 6 | Reference Documentation | Rakudo