Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2012-11-29

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:13 lue huf: interesting. Never would have thought of needing to do that.
00:13 huf it's so frequent a question on #perl that there are multiple modules to make it nicer :)
00:14 lue (although I fear I don't quite get what use lib is trying to do. /me researches use lib)
00:14 huf BEGIN { push @INC, '....' }
00:15 lue Ah. I'm getting the feeling a lot of my confusion stems from me not having ever needed to do this.
00:16 huf :)
00:20 lue Huh. running perl6 -e "require 'lib/lib.pm6';" in the base of rakudo's repo generates a "could not find" error.
00:25 lue Which, IIRC, is wrong.
00:33 lue .oO(And all this talk today when I thought I'd never have problems with P6's module system again...)
00:52 anuby joined #perl6
00:58 ServerSage joined #perl6
01:09 hypolin joined #perl6
01:10 pjcj joined #perl6
01:16 benabik joined #perl6
01:34 tokuhiro_ joined #perl6
01:38 Pleiades` joined #perl6
01:47 [Coke] r: say (1..20).grep(*.is-prime)
01:47 p6eval rakudo 247f87: OUTPUT«2 3 5 7 11 13 17 19␤»
01:47 [Coke] huh. that's failing locally.
01:47 [Coke] r: say (1..100).grep(*.is-prime)
01:47 p6eval rakudo 247f87: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97␤»
01:56 [Coke] even weirder, the spectest fails, but the code
01:56 [Coke] snippet works in the REPL
01:59 anuby_ joined #perl6
02:11 tokuhiro_ joined #perl6
02:32 jerome joined #perl6
02:45 diakopter [Coke]: that's subprime
02:50 jerome joined #perl6
02:51 cyclick2 joined #perl6
03:01 hash_table joined #perl6
03:03 japhb phenny, tell TimToady I've been tanking on S11 and the big discussion we had this afternoon.  I'm beginning to wonder if the 'use' statement isn't the right place to be specifying auth and ver, or at least not the only place to make that choice.
03:03 phenny japhb: I'll pass that on when TimToady is around.
03:05 japhb phenny, tell TimToady I understand the desire to say "I need this API version from my dependency." But it's the wrong place to specify it when you want to bundle, i.e. specify "These packages of software have been tested together", where you want to make a more project-global choice
03:05 phenny japhb: I'll pass that on when TimToady is around.
03:06 japhb phenny, tell TimToady I'll continue to tank on it, but input from your highly non-linear thinking patterns desired.
03:06 phenny japhb: I'll pass that on when TimToady is around.
03:06 japhb .oO( TimToady as non-deterministic oracle )
03:09 noggle_ joined #perl6
03:09 cyclick2 is perl6 better then perl5 at character encodings?
03:11 sorear yes in some regards
03:11 sorear character strings and byte strings are no longer interchangable, so it's no longer possible to accidentally not handle encoding or handle it twice
03:12 sorear however the encoding menu is much smaller, no current p6 impl supports EBCDIC
03:15 cyclick2 is it possible to say all strings are utf-8 for example, no matter which module your are using?
03:15 spiderweb joined #perl6
03:16 spiderweb lo
03:17 japhb cyclick2, character strings by default follow Unicode semantics in Perl 6.
03:17 sorear cyclick2: All strings can contain any Unicode character.
03:17 sorear cyclick2: The encoding is an implementation detail - Niecza uses UTF-16
03:18 cyclick2 good
03:18 hypolin joined #perl6
03:26 mikemol joined #perl6
03:28 sorear mikemol! \o/
03:34 mikemol o/
03:44 orafu joined #perl6
03:44 cyclick joined #perl6
03:55 jaldhar joined #perl6
04:05 gootle joined #perl6
04:07 jaldhar joined #perl6
04:42 Quix0te joined #perl6
04:54 Quix0te left #perl6
05:25 SamuraiJack joined #perl6
05:32 kaleem joined #perl6
05:36 telex joined #perl6
05:36 sirrobert joined #perl6
05:37 sirrobert lo
05:46 raiph joined #perl6
05:46 raiph hi sirrobert
05:48 Kharec joined #perl6
05:56 moritz \o
06:14 birdwindupbird joined #perl6
06:19 GlitchMr joined #perl6
06:24 daniel-s joined #perl6
06:26 daniel-s_ joined #perl6
06:28 anuby joined #perl6
06:39 SamuraiJack joined #perl6
06:49 chee joined #perl6
06:50 daniel-s_ joined #perl6
07:07 Kharec joined #perl6
07:25 nebuchadnezzar joined #perl6
07:30 cognominal joined #perl6
07:36 FROGGS joined #perl6
07:36 FROGGS good morning
07:43 sorear good morning FROGGS
07:47 wamba joined #perl6
07:54 tadzik good morning #perl6
07:56 FROGGS tadzik, sorear o/
07:59 Psyche^ joined #perl6
08:00 atrodo joined #perl6
08:03 brrt joined #perl6
08:08 DarthGandalf joined #perl6
08:27 cognominal joined #perl6
08:43 araujo joined #perl6
08:43 araujo joined #perl6
08:49 wamba joined #perl6
08:58 sqirrel joined #perl6
08:59 grondilu joined #perl6
09:02 cyclick2 joined #perl6
09:10 rindolf joined #perl6
09:13 dayangkun joined #perl6
09:21 cognominal joined #perl6
09:32 sqirrel so, spüler läuft
09:32 sqirrel oh sorry wrong window
09:35 hoelzro hehe
09:35 hoelzro #housework?
09:35 hoelzro rather, #housework_de
09:38 FROGGS ;o)
09:38 FROGGS sqirrel++
09:41 mtk joined #perl6
09:45 Su-Shee I'm guessing either company inhouse irc or family available via chat ;)
09:45 huf :)
09:47 sqirrel yep, my life is boring *gg*
09:48 FROGGS hey, just because I'm not at home :o)
09:48 dakkar joined #perl6
09:48 sqirrel yes froggs busy earning money for sqirrels expensive hobbies *gg* froggs++
09:50 FROGGS *sigh*, there something tru 'bout that
09:50 wamba joined #perl6
09:51 kresike joined #perl6
09:51 kresike hello all you happy perl6 people
09:51 FROGGS kresike o/
09:52 fhelmberger joined #perl6
09:55 Kharec joined #perl6
09:58 Kharec joined #perl6
09:59 kresike FROGGS, o/
10:01 Kharec joined #perl6
10:06 Rix joined #perl6
10:09 flue joined #perl6
10:09 flue left #perl6
10:21 dalek Perlito: a9955f1 | (Flavio S. Glock)++ | t5/op/index.t:
10:21 dalek Perlito: Perlito5 - another 42 standard tests pass
10:21 dalek Perlito: review: https://github.com/fglock/​Perlito/commit/a9955f1985
10:29 bowtie joined #perl6
10:33 toebu joined #perl6
10:34 dalek Perlito: 73d6564 | (Flavio S. Glock)++ | t5/ (3 files):
10:34 dalek Perlito: Perlito5 - another ~200 standard tests
10:34 dalek Perlito: review: https://github.com/fglock/​Perlito/commit/73d65642dd
10:39 PacoAir joined #perl6
10:44 cognominal joined #perl6
10:52 dalek Perlito: db3046c | (Flavio S. Glock)++ | / (4 files):
10:52 dalek Perlito: Perlito5 - js2 - bit shift overflow fix
10:52 dalek Perlito: review: https://github.com/fglock/​Perlito/commit/db3046c4b2
10:59 rindolf joined #perl6
11:05 Su-Shee_ joined #perl6
11:08 wamba joined #perl6
11:12 Rix joined #perl6
11:17 ircmouser joined #perl6
12:17 brrt joined #perl6
12:56 tokuhiro_ joined #perl6
13:34 cognominal joined #perl6
13:35 sftp joined #perl6
13:37 evgi joined #perl6
13:38 evgi hi
13:38 evgi I have a question
13:39 evgi When will Perl 6 be released?
13:41 moritz evgi: there are two Perl 6 compilers that already have monthly releases
13:41 moritz so in some sense, it already has been released
13:41 [Coke] and a distribution built on top of one of them that has regularly (but not necessarily monthly) releases.
13:41 [Coke] Is there a particular feature you're looking for?
13:41 moritz though we haven't snapshotted a version of the spec yet and called it "Perl 6.0.0"
13:42 [Coke] *regular, whoops.
13:43 evgi I am working currently with Perl 5
13:43 evgi I have a lot of code there
13:43 evgi How stable Perl 6?
13:43 evgi How I can convert all my code (Perl 5) to Perl 6?
13:43 moritz first of all if you have lots of Perl 5 code, it doesn't make any sense to port it to Perl 6
13:44 moritz there's no automagic conversion, and Perl 5 won't go away
13:44 moritz how stable is not? not as stable as Perl 5, by far
13:44 evgi but Perl 6 has a many updates/improvments
13:44 evgi like obkects and many more
13:45 moritz yes
13:45 evgi objects
13:45 moritz but updating your code from Perl 5 to Perl 6 would be nearly complete rewrite
13:45 evgi ohh
13:46 noggle_ joined #perl6
13:46 moritz so, don't do that
13:47 moritz use Perl 6 for new software, if it doesn't need to be tightly coupled with the existing code
13:47 shachaf joined #perl6
13:47 evgi ok, i'll try
13:47 evgi thanks a lot
13:48 moritz and note that if you do that, you're an early adaptor
13:48 gugod joined #perl6
13:48 twinshadow joined #perl6
13:48 ingy joined #perl6
13:49 evgi left #perl6
13:49 [Coke] no, you are.
13:50 dalek rakudo/nom: 7ede6cd | moritz++ | / (2 files):
13:50 dalek rakudo/nom: remove ~/.perl6/lib from default @*INC
13:50 dalek rakudo/nom:
13:50 dalek rakudo/nom: as the prophe^W deprecation notes predicted
13:50 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/7ede6cdbcf
13:50 pecastro joined #perl6
13:52 [Coke] moritz: any idea why "is-prime" was giving me different answers locally?
13:53 moritz [Coke]: not really :(
13:53 moritz [Coke]: what test(s) fail?
13:53 [Coke] REPL and ./perl6 seem fine. but that spec test was still failing. weird.
13:53 [Coke] t/spec/S32-num/is-prime.t
13:54 moritz [Coke]: first, check if t/spec/ might be stuck
13:54 geekosaur joined #perl6
13:54 [Coke] r: use Test; is (1..10).grep(*.is-prime), (2,3,5,7);
13:54 p6eval rakudo 247f87: OUTPUT«ok 1 - ␤»
13:54 moritz [Coke]: second, run ./perl6 -Ilib t/spec/S32-num/is-prime.
13:54 moritz erm,
13:54 moritz [Coke]: second, run ./perl6 -Ilib t/spec/S32-num/is-prime.t
13:55 immortal joined #perl6
13:55 immortal joined #perl6
13:55 moritz (just to check if it's a problem with the test harness, or with rakudo)
13:55 moritz latest roast commit is 127f694a94d4e1087f94fc8b70bce6bffc1ff69f
14:04 dalek nqp/gh67-threads: 6806339 | (Tobias Leich)++ | src/pmc/nqplexpad.pmc:
14:04 dalek nqp/gh67-threads: checking for enum_class_Proxy instead of PARROT_HAS_THREADS
14:04 dalek nqp/gh67-threads:
14:04 dalek nqp/gh67-threads: In parrot 4.4.0 enum_class_Proxy isnt defined but PARROT_HAS_THREADS is. So enum_class_Proxy would be added to the condition.
14:04 dalek nqp/gh67-threads: review: https://github.com/perl6/nqp/commit/6806339701
14:05 ker2x joined #perl6
14:05 dalek joined #perl6
14:05 ker2x friendly greetings!
14:06 atrodo joined #perl6
14:06 moritz hi ker2x
14:07 noggle joined #perl6
14:07 tokuhiro_ joined #perl6
14:11 [Coke] moritz: updated, verified, spec test still fails here.
14:11 [Coke] weird.
14:12 moritz [Coke]: does it also fail when you run ./perl6 -Ilib t/spec/S32-num/is-prime.t ? if yes, please nopaste the output
14:20 [Coke] https://gist.github.com/4169395
14:25 GlitchMr joined #perl6
14:26 [Coke] urk? https://gist.github.com/4169433 - moritz, you'll LOVE that one.
14:26 [Coke] Note that the first 2 tests are a cut and paste, and yet one passes and one fails.
14:27 [Coke] (er, sorry, that one ended up with all 3 tests the same, first failing, 2nd passing.
14:27 [Coke] (+3rd)
14:28 moritz that is 'urk?' indeed
14:28 jnthn o/ from vilnius
14:28 phenny jnthn: 28 Nov 23:22Z <sirrobert> tell jnthn I think the perl6-debug in rakudo-star-2012.11-RC1 steps over method/function calls by default, instead of stepping into them.
14:29 jnthn phenny: tell sirrobert be sure you're not trying to step into a pre-compiled module (ideally it would never use pre-compiled things and always compile debug versions...I'll make it do that eventually...)
14:29 phenny jnthn: I'll pass that on when sirrobert is around.
14:29 * jnthn yawns
14:31 moritz \o
14:31 hash_table joined #perl6
14:35 isBEKaml joined #perl6
14:35 moritz [Coke]: weird, the tests passes for me, even when I run the thing three times
14:38 isBEKaml 'evening, #perl6!
14:38 colomon joined #perl6
14:40 moritz \o isBEKaml
14:42 benabik joined #perl6
14:42 isBEKaml hello, moritz
14:43 isBEKaml [Coke]: I don't see failures here as well. That's after running it 5 times over seq.
14:43 isBEKaml (win7 64 bit cygwin)
14:46 moritz [Coke]: are you on a 32bit machine?
14:51 bluescreen10 joined #perl6
14:52 dalek Perlito: 3b33471 | (Flavio S. Glock)++ | t5/op/arith.t:
14:52 dalek Perlito: Perlito5 - add standard test t/op/arith.t
14:52 dalek Perlito: review: https://github.com/fglock/​Perlito/commit/3b334713eb
14:52 dalek Perlito: 283ea50 | (Flavio S. Glock)++ | / (6 files):
14:52 dalek Perlito: Perlito5 - parser - precedence fix; Tests=863
14:52 dalek Perlito: review: https://github.com/fglock/​Perlito/commit/283ea501a5
14:57 stopbit joined #perl6
15:03 kaare__ joined #perl6
15:09 cognominal joined #perl6
15:16 BillySeth joined #perl6
15:16 BillySeth left #perl6
15:20 dalek Perlito: 1fdfe81 | (Flavio S. Glock)++ | t5/01-perlito/ (5 files):
15:20 dalek Perlito: Perlito5 - tests - "make test" is clean; Tests=876
15:20 dalek Perlito: review: https://github.com/fglock/​Perlito/commit/1fdfe81f2b
15:26 [Coke] moritz: 64bit mac.
15:32 skids joined #perl6
15:37 hash_table joined #perl6
15:39 [Coke] the rakudo build I have right now is installed locally, and before I built it, I rm'd install/ nqp/ parrot/ . ('sweird.)
15:40 wamba joined #perl6
16:01 Kharec joined #perl6
16:05 benabik joined #perl6
16:10 skids r: for (-8..-1) { printf("%x » %x / ",$_ +& 15, ($_ +> 1) +& 15) } # needs libtomath repr love
16:10 p6eval rakudo 7ede6c: OUTPUT«8 » c / 9 » d / a » d / b » e / c » e / d » f / e » f / f » 0 / »
16:12 skids The other ops have compensation for negatives; this one doesn't.
16:14 skids (So this is different than the other int bitops problems up near the high bits.)
16:17 brrt joined #perl6
16:19 moritz which op? +& or +> ?
16:20 skids +>
16:20 * moritz not very well concentrated
16:20 moritz is there a bug report for that already?
16:20 moritz if not, please open one
16:21 skids Not sure.  I'm more saying I dug in enough to see that the fix for this one is probably easier than that other stuff.
16:22 bluescreen10 joined #perl6
16:23 FROGGS joined #perl6
16:26 moritz a fix would also be welcome :-)
16:26 benabik joined #perl6
16:27 skids Oh.  It turns out you are more concentrated than me.
16:27 skids r: for (-8..-1) { printf("(%i) %x » (%i) %x / ",$_, $_ +& 15, ($_ +> 1), ($_ +> 1) +& 15) }
16:27 p6eval rakudo 7ede6c: OUTPUT«(-8) 8 » (-4) c / (-7) 9 » (-3) d / (-6) a » (-3) d / (-5) b » (-2) e / (-4) c » (-2) e / (-3) d » (-1) f / (-2) e » (-1) f / (-1) f » (0) 0 / »
16:27 skids it is +&.
16:31 erkan joined #perl6
16:31 erkan joined #perl6
16:33 kresike bye folks
16:37 skids .oO(These are made harder to hunt given the limited range of fmt/printf directives.)
16:40 moritz why not just use say() ?
16:41 moritz r: say (-8).base(2)
16:41 p6eval rakudo 7ede6c: OUTPUT«-1000␤»
16:41 moritz r: say (-8 +> 1).base(2)
16:41 p6eval rakudo 7ede6c: OUTPUT«-100␤»
16:41 moritz r: say (-2 +> 1).base(2)
16:41 p6eval rakudo 7ede6c: OUTPUT«-1␤»
16:42 moritz r: given -2 { say .base(2); say ($_ +> 1).base(2) }
16:42 p6eval rakudo 7ede6c: OUTPUT«-10␤-1␤»
16:42 moritz r: given -1 { say .base(2); say ($_ +> 1).base(2) }
16:42 p6eval rakudo 7ede6c: OUTPUT«-1␤0␤»
16:42 moritz r: given -3 { say .base(2); say ($_ +> 1).base(2) }
16:42 p6eval rakudo 7ede6c: OUTPUT«-11␤-1␤»
16:43 moritz nr: given -3 { say .base(2); say ($_ +> 1).base(2) }
16:43 p6eval rakudo 7ede6c: OUTPUT«-11␤-1␤»
16:43 p6eval ..niecza v22-44-g0247fe5: OUTPUT«-11␤-10␤»
16:45 skids r: for (1..8) { my $a = $_ - 0x10000000000000000; my $b = $a +> 1; $b = $b + 0x8000000000000000; printf("%x » %x / ", $a, $b); }
16:45 p6eval rakudo 7ede6c: OUTPUT«1 » 1 / 2 » 1 / 3 » 2 / 4 » 2 / 5 » 3 / 6 » 3 / 7 » 4 / 8 » 4 / »
16:46 skids There's the one, I think.  Unless I am being deceived again.  I'll try .base, thanks moritz++.
16:51 skids (So to rescind my above statement, this might be tied to the longstanding higher bit problems after all.)
17:04 wamba joined #perl6
17:08 masak oh hai, #perl6
17:08 tadzik hai masak
17:09 masak I am tired. just one more work day left.
17:09 masak until le weekend, I mean.
17:10 mathw joined #perl6
17:10 ker2x joined #perl6
17:10 takesako joined #perl6
17:11 stanley joined #perl6
17:11 lorn joined #perl6
17:12 bapa joined #perl6
17:17 masak phenny: de en "so, spüler läuft"?
17:17 phenny masak: "so dishwasher runs" (de to en, translate.google.com)
17:17 masak hehe, "spüler" :)
17:19 Kharec joined #perl6
17:19 masak to a Swedish ear, that word sounds like "a thing that flushes". which I guess is actually the correct etymology.
17:19 masak but I wouldn't use that word for a dishwasher, I'd use it for something like the thing that cleans the windshield of a car by squirting liquid on it.
17:20 perigrin windshield washer
17:20 perigrin English is only inventive when it's inconvient for learning.
17:21 masak ker2x: friendly greetings back!
17:22 FROGGS masak++
17:22 FROGGS *gg*
17:22 masak perigrin: both "dishwasher" and "windshield washer" seem like totally cromulent words to me.
17:23 FROGGS phenny: de en "Spritzdüse"
17:23 FROGGS phenny: de en "Spritzdüse"?
17:23 phenny FROGGS: "nozzle" (de to en, translate.google.com)
17:23 FROGGS masak^^
17:24 perigrin masak: it's what we get from ebiggening teh language
17:24 FROGGS btw, a dishwasher is squirting water to clean the dishes
17:25 colomon joined #perl6
17:26 FROGGS phenny: en de "cromulent"?
17:26 phenny FROGGS: "cromulent" (en to de, translate.google.com)
17:26 FROGGS wow, thanks -.-
17:32 masak FROGGS: "cromulent" is a Simpsons neologism.
17:33 masak https://en.wiktionary.org/wiki/cromulent
17:57 pmurias joined #perl6
18:01 pmurias japhb: re version and auth and bundles, picking the right versions of module without declarations of dependencies seems to be hard
18:02 pmurias japhb: unless we just take the newest version of everything
18:03 colomon joined #perl6
18:03 colomon cromulent is a wonderful word.
18:04 nebuchadnezzar joined #perl6
18:08 cognominal joined #perl6
18:20 Pompel joined #perl6
18:28 Moukeddar joined #perl6
18:29 Moukeddar Hello \o/ #perl6
18:29 PerlJam o/
18:30 Moukeddar how you doin' ?
18:32 PerlJam a bit of a headache and a little hungry,  but otherwise fine  :)
18:34 Moukeddar get well soon, i have the cold
18:34 * PerlJam munches on some almonds.
18:36 Moukeddar can i ask something, it's not really related to perl6
18:39 FROGGS well, can't hurt to ask
18:41 Moukeddar we're studying HTML and CSS, it's fun, but all i think is : it's for designers,and with all what's popping in the web, what you learn today seems obsolete tomorrow, how to deal with it ?
18:42 FROGGS well, I need to know HTML and CSS for example, cant say that it is obsolete, and I'm not a designer
18:42 masak Moukeddar: learn principles that won't be obsolete tomorrow.
18:43 PerlJam Moukeddar: the same way you deal with any technical knowledge ... continue to learn new things every day  :-)
18:43 PerlJam Moukeddar: but like masak says, the principles are the important thing
18:44 masak yes, I'm afraid constant learning is inevitable.
18:44 masak but there are nice underlying currents to all the trivial details of web dev.
18:44 masak like "why does IE6 suck att DOM node removal?" -- "because it uses a refcounting GC and its devs didn't realize that was a problem until the web started using Ajax".
18:45 FROGGS btw, knowing HTML and CSS will help you writing a template engine for example, you always can reuse knowledge
18:45 masak very true.
18:45 masak and a template engine is just a compiler :)
18:45 Moukeddar the one thing that frightens me, i'm not exactly a web dev guy, and the whole world seems to be moving faster than ever toward that
18:45 PerlJam *everything* is a compiler at some level  ;)
18:46 FROGGS Moukeddar: what typ'o'guy are you?
18:47 Moukeddar FROGGS, i don't know the definition, but i prefer to work on the back-end of something than the front end, even if it was a store lol
18:47 FROGGS k
18:47 huf lots of people are like that who are stuck working in web :)
18:47 * moritz is sa text munging guy
18:47 moritz *is a
18:48 FROGGS but its always good to know a bit about what the ppl are doing you are working with
18:48 huf my ideas about UI-s are so incompatible with what the r=1 users have been molded to want
18:48 PerlJam moritz: you should have left it as an unintentional pun ;)
18:48 moritz PerlJam: :-)
18:48 huf that i shouldnt be put in a UI-related position anyway :D
18:50 * masak likes to process heterogenous data
18:50 Moukeddar lol, i'm not a complete ignorant in HTML or CSS, i know some basics, and i can build a good ui when i have to,not a website, but this whole HTML/CSS, and HTML5/CSS3 is just insane
18:50 * FROGGS cant really describe what he likes
18:51 huf Moukeddar: yep.
18:51 huf i agree 100%
18:51 FROGGS CSS is an improvement to HTML, a great one
18:51 huf but i also see why it's spinning so fast
18:51 huf 0deploy apps used to be a dream
18:51 huf now they're commonplace with webapps
18:51 * PerlJam never did grok CSS on a deep level.
18:52 masak quote from the course I gave today: "it's 2012 and HTML forms still don't have a date picker."
18:52 huf PerlJam: it's like sql :D
18:52 masak PerlJam: selector { declaration; declaration }
18:52 masak PerlJam: there, done :)
18:52 Moukeddar PerlJam, i never had the desire to go any deeper than basics too
18:52 huf (worst simile of the day)
18:52 PerlJam :-P
18:53 huf SELECT DOM NODES FROM POO { UPDATE THEM HERE } :)
18:53 PerlJam Mainly I can do simple components (with CSS), but when it comes to putting them together, I tend to fail  (it never looks like I expect or want)
18:53 Moukeddar huf, wish there was some script like that
18:53 huf PerlJam: oh, that.
18:53 masak PerlJam: no-one knows how to do that :P
18:53 huf that's because you dont know about a million weird rules that are nowhere documented
18:54 huf except in the source code of each render engine
18:54 Moukeddar lol
18:54 PerlJam masak: Eric Meyer does!  :)
18:54 huf and are of course weirdly incompatible AND somewhat illogical
18:54 * masak hugs Eric Meyer
18:54 huf you could say, historical
18:55 masak huf: I thought CSS had gotten lots better at those things.
18:55 masak what with ACID2 and all those tests.
18:55 Moukeddar seems like we got caught in the cross-fire inside a browser war
18:55 masak several.
18:55 masak the fact that JavaScript and CSS aren't better integrated than they are is a crying shame.
18:57 FROGGS damn it, I have to 'bout that the whole day... can we talk about something cooler now please? :o)
18:57 Moukeddar does it get on your nerves too ?
18:57 FROGGS ya, 8 hours is enough
18:58 PerlJam FROGGS: I'm contemplating making a workbench for my garage.
18:58 PerlJam FROGGS: and I'm going to make a simple treadle lathe at some point.
18:58 Moukeddar PerlJam to the rescue :)
18:59 PerlJam (I made a spring-pole lathe completely out of wood (except for the rope) when I was a teenager, but I haven't really done much wood working in the last decade or so)
18:59 FROGGS PerlJam: I like to build things too (a server case for a storage system lately, but also stuff for my kids with wood)
19:00 Moukeddar server case with wood ?
19:00 * masak .oO( why do only the kids with wood get stuff? )
19:00 FROGGS no, aluminum
19:00 FROGGS *gg*
19:00 * masak slaps FROGGS with a British spelling
19:01 Moukeddar oh
19:01 FROGGS build a kitchen out of beechwood, with a illuminated glass hob
19:01 FROGGS masak: wont help :o/
19:10 MayDaniel joined #perl6
19:11 masak from https://en.wikipedia.org/wiki/Css#CSS_4 -- "There is no such thing as CSS4." o.O
19:11 masak I... never claimed there was...
19:11 tadzik haha
19:11 tadzik this paragraph almost says "There's no santa, kiddo"
19:13 masak I feel like I'm reading The Hitchhiker's Guide when they have a *section* about something starting with "there is no such thing".
19:19 Woodi hallo today :)
19:19 spider-mario joined #perl6
19:20 Woodi here something about possible versioning:  http://pastebin.com/e3CqAMQs
19:21 [Coke] SFW perl tattoo? https://i.chzbgr.com/completestore/1​2/11/25/TsjQMvkFTE--N-I2YhU7LQ2.jpg
19:22 masak one thing I've been thinking of: would it be possible to build a small simulation of module handling, including auth and ver and stuff.
19:22 masak just outside of either of the compilers.
19:22 masak sometimes it helps to prototype stuff outside of the compilers, just to make sure they're sane.
19:22 sorear good * #perl6
19:23 masak sorear! \o/
19:27 PerlJam masak: until we get some implementations going *somehow* I fear we'll just have lots of guesses and about the pros and cons of various approaches and very little actual experiences.
19:28 masak that's what I mean.
19:28 masak a prototype might be the easiest way to gain some insight in how it'd all hang together.
19:29 masak just the simplest thing that could possible emulate module authing and versioning, and where everything is to be stored.
19:29 masak with proper tests to back it up and make sure it does what we want.
19:29 [Coke] (always learning) I had an old boss who told me once you stop learning, you're dead. (and that's from when I stocked shelves at the grocery store!)
19:31 PerlJam masak: I guess it would help if we had an API for the types of operations we'd expect of "module handling" with auths + versions
19:32 masak yeah.
19:34 japhb As I tanked on this yesterday evening, it struck me that there are a lot of *conflicting* ways that people want to use versioned modules/packages
19:35 japhb I sketched out a few different designs, but each has drawbacks -- not least of which is that the design that handled the most use cases was also the most complex.
19:36 sorear worst part of being a language implementor: conflicting user demands
19:36 japhb Because you don't get too far before you are taking about managed symlink farms (and not just the 'Any' stuff I was talking about yesterday, more like alternatives + more)
19:36 japhb and other such complexities.
19:37 japhb Lord help you if you want people to make S11 happy and FHS happy at the same time
19:38 japhb pmurias: I'm not talking about not declaring dependencies.  Just that declaring dependencies per comp unit seems inappropriately granular.
19:38 PerlJam japhb: I just want some sufficiently advanced technology so that I don't have to think so much about FS issues  :)
19:38 sorear japhb: THANK YOU I HAVE BEEN SAYING THAT FOR YEARS
19:39 japhb PerlJam: I think we can get close to that sufficiently advanced technology.  But my current designs unfortunately trade complexity, fragility, and limited expandability to get there.
19:39 japhb Which sucks.
19:40 japhb sorear, heh.  :-)
19:40 sorear japhb: at this point I'm more than happy to let you think about this instead of me :p
19:40 Woodi btw. 'auth' is branch like in git too.
19:41 japhb sorear, *exaggerated sigh*  Oh good, I've proven myself just competent enough to get assigned the task.  ;-)
19:41 Woodi maybe we should think about modules as sub-languages, which are lexically scoped
19:43 japhb Woodi, one of the big discussion items yesterday was that modules do not exist in a lexical vacuum in the real world.  They have other modules, share directories, scripts, and so on, that bundle them into a package.
19:44 masak yes, 'package' (or 'project', or whatever) is an under-specified abstraction level.
19:44 masak CPAN calls it "distribution", no?
19:44 japhb ya
19:47 Woodi but { use Module; code; code; } code; # is a way of seeing packages/distributions usage
19:53 [Coke] r: say so but do not;
19:53 p6eval rakudo 7ede6c: OUTPUT«===SORRY!===␤Confused␤at /tmp/8Um_ZVCEQ1:1␤»
19:53 [Coke] awww.
19:53 japhb [Coke], attempting to become the new Perl 6 Poet Laureate?
19:54 masak [Coke]: 'so' is a listop, and 'but' cannot appear in term position.
19:55 masak r: say Whatever but do not Whatever;
19:55 p6eval rakudo 7ede6c: OUTPUT«Whatever+{<anon>}()␤»
19:55 masak r: say Rat but do not Cool;
19:55 p6eval rakudo 7ede6c: OUTPUT«Cannot look up attributes in a type object␤  in block  at src/gen/CORE.setting:887␤  in method reify at src/gen/CORE.setting:5488␤  in method reify at src/gen/CORE.setting:5390␤  in method reify at src/gen/CORE.setting:5390␤  in method gimme at src/gen/CORE.setting…
19:55 masak tee hee
19:56 sorear r: die Rat
19:56 p6eval rakudo 7ede6c: OUTPUT«use of uninitialized value of type Str in string context  in block  at /tmp/TJn1SfZWwM:1␤␤use of uninitialized variable  of type Rat in string context  in block  at /tmp/TJn1SfZWwM:1␤␤use of uninitialized value of type Str in string context␤use of uninitialized var…
19:56 sorear r: der Rat
19:56 p6eval rakudo 7ede6c: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&der' called (line 1)␤»
19:56 Woodi maybe we just trying to avoiding work ? :) if we have versions then we must use defaults (newer or official-by-marked-somewhere) as defaults and specify exceptions or always/everywhere use version numbers
19:56 sorear note, masak, that Perl 6 actually insists on bad grammar. :p
19:56 Woodi same with auth, just next letter to permutate
19:58 japhb Woodi, I'm having a hard time trying to understand what you are saying that is different from S11.
19:59 Chillance joined #perl6
19:59 Woodi I am saying that adding versions to modules make system little more complicated and we looking for magic recipe. same with auth - next bit of complication.
19:59 Woodi if we keep that additional bits we must manage them
20:00 Woodi one level of abstraction can help, eg. sets
20:00 japhb Woodi: Yes, I think we understand that, but I think you are underestimating the complexity, because there are a lot of use cases that want different things to be easy.
20:00 Woodi very possible
20:00 japhb In the waterbed theory of complexity, there are a lot of people pushing down on different spots on the bed.
20:00 japhb Building a fair amount of pressure.  :-)
20:02 diakopter heavy waterbed is underwater
20:02 Woodi I think so far we wanted to resolve everything via directory hierarchies. some middle-something is needed
20:02 sorear japhb: what does springing a leak and/or bursting correspond to in the analogy?
20:03 PerlJam Enumerating the use-cases would at least allow us to group the compatible ones or segregate the incompatible ones.
20:04 japhb Woodi, I think we pretty much agreed that directory hierarchies are not enough, that there needs to be some form of database (in the general sense) that disambiguates -- if nothing else, literally because filesystems suck compared to Perl 6.
20:04 PerlJam japhb: aye
20:04 japhb sorear, Someone running away screaming things about Elder Gods?
20:04 Woodi japhb: ok. so for me way to prototype is open :)
20:06 japhb PerlJam, I had sketched various use cases last evening, but nothing terribly coherent.  Hmmm, perhaps I should just post what I have.
20:06 japhb Woodi, prototype away.  But I'd encourage you to post your design for us to look at before you spend a bunch of time on something that won't work for what we need.
20:06 * japhb is very happy he didn't just go and implement his initial design.
20:08 Woodi japhb: I think writting all requirements on web page is needed. you constantly scarry me and I now don't know what I don't know :)
20:10 japhb https://gist.github.com/4171575
20:10 japhb ^ my notes from yesterday evening
20:10 japhb Woodi, It's always good to try to convert unknown unknowns into known unknowns.  ;-)
20:10 Woodi wow :) I did not see this yet
20:14 kurahaupo joined #perl6
20:19 Woodi for me 'use Module;' just uses module from fs, everybody happy. now: how bin/* thingies make problems ? 'distributions' are installable things and bin/* components do not realy belong to "module-version-auth" problem, IMO...
20:20 Woodi they can installable outside of modules root
20:20 japhb What if program A wants to run() a script from distribution B?  How does it specify which version of that script to run?
20:20 Woodi scripts do not have versions
20:21 japhb o.O
20:21 Woodi distros have
20:21 japhb OK, so how do you find the script that matches a certain version of a distro?
20:22 Woodi installation location for bin/ and share/ need to be known
20:22 Woodi I would like to keep things separate
20:23 japhb Yes, so: how many different scripts of the same name are you planning to store in that bin?  ;-)
20:25 Woodi ls /opt/lib/graft/
20:25 Woodi graft-2.4  log parrot-4.6.0  parrot-4.9.0  star-2012.08  star-2012.09  star-2012.10
20:25 japhb Ah, I see.  What about auth?  Just another hyphenated section of the name?
20:28 Woodi for modules in modules repo yes: root/Foo/Bar.pm.0.98.auth
20:28 japhb scripts and share directories also need to obey both auth and ver.
20:28 japhb And we still want to have a PATH that doesn't explode a la rubygems
20:30 LordVorp joined #perl6
20:31 Woodi http://pastebin.com/zqDH7DhD :)
20:31 Woodi we exploit fs or write intermediate vitual fs or db...
20:33 kurahaupo Woodi: the point of "bin" directories is that the files therein can be invoked without the full pathname, because the directories are in the user's $PATH setting. If versions are that important, then keep the file in the package's "lib" or "libexec" directory with a version suffix in the name.
20:35 * japhb trundles off to retrieve food ...
20:36 kurahaupo Generally, if version N of script S in version N' of package X must invoke version M of script Q in version M' of package Y, then version N' package X should specify the a dependency on version M' of package Y.
20:37 Woodi I refuse to parse... :)
20:37 kurahaupo Short version: it becomes a package versioning problem, as a proxy for the script versioning problem
20:39 Woodi kurahaupo: for me user can just switch used distribution... but probably we talking about few distros used at once...
20:41 Woodi maybe just cloning like in git resolve this. user should build his working env. and for others he should deplay things...
21:01 Woodi I think deriving personal packages should be easy: install two packages, derive personal from it, add things from second
21:03 colomon has -I gone back to not working in Rakudo?
21:03 colomon (command line -I, that is)
21:04 colomon never mind, I think I just got the command line order wrong.
21:04 moritz worked earlier today
21:04 * colomon is having a long, if productive, day
21:07 * moritz finishes the day by producing coconut macraoons
21:12 masak moritz: yum.
21:13 bruges_ joined #perl6
21:49 colomon joined #perl6
21:54 masak 'night, #perl6
21:54 sorear knight masak
21:56 tadzik good evenin
21:57 sorear o/ tadzik!
21:58 tadzik sorear! \o/
21:59 ryutin joined #perl6
22:00 tadzik I have to get up so early that going to sleep seems to be a waste of time
22:00 tadzik but the experience says "no, go o sleep tadzik"
22:11 dalek perl6-roast-data: 5b135d0 | coke++ | / (4 files):
22:11 dalek perl6-roast-data: today (automated commit)
22:11 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/5b135d0266
22:12 tadzik oh wow
22:12 tadzik this Invalid Pointer in free() sounds scary
22:28 [Coke] it's happening sporadically in rakudo.
22:28 [Coke] (check the history of that file over the past week.)
22:28 cognominal joined #perl6
22:46 am0c joined #perl6
22:55 grondilu joined #perl6
22:55 grondilu rn: my @a = ^2; for @a -> $x is rw { $x++ }; say @a;
22:55 p6eval rakudo 7ede6c, niecza v22-44-g0247fe5: OUTPUT«1 2␤»
22:56 grondilu eval:  my @a = (0, 1); for my $x (@a) { $x++ }; print join ' ', @a;
22:56 buubot_backup grondilu: 1 21
22:56 grondilu hum
22:57 grondilu hang on
22:58 grondilu eval:  my %a = (foo => 0, bar => 1); for my $x (map { $a{$_} } qw(foo bar)) { $x++ }; %a;
22:58 buubot_backup grondilu: 2/8
22:59 grondilu eval:  my %a = (foo => 0, bar => 1); for my $x (map { $a{$_} } qw(foo bar)) { $x++ }; values %a;
22:59 buubot_backup grondilu: 2
22:59 grondilu eval:  my %a = (foo => 0, bar => 1); for my $x (map { $a{$_} } qw(foo bar)) { $x++ }; join ' ', values %a;
22:59 buubot_backup grondilu: 1 0
23:00 grondilu rn: my %a = (foo => 0, bar => 1); for %a<foo bar> -> $x is rw { $x++ }; say %a.perl;
23:00 p6eval niecza v22-44-g0247fe5: OUTPUT«{"bar" => 2, "foo" => 1}.hash␤»
23:00 p6eval ..rakudo 7ede6c: OUTPUT«("foo" => 1, "bar" => 2).hash␤»
23:00 grondilu rn: my %a = (foo => 0, bar => 1); for %a<foo bar> -> $x is rw { $x++ }; say %a.values;
23:00 p6eval rakudo 7ede6c, niecza v22-44-g0247fe5: OUTPUT«1 2␤»
23:01 grondilu eval:  my %a = (foo => 0, bar => 1); for my $x (map { $a{$_} } qw(foo bar)) { $x++ }; join ' ', values %a;
23:01 buubot_backup grondilu: 1 0
23:02 grondilu rn: my %a = (foo => 0, bar => 1); for %a<foo bar>.clone -> $x is rw { $x++ }; say %a.values;
23:02 p6eval niecza v22-44-g0247fe5: OUTPUT«Unhandled exception: cannot increment a value of type Parcel␤  at <unknown> line 0 (ExitRunloop @ 0) ␤  at /home/p6eval/niecza/lib/CORE.setting line 270 (Mu.succ @ 9) ␤  at <unknown> line 0 (ExitRunloop @ 0) ␤  at /tmp/OWJGVqv_6V line 1 (mainline @ 8) ␤  a…
23:02 p6eval ..rakudo 7ede6c: OUTPUT«1 2␤»
23:03 grondilu rn: my %a = (foo => 0, bar => 1); for [%a<foo bar>] -> $x is rw { $x++ }; say %a.values;
23:03 p6eval niecza v22-44-g0247fe5: OUTPUT«Unhandled exception: cannot increment a value of type Array␤  at <unknown> line 0 (ExitRunloop @ 0) ␤  at /home/p6eval/niecza/lib/CORE.setting line 270 (Mu.succ @ 9) ␤  at <unknown> line 0 (ExitRunloop @ 0) ␤  at /tmp/bERBUAAlpt line 1 (mainline @ 7) ␤  at…
23:03 p6eval ..rakudo 7ede6c: OUTPUT«No such method 'succ' for invocant of type 'Array'␤  in sub postfix:<++> at src/gen/CORE.setting:1611␤  in block  at /tmp/yTpKpSFcPm:1␤␤»
23:03 grondilu ahh I give up
23:05 grondilu oh I know
23:05 grondilu rn: my %a = (foo => 0, bar => 1); for %a.map: <foo bar> -> $x is rw { $x++ }; say %a.values;
23:05 p6eval rakudo 7ede6c: OUTPUT«No such method 'count' for invocant of type 'Parcel'␤  in method reify at src/gen/CORE.setting:5467␤  in method reify at src/gen/CORE.setting:5390␤  in method reify at src/gen/CORE.setting:5390␤  in method reify at src/gen/CORE.setting:5480␤  in method reify at src…
23:05 p6eval ..niecza v22-44-g0247fe5: OUTPUT«Unhandled exception: Unable to resolve method postcircumfix:<( )> in type Parcel␤  at <unknown> line 0 (KERNEL map @ 2) ␤  at /home/p6eval/niecza/lib/CORE.setting line 363 (Any.map @ 4) ␤  at /tmp/VF7G3zBpPf line 1 (mainline @ 4) ␤  at /home/p6eval/niecza/…
23:06 grondilu rn: my %a = (foo => 0, bar => 1); for %a.map(<foo bar>) -> $x is rw { $x++ }; say %a.values;
23:06 p6eval rakudo 7ede6c: OUTPUT«No such method 'count' for invocant of type 'Parcel'␤  in method reify at src/gen/CORE.setting:5467␤  in method reify at src/gen/CORE.setting:5390␤  in method reify at src/gen/CORE.setting:5390␤  in method reify at src/gen/CORE.setting:5480␤  in method reify at src…
23:06 p6eval ..niecza v22-44-g0247fe5: OUTPUT«Unhandled exception: Unable to resolve method postcircumfix:<( )> in type Parcel␤  at <unknown> line 0 (KERNEL map @ 2) ␤  at /home/p6eval/niecza/lib/CORE.setting line 363 (Any.map @ 4) ␤  at /tmp/Yatv3Fhadp line 1 (mainline @ 4) ␤  at /home/p6eval/niecza/…
23:06 grondilu silly
23:07 grondilu rn: my %a = (foo => 0, bar => 1); for map { %a{$_} }, %a.keys -> $x is rw { $x++ }; say %a.values;
23:07 p6eval rakudo 7ede6c, niecza v22-44-g0247fe5: OUTPUT«1 2␤»
23:07 grondilu ??
23:08 grondilu eval:  my %a = (foo => 0, bar => 1); for my $x (map { $a{$_} } qw(foo bar)) { $x++ }; join ' ', values %a;
23:08 buubot_backup grondilu: 1 0
23:08 grondilu see what's wrong here??
23:09 grondilu rn: my %a = (foo => 0, bar => 1); for map { +%a{$_} }, %a.keys -> $x is rw { $x++ }; say %a.values;
23:09 p6eval rakudo 7ede6c: OUTPUT«Cannot assign to a non-container␤  in sub postfix:<++> at src/gen/CORE.setting:3077␤  in block  at /tmp/4IyeXiUq9k:1␤␤»
23:09 p6eval ..niecza v22-44-g0247fe5: OUTPUT«Unhandled exception: Writing to readonly scalar␤  at /tmp/8PhUi6niYt line 1 (mainline @ 9) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4211 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4212 (module-CORE @ 578) ␤  at /home/p6eval/niecza/li…
23:10 grondilu rn:  for ^2 -> $x is rw { $x++ }
23:10 p6eval niecza v22-44-g0247fe5:  ( no output )
23:10 p6eval ..rakudo 7ede6c: OUTPUT«Cannot assign to a non-container␤  in sub postfix:<++> at src/gen/CORE.setting:3077␤  in block  at /tmp/6qPaqNLVpi:1␤␤»
23:11 pmurias japhb: I agree, but that seems like something that is a responsibility of a tool like carton for p5
23:18 * japhb has never used Carton ... (looks it up) ... oh of *course* it's Miyagawa!  Who else would it be?  :-)
23:41 bluescreen10 joined #perl6

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

Perl 6 | Reference Documentation | Rakudo