Camelia, the Perl 6 bug

IRC log for #perl6, 2008-12-04

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs

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

All times shown according to UTC.

Time Nick Message
00:00 thei0s a solution would be the old perl5 style: ${/<single><sth>}
00:06 [particle]1 thei0s: you can look at http://feather.perl6.nl/~azawaw​i/html/spec/S05-mass/rx.t.html to see how STD.pm parses it
00:07 lambdabot Title: ../../t/spec/S05-mass/rx.t
00:07 [particle]1 it'll take me a while to load 7.2MB over my mobile phone
00:07 thei0s huh, you are on the net over a mobile?...
00:07 [particle]1 yes
00:09 [particle]1 getting an iv infusion atm in a hospital
00:09 [particle]1 sounds more dramatic than it is, though
00:09 thei0s yes, it sounds dramatic... :)
00:10 rakudo_svn r33469 | jonathan++ | [rakudo] Track changes in namespace/class symbol handling in Parrot in code to convert multis to Perl6MultiSub.
00:10 rakudo_svn r33470 | jonathan++ | [rakudo] Implement some missing methods and add some documentation to the Perl6MultiSub PMC.
00:10 thei0s it looks like this is not covered
00:19 [particle]1 pmichaud: still around?
00:20 rakudo_svn r33471 | jonathan++ | [rakudo] Re-enable multi dispatch caching, tests continue passing.
00:34 meppl gute nacht
00:46 bacek joined #perl6
00:59 pmichaud following the postcircumfix notion, $/<single> would likely parse as $<single> and not [$/]<single>
00:59 pmichaud however, I think it'd be rare to want to match $/ in a regex, and as you can see, it's easy to use brackets to disambiguate.
01:00 rakudo_svn r33473 | jonathan++ | [rakudo] .perl for proto-objects.
01:00 rakudo_svn r33474 | jonathan++ | [rakudo] First cut of .perl for Signature.
01:00 rakudo_svn r33475 | jonathan++ | [rakudo] When we have an ambiguous dispatch, report the signatures of the candidates that were ambiguous.
01:00 pmichaud [particle]: I'm back.  :-)
01:00 pmichaud will have to leave again shortly.
01:01 pmichaud jnthn: wouldn't .perl on a protoobject need to return its longname?
01:02 jnthn pmichaud: Hmm. Yes. :-)
01:02 jnthn But we haven't done .WHERE yet, I don't think.
01:02 jnthn No
01:02 jnthn .WHO
01:03 [particle]1 pmichaud: i'm trying to figure the best tree to represent cmdline options
01:03 [particle]1 actually, right now i'm testing my grammar, which parses everything in S06 so far
01:03 [particle]1 don't know if it parses correctly, but it does parse
01:03 pmichaud I tend to think of cmdline options as array-like things and not tree-like things.
01:04 lambdabot joined #perl6
01:04 pmichaud unless, of course, you mean array of tree-like things :-)
01:06 [particle]1 well, i need to differentiate between options and resources
01:06 [particle]1 but still get them in order
01:07 [particle]1 rule arguments {    [ $<argument>=[ <option> | <resource> ] ]* ... }
01:07 jnthn pmichaud: I've added a comment to the ticket for .WHO saying .perl for Proto Object should be updated when it's done.
01:07 [particle]1 is what i have now
01:08 [particle]1 that gives me in-order, and by-type arrays
01:09 pmichaud I'm not sure you can tell for each element if it's an option or a resource, though.
01:09 pmichaud $<argument> will just be an array of text-like objects with no submatches.
01:10 pmichaud have to run -- bbl
01:10 [particle]1 ok, i've modified to something better
01:10 [particle]1 thanks
01:14 DemoFreak joined #perl6
01:18 jnthn [particle]1: How's the S16 spec going?
01:19 jnthn .oO( I think it's S16... )
01:19 [particle]1 19
01:19 jnthn Ah. :-)
01:19 jnthn The one you're working on.
01:19 [particle]1 i'm writing tests for my grammar
01:19 jnthn Ah, OK. Code work. :-)
01:19 [particle]1 if i can keep a nice regression suite, i can make sure i haven't missed anything
01:20 jnthn Yes.
01:20 [particle]1 the grammar is still in progress, and will be the basis for my synopsis
01:20 [particle]1 it's code-first-speccing
01:20 jnthn Aha, good plan.
01:20 [particle]1 er, test-first-design?
01:20 [particle]1 anyway... :)
01:20 [particle]1 i'm agile like that.
01:23 alester_ joined #perl6
01:36 [particle]1 gotta run &
01:49 smg joined #perl6
02:24 eternaleye joined #perl6
02:40 [particle]1 joined #perl6
02:54 sail0r joined #perl6
03:00 sail0r left #perl6
03:06 drbean joined #perl6
03:08 cookys joined #perl6
03:14 Psyche^ joined #perl6
03:31 ChrisDavaz joined #perl6
03:39 alc joined #perl6
04:05 wayland76 joined #perl6
04:05 wayland76 Well, I'm back, but my screen's only 800x600 (anyone else hate Xorg sometimes?) :)
04:06 Patterner I violate the spirit of linux and use a binary gfx driver. but I have 2048x1536. Har.Har.Har.
04:21 alester_ joined #perl6
04:34 pnu_ joined #perl6
05:27 ChrisDavaz joined #perl6
05:35 drbean_ joined #perl6
05:37 pugs_svn r24148 | hinrik++ | [util/perl6.vim] fix interpolated contextualizers, add more keywords, more todo items
05:40 xinming joined #perl6
06:00 justatheory joined #perl6
06:02 lambdabot joined #perl6
06:08 drbean joined #perl6
06:13 ab5tract_ joined #perl6
06:13 drbean__ joined #perl6
06:13 [particle]1 left #perl6
06:16 ab5tract_ joined #perl6
06:19 drbean joined #perl6
06:26 drbean joined #perl6
06:38 jrockway joined #perl6
06:38 wayland76 Patterner: I installed the binary driver, and it locked up my whole computer when I tried to get both screens going :)
06:39 Patterner ouch
06:39 Patterner Evil ATI or Good Nvidia?
06:40 xinming_ joined #perl6
06:45 buubot joined #perl6
06:49 drbean_ joined #perl6
06:59 literal damn you Perl 6
07:00 literal vim syntax highlighting of unicode stuff and unspace is quite impossible
07:00 literal :)
07:13 stephens joined #perl6
07:13 [particle]2 joined #perl6
07:16 adc_penner joined #perl6
07:39 DemoFreak joined #perl6
08:09 ashizawa joined #perl6
08:12 adc_penner joined #perl6
08:16 literal perl6: say "yay" if 1<3
08:16 p6eval elf 24148, pugs, rakudo 33477: OUTPUT[yay␤]
08:32 pugs_svn r24149 | hinrik++ | [util/perl6.vim] literal string fixes, support hyperoperators, keyword reorganization, more todo items
08:38 iblechbot joined #perl6
08:41 literal perl6: 050.perl
08:41 p6eval rakudo 33477: OUTPUT[Leading 0 does not indicate octal in Perl 6␤]
08:41 p6eval ..pugs: RESULT["50"]
08:41 p6eval ..elf 24149: RESULT["40"␤]
08:41 moritz_ rakudo: say 050
08:41 p6eval rakudo 33477: OUTPUT[Leading 0 does not indicate octal in Perl 6␤50␤]
08:41 moritz_ elf FAIL
08:42 literal very much so
08:42 drbean joined #perl6
08:42 moritz_ rakudo's output didn't include the '50' due to a p6eval weirdness
08:43 moritz_ perl6: print 050.perl
08:43 p6eval pugs: OUTPUT[50]
08:43 p6eval ..rakudo 33477: OUTPUT[Leading 0 does not indicate octal in Perl 6␤50]
08:43 p6eval ..elf 24149: OUTPUT[40]
08:47 literal perl6: say 1e2
08:47 p6eval elf 24149, pugs, rakudo 33477: OUTPUT[100␤]
08:49 literal perl6: say 1E1
08:49 p6eval elf 24149, pugs, rakudo 33477: OUTPUT[10␤]
08:51 literal perl6: say 1.
08:51 p6eval rakudo 33477: OUTPUT[Statement not terminated properly at line 1, near "."␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤]
08:51 p6eval ..pugs: OUTPUT[*** ␤    Unexpected end of input␤    expecting ".", "\187", ">>", "=", "^", operator name, qualified identifier, variable name, "...", "--", "++", "i", array subscript, hash subscript or code subscript␤    at /tmp/X0aIBBAze0 line 1, column 7␤]
08:51 p6eval ..elf 24149: OUTPUT[Parse error in: /tmp/p6CVKUYqSH␤panic at line 1 column 0 (pos 0): Can't understand next input--giving up␤WHERE: say 1.␤WHERE:/\<-- HERE␤  STD_red/prelude.rb:99:in `panic'␤  STD_red/std.rb:76:in `scan_unitstopper'␤  STD_red/std.rb:224:in `comp_unit'␤  STD_red/std.rb:210:in
08:51 p6eval ..`_UNIT'...
08:52 moritz_ I think that's actually correct (that it's a parse error)
08:52 literal ok
08:58 bacek joined #perl6
09:07 elmex joined #perl6
09:13 kane_ joined #perl6
09:23 jferrero joined #perl6
09:29 pugs_svn r24150 | hinrik++ | [util/perl6.vim] improved highlighting of numbers, <> quote, and variables
09:49 nothingmuch joined #perl6
09:55 thei0s joined #perl6
09:56 wayland76 left #perl6
10:46 zamolxes joined #perl6
10:55 smg joined #perl6
11:10 knollix_ru joined #perl6
11:42 masak joined #perl6
11:44 ChrisDavaz joined #perl6
11:51 bacek pugs: my ($a,$b,@c) = 1..5; say ~@c
11:51 p6eval pugs: OUTPUT[3␤]
11:51 bacek hmmm
11:52 bacek pugs: my ($a,@b,$c) = 1..5; say ~@b
11:52 p6eval pugs: OUTPUT[2␤]
11:53 aindilis joined #perl6
11:54 bacek pugs: my ($a,@b,$c) = 1..5; say $c
11:54 p6eval pugs: OUTPUT[3␤]
11:54 * bacek prefer to have $c eq 5...
11:56 masak bacek: how would @b know to take exactly three elements?
11:56 bacek masak: how [*-1] works?
11:56 masak what if there were two arrays, `my ($a, @b, $c, @d)`
11:57 masak bacek: I've been thinking about that too ([*-1])
11:57 masak I honestly don't know.
11:57 bacek masak: good question :)
11:57 masak seems like magic to me
11:59 * bacek vote for postcircumfix:<?>
12:00 bacek ($a,@b,$c,@d) = 1..5; will produce @b ~~ (2,3,4)
12:01 bacek ($a,@b?,$c,@d) = 1..5; will produce @b ~~ (2) :)
12:01 masak postcircumfix ? in parameter lists already has other semantics.
12:07 Lorn joined #perl6
12:09 bacek pugs: grammar Foo { rule Token { \w+ }; rule ws { \s+| 'W'+ }; rule Tokens { <Token> <Token>}; }; say "foo WWW bar" ~~ Foo::Tokens
12:09 p6eval pugs: OUTPUT[*** No compatible multi variant found: "&Foo::Tokens"␤    at /tmp/DrOz8QlL3O line 1, column 97 - line 2, column 1␤]
12:09 bacek pugs: grammar Foo { rule Token { \w+ }; rule ws { \s+| 'W'+ }; rule Tokens { <Token> <Token>}; }; say "foo WWW bar" ~~ Foo
12:09 p6eval pugs: OUTPUT[␤]
12:09 bacek pugs: grammar Foo { rule Token { \w+ }; rule ws { \s+| 'W'+ }; rule TOP { <Token> <Token>}; }; say "foo WWW bar" ~~ Foo
12:09 p6eval pugs: OUTPUT[␤]
12:10 bacek pugs: grammar Foo { rule Token { \w+ }; rule ws { \s+| 'W'+ }; rule TOP { <Token> <Token>}; }; say "foo" ~~ Foo
12:10 p6eval pugs: OUTPUT[␤]
12:10 bacek pugs: grammar Foo { rule Token { \w+ }; rule ws { \s+| 'W'+ }; rule TOP { <Token> <Token>}; }; say ("foo" ~~ Foo).perl
12:10 p6eval pugs: OUTPUT[Bool::False␤]
12:10 bacek interesting...
12:10 masak bacek: it's not a good idea at this time to call a rule 'ws', methinks.
12:10 masak it's a bug in PGE.
12:10 bacek it's 'rule'
12:11 masak yes. so?
12:11 bacek masak: 'rule' should use 'ws' rule between tokens
12:11 bacek But I agreed with 'bug in PGE' :)
12:12 bacek s/d//
12:15 bacek pugs: grammar Foo { rule Token { \w+ }; rule ws { [\s| 'W']+ }; rule TOP { <Token> <Token>}; }; say ("foo" ~~ Foo).perl
12:15 p6eval pugs: OUTPUT[Bool::False␤]
12:15 bacek pugs: grammar Foo { rule Token { \w+ }; rule ws { [\s| 'W']+ }; rule TOP { <Token>+}; }; say ("foo" ~~ Foo).perl
12:15 p6eval pugs: OUTPUT[Bool::False␤]
12:16 bacek it's... weird
12:16 bacek perl6: grammar Foo { rule Token { \w+ }; rule ws { [\s| 'W']+ }; rule TOP { <Token>+}; }; say ("foo" ~~ Foo).perl
12:16 p6eval rakudo 33477: OUTPUT[maximum recursion depth exceeded␤current instr.: 'parrot;PGE;Match;new' pc 25 (compilers/pge/PGE/Match.pir:54)␤]
12:16 p6eval ..pugs: OUTPUT[Bool::False␤]
12:16 p6eval ..elf 24150: OUTPUT[Can't locate object method "cb__RegexDef" via package "EmitSimpleP5" at ./elf_f line 1612.␤]
12:27 jimmy__ joined #perl6
12:27 jimmy__ rakudo: grammar Foo { rule Token { \w+ }; rule ws { \s+| ''W''+ }; rule Tokens { <Token> <Token>}; }; say "foo WWW bar" ~~ Foo::Tokens;
12:27 p6eval rakudo 33477: OUTPUT[foo WWW ␤]
12:28 jimmy__ rakudo: grammar Foo { rule Token { \w+ }; rule ws { \s+| ''W''+ }; rule Tokens { <Token><.ws><Token>}; }; say "foo WWW bar" ~~ Foo::Tokens;
12:28 p6eval rakudo 33477: OUTPUT[foo WWW ␤]
12:28 jimmy__ rakudo: grammar Foo { rule Token { \w+ }; rule ws { \s+| ''W''+ }; rule Tokens { <Token><ws><Token>}; }; say "foo WWW bar" ~~ Foo::Tokens;
12:28 p6eval rakudo 33477: OUTPUT[foo WWW ␤]
12:28 jimmy__ rakudo: grammar Foo { rule Token { \w+ }; rule ws { \s+| ''W''+ }; rule Tokens { <Token> }; }; say "foo WWW bar" ~~ Foo::Tokens;
12:28 p6eval rakudo 33477: OUTPUT[foo ␤]
12:28 jimmy__ rakudo: grammar Foo { rule Token { \w+ }; rule ws { \s+| ''W''+ }; rule Tokens { <Token><ws> }; }; say "foo WWW bar" ~~ Foo::Tokens;
12:28 p6eval rakudo 33477: OUTPUT[foo ␤]
12:28 jimmy__ rakudo: grammar Foo { rule Token { \w+ }; rule ws { \s+| 'W'+ }; rule Tokens { <Token><ws> }; }; say "foo WWW bar" ~~ Foo::Tokens;
12:28 p6eval rakudo 33477: OUTPUT[foo ␤]
12:29 [particle] joined #perl6
12:29 [particle]1 joined #perl6
12:29 jimmy__ rakudo: grammar Foo { rule Token { \w+ }; rule ws { \s+| 'W'+ }; rule Tokens { <ws> }; }; say "foo WWW bar" ~~ Foo::Tokens;
12:29 p6eval rakudo 33477: OUTPUT[␤]
12:31 moritz_ rakudo: Grammar A { rule stuff { \d+ \d+}; token ws { 'A'+ } } say "3A4" ~~ A::stuff
12:31 p6eval rakudo 33477: OUTPUT[Statement not terminated properly at line 1, near "A { rule s"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤]
12:31 moritz_ rakudo: Grammar A { rule stuff { \d+ \d+ }; token ws { 'A'+ } }; say "3A4" ~~ A::stuff
12:31 p6eval rakudo 33477: OUTPUT[Statement not terminated properly at line 1, near "A { rule s"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤]
12:31 moritz_ rakudo: grammar A { rule stuff { \d+ \d+ }; token ws { 'A'+ } }; say "3A4" ~~ A::stuff
12:31 p6eval rakudo 33477: OUTPUT[␤]
12:31 ejs joined #perl6
12:31 moritz_ so overriding <ws> from rakudo doesn't work.
12:31 jimmy__ yes
12:34 masak ...and calling something 'ws' is not a good idea from the beginning, due to ticket number <mumble mumble>.
12:35 moritz_ masak: calling something 'text' is equally confusing, as you surely remember :-)
12:35 masak yes, that's the bug.
12:35 masak too bad, 'text' is a useful, common name
12:36 moritz_ rename that to ____pge__text__should__never__be__exposed, and nobody will ever find out with black box testing :-)
12:38 masak I've been thinking along similar thoughts.
12:38 masak but I'll leave that in pmichaud++'s competent hands.
12:58 smg joined #perl6
13:02 pmichaud the problem is that PGE confuses Match's .text with a grammar's .text
13:02 pmichaud in the case of <ws>, the issue is that the thing that invokes a rule isn't creating the correct sort of Match object.
13:04 pmichaud right now Match objects are intimately tied to grammars
13:04 pmichaud (so that inheritance can work)
13:05 masak pmichaud: are you blocking on the spec in some regards?
13:05 jimmy__ <.ws> and <ws> are the same ?
13:05 pmichaud same rule, yes.  <.ws> is non-capturing, <ws> is capturing.
13:05 masak jimmy__: the latter "captures" into $/, the former doesn't.
13:06 jimmy__ i saw.
13:06 jimmy__ so, PGE should be improved ?
13:07 moritz_ I think it's the rakudo "bindings" to PGE
13:07 moritz_ (but PGE also needs to be improved, it needs to handle longest-token-matching)
13:07 pmichaud the immediate issue is rakudo's interfacing with PGE
13:07 masak ...and Perl 6 closures.
13:08 pmichaud the next issue after that is that we don't have the official syntax for invoking a rule from a grammar
13:08 pmichaud and the issue after that is that PGE needs to be improved to follow the spec (whatever it ends up being)
13:08 moritz_ and the spec might change a bit, if $/.tokens or some such will be incorporated
13:13 pmichaud (that could actually be a somewhat significant change)
13:15 apeiron joined #perl6
13:20 rakudo_svn r33478 | pmichaud++ | [rakudo]: spectest-progress.csv update: 227 files, 4703 passing, 0 failing
13:23 vixey joined #Perl6
13:32 masak ($/.tokens)++
13:42 [particle]1 joined #perl6
13:49 cognominal joined #perl6
13:53 cognominal joined #perl6
13:56 [particle]2 joined #perl6
13:56 [particle]3 joined #perl6
14:02 cdavaz joined #perl6
14:07 Lunchy joined #perl6
14:09 Lunchy joined #perl6
14:19 lisppaste3 joined #perl6
14:23 ejs1 joined #perl6
14:36 jhorwitz joined #perl6
14:41 zamolxes joined #perl6
14:49 hercynium joined #perl6
15:00 * ruoso later &
15:20 rindolf joined #perl6
15:21 apeiron joined #perl6
15:23 Exodist joined #perl6
15:24 masak joined #perl6
15:33 DemoFreak joined #perl6
15:35 masak rakudo: say "我想租一辆自行车。"
15:35 p6eval rakudo 33480: OUTPUT[我想租一辆自行车。␤]
15:35 masak impressive.
15:55 masak rakudo: say "我想租一辆自行车。".chars
15:55 p6eval rakudo 33481: OUTPUT[9␤]
15:55 masak yup.
15:57 * moritz_ didn't know that p6eval was that unicode safe :-)
15:57 masak no, it's a positive surprise :)
16:05 stephens joined #perl6
16:06 ejs joined #perl6
16:17 thei0s1 joined #perl6
16:20 zamolxes_ joined #perl6
16:23 thei0s1 joined #perl6
16:26 smg joined #perl6
17:00 [particle]3 http://creativecommons.org/weblog/entry/11045
17:00 lambdabot Title: Non-Commercial study questionnaire - Creative Commons
17:09 ejs joined #perl6
17:12 masak [particle]: I agree with Stephen, the last to comment on that post. the 21% question is impossible to answer correctly.
17:15 [particle] i'm at 56% and sick of this survey already
17:16 masak I dropped it at 21%, but even before that I didn't like the alternatives.
17:16 [particle] agreed
17:17 [particle] i wonder how much of pugs's CGI.pm rakudo can parse now
17:18 [particle] and the other ext/ modules
17:18 [particle] it'd be a good test for STD.pm, too, i suppose
17:18 masak aye
17:18 [particle] masak: do you have highlighted November code yet?
17:19 masak [particle]: define 'highlighted November code'.
17:19 [particle] azawawi's script
17:19 [particle] run against November codebase
17:19 masak [particle]: haven't tried that, no.
17:19 masak what should I expect to find?
17:20 [particle] wonder if there's a way to plug in a highlight engine to SVN::Web etc
17:20 masak would be neat.
17:21 masak [particle]: but don't go revive CGI.pm before we can develop a decent Web.pm :)
17:21 * masak has to go now
17:27 buu joined #perl6
17:36 alester joined #perl6
17:37 jan_ joined #perl6
17:46 meppl joined #perl6
17:49 ejs joined #perl6
17:56 M8YCjC56g joined #perl6
17:59 diakopter joined #perl6
18:02 charsbar_ joined #perl6
18:04 diakopter joined #perl6
18:07 araujo joined #perl6
18:09 shachaf joined #perl6
18:20 schmalbe joined #perl6
18:21 apeiron joined #perl6
18:22 [particle]1 joined #perl6
18:22 smg joined #perl6
18:29 diakopter joined #perl6
18:34 sri_kraih joined #perl6
18:54 adc_penner joined #perl6
18:55 thepler rakudo: (undef === undef).say
18:55 p6eval rakudo 33482: OUTPUT[0␤]
19:10 rakudo_svn r33483 | bernhard++ | [codingstd] Fix some glitches: c_parens, tabs and trailing_space
19:15 PerlJam Are most of the perl6 modules that people have developed in the pugs repo?
19:16 moritz_ PerlJam: either there, or in the repo of the November wiki
19:17 ryanc joined #perl6
19:21 ejs joined #perl6
19:24 renormalist joined #perl6
19:26 [particle]1 joined #perl6
19:40 [particle] joined #perl6
20:10 rakudo_svn r33487 | pmichaud++ | [rakudo]:  Add inline PIR to Perl 6 subroutines.
20:10 mncharity joined #perl6
20:16 mncharity no coding yesterday.  did see a fun talk - Saul Greenberg's Enhancing Creativity Through Toolkits  http://crpit.com/confpapers/CRPITV28Greenberg.pdf  http://www.google.com/url?sa=U&amp;start=1&amp;q=h​ttp://www.researchchannel.org/prog/displayevent.as​px%3FrID%3D11041%26fID%3D569&amp;ei=aDk4SasZjLL0BM​ydoZgP&amp;usg=AFQjCNHmwemUxuzeuImMSxqaAYs-qQ8oGg
20:16 lambdabot Title: Enhancing Creativity with (Groupware) Toolkits
20:16 mncharity (I don't know that the web video matches the talk)
20:17 mncharity sigchi committee in town.  lots'o talks. :)
20:20 rakudo_svn r33488 | moritz++ | [cage] trailing whitespace
20:21 mncharity cage?
20:21 moritz_ mncharity: cage cleaning for the bird
20:22 mncharity :)
20:31 Hinrik joined #perl6
20:31 Hinrik perl6: say 'a\\b\''
20:31 p6eval pugs, rakudo 33489: OUTPUT[a\b'␤]
20:31 p6eval ..elf 24150: OUTPUT[a\b\'␤]
20:32 moritz_ mncharity: do you haev a bug tracker for elf? ;-)
20:33 moritz_ perl6: say 'a\\b\''.chars
20:33 p6eval pugs, rakudo 33489: OUTPUT[4␤]
20:33 p6eval ..elf 24150: OUTPUT[5␤]
20:35 pugs_svn r24151 | moritz++ | [t/spec] add a quoting test by Hinrik++
20:43 mncharity moritz_: ???  To report to the mncharity and TimToady of early Spring that some combination of STD and STD_red is being sloppy with strings?  Known.  Even at the time.  Since then, the STD.pm string code has been rewritten, and STD_red is in maintenance-only mode, awaiting disposal.  Since it turned out no one was interested in writing p6 prelude, elf has been about bootstrap, not bits and pieces.
20:44 mncharity So a bug tracker wouldn't help.  They're for coordinating work among developers, or tracking details.  No developers, and nothing to track just now.  Just a clearly staked path to bootstrap, being walked slowly...
20:45 moritz_ mncharity: it seems you didn't notice the smiley ;-)
20:45 mncharity ;)
20:45 PerlJam moritz_: when you hit that close to a nerve, smileys don't help  :)
20:46 mncharity I think it was an "elf FAIL" comment in the backlog seen a few minutes ago. :)
20:46 moritz_ sorry for that
20:47 moritz_ not important for bootstrap, that one
20:47 mncharity oh, so very no problem.  #re sorry
20:47 moritz_ mncharity: what's your plan right now? first bootstrap, then try to get more developers on board? or the other way round?
20:48 mncharity re string, STD_blue is giving a valid string parse, so modulo my having a very incomplete understanding of nibbles and quibbles and etcbibbles, escaping should be working in elfblue.
20:48 moritz_ lol # etcbibbles
20:50 ruoso joined #perl6
20:50 mncharity re plan, bootstrap and then developers.
20:51 ruoso Hello!
20:51 mncharity Though I'd welcome anyone working on prelude written in p6, either on elf or rakudo.
20:51 mncharity hi ruoso :)
20:51 ruoso hi mncharity
20:51 ruoso I must say I make mine your words
20:51 ruoso SMOP will need prelude written in Perl 6
20:51 ruoso also
20:52 ruoso mncharity, are you aware of v6/smop/src-s1p/*pm
20:52 ruoso ?
20:53 mncharity there are bits of prelude scattered around.  in pugs, elf, elsewhere(?).  any thoughts on how to attract someone to work on prelude?
20:53 mncharity re aware, no, well, not sure, looking...
20:53 mncharity http://svn.pugscode.org/pugs/v6/smop/src-s1p/
20:53 lambdabot Title: pugs - Revision 24151: /v6/smop/src-s1p
20:54 ruoso mncharity, the src-s1p directory is my intended place for prelude in SMOP
20:58 mncharity neat.  the ClassHOW and Object code aren't things elf can handle any time soon.  But once you start putting in things like pug's Prelude.pm, some of that should be elfable.
20:58 sjwalter joined #perl6
20:58 mncharity http://svn.pugscode.org/pugs/src/perl6/Prelude.pm
21:00 mncharity elf currently has http://svn.pugscode.org/pugs/​misc/elf/elf_h_src/Prelude.pm  http://svn.pugscode.org/pugs/mis​c/elf/elf_h_src/PrimitivesP5.pm  http://svn.pugscode.org/pugs/misc​/elfish/on_sbcl/PrimitivesSBCL.pm
21:00 mncharity there's various other stuff scattered around.  I'm not sure how much of it is worth salvaging, if any.
21:00 eternaleye joined #perl6
21:02 ruoso mncharity, SMOP is not able to run all of ClassHOW and Object yet...
21:02 ruoso we're working on mildew to make it work
21:03 ruoso I started by ClassHOW and Object just because SMOP is all about OO, so that was pretty much the most reasonable place to start
21:03 mncharity nod
21:03 ruoso mncharity, take a look at mildew/t
21:03 ruoso you'll see what it can run alreadey
21:04 ruoso I need the map operator now, so I can implement ClassHOW.can
21:04 ruoso and then I can have ClassHOW.dispatch
21:05 mncharity One suggestion, as I notice the empty Any.pm.  I'd suggest organizing core methods by method, rather than scattering them across class files.  Spec and spec changes tend to impact methods, rather than classes.  And there are a _lot_ of classes.
21:06 ruoso mncharity, take a look at v6/mildew/Map.pm
21:06 mncharity http://svn.pugscode.org/pugs/v6/mildew/t/
21:06 ruoso that's a sketch on how the map implementation would be added to the prelude
21:06 lambdabot Title: pugs - Revision 24151: /v6/mildew/t
21:06 mncharity http://svn.pugscode.org/pugs/v6/mildew/Map.pm
21:07 mncharity knowhow?
21:07 apeiron joined #perl6
21:07 ruoso mncharity, yes... a Pure Prototype Object
21:07 ruoso (it should appear in S12 sometime)
21:07 ruoso *someday
21:08 ruoso knowhow is important for bootstrap reasons
21:08 ruoso and is important for prototype-based oo
21:09 ruoso mncharity, I should leave now...
21:09 * ruoso later &
21:09 * ruoso will try to backlog...
21:09 mncharity cheers :)
21:09 mncharity thanks for the pointers
21:15 mncharity moritz_: re bootstrap and then developers: though, there are a couple of people around Boston who are interested in pushing p6, so the idea has been floating around of doing a sprint.  Boston ruby group manages a weekly meeting (to hack; normal meeting; bar; ruby movies).  Perhaps could try for a 1st and 3rd Wednesday of the month, elf and p6 hacking at <someplace with wifi and free food>.
21:16 mncharity Or perhaps at some perl conference, though my travel budget approximates zero at the moment.
21:19 Tene Anyone here feel like looking up some Perl 6 specs for me about the behavior of try {} blocks?
21:22 mncharity /me shakes head, sorry, hacking...
21:25 zamolxes joined #perl6
21:25 bacek joined #perl6
21:26 moritz_ "A try block by default has a CATCH block that handles all exceptions by ignoring them."
21:26 moritz_ http://perlcabal.org/syn/S​04.html#Exception_handlers
21:26 lambdabot Title: S04
21:28 moritz_ that seems to imply control exceptions, IMHO
21:28 moritz_ (and notices that in Perl 5 an eval { ... } block also catches return() )
21:28 moritz_ $ perl -wle 'sub f { eval { return; }; print "alive" }; f()'
21:28 moritz_ alive
21:40 apeiron joined #perl6
21:42 pugs_svn r24152 | hinrik++ | [t/spec] remove Windows line endings from S06-signature/type-capture.t
21:50 mncharity moritz_: any thoughts on attracting developers?
21:53 moritz_ mncharity: docs & blog
21:54 moritz_ a nice architectural overview (that explains all abbreviations ;-), and a short hacking howto (where to change what, what do you need to rebuild, how to test etc.)
21:57 mncharity good thoughts.  blog is raw exposure.  the others are about visitor experience.  have we been getting visitors, and if so, do we have feedback on what made it more/less likely they would contribute?
21:59 moritz_ when I mention elf on perlmonks, I often got asked "elf? what's that?", so I gave them the link
22:00 moritz_ I thought about contributing to elf as well, but I found it a bit deterrent (description used abbreviations I didn't know, so little docs, no overview what features are working)
22:01 moritz_ mncharity: my offer for you is: I can try to set up a 'make spectest' for elf similar to that of rakudo
22:02 moritz_ mncharity: ie a test target that's supposed to run cleanly, and for which it's worth to fudge tests
22:02 moritz_ mncharity: I don't know if that makes sense in the current stage, but as soon as it does, tell me
22:08 mncharity re deterrent, ah, good to know.   re offer/spectest, nice.   re "as soon as it does, tell me", will do.
22:08 mncharity is parrot ready to accept prelude code written in p6?
22:10 moritz_ not yet (you knew that, didn't you?)
22:10 moritz_ though most pieces are in place now
22:10 moritz_ class Array is also { ... } works now
22:11 moritz_ and today pmichaud added inline PIR to rakudo
22:11 moritz_ don't know what else is needed right now
22:11 pmichaud we're getting closer
22:11 pmichaud we need to fix method generation (jonathan++'s version is a little heavier weight than I want)
22:11 PerlJam what does prelude need that rakudo doesn't have?
22:12 pmichaud PerlJam: depends on what we expect to write in Prelude.
22:12 pmichaud rakudo will be moving nearly all of its builtins to Perl 6 wrappers, at least.
22:12 pmichaud if not actual Perl 6 code within the wrappers.
22:13 moritz_ (and if it's just to get the signatures right ;-)
22:13 mncharity moritz_: re "(you knew that, didn't you?)", no.  last I heard, a few months back, was rsn.
22:13 pmichaud well, signatures and multidispatch are one of the driving reasons for it, yes.
22:13 pmichaud jonathan and I both took vacations.
22:13 PerlJam well, I'm looking at pugs/src/perl6/Prelude.pm.  I think whatever is in there is what we expect to write  :)
22:13 moritz_ mncharity: that's wasn't entirely serious as well :)
22:14 * PerlJam doesn't know what "is builtin" or "is unsaef" are supposed to do
22:14 moritz_ PerlJam: 'is unsafe' was obviously used for pugs' safe mode
22:14 moritz_ PerlJam: and I guess 'is builtin' is a pugs hack
22:14 pmichaud I expect rakudo's Prelude to be much longer than the current Prelude.pm
22:15 pmichaud and fairly different, as well.
22:15 pmichaud we'll have a lot of     class Any is also { method ... ; method ... ; method ... ;}
22:15 pmichaud to define most of the builtins
22:15 mncharity yay
22:16 PerlJam feels like monkey patching, but official  :)
22:16 pmichaud well, when we can write the methods in pure Perl 6, we will; when we can't, we'll use inline PIR.
22:16 pmichaud eventually I suspect we'll separate out the pure Perl 6 parts and leave them with stubs to be filled in with the PIR
22:17 pmichaud so that we'll have a pure Perl 6 Prelude.pm with { ... } for the parts we didn't write in Perl 6
22:17 pmichaud and a PIR-Prelude.pm with q:PIR for the parts we wrote in PIR (filling in the parts we didn't write in Perl 6)
22:17 pmichaud all of which will be Perl 6 syntax.
22:18 pmichaud I better disappear for a while or I won't have my presentation finished.  bbl.
22:19 mncharity thanks for the update.:)
22:24 Hinrik perl6: say 1 if 'a'eq'a'
22:24 p6eval elf 24152, pugs, rakudo 33492: OUTPUT[1␤]
22:29 Limbic_Region joined #perl6
22:32 b_jonas joined #perl6
22:33 b_jonas hello! are the slides (or transscript or something) of the 2008 state of the onion keynote up somewhere?
22:35 b_jonas ah, irc logs say from 2008-10-08 that "TimToady probably won't publish this year's since it was a tech talk/demo of STD.pm
22:35 b_jonas "
22:35 b_jonas pity
22:49 meppl good night
22:50 b_jonas by the way, thanks for fixing the search feature of the irc logs page
22:50 b_jonas -> moritz
22:51 moritz_ b_jonas: I didn't fix anything, the failures seem random :/
22:56 [particle] ok, my grammar successfully parses all cmdline options in S06. baseline complete. now to attack my additions.
22:58 b_jonas heh
23:01 Limbic_Region b_jonas - Larry emailed me back - said that a very similar talk is available via youtube at his google chicago talk
23:01 Limbic_Region b_jonas - http://www.youtube.com/watch?v=JzIWdJVP-wo
23:01 lambdabot Title: YouTube - Larry Wall Speaks at Google
23:02 b_jonas ah, thanks
23:03 b_jonas I wonder what should go to [id://153]
23:03 Tene Can someone tell me what $! should be after a try{} that threw an exception?
23:03 Limbic_Region b_jonas - btw, your homenode pic on perlmonks no longer matches the caption
23:04 b_jonas I still have a caption?
23:04 b_jonas I thought I removed that so I don't have to change it
23:04 b_jonas let me see
23:04 b_jonas uh huh
23:05 b_jonas ok, this time I'll just remove it and don't add what this picture is (chrismas decoration on a shop in Wien)
23:06 Tene ah, there's the spectest
23:06 moritz_ Tene: which could be wrong :/
23:06 Tene t/spec/S02-magicals/dollar_bang.t
23:06 dr_df0 joined #perl6
23:07 dr_df0 pugs: my @a=<1 2 3 4>; my @b=<1 2 3 4>; my @c = @a >>+<< @b; say @c
23:07 p6eval pugs: OUTPUT[2468␤]
23:07 Limbic_Region pugs: my @a=<1 2 3 4>; my @b=<1 2 3 4>; my @c = @a >>+<< @b; say @c.perl
23:07 p6eval pugs: OUTPUT[\(2.0, 4.0, 6.0, 8.0)␤]
23:07 Tene I guess it makes sense to have Exception always return 1 for get_bool...
23:08 moritz_ yes, because if $! isn't set it's undef, not an Exception
23:08 Tene Are there any false exception values?
23:08 Tene I would'nt think there were...
23:08 adc_penner2 joined #perl6
23:08 [particle] Exception;false
23:08 [particle] :)
23:08 moritz_ die 0;
23:08 [particle] does exit throw an exception?
23:09 Tene so: try { die 0; }; say 'True' if $!;
23:09 moritz_ dunno if it should act as a proxy for capture that die() saw
23:09 Tene Does that print?
23:09 moritz_ dunno
23:09 Tene Maybe it should call get_bool on its payload?
23:12 moritz_ "If you test a C<Failure> for C<.defined> or C<.true>, it causes C<$!>
23:12 moritz_ to mark the exception as I<handled>;"
23:12 moritz_ "Because the contextual variable C<$!> contains all exceptions collected in the
23:12 moritz_ current lexical scope [...]"
23:12 moritz_ uhm.
23:12 Tene moritz_: could I get you to review dollar_bang.t and update it?
23:13 moritz_ So $! is a collectioin of exceptions...
23:13 PerlJam moritz_: but you probably only get to look at the "top" one in item context.
23:13 moritz_ Tene: I tried that not-so-long-ago, and couldn't find the appropriate specs
23:13 moritz_ Tene: maybe I need to send a mail to p6l for clarification, so that might take a while.
23:14 [particle] assume it means something that's reasonable to implement until we know what it really means
23:15 moritz_ actually just having Excpetion.get_bool returning always 1 would give us a few more passing tests
23:19 Tene I'll have it always return 1 for now.  Can you ask p6l?
23:20 moritz_ I'm writing the mail right now.
23:22 moritz_ Tene: actually dollar_bang.t looks saner than I remembered ;-)
23:23 Tene make sure to ask for clarification of what the default argument referred to in test 8 is.
23:24 Tene or 9 or whichever
23:24 Tene the end
23:26 moritz_ will do.
23:27 pugs_svn r24153 | tene++ | Unfudge a few tests to track new passes in Parrot.
23:30 moritz_ mail sent, going to bed now.
23:34 Tene Thank you.
23:50 Hinrik perl6: say 1x2
23:50 p6eval pugs: OUTPUT[*** ␤    Unexpected "x2"␤    expecting "_", fraction, exponent, term postfix, operator, ":" or ","␤    at /tmp/1f70jZkvX3 line 1, column 6␤]
23:50 p6eval ..elf 24153: OUTPUT[11␤]
23:50 p6eval ..rakudo 33494: OUTPUT[Syntax error at line 1, near "say 1x2"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤]
23:50 Hinrik perl6: say 1 x 2
23:50 p6eval elf 24153, pugs, rakudo 33494: OUTPUT[11␤]
23:52 Hinrik perl6: say "ok" if 1eq1
23:52 p6eval elf 24153: OUTPUT[ok␤]
23:52 p6eval ..pugs: OUTPUT[*** ␤    Unexpected "eq1"␤    expecting "_", "-", "+", exponent, term postfix or operator␤    at /tmp/ZSuf0Brfre line 1, column 14␤]
23:52 p6eval ..rakudo 33494: OUTPUT[Syntax error at line 1, near "if 1eq1"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤]
23:52 Hinrik I see

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

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs