pugscode.org/ | nopaste: sial.org/pbot/perl6 | pugs: [~] <m oo se> (or rakudo:, kp6:, smop: etc.) || We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by TimToady on 25 January 2008.
literal hi 01:47
is this book still up-to-date? www.amazon.com/Perl-6-Parrot-Essent...amp;sr=1-2
lambdabot tinyurl.com/2mc43c
mugwump literal: that book is good, and much of what is in there is still current 02:05
literal ok
mugwump the parrot stuff in there is getting out of date, but should still mostly work
just not current practice
literal do you know anything about this one? www.amazon.com/exec/obidos/tg/detai...p;v=glance
mugwump there have been some design decisions; another edition is probably due once STD-6.0 finalizes IMHO
lambdabot tinyurl.com/2kk7cq
mugwump no idea 02:06
wow, published this yes
year, even
literal yeah
mugwump I don't know the authors though 02:07
literal one other thing
mugwump maybe someone else here does ... I'm quite out of touch comparatively :)
literal are you still maintaining Lingua::Translate?
literal hasn't had a release in 3 years, I was wondering of writing a few patches for it 02:08
s/wondering/thinking/ 02:11
mugwump Oh, sure
Someone asked for co-maint on it once, but nothing came of it 02:12
If you can make a new release I'll be happy to give you co-maint on it
literal awesome
mugwump you can clone git://utsl.gen.nz/Lingua-Translate if you like 02:13
literal ok
mugwump not exactly the tidiest history mind ;)
literal heh 02:14
mugwump I don't have a copy of the SysTrans software any more to test that part of it
literal would you consider this eligible for inclusion? search.cpan.org/~hinrik/Lingua-Tran...terTran.pm 02:15
lambdabot Title: Lingua::Translate::InterTran - A Lingua::Translate backend for InterTran. - sear ..., tinyurl.com/2ysps3
literal <-- hinrik
mugwump oh, nice 02:16
that'll do, have co-maint and try not to break anything ;) 02:18
literal will do
pugs_svnbot r20077 | putter++ | [redsix]: redsix is now the code, and redsix.rb a symlink 'hack to make emacs happier'. Rather than the other way around. 14:47
diff: dev.pugscode.org/changeset/20077
mncharity Sigh. Fyi, r20077 has made svn unhappy. Attempting cleanup. 14:53
WARNING - do not svn up to r20077 ! 14:56
PerlJam unhappy how? 14:58
cognominal_ mncharity, found how to hook the yaml dumper. but apparently I need to tweak it 14:59
all I get so far is rakudo --target=parse t.p6 15:00
---
{
"VAR1" : !Perl6::Grammar { ... },
}
apparently I need a specail libsyck 15:04
pugs_svnbot r20078 | putter++ | Yet another attempt at repairing svn after r20077. 15:14
diff: dev.pugscode.org/changeset/20078
pugs_svnbot r20079 | putter++ | Continuing r20077 cleanup: redsix is gone, now deleting redsix.rb. 15:16
diff: dev.pugscode.org/changeset/20079
mncharity Ok, if you svn'ed up after r20077 but before r20079, to unwedge you svn, do: rm -rf misc/pX/Common/redsix; svn cleanup; svn up 15:18
*your
Anyone know what svn reflectors are running? darcs, whatever? Or by whom? They may be wedged now. 15:20
Every once in a while, I have a flash of wondering if I'm being excessively paranoid - a moment of, well, not trusting, but of giving my tools the benefit of the doubt. This error in judgment usually self corrects quite rapidly. :/ 15:24
Sorry folks. 15:25
mncharity cognominal_: hi. sorry, distracted. 15:25
mncharity re rakudo, you saw misc/winter_jig/ast_rakudo ? 15:27
oh, if you got that far, nevermind. 15:28
PerlJam: svn up -r20077 both fails, and leaves the local svn copy broken - can't svn cleanup, and can't svn rm --force the symlink file it's broken on. The workaround is to svn delete URL/path-to-file , bypassing the local copy entirely. 15:32
mncharity *The workaround to fix the svn repository. 15:33
To fix the svn local filesystem if it's become thus wedged: rm -rf misc/pX/Common/redsix; svn cleanup; svn up . 15:34
pmurias mncharity: hi 15:35
pmurias what do the "fates" in STD5.pm refere to? 15:35
moritz_ re 15:37
mncharity cognominal_: I ran rakudo --target=parse (non-yaml) against pugs's t/spec/. Using misc/winter_jig/ast_rakudo/parse_t_spec.t. There more failures than I expected. Which leaves me wondering,
avar yes the parser fails on all sorts of strange stuff 15:39
[particle] what does it leave you wondering? rakudo is far from complete 15:40
mncharity pmurias: re fates, my understanding is fuzzy, since STD_red skipped that part, but I believe it is how TimToady's "longest matching DFA engine" (for some reason the name keeps not sticking with me), after a successful matching path is found, tells all the regex parts which way to go (no search required - already done), as they are rerun for sideeffects. or something like that.
[particle] and the pugs tests are not designed for partial implementations of perl 6
moritz_ avar: you're free to wonder why it fails once STD.pm runs 15:41
mncharity re wondering, sorry, "interleave quick reply" wasn't, let's see... 15:41
mncharity there are several (well, lots) of rakudo tasks which might be interesting: () getting a feel for why the parses are failing, () fixing them, () getting yaml ast dumping working, () getting yaml dumping of pge match results working. 15:43
without having a feel for why the parses are failing (I haven't looked yet), I'm unclear on whether getting an ast dump of them working will be of immediate use or not. 15:44
I basically have no idea how much/little rakudo is parsing, compared with say kp6.
[particle] i'm not sure of that, either
i guess you could try running kp6 on t/spec 15:45
pmurias mncharity: kp6 dosn't parse expressions correctly
mncharity err, rephrase, I'm unsure whether the dialect rakudo parses is currently large enough, or includes the "right things", to permit throwing kp6-style "restricted dialect but still usable code" at it. 15:48
[particle] in many of the spec tests, there are directives to skip lines that rakudo can't parse 15:50
#?rakudo skip "parsefail"
mncharity eg, yes, I expect its operator precedence parser gives it much more real expressions than kp6 has. but if, for silly example instance, it can't parse "class" of a class definition, then you couldn't use it to parse "class A {...}", and it wouldn't yet be of help as a frontend.
[particle] timtoady++ created t/spec/fudge* to handle this preprocessing 15:51
so far, kp6 hasn't taken advantage of fudge
rakudo works very well with classes
mncharity I've no idea whether the parse failures are of the (very common) form of "wierd little thing one really doesn't need in practice, but is used in the test file, blows out the test". Or whether, to be silly again, scalars don't parse. 15:52
[particle] iirc named parameter support isn't all there yet 15:53
otherwise, most common things are parsed
mncharity If the latter, yaml ast dumping isn't of immediate use. If the former, one simply "just don't *do* that"'s.
re "most common things are parsed", yay :) 15:54
An interesting test could be to run rakudo against the p6 files kp6 can parse. Eg, it's test cases. And see if any parsefail for reasons which are worrying. 15:55
mncharity v6/v6-KindaPerl6/t/kp6/*.t 15:56
cognominal_ mncharity, I have submitted a patch that support part of the pair syntax 15:57
pasteling "[particle]" at 24.19.3.148 pasted "rakudo run against kp6 tests" (2091 lines, 100K) at sial.org/pbot/30459 15:58
[particle] everything fails to parse 15:59
mncharity looking...
[particle] however, most failures are likely related to a few root causes
mncharity cognominal_: progress :)
re few root causes, indeed. 16:00
cognominal_ I woud really like to use a grammr derived from rakudo to template a small web site. 16:01
mncharity :) 16:02
moritz_ cognominal_: actually I thought about a PGE based template system as well ;-) 16:03
I'm currently updating dev.perl.org/perl6/ - any thoughts what's most urgent? 16:04
avar [particle]: I did that a few months ago and some of it parsed
avar [particle]: A lot of those are due to anonymous =begin/=end blocks. I submitted a patch to fix that but it was rejected 16:04
[particle] 1) i've been working on a pod6 grammar, kp6 could steal that 16:05
2) i've been changing t/spec tests to pod6 16:06
[particle] so, i think kp6 should migrate from pod5 to pod6, and it's definitely possible now 16:06
mncharity cognominal_: so I'm wondering if () get rakudo to parse the kp6 tests (or to only not parse for reasons which can be workaround when using it as a frontend), () add yaml match dumping to pge, to give the world access to pge, might be "more useful in themselves" short term goals than the ast yaml dump? thoughts?
mncharity the first seem at least something vaguely like a prerequisite for then actually using ast yaml dumps as a frontend. 16:07
[particle] pod6 grammar: svn.perl.org/parrot/trunk/languages...grammar.pg
cognominal_ mncharity, I just mailed the author of the yaml dumper to make availaible his libsyck library tuned for parrot
[particle] what yaml dumper?
why modify libsyck? 16:08
i wrote a yaml dumper written in pir
it's just not finished yet
mncharity goes to finish cleanup up from svn mess.
cognominal_ I am speaking of library/YAML/Dumper.pir which used a modified libsyck library 16:10
[particle] ok. i wrote that. 16:11
i copied it from parrot's Dumper.pir
never modified the author info
cognominal_ so what's the story about the modified libsyck? 16:12
[particle] i don't know what you're talking about 16:13
pugs_svnbot r20080 | putter++ | Continuing r20077 cleanup: recover r20076's redsix.rb as redsix. 16:13
diff: dev.pugscode.org/changeset/20080
cognominal_ [particle], maybe you wrote yaml_dumper.pir 16:15
mncharity sigh. and no time for actual work now. ah well.
cognominal_ there may be two yaml parrot libraries
[particle] i wrote yaml_dumper.pir, YAML/Dumper.pir, YAML/Dumper/Default.pir and YAML/Dumper/Base.pir
all of which are compiled down to yaml_dumper.pbc 16:16
cognominal_ the library I used call libsyck
or so it says
ho, that the parser... 16:17
I am confused
mncharity To recap, svn co from r20077 until r20079, would leave a broken local svn filesystem. Folks running, or knowing of, svn mirrors for pugs, please try to "svn up", to see if they got wedged. Sorry for the hassle. Thank you. 16:18
cognominal_ it gets confusing when navigating too much sources
mncharity It's sufficient to "svn up"(or whatever) in misc/pX/Common, if you don't want to affect the rest of your world. 16:19
bbl &
[particle] yes, the parser may use libsyck 16:20
cognominal_ www.yamli.com/ # oops
pugs_svnbot r20081 | moritz++ | [STD.pm] more perlhints 16:31
diff: dev.pugscode.org/changeset/20081
mncharity two unexpected hours for hacking! :) 17:12
pugs_svnbot r20082 | putter++ | redsix/redsix: now emacs modes and syntax highlighting "just work". 17:31
diff: dev.pugscode.org/changeset/20082
lambdabot Title: Changeset 20082 - Pugs - Trac
pmichaud perl6: say True.perl(); 17:52
exp_evalbot kp6: OUTPUT[DISPATCH: calling perl on invalid object:$VAR1 = undef;␤␤ at compiled/perl5-kp6-mp6/lib/KindaPerl6/Runtime/Perl5/MOP.pm line 128␤ main::DISPATCH('undef', 'perl') called at - line 11␤] 17:53
..pugs: OUTPUT[Bool::True␤]
..rakudo: OUTPUT[Null PMC access in find_method()␤current instr.: '_block10' pc 25 (EVAL_10:14)␤called from Sub 'parrot;PCT::HLLCompiler;eval' pc 785 (src/PCT/HLLCompiler.pir:458)␤called from Sub 'parrot;PCT::HLLCompiler;evalfiles' pc 1059 (src/PCT/HLLCompiler.pir:585)␤called from Sub
..'parrot;PCT::HLLCompiler;command_line' pc 1238 (src/PCT/HLLCompiler.pir...
cognominal_ what is the perl6 equivalent of perl5 prefix << (here-document) 18:03
[particle] q:to:END 18:05
iirc
[particle] nqp supports this syntax 18:06
cognominal_ yes, described in S02. tried to match here-document but it is described as here-doc :(
cognominal_ it supports it using PIR inlining 18:15
cognominal_ I wonder what S02 means by "identifier extension", it does not grep in any other synopsis 18:27
pugs_svnbot r20083 | ruoso++ | [smop] implementing low-level scalar... not working yet... I am commiting in the hope that someone tries to understand why test/11_lowlevel_scalar.sm0p is not being match by sm0p/KP6sm0p.pl
diff: dev.pugscode.org/changeset/20083
lambdabot Title: Changeset 20083 - Pugs - Trac
ruoso mncharity, I'm using a KP6 grammar in sm0p, but it's failing to match a file and I can't see why... if you have some time, could you please take a look at the files referred above and see if you can see what I am missing? 18:29
ruoso home & 18:35
cognominal_ how can a capture refers to its parent? 18:47
pugs_svnbot r20084 | putter++ | [redsix] Blowing dust off redsix. t/01-sanity almost passes. 18:48
diff: dev.pugscode.org/changeset/20084
lambdabot Title: Changeset 20084 - Pugs - Trac
mncharity ruoso: certainly, will do. but i'm afraid it will be this now+8ish hours, or tomorrow. 18:49
cognominal_: parent? 18:50
mncharity rides here. bbl & 18:52
cognominal_ in a rakudo grammar, 2<3> matches I would like in the postcircumfix action to test if the expression in the left is not a scalar litteral to rule out that case 18:57
hum 18:58
cognominal_ well I could do it in the action that corresponds to the parent rule 18:59
TimToady cognominal_: that should be driven by the role/class system, not the compiler 20:02
cognominal_ you mean it's ok to whine at execution versus compilation? 20:05
TimToady you can always whine at compilation if you can prove something will whine at run time
cognominal_ here, I can't? 20:06
TimToady or if you pragmatically ask for it to assume that certain things won't change
such as adding the Hash role to integers :)
cognominal_ ho boy
ok, obviously, I am still in a compiled language mindset 20:07
with better errors messages I can live with that. 20:08
pugs_svnbot r20085 | buchetc++ | [t/spec] document coroutines 23:20
diff: dev.pugscode.org/changeset/20085
lambdabot Title: Changeset 20085 - Pugs - Trac
pugs_svnbot r20086 | gwern++ | Update HsJudy to build against the system's installation of Judy and not the version taking up space in thirdparty/ 23:32
diff: dev.pugscode.org/changeset/20086
lambdabot Title: Changeset 20086 - Pugs - Trac
pugs_svnbot r20087 | gwern++ | Mv all the Judy modules to Data.Array.Judy to better fit in the hierarchical modules. 23:38
diff: dev.pugscode.org/changeset/20087
lambdabot Title: Changeset 20087 - Pugs - Trac

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!