pugscode.org/ | nopaste: sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by diakopter on 11 July 2007.
00:01 polettix left 00:04 thoughtpolice left 00:05 theorb joined 00:06 REPLeffect left 00:07 REPLeffect joined 00:08 luke-jr left 00:17 theorbtwo left 00:28 weinig left, justatheory left 00:38 weinig joined 00:39 dvorak joined 00:40 devogon left 00:44 amnesiac left 00:46 weinig left 01:19 diotalevi joined
pugs_svnbot r18492 | avar++ | the functionality of the script/run_* scripts has been merged into script/kp6 which now supports executing in-place. It's a bit messy at the moment, argv isn't supported properly and some other things probably don't work, but better to commit and fix later 01:21
diff: dev.pugscode.org/changeset/18492
lambdabot Title: Changeset 18492 - Pugs - Trac
avar sh-3.1$ perl script/kp6 -Bperl5 -e 'say "foo"' 01:23
foo
sh-3.1$ perl script/kp6 -Bcl-ecl -e 'say "foo"'
foo
Not a condition type: KP6-LISP::NOT-IMPLEMENTED
Broken at MAKE-CONDITION.
KP6-LISP-USER>>
I'll add support for making stand-alone binaries with the lisp backend later 01:24
-C'cl -ecl -o my.exe' or something
pugs_svnbot r18493 | avar++ | * [kp6-lisp] ecl support in (quit) 01:34
diff: dev.pugscode.org/changeset/18493
lambdabot Title: Changeset 18493 - Pugs - Trac
pugs_svnbot r18494 | avar++ | (quit 0) not (quit) 01:37
diff: dev.pugscode.org/changeset/18494
lambdabot Title: Changeset 18494 - Pugs - Trac
01:52 weinig joined 02:12 weinig left, weinig joined, REPLeffect left 02:19 arcady_ joined 02:26 arcady__ joined 02:32 arcady left 02:40 fax left 02:42 arcady joined 02:43 arcady_ left 02:52 buu left, buu joined 02:56 arcady__ left 03:00 buu left, buu joined 03:04 elmex_ joined 03:20 elmex left 03:32 Limbic_Region left 03:47 daxim joined 03:53 daxim_ left 03:58 scent joined 04:03 amnesiac joined 04:08 justatheory joined 04:35 REPLeffect joined 04:50 snearch joined 05:06 scent left, REPLeffect left 05:25 jisom joined 05:29 devogon joined 05:34 mncharity left 05:43 jisom_ joined 06:00 jisom left 06:03 devogon left, devogon joined 06:09 d4l3k_ joined 06:10 dalek left 06:11 d4l3k_ is now known as dalek 06:18 funktio joined 06:39 amnesiac left 06:51 diotalevi left 06:52 gorax joined
meppl good morning 06:56
agentzh morning, meppl :) 06:58
meppl good morning agentzh 06:59
agentzh :)
07:15 justatheory left 07:23 dduncan left 07:28 Southen joined 07:35 Aankhen`` joined 07:47 Southen_ left 07:50 DarkWolf84 left, funktio left 07:58 marmic left 08:45 devogon left, kcwu left 08:47 drrho joined 08:59 masak joined
masak g'localtime(), people 08:59
09:02 iblechbot joined
spinclad good 05:13Z-0400, @all 09:13
mornin, masak 09:14
masak spinclad: mornin'
so this is the weekend when I'm going to look into a kp6->parrot backend 09:19
I'll just perform a few other urgent tasks, then I'll get right on it 09:20
09:21 polettix joined 09:31 barney joined 09:43 fglock joined 09:48 kcwu joined 09:55 iblechbot left 10:01 marmic joined
fglock hi from Open aLANtejo|07 alantejo.uevora.pt/2007/ 10:08
lambdabot Title: Delta aLANtejo|07
10:24 jisom_ left 10:41 devogon joined
pugs_svnbot r18495 | fglock++ | [kp6] added Signature options 10:46
diff: dev.pugscode.org/changeset/18495
lambdabot Title: Changeset 18495 - Pugs - Trac
10:56 Aankhen`` left, Aankhen`` joined
spinclad kp6 thoughts: 11:01
is kp6 MOP ready to fold constants here, as moose does?: (dispatch (dispatch (dispatch (dispatch $o :how) :lookup $methname) :choose-given $capture) :apply $capture) 11:03
kp6-cl indent: print to stream, which counts indents within itself. kill $indent.
kp6-parrot: could this go via a minimal rewrite to nqp? 11:04
( masak: talk with pmichaud? )
masak spinclad: 's a good idea
I'm still busy with pesky $work stuff, but after that, I will 11:05
fglock spinclad: re nqp, it's possible - MiniPerl6 had a p6parrot backend (maybe still works) 11:06
not sure what you mean with fold constants? 11:07
spinclad i'm hoping nqp is nearly ready -- it was waiting on the new pdd15 object model, which has just been merged to parrot trunk
fglock fold constants == compile time optimization? 11:08
pasteling "fglock" at 213.13.106.82 pasted "avar: strange behaviour of script/kp6" (27 lines, 963B) at sial.org/pbot/28135 11:10
spinclad constant folding: $o is of known type / :how; its :how is closed (immutable) so its :lookup $methname is known; this is a known multi, and can :choose a signature against $capture, giving a particular sub; leaving only the :apply for runtime. 11:11
fglock spinclad: yes, it could - but we don't have closed classes yet 11:12
spinclad this depends of course on known type, closed class, closed multi.
i could try adding 'no Moose' to the language maybe :) 11:13
11:14 devogon left
spinclad first though i think i'll fix up simple-minded indenting in kp6-cl 11:15
(besides kibitzing/joining in on kp6-parrot, if masak doesn't mind) 11:16
masak spinclad: 'course not 11:17
fglock spinclad: this can be added as a Visitor module, there is probably one you could use as a template
masak just glad to have company / someone to discuss with
spinclad excellent. good fun to be had here, then, on several fronts. 11:18
masak :)
spinclad but now, to bed. good dawn to @all! 11:19
masak spinclad: g'dawn
spinclad &
fglock can someone try to out script/kp6? I'm having problems with the new command line
spinclad .oO{ try anything particular? } 11:21
11:22 masak left
pugs_svnbot r18496 | fglock++ | [kp6] more Signature options 11:25
diff: dev.pugscode.org/changeset/18496
fglock spinclad: sial.org/pbot/28135
lambdabot Title: Changeset 18496 - Pugs - Trac
Title: Paste #28135 from "fglock" at 213.13.106.82
fglock it fails to compile code from src/KindaPerl6/Runtime/Perl6/* 11:26
I'm just using an older version for now
11:29 kanru left
fglock brb & 11:29
spinclad $ echo say 123 | script/kp6 -Bperl5 11:32
123
yes, i get the same as you... 11:36
11:37 garfield_cat joined 11:41 fglock_ joined
spinclad Perl::Tidy::perltidy is unhappy with { source => $out, ... } 11:42
$out should be a file name but is the program text instead?
11:44 fglock left, fglock_ is now known as fglock
spinclad from the Perl::Tidy manpage, that should be source => \$out, maybe: a ref to a string. trying... 11:46
pugs_svnbot r18497 | fglock++ | [kp6] prepare to add parameter options to the grammar
diff: dev.pugscode.org/changeset/18497
lambdabot Title: Changeset 18497 - Pugs - Trac
fglock lunch & 11:48
spinclad fglock: yes, changing the call to perltidy to { source => \$out, ... } fixes it. 11:58
11:58 masak joined, iblechbot joined
spinclad trying to svn ci: it asks for my password... this is new since last year... i'll have to figure it out, after sleep. 11:59
fglock: i leave a one-char change in your care. ^ 12:01
spinclad &
(really!)
12:03 fglock left
spinclad (script/kp6, line 423) 12:03
12:04 garfield_cat left
spinclad avar: see above re script/kp6 12:07
12:09 amnesiac joined 12:10 amnesiac left 12:11 amnesiac joined
integral wonders why that eval just eats the error message 12:14
12:14 amnesiac_ joined
integral oh, I see. it's meant to only run when Perl::Tidy is present. 12:14
12:16 chris2 joined, amnesiac__ joined
pugs_svnbot r18498 | bsmith++ | [kp6] Fix script/kp6's call to perltidy. 12:22
diff: dev.pugscode.org/changeset/18498
lambdabot Title: Changeset 18498 - Pugs - Trac
12:24 IllvilJa left 12:27 amnesiac left 12:31 amnesiac_ left 12:33 amnesiac__ left 12:36 dolmen joined 13:03 snearch left 13:06 gorax left 13:25 cmarcelo joined 13:42 fglock_ joined, fglock_ is now known as fglock 13:48 lichtkind joined 13:55 rgs left 14:01 fglock left
pugs_svnbot r18499 | fglock++ | [kp6] move Signature parsing to it's own module at KindaPerl6::Grammar::Signature 14:07
diff: dev.pugscode.org/changeset/18499
lambdabot Title: Changeset 18499 - Pugs - Trac 14:08
14:24 Limbic_Region joined 14:26 rgs joined, mr_ank joined
pugs_svnbot r18500 | fglock++ | [kp6] more Signature syntax; ( does :x<1> in a sig implies it is a named-only parameter? ) 14:29
diff: dev.pugscode.org/changeset/18500
lambdabot Title: Changeset 18500 - Pugs - Trac
14:30 masak left 14:31 fglock joined
pugs_svnbot r18501 | fglock++ | [kp6] updated AST with signature options 14:47
diff: dev.pugscode.org/changeset/18501
lambdabot Title: Changeset 18501 - Pugs - Trac
fglock trying to figure out what is the data structure of a sig with named positionals 14:59
Aankhen`` Treat them like normal positionals with a special flag indicating they're named-only? 15:05
I don't really understand the point of named positionals though.
fglock hmm 15:07
15:09 jdv79 joined
fglock you mean s/named-only/named-too/ ? 15:10
it's probably just an array of parameters and their descriptions 15:11
15:12 pjcj left
fglock complex sigs need to be interpreted at run-time anyway 15:12
15:12 pjcj joined
Aankhen`` I'm not sure I understand now. :-S 15:15
What are named positionals? 15:16
fglock sub x ($a) {...}; x(10); x( :a<10> ); 15:17
the second form 15:18
sub x ( :$a ) {...} # named only
Aankhen`` So you're talking about the arguments passed rather than the items in the signature? 15:19
fglock I was representing the sig as a capture-ish thing (a scalar, an array, and a hash) 15:20
but it's more complex than that
Aankhen`` Right.
fglock it's probably an array of "statements" instead 15:21
Aankhen`` In KP6-LISP, I had separate slots for positional, optional and named parameters, as well as slurpy hash, block, and array.
fglock to be resolved at runtime
Aankhen`` I think I had decided to punt on named positionals.
Just didn't seem to me to be that important. 15:22
fglock resolved == when it is bound to a capture
Aankhen`` I mean, what is the use case over a regular named parameter?
Wait, I'm off on a tangent again.
x(:a<10>) # :a<10> is a named positional?
fglock this is needed according to the spec 15:23
yes, if the $a argument was not "named-only"
Aankhen`` OK, yeah, that was handled in KP6-LISP too. I think.
Aankhen`` checks.
fglock I'm just adding it to the AST; check the output of 15:24
$ perl script/kp6 -Cast <<< 'sub x ( *$a => 42 ) { 123 };'
Aankhen`` I'm not familiar with that notation. What does the * indicate? 15:25
Isn't it used for slurpy parameters or unpacking parameters?
fglock slurpy (expands lists into @_)
Aankhen`` Okay. 15:26
Yeah, unpacking parameters is completely unimplemented in KP6-LISP.
fglock brb & coffee break 15:27
15:32 funktio joined
fglock b 15:41
15:41 kanru joined 15:50 wilx left, wilx joined
fglock i've got an impl, but it breaks the runtime - should probably fix the runtime before committing 15:52
15:57 rindolf joined
pugs_svnbot r18502 | fglock++ | [kp6] large change to signature code to accomodate named-positionals and such; most tests still pass 16:08
diff: dev.pugscode.org/changeset/18502
lambdabot Title: Changeset 18502 - Pugs - Trac
16:10 jdv79_ joined, jdv79 left 16:16 LimbicRegion joined, Limbic_Region left, LimbicRegion is now known as Limbic_Region
lumi Hi, what's new in p6land? 16:26
pugs_svnbot r18503 | fglock++ | [kp6] reverted Lit::NamedArgument for use in Capture; now use Lit::SigArgument for Signature; "fixed" the Lisp impl too
diff: dev.pugscode.org/changeset/18503
lambdabot Title: Changeset 18503 - Pugs - Trac
pugs_svnbot r18504 | fglock++ | [kp6] fixed SigArgument object 16:29
diff: dev.pugscode.org/changeset/18504
lambdabot Title: Changeset 18504 - Pugs - Trac
16:34 IllvilJa joined
pugs_svnbot r18505 | fglock++ | [kp6] killed Sig.named (might be replaced by a method later) 16:34
diff: dev.pugscode.org/changeset/18505
lambdabot Title: Changeset 18505 - Pugs - Trac
fglock lumi: we've got some new stuff in kp6, see latest pugs blogs - I'm now rewriting signatures to support more Perl 6 features 16:38
lumi Oh nice, kp6 on lisp! Does that mean I can write my emacs macros in Perl 6 soon? :) 16:43
fglock I guess so :) 16:46
16:48 jisom joined
fglock CAPTURIZE() is broken (in Perl5::MOP) because of the sig changes 17:04
pugs_svnbot r18506 | fglock++ | [kp6] 'is rw', 'is copy' in Signature
diff: dev.pugscode.org/changeset/18506
lambdabot Title: Changeset 18506 - Pugs - Trac
17:14 drrho left
funktio just wondering, what does do { [] }; do? 17:19
17:20 rindolf left
lumi Wouldn't it create a new array, but in void context, so lose it immediately? 17:21
QtPlatypus lumi: Its common list rather then elisp.
lumi I might be wrong, I've been some months out of the loop
integral www.lisp.se/emacs-cl/ # But it doesn't support CLOS 17:22
lambdabot Title: Emacs Common Lisp
lumi integral: So you write a mode in P6, compile it via kp6 to CL, and then run it in emacs-cl, win 17:23
17:23 Psyche^ joined
integral I had the decided impression that we were using CLOS... 17:23
lumi Oh, um.
integral prefers scheme personally 17:24
lumi aol
It's somehow neater
17:24 [particle1 joined
integral TBH I'd probably be happier with strict haskell and CLOS, but that doesn't exist :-( 17:25
lumi Strict Haskell? That kinda spoils it... 17:26
Although I guess it's nice to be able to reason about space 17:27
pugs_svnbot r18507 | fglock++ | [kp6-perl5] created ROUTINE env variable, so that a routine can get to it's signature at runtime 17:28
diff: dev.pugscode.org/changeset/18507
lambdabot Title: Changeset 18507 - Pugs - Trac
fglock CLOS is a shortcut, we can use the Perl 6 native MOP when it gets written 17:29
17:34 pmurias joined
pmurias b 17:34
fglock pmurias: i broke parameter passing, fixing 17:38
pugs_svnbot r18508 | fglock++ | [kp6-perl5] fixed the param binding code
diff: dev.pugscode.org/changeset/18508
lambdabot Title: Changeset 18508 - Pugs - Trac
fglock pmurias: I added richer signatures, trying to get the runtime up to it 17:39
I may have done something that broke CAPTURIZE()
pmurias fglock: hi 17:40
fglock hi :)
17:40 Patterner left, Psyche^ is now known as Patterner, [particle] left
fglock pmurias: see $ perl script/kp6 -Cast <<< 'sub x ( *$a => 42 is rw ) { 123 };' 17:41
actually, $ perl script/kp6 -Cast -e 'sub x ( *$a => 42 is rw ) { 123 };'
pmurias fglock: nice 17:43
fglock brb 17:46
17:49 lisppaste3 left 17:56 lisppaste3 joined, devogon joined, [particle] joined 18:03 fglock left 18:05 funktio left 18:07 jisom left 18:12 [particle1 left 18:16 chris2 left
pmurias should script/kp6 -Cperl5 t/kp6/01-tap.t work? 18:19
it dosn't
script/kp6 -Cperl5 < t/kp6/01-tap.t works 18:20
18:24 Khisanth left 18:31 Torment joined 18:33 wilx left 18:37 wilx joined 18:48 Jedai left 18:51 barney left 18:54 jdv79_ left 18:57 Limbic_Region left 19:12 cathyal joined 19:16 Khisanth joined 19:17 lichtkind left, cathyal left 19:28 ST47 joined
pmurias @tell fglock EmitPerl5 emits Signatures incorectly as it uses Main::perl (via .perl) on Lit::SigArgument(in .positionals) 19:34
lambdabot Consider it noted.
19:42 rindolf joined
pmurias avar++ #the new CLI interface is handy 19:43
19:53 weinig left
avar pmurias: yeah but I apperently screwed up some stuff:) 20:01
20:07 pjcj_ joined
pmurias avar: i still like it anyway ;) 20:11
metaobjects++ #fun 20:14
20:17 kanru left
integral pmurias++ # fixed that bug 20:18
pmurias integral: which bug? i didn't fix anything 20:20
pugs_svnbot r18509 | bsmith++ | [kp6] Fix a bug in calling Perl::Tidy when source files are padsed on the command line instead of via STDIN.
diff: dev.pugscode.org/changeset/18509
lambdabot Title: Changeset 18509 - Pugs - Trac
integral ^^^ that one
20:21 kanru joined
pugs_svnbot r18510 | pmurias++ | [kp6] 20:21
r18510 | pmurias++ | - kp6.sh (pointless now)
r18510 | pmurias++ | added .attributes (should return descriptors instead of method names)
r18510 | pmurias++ | + an example using the metaobject to generate an wrapper class
diff: dev.pugscode.org/changeset/18510
lambdabot Title: Changeset 18510 - Pugs - Trac
20:23 _Helios is now known as Helios-
pmurias how should method and attribtors be called? (class names) 20:25
20:53 baest left, baest joined 20:58 rfordinal_ joined 21:03 pmurias left 21:05 pmurias joined
pugs_svnbot r18511 | andara++ | [runpugs] 21:21
r18511 | andara++ | Clean-up of internals.
diff: dev.pugscode.org/changeset/18511
lambdabot Title: Changeset 18511 - Pugs - Trac
21:23 rfordinal left
PerlJam Is it safe to say that a regex always has an implicit .* at the front of it? For instance, just as "foobar" ~~ /bar/ matches, so would regex foo { <bar> } regex bar { "bar" } "foobar" ~~ /<foo>/ ? 21:29
pugs_svnbot r18512 | bsmith++ | [kp6] Updated manifest.
diff: dev.pugscode.org/changeset/18512
lambdabot Title: Changeset 18512 - Pugs - Trac
21:32 Aankhen`` left 21:33 rindolf left
PerlJam Hmm. that's a poor example of what I'm trying to get at. 21:35
regex foo { "f" } regex bar { "bar" } "foobar" ~~ /<foo>/; # Matches or not? 21:36
blah 21:37
regex foo { "f" <bar> } regex bar {'bar' } I meant
pugs_svnbot r18513 | Darren_Duncan++ | ext/Muldis-DB/ : removed the files Operators.pm and PhysType.pm but they will be replaced later 21:44
diff: dev.pugscode.org/changeset/18513
lambdabot Title: Changeset 18513 - Pugs - Trac
21:52 pmurias left
spinclad PerlJam: definitely not, as far as i can tell 21:54
"foobar" !~~ / 'f' <bar> /
21:55 [particle1 joined
spinclad i don't have a pointer to spec for it atm tho 21:56
PerlJam spinclad: I didn't used to think so either, but this sentence from the section that talks about :ratchet implies that it is so "When a ratcheted regex is called as a subrule, the front is anchored to the current position ..."
at least to me.
(it implies that a non-ratcheted regex is not anchored to the current position)
spinclad hmm 21:57
Khisanth it implies it might or might not be anchored :P
PerlJam Well, I'll ask TimToady next time I see him or p6l or something 21:58
(But right now I've got some kids to take care of)
PerlJam &
21:59 [particle] left
spinclad i'd say practice depends on front anchoring, and spec likely must follow... 21:59
but idunno really 22:00
afk, back later & 22:01
22:06 faxathisia joined
integral wouldn't that be the equivalent of .*? anyway not .*? 22:14
oops. ".*?" and ".*" 22:15
pugs_svnbot r18514 | Darren_Duncan++ | ext/Muldis-DB/ : this is the Perl 6 equivalent of what will end up on CPAN as Muldis::DB version 0.5.0 for Perl 5 22:20
diff: dev.pugscode.org/changeset/18514
lambdabot Title: Changeset 18514 - Pugs - Trac
22:35 weinig joined, weinig left 22:49 iblechbot left 22:52 weinig joined 23:01 pjcj left, pjcj_ left 23:07 pjcj joined 23:12 weinig left, weinig joined 23:15 devogon left 23:27 polettix left 23:53 weinig left 23:56 [particle1 is now known as [particle] 23:58 slango joined
slango is perl 6 ridiculously different? 23:58