pugscode.org/ | nopaste: sial.org/pbot/perl6 | pugs: [~] <m oo se> (or rakudo:, kp6:, elf: etc.) (or perl6: for all) | irclog: irc.pugscode.org/
Set by Tene on 29 July 2008.
meppl good night 00:03
masak Wikipedia++ # This article is about the function transformation technique. For the dish, see Curry. 00:06
pugs_svn r22908 | bacek++ | [spec] Replace some skip with todo for rakudo 02:37
pugs_svn r22909 | lwall++ | [gimme5] move more boilerplate out of normal matcher methods 04:20
bacek_ perl6: say Object ~~ Any; say Any ~~ Object 04:21
p6eval elf 22847: OUTPUT[Can't call method "Str" on an undefined value at ./elf_f line 649.␤ at ./elf_f line 3861␤]
..pugs: OUTPUT[␤1␤]
..rakudo 32404: OUTPUT[Null PMC access in get_string()␤current instr.: 'parrot;P6metaclass;add_parent' pc 95 (runtime/parrot/library/P6object.pir:126)␤]
literal did someone answer my question while my connection was out? 04:59
pmichaud literal: I didn't see the question. 05:05
literal here it is: does the language standard define some way of getting at the name of the implementation? like $?RUNTIME or something
pmichaud possibly $?COMPILER 05:06
pmichaud moritz_ / masak: the tests for RT #58526 in S29-list/pick.t look wrong to me. 05:09
pugs_svn r22910 | pmichaud++ | [spec]: Fix test in S29-list/pick.t . 05:42
pugs_svn r22911 | lwall++ | [tryfile] needs chomp when files from stdin 05:56
pugs_svn r22912 | pmichaud++ | [spec]: Add tests that "$a" and "{3}" return Str. 06:18
moritz_ rakudo: say 'test'; 07:20
moritz_ I've just switched the evalbot to use a local parrot mirror. It seems it didn't work ;) 07:21
moritz_ rakudo: say "another test" 07:32
p6eval rakudo 32416: OUTPUT[another test␤]
moritz_ rakudo: say "{1}".WHAT 07:37
p6eval rakudo 32416: OUTPUT[Int␤]
moritz_ rakudo: say "a {1}".WHAT 07:38
p6eval rakudo 32416: OUTPUT[Str␤]
pmichaud I have that fixed in the rakon branch, but not in trunk yet. 07:41
Feel free to skip/todo the tests until we merge the branch. 07:42
rakudo_svn r32417 | pmichaud++ | [rakudo]: partial spectest-progress.csv update 07:42
moritz_ pmichaud: done. 07:43
pugs_svn r22913 | moritz++ | [t/spec] TODO'ed two new tests for rakudo
pmichaud we should be able to merge the branch tomorrow sometime. It fixes most of the array issues we've been having.
moritz_ nice
sadly I'm not online during the weekend, so I can't witness it first hand (or help with testing) :( 07:44
pmichaud > say [ [1,2], [3,4] ].perl
[[1, 2], [3, 4]]
lambdabot Not in scope: `say'Not in scope: `perl'
moritz_ but I volunteer to clean up any passing TODO tests on Sunday.
pmichaud rakudo: say [ [1,2], [3,4] ].perl
p6eval rakudo 32416: OUTPUT[[[1, 2], [3, 4]]␤]
pmichaud oh, that was working already. 07:45
oh well. :-)
moritz_ it wasn't working when stored in a hash.
pmichaud rakudo: say [ [1,2], [3,4] ].pick(*)
p6eval rakudo 32417: OUTPUT[341␤]
moritz_ rakudo: my %h = (a => [[1, 2], [3, 4]]); say %h<a>.perl
p6eval rakudo 32417: OUTPUT[[1, 2, 3, 4]␤]
pmichaud right, that should be fixed in the new branch.
let's see 07:46
> my %h = (a => [[1, 2], [3, 4]]); say %h<a>.perl
[[1, 2], [3, 4]]
lambdabot <no location info>: parse error on input `='
moritz_ nice.
pmichaud this fixes lots of the "little" issues that have been surrounding arrays and hashes 07:47
moritz_ that should also reduce the amount of undef warnings in tests that are now marked as todo
pmichaud yes, likely.
anyway, I need sleep -- back in a while.
moritz_ good night (and great work)
ludan hi 09:15
moritz_ hi there
jauaor Hello 11:58
wayland76 jauaor: Hello :) 12:09
jauaor Hi wayland76 :-) 12:10
breakfast, brb
wayland76 bedtime for me & 12:14
jauaor back 13:13
jauaor finally compiling parrot 14:26
pmurias jauaor: re finally, did you have any problems with compiling parrot? 14:35
lambdabot pmurias: You have 3 new messages. '/msg lambdabot @messages' to read them.
jauaor pmurias: no .... just no free time 14:37
hah
pmurias: I am compiling for mac though, running a fulltest now
bacek pugs: say Bool.pick 15:08
p6eval pugs: OUTPUT[*** pick not defined: VType (mkType "Bool")␤ at /tmp/8CzicbeaQR line 1, column 5 - line 2, column 1␤]
bacek pugs: say Bool.pick(*) 15:09
p6eval pugs: OUTPUT[*** pick not defined: VType (mkType "Bool")␤ at /tmp/yLV3OGGQQQ line 1, column 5 - line 2, column 1␤]
TimToady that is a very recent spec change 15:11
I would not expect anyone to have implemented it yet 15:12
it's also possible it's a bad design, and there should be a way of getting a list of the types values instead 15:14
*type's
pmichaud an interesting artifact of Bool.pick is that ?Bool.pick will always return Bool::True . 15:28
[particle] why is that? 15:29
pmichaud because .pick returns a list of one element.
it does not return one element. :-)
[particle] ok, so you need ?$(Bool.pick)
no...
pmichaud no
Bool.pick[0]
[particle] you need yep
pmichaud or I've been writing Bool.pick.[0] 15:30
[particle] do lists have a .head method? 15:30
pmichaud .first, I think.
jauaor mmm it seems like a fulltest is indeed a fulltest
pmichaud so Bool.pick.first
[particle] so ?Bool.pick.first works
pmichaud rakudo: say [1,2,3].pick.perl; 15:31
p6eval rakudo 32423: OUTPUT[[2]␤]
[particle] jauaor: that's no lie
pmichaud rakudo: say [1,2,3].pick.first.perl;
p6eval rakudo 32423: OUTPUT[No applicable methods.␤␤current instr.: '_block11' pc 66 (EVAL_11:20)␤]
TimToady maybe pick should be returning a capture, not a list/array 15:31
pmichaud ENOFIRST
[particle] rakudo: say [1,2,3].pick.first
p6eval rakudo 32423: OUTPUT[No applicable methods.␤␤current instr.: '_block11' pc 66 (EVAL_12:20)␤]
[particle] yep, no .first
TimToady I think returning an array is bogus 15:32
pmichaud could be.
TimToady is like saying that map's closure always returns an array 15:33
[particle] what would prefix:? interrogate in the Capture returned from .pick?
TimToady you mean prefix:<?> I presume. this is #perl where that notation is forbidden :) 15:34
#perl6 even
[particle] :P
TimToady prefix:<?> takes a scalar argument, and if you bind a capture with one positional argument, it just binds that one argument 15:35
pmichaud yes, I've been thinking about ways to fix that in Rakudo (the prefix:? notation)
TimToady so ?Bool.pick ought to be false half the time, I think 15:36
pmichaud same for something like foo() if Bool.pick; ? 15:37
is there a binding there?
TimToady presumably 15:38
[particle] doesn't 'if' imply prefix:<?> ?
pmichaud I think it just implies a test for boolean-ness
that doesn't _have_ to be prefix:<?>
[particle] ok, yes, that's correct. 15:39
TimToady it implies item context on the capture returned by the pick, so that it can call the appropriate low-level method
which at the moment is .true still
jauaor mm.. seems like failed 3 tests out of 66
pmichaud I have to finish preparing for a meeting in 20 mins -- will be back in 50. 15:40
(actually, I'm here the whole time -- will start paying attention again in 50)
jauaor mm.. seems like the test got stuck or something ... 15:54
rakudo_svn r32425 | pmichaud++ | [rakudo]: spectest-progress.csv update: 211 files, 4413 passing 16:10
pmurias TimToady: congrats on getting rid of TRE ;)
jauaor ended up interrupting the test 16:11
TimToady 'course, that means I have to write a real lexer now :)
the current one is just the TRE workaround that happened to be good enough to workaround TRE entirely 16:12
jauaor pastebin.com/m59d4751 16:27
buubot jauaor: The paste m59d4751 has been copied to erxz.com/pb/13317
jauaor anybody got an idea?, trying to run 'parrot perl6.pbc'
parrot 0.8.0
jauaor mmm.. interesting 16:48
it works if I 'make perl6'
pmichaud jauaor: is it actually Parrot 0.8.0 or the current head in the svn repo? 16:49
(i.e., the 0.8.0 release?)
jauaor pmichaud: the parrot 0.8.0 .. from the tarball itself
correct
pmichaud ah. I'm not sure what the issue is there. 16:50
rakudo_svn r32428 | pmichaud++ | [rakudo]: Remove -G flag from test_summary.pl, 16:50
r32428 | pmichaud++ | as it results in "out of memory" errors on my machine
r32428 | pmichaud++ | for the S05-mass/rx.t tests.
pmichaud it looks kinda weird to me.
jauaor yeah .. in principle, it looks weird 17:19
pmurias moritz_: the irc log for today is blank 17:20
masak oh! pair query methods are gone. fancy that. 18:48
masak what's this TRE people seem so happy that TimToady got rid of? 18:53
also, FWIW, I liked Bool.pick the moment I saw it. :)
TimToady well, if we don't break that you can now say Day ~~ :pick<Tue> to get a 1/7th chance :) 19:00
or maybe that's :pick(Tue) 19:01
PerlJam or both 19:02
TimToady since Tue is acting as a typename
yes, but it just seems like more work to do Tue ~~ 'Tue' than Tue ~~ Tue
and that's assuming types stringify to their names, which is still in doubt 19:03
PerlJam What else would they stringify to? A structure representation? A composition representation? ? 19:04
[particle] 0x5afecafe 19:05
TimToady or maybe a package name with a :ver and :auth
PerlJam Hmm
TimToady or maybe they're undef like the protoobjects of the same name :)
[particle] that sounds like a Capture
TimToady so maybe they stringify to a warning 19:06
or maybe '' but Failure 19:07
PerlJam "ope' not this Class..."
~Typename ~~ Failure seems a little harsh :) 19:08
[particle] ~Bool ~~ /you have been eaten by a grue/
PerlJam I have got to get a better font. All of the tildes look like dashes. 19:09
TimToady @tell azawawi S_s_h seems to be calling is_type and is_routine on all sorts of things for which it makes no sense, such as the insides of strings like "True" and the various bits of longnames like statement_control:PRE 20:06
lambdabot Consider it noted.
TimToady @tell azawawi the problem is you can't just scan for /identifier/; you need to pick color based on the *last* context you care about, like nibbler or deflongname 20:22
lambdabot Consider it noted.
TimToady @tell azawawi so in the case of push @basetypenames, "True" the True is currently blue because the push is an identifier, which is bogus, since True is just a nibble 20:30
lambdabot Consider it noted.
ruoso hmmm... irc log seems offline 21:37
pmurias ruoso hi 21:42
ruoso hi pmurias
have you seen my changes in mildew?
I took the liberty to do it, as you were offline for a few days 21:43
pmurias ruoso: feel free to change anything you want i don't belive in code "ownership"
ruoso ;P
but did you agree with removing $AST::CODE? 21:44
pmurias looking...
i was just attempting to make emitting a bit more consise 21:46
it was propably a bit to clever
ruoso: sm0p has bit rotted 21:47
ruoso pmurias, in which sense?
pmurias it doesn't compile with a new gimme5 21:48
pmurias does anything recurse in DESTROYALL atm? 21:48
ruoso oh... the sm0p compiler..
ruoso pmurias, we might just implement the 'free' operator in mold 21:49
and re-implement it in mold correctly
and it does recurse today, yes
because one object's destroyall can end up in another destroyall call
pmurias ahh 21:50
implementing the 'free' operator in mold isn't hard 21:51
ruoso pmurias, then we have just some tests using slime, right? 21:54
pmurias ruoso: i wanted to remove slime to have an easier task of hunting down NULLs which are a problem with mold
ruoso: tests 2,9,12,16,17
ruoso yes.... I agree completely...
slime already played its role
the most important of it being proving the concept that SMOP can have more than one interpreter cooperating at the same time 21:55
pmurias ruoso: i'll remove those tests
ruoso pmurias, just comment them on the CMakeLists.txt 21:56
pmurias, we can rewrite them using mold
later
pmurias ok
ruoso dinner & 21:58
pmurias ruoso: the remaining tests pass with a SMOP_DISPATCH DESTROYALL 22:00
pugs_svn r22914 | azawawi++ | [STD_syntax_highlight] removed check on 'lex' directory 22:01
rakudo_svn r32441 | pmichaud++ | [rakudo]: merge rakon branch to trunk 22:10
pugs_svn r22915 | pmurias++ | [smop] 22:14
r22915 | pmurias++ | removed SLIME
r22915 | pmurias++ | commented out slime using tests
r22915 | pmurias++ | commented out test 1 as it tests slightly inlegal things and is duplicated by higher level stuff
r22915 | pmurias++ | test 4 fails as we are doing DESTROYALL incorrectly now
pugs_svn r22916 | bacek++ | [spec] Made Incrementor/Decrementor behave according to S03. 22:34
DarkArchon moritz_ doesn't happen to be around does he? 23:03
pugs_svn r22917 | azawawi++ | [STD_syntax_highlight] added --snippet-html 23:06
masak DarkArchon: well, he's usually gone on weekends.
DarkArchon :( damn
masak yeah, can't everyone be 100% dedicated to Perl 6 all the time? would be much easier that way. 23:07
DarkArchon it would, alas, the world isn't perfect 23:08
masak ...yet >:) 23:09
DarkArchon some evil death ray that will make us all dedicated to perl 6 in mind? 23:10
I don't suppose anyone else on knows a lot about the ilbot? 23:15
DarkArchon how about perl in general? 23:19
masak what about it? 23:20
DarkArchon I'm having trouble getting ilbot to run 23:21
it keeps failing on line 22
masak did you mean "how about some evil death ray that will make us all dedicated to perl in general in mind?" or "I don't suppose anyone else on knows a lot about perl in general" :)
DarkArchon use IrcLog qw(get_dbh gmt_today); <-- that
which means I'm missing something
(unless I'm way off)
the question is what?
masak DarkArchon: sorry, I wish I could be of any assistance, but your guess is as good as mine in this case. 23:22
DarkArchon damn
pmichaud std: my %h; %h{"key"} = "value"; say %h\{"key"}; 23:34
p6eval std 22847: OUTPUT[parse failure␤]
masak :) 23:38
even std is wrong. 23:39
Khisanth DarkArchon: it means it's probably reporting the wrong line number :) 23:40
DarkArchon it reported correct line numbers for other, similar, errors. If so though, then how do I find out what the problem is? 23:41
masak there's a Perl 6 authority food chain nowadays: Sxx > std > spectests > tests > implementations
diakopter food is right. in what direction does the eating occur? 23:44
DarkArchon ? 23:45
diakopter chomps 23:47
masak diakopter: chomps always work from the right, IIRC. this goes for both Perl 5 and 6. 23:48

Powered by Cro and the Raku® Programming Language.

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

Thank you!