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.
ruoso good night 00:55
lambdabot ruoso: You have 1 new message. '/msg lambdabot @messages' to read it.
ruoso @tell pmurias, not immediatly, "goto" would return immediatly, and when the interpreter would execute the next step, it would then call $io.print 00:58
lambdabot Consider it noted.
jdv79 is there a transcript of the state of the onion yet? 02:26
Tene jdv79: www.perlfoundation.org/perl6/index....on_12_2008 -- looks like no 03:10
lambdabot Title: State of the Onion / Perl 6, tinyurl.com/5vxltk
pugs_svn r21602 | putter++ | [rx_on_re] Started moving AST node definition to p6. Mostly relocated MatchN p5 unchanged into emit5.pm. 03:16
pugs_svn r21603 | putter++ | misc/pX/Common/yet_another_regex_engine/Regexp_ModuleA.pm: Replaced a split(), called via a /(?{ })/. Thus avoiding perl memory panic segfaults. 03:22
ingy hello! 04:02
Tene Hi! 04:03
pugs_svn r21604 | lwall++ | .** isn't a valid regex 04:30
r21605 | lwall++ | [STD] fix angle parsing problems
Auzon doesn't know where he got .** from instead of .+ 04:31
TimToady I changed it to .**2, but feel free to change it
Auzon The actual regex wasn't important, I don't think. 04:32
TimToady well, you don't want it to match as a string is the main thing
Auzon Yeah, so pretty much anything with a metacharacter is good
hm, except /.+/ would match '.+' 04:33
TimToady nodnod
Auzon Thanks for catching it, TimToady++
TimToady anyway, it parses now :)
TimToady blames STD 04:34
Auzon Speaking of parsing, I'll give tryfile a try now
TimToady well, the test also found some STD bugs, to be sure
didn't have the foggiest idea how to parse <::(...)>
as it is, ::() is rather ambiguous in the standard grammar 04:35
is it a symbolic reference, or a cast to :: sigil type?
which might be a good reason for not consiring :: a sigil... 04:36
Auzon My use quoted S05 claiming that it was a symbolic reference
TimToady *considering
I think that's how people would usually think of it in that context
though <."$ref"()> would probably be clearer 04:37
maybe that's an argument for ::"$name" instead of ::($name) though
Auzon I wonder how long it'd take to run tryfile on all of t. 04:41
well, t/spec first.
TimToady it's faster than it used to be, and not fast enough :)
Auzon Is that what you're doing (running it against all files)? 04:42
TimToady make test compiles STD.pm and t/spec
takes on the order of 20 minutes on my machine
Auzon ah, cool. 04:43
TimToady but I haven't timeed it exactly.
my goal is for STD to compile itself roughly at the same speed that gimme5 does, but that's gonna be a while yet...
need about a 120x speedup for that
Auzon What does STD do when it encounters a new operator definition? 04:44
TimToady nothing currently
it calls .add_macro which does nothing yet
TimToady I cheated on postfix:<!> 04:45
Auzon No fair ;)
TimToady pmurias was thinking about working on the macros though 04:45
and then it would use the definition in the test
one of the remaining failures is due to lack of infix:<>
Auzon Isn't that just nothing? 04:46
TimToady STD compiles itself in 3 minutes with a full cache rebuild, and in 2 minutes with a loaded cache
ingy speed killz 04:47
TimToady would like to get in the ballpark of a second, but it's already good enough for correctness testing
but that'll take a real dfa engine dedicated to doing p6 regex fast
Cursor spends far too much time babysitting TRE at the moment 04:48
ingy :)
someone should fix tre
TimToady feel free
ingy feels free 04:49
fr33 slaps ingy 04:49
_lidlpenner93_7 ICH WEIS WO MAN VIDEOS GUCKEN KANN UND WENN MAN WILL KANN MAN DANN DIE VIDEOS KOSTENLOS AUF SEIN HANDY SCHICKEN LASSEN WENN IHR WOLLT GEHT DA DRAUF www.extrem-geile-videos.de.vu 06:22
moritz_ Tene++ 06:24
Tene Channel ban list is full? 06:29
Oh, it is. 06:30
moritz_ too bad
Tene All set at the same time over three months ago by ballard.freenode.net 06:31
pugs_svn r21606 | fglock++ | [v6.pm] fixed bare_sigil 09:41
moritz_ @tell Auzon $handle.eof now works in rakudo 10:42
lambdabot Consider it noted.
moritz_ does anybody here no diakopter's real name? I'm working on an article for the perl review, and don't know how to attribute him 12:15
pugs_svn r21607 | fglock++ | [v6.pm] fixed operators, warnings 12:23
pmurias moritz_: what's your article about? 12:24
moritz_: see AUTHORS 12:25
moritz_ pmurias: it's an updated version of "getting involved with perl 6"
thanks (re AUTHORS) 12:26
brian d foy read it on perlmonks and asked if he could use it for TPR
zamanfou is now away: In the rain my tears are forever lost. 12:42
pugs_svn r21608 | pmurias++ | [smop] nested sm0p blocks 13:21
pmurias ruoso: just calling $interpreter.goto({$io.print("ok 3\n")}) doesn't seem to print anything 13:22
ruoso pmurias, exactly... 13:23
you need to call a loop afterwards
pmurias how am i supposed to do this the nodes after goto aren't executed? 13:25
ruoso oh... 13:26
this is inside a frame that is being run... 13:27
pugs_svn r21609 | pmurias++ | [smop] forgot a initial empty node 13:34
pmurias ruoso: should i replace you Code implemenationn with a sm0p frame taking variant or layer the variant on top of it 13:43
ruoso pmurias, I think we can call the Code I implemented as CCode and let the variant to be "Code" 13:44
pmurias we could even have CCode.new("c_function_name") 13:45
ruoso indeed 13:48
pugs_svn r21610 | pmurias++ | [smop] renamed Code to CCode 13:51
pugs_svn r21611 | pmurias++ | [smop] fixed previous commit 14:05
moritz_ @tell Auzon $handle.eof works - kind of. But instead I discovered that Str.split(Str) actually works in rakudo 14:09
lambdabot Consider it noted.
moritz_ rakudo: say "abc".split("b").perl
p6eval rakudo 29834: OUTPUT[["a", "c"]␤]
pmurias pugs: our $foo=3;say $foo;say $*foo; 14:41
p6eval pugs: OUTPUT[3␤␤]
moritz_ our $foo; creates $main::foo, not $GLOBAL::foo 14:42
pmurias moritz_: godd 15:02
* good
pugs_svn r21612 | moritz++ | [spec] one more test in S16-io/basic-open.t, and fudged for rakudo 15:06
pugs_svn r21613 | moritz++ | [spec] more unfudgind, autounfudge++ 15:28
Alias_ tries to build Vanilla Parrot 15:29
(again)
3 hours of cpu later... 15:30
moritz_ why does it need so long for you? weird platform?
Alias_ Windows 15:31
I have to build all of strawberry first
Because the win32 version only builds inside of a strawberry install
moritz_ then I guess windows qualifies as a "weird platform" ;-) 15:33
pmurias ruoso: why does Code use call instead of postcircumfix:<( )>? 15:34
pmurias (in CCode) 15:56
pugs_svn r21614 | lwall++ | [STD] buglets 16:21
ruoso @tell pmurias, I usually forget about this convention.... 17:17
lambdabot Consider it noted.
viklund I think I might have found a bug in rakudo. Classes declared with extra namespaces are empty 17:56
class A {}; # works 17:57
class A::A {}; # doesnt
Tene what does it do?
moritz_ viklund: I think that's a known issue, and there's a patch for it waiting in RT that needs review
viklund ok, great
Tene is there a failing test? 17:58
moritz_ dunno 17:59
seems like, yes 18:00
spec/S12-class/inheritance.t
class Foo::Bar is Foo {
spec/S04-statements/try.t
130: class Naughty::Specific is Naughty {};
moritz_ ok, the patch is not quite for that I think: 18:02
rt.perl.org/rt3/Ticket/Display.html?id=56700
lambdabot Title: #56700: [PATCH] first pass at interpolated namespaces
moritz_ viklund: so it won't hurt to open a new ticket
viklund moritz_: Can I do that by mailing rakudobug at perl.org or do I need an account on RT or both (probably should get one anyway)? 18:08
moritz_ viklund: mailing rakudobug 18:09
pugs_svn r21615 | jnthn++ | [spectest] Unskip a test for Rakudo, and remove some tests for the now-removed sub fallback semantics. 18:18
r21616 | jnthn++ | [spectest] Unskip a couple of tests that Rakudo can now pass. 18:19
viklund ehm, my email got rejected 18:31
moritz_ any reasons given?
viklund 550 550 sorry, your envelope sender is in my badmailfrom list (state 14). 18:32
I'm using gmail (I have already tried with another account)
which hasn't been rejected yet 18:33
ahh, it worked 18:34
rt.perl.org/rt3/Ticket/Display.html?id=57388
lambdabot Title: #57388: Namespaces don't work in class declarations
viklund my first bugreport ;)
[particle]1 viklund++
moritz_ welcome to the club viklund ;) 18:35
pugs_svn r21617 | fglock++ | [v6.pm] 0.030 # ~6200 tests pass 18:58
ruoso home & 19:13
ludan hi 19:21
moritz_ hi ludan 19:22
meppl good night 19:33
pmurias TimToady: any guesses what reverts the cursor back to Perl? macros work when the cursor is reblessed from comp_unit, but are ignored when reblessed from deflongname 20:00
smop: $*OUT.print("hello world\n") 20:02
moritz_: the channel title is misleading... ;) 20:03
s/title/topic
moritz_ pmurias: yes 20:08
pmurias: but only channel ops can change the topic
pmurias why do we have so few of them? 20:10
adding smop: to the evalbot would make it self misleading
;)
moritz_ can smop actually run anything that a human can enter in less than 30s? 20:34
Tene pmurias: what would you like the topic set to? 20:36
moritz_ s/smop/elf/ perhaps?
Tene pugscode.org/ | nopaste: sial.org/pbot/perl6 | pugs: [~] <m oo se> (or rakudo:, kp6:, elf: etc.) || We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/ 20:36
Tene > reverse . show $ foldl1 (*) [1..4] 20:37
lambdabot "42"
Tene Are there any haskellers in here still?
moritz_ seldom
perhaps just delete that part as well :/
Tene pugscode.org/ | nopaste: sial.org/pbot/perl6 | pugs: [~] <m oo se> (or rakudo:, kp6:, elf: etc.) (or perl6: for all) | irclog: irc.pugscode.org/ 20:38
Tene Is pugscode.org still the best url for this channel? 20:39
moritz_ dunno
what else could we chose?
Tene dunno
moritz_ must discussion here is related to projects in the pugs repo 20:40
so it's not such a bad choice after all
Tene The only discussions I pay attention to in here are spec questions and about tests, usually.
[particle] yes, pugscode.org is best 20:41
Tene Does pugs even compile with current ghc? i think I heard that it didn't? 20:42
Auzon You can use cabal to install it.
lambdabot Auzon: You have 2 new messages. '/msg lambdabot @messages' to read them.
moritz_ Tene: it does, when you compile/install it via cabal
Tene And that information is on the website somewhere?
Okay.
moritz_ Tene: no, in INSTALL
Tene Ah.
Auzon rakudo: my $foo = "a b c d e"; $foo.split(' ').perl.say
p6eval rakudo 29834: OUTPUT[["a", "b", "c", "d", "e"]␤]
Tene Okay. 20:43
Auzon rakudo: my $foo = "a b c d e"; $foo.split(/\s+/).perl.say
p6eval rakudo 29834: OUTPUT[["a b c d e"]␤]
Auzon It used to parsefail on the latter one. So something changed recently.
Thanks for finding that moritz_++
moritz_ you're welcome
I'd suggest you use slurp, because .eof behaves weird (it only is true after reading the first line after the end of file) 20:44
Auzon Yeah, I backlogged in #parrot and saw that 20:45
pmurias TimToady: one of the things which stop the changed grammar propagation is _SUBSUME, not sure what's the right way to handle it 20:47
TimToady: _STARr and _PLUSr do it to, btw. shouldn't _PLUSr be defined in terms of _STARr? 20:52
pugs_svn r21618 | jnthn++ | [spectest] Unskip some more Range tests that now pass in Rakudo. 20:56
pugs_svn r21619 | fglock++ | [PCR] 0.36 22:10
r21620 | fglock++ | [PCR] fixed last commit 22:17
pugs_svn r21621 | fglock++ | [v6.pm] require newer PCR 23:03