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.
pugs_svn r22036 | ruoso++ | [smop] starting the sketch on the sample compiled s1p code... 01:10
evan could anyone point me to any info on how parrot does/will do profiling? 04:31
dhruvakm I am still not able to build a usable version of perl6 on windows XP. I earlier had issues on using ActiveState perl as it uses msvc compiler. I am not using strawberry perl and the build goes through but fails on running a simple inline script 05:23
dhruvakm The error I get: .\perl6.exe -e"say 'Hello, world.'" 05:37
Null PMC access in get_string()
dhruvakm i guess no help here! 05:52
moritz_ @tell dhruvakm try #parrot on irc.perl.org 06:25
lambdabot Consider it noted.
masak @tell moritz_ all tests pass now, except for that one with the ESCAPE attribute. I might have some time late tonight to expand/implement the tests enough to make the new HTML::Template a viable replacement for the old one. 08:35
lambdabot Consider it noted.
moritz_ is back from the dentist 08:37
lambdabot moritz_: You have 1 new message. '/msg lambdabot @messages' to read it.
masak Perl 6 grammars are already very nice to work with 08:39
there's pain avoided in every efficient line of code, compared to writing one's own lexer/AST generator 08:40
with the advent of {*}, it will be twice as wonderful
moritz_ but it would be nice if some of the errors produced nicer messages
masak indeed. if you have time to experiment, I can paste you the latest version of the module 08:41
moritz_ or just use a branch in git 08:41
whatever you prefer ;)
masak branch-in-git is clearly the Right Thing 08:42
I'll try :)
moritz_ btw I've got an email from someone in Belarus asking for permission to translate my blog posts to Russian ;)
masak might be the same person whose Russian Perl 6 blog I subscribe to 08:44
moritz_ Andrew Shitov 08:46
masak same one. 08:47
Андрей Шитов
Russian++ # much saner consonants 08:48
actually, Cyrillic++, rather
masak moritz_: github.com/viklund/november/tree/ne...l-template 09:03
lambdabot Title: viklund's november at new-html-template — GitHub
moritz_ how do I fetch remote branches with git? 09:05
masak moritz_: git checkout new-html-template 09:06
(I think)
no, that will probably not work... hold on 09:07
moritz_ error: pathspec 'new-html-template' did not match any file(s) known to git.
and git-branch doesn't show that branch
masak moritz_: try git checkout -b new-html-template
you might need to `git pull` before
or maybe even `git pull origin new-html-template` 09:08
I dunno
moritz_ error: Entry 'p6w/t/html-template.t' would be overwritten by merge. Cannot merge. 09:09
masak moritz_: you working on an unclean tree again? :)
moritz_ dunno
masak time to test that `git clean` you mentioned yesterday, methinks
`git status` answers questions about the state of files 09:10
masak seems to me a simple `rm -r p6w/t` would solve your current problem, however 09:10
moritz_ ok, after a 'git reset ...' I could pull
masak ah.
git embraces TMTOWTDI, for sure 09:11
moritz_ but after 'git pull origin new-html-template' I still don't see a new-html-template branch
masak does `git checkout new-html-template` do anything useful? 09:12
moritz_ same as before
masak hm :/
what about if you clone the new branch directly?
moritz_ in a new repository? 09:13
masak yes 09:14
not optimal, but until we grok git :P
moritz_ I'll ask in #git
masak good idea. 09:15
IRC++
moritz_ omg, #git is even larger than #perl6 09:22
(but very helpful)
masak just as #perl6 :) 09:23
moritz_ ;)
masak moritz_: actually, I would guess that the bigger number of inhabitants in #git compared to #perl6 is commensurate with the number of users of the respective softwares -- or if not, probably to folks-on-#perl6's advantage 09:29
moritz_ digests the sentence 09:37
probably, yes ;)
and we have more bots ;) 09:38
masak I, for one, welcome our new git-using, Perl 6-driven, IRC-dewlling overlords 09:41
moritz_ ;) 09:42
if I manage to understand git at some point, I will too 09:43
have you thought about a simple test harness already?
masak moritz_: besides Test.pm? no. 09:44
moritz_ if prove can be made to run a different program than perl, that's all we need 09:45
masak there's prove6 in the Pugs repo
...perhaps indication that prove can't be made to run a different program 09:46
s/indication/indicating/
moritz_ prove6 is a bit pugs specific
but certainly a good place to start stealing from ;)
masak :) 09:48
moritz_ dev.perl.org/perl6/faq.html talks about "active metadata on values, variables, subroutines, and types" - any idea what's meant by that? 10:14
lambdabot Title: Perl 6 FAQ - perl6
moritz_ is that the .WHAT and .HOW stuff?
pmurias ruoso: hi 12:50
ruoso hi pmurias
pmurias ruoso: are you blocking on anything? 12:54
ruoso: i yhink is should add context support to mold, is adding it to the method call a good idea? 12:56
pmurias is it possible to call a function with arbitary context in Perl 6? 13:03
moritz_ I think so, yes
pmurias moritz_: how? 13:04
my $context=:($ is rw); ???
moritz_ pmurias: by using the appropriate listop
or prefix, or whatever it is 13:05
pmurias moritz_: i meant in a generall way instead of using a (propably infinite if chain) or eval 13:06
* general 13:07
moritz_ pmurias: that I don't know 13:09
moritz_ masak: #58372 is actually "autovivification doesn't work" or something along these lines 14:04
masak moritz_: you're right. 14:08
feel free to add that
masak or even change the bug title 14:08
moritz_ otoh I don't know if $/ is allowed to autovivify, since it might be immutable (unless the regex was matched with 'is rw') 14:09
masak moritz_: btw, are you sure that long hashing chains autovivify? 14:10
moritz_ I'll just reply to the ticket
in perl 5 $h{a}{b} autovivifies $h{a}
it doesn't make sense to autovivify on the last indexing level, since you can always just assign to it 14:11
oh wait, I think in Perl 6 only write accesses autoviv
moritz_ so 'defined $/<a><b>' shouldn't, but '%h<a><b> = $stuff' should 14:12
masak moritz_: ISTR that too
what it shouldn't do is halt the program
hence the current title of the bug
moritz_ agreed
moritz_ is quite suprised, because he's seen some autovivification code in actions.pm 14:13
perhaps one of the features that weren't tested, and thus bitrotted 14:14
masak moritz_: but actions.pm is nqp, innit? 14:16
moritz_ masak: I mean PAST initialization code which handles autovivification 14:17
just search for :vivibase 14:18
masak ach so
yes, I also recall previous discussions about it 14:19
ruoso @tell pmurias, I'm starting to sketch the output of s1p... I'm kinda blocking on that because I'd use some help... you can see a file in doc/ 14:47
lambdabot Consider it noted.
dr_df0 rakudo: my @a=(1,2,3); my @b=(0,2); say @a[@b] 14:57
p6eval rakudo 29834: OUTPUT[3␤]
moritz_ rakudo doesn't do slices yet :( 14:57
dr_df0 :( 14:58
masak the implementor in me says "bah -- slices are sugar. they're nice, but you can work around not having them" 14:59
that said, they would sure be nice to have
moritz_ the implementor in me has looked at the code generation for slices, and sees that it might be a PCT thing
but I'm not quite sure 15:00
dr_df0 what's PCT?
moritz_ Parrot Compiler Tools
masak what does it mean, in this case, that it might be a PCT thing? 15:01
moritz_ dr_df0: the IRC logs at irclog.perlgeek.de/perl6/today automatically anotate some abbreviations with their meaning (just FYI, it's fine to ask here also) 15:01
lambdabot Title: IRC log for #perl6, 2008-08-26
moritz_ masak: method postcircumfix looks like that to me (but not sure) 15:02
masak oki
moritz_ it just generates a a PAST::Var node 15:04
from the semilist inside the subscript
but I don't find the place where that is actually turned into a hash or array lookup 15:05
the term postcircumfix only appears actions.pm, grammar.pg and the PIR files that are generated from those two 15:06
moritz_ wonders if it should be done in method variable 15:08
masak I wouldn't know. 15:09
moritz_ probably not, the only postcircumfix stuff it handles is $<foo> (ie acceess to $/)
masak ah. 15:10
moritz_ @tell TimToady S09:1209 uses exists as a function. Maybe the example should use defined() in both p5 and p6 15:43
lambdabot Consider it noted.
bennymack-work this is probably a silly question but "Int"s do not get automatically converted to "Num"s? 16:26
moritz_ they should, I think
bennymack-work I had a sub that takes a Num but I sent it a 3 and it died. Change the prototype to Int and it worked 16:27
moritz_ that's a bug in rakudo
bennymack-work rakudo: sub foo(Num $x) { say $x; } foo(2};
p6eval rakudo 29834: OUTPUT[Statement not terminated properly at line 1, near "foo(2};"␤current instr.: 'parrot;PGE::Util;die' pc 120 (runtime/parrot/library/PGE/Util.pir:82)␤]
bennymack-work oh ok 16:28
moritz_ you need a ; after the } if it's not followed by a newline
bennymack-work rakudo: sub foo(Num $x) { say $x; }; foo(2};
p6eval rakudo 29834: OUTPUT[Statement not terminated properly at line 1, near "(2};"␤current instr.: 'parrot;PGE::Util;die' pc 120 (runtime/parrot/library/PGE/Util.pir:82)␤]
moritz_ rakudo: sub foo(Num $x) { say $x; }; foo(2) 16:29
p6eval rakudo 29834: OUTPUT[Parameter type check failed␤current instr.: 'foo' pc 132 (EVAL_15:55)␤]
bennymack-work ahh, my bad
moritz_ it's a known rakudo bug
problem is, the numeric types don't fit well into the general type system
bennymack-work I just tried to run this from your perl-5-to-6 blog: say "Rest: @rest[]"; and it printed "Rest: @rest[]" 16:35
moritz_ that's also a known bug in rakudo :(
pugs: my @a = 1, 2, 3; say "@a[]"
p6eval pugs: OUTPUT[1 2 3␤]
moritz_ pugs does it right 16:36
(the tests are in t/spec/S02-literals/array-interpolation.t )
pugs_svn r22037 | moritz++ | [t/spec] more tests for hash interpolation 16:45
rindolf pugs: my ($x, $y) ; $x, $y = 5, 6; [ $x, $y] 16:54
p6eval pugs: RESULT[\(undef, 5)]
bennymack-work rakudo: multi sub frob(Str $s) { say "Frobbing String $s"; }; frob("x"); 17:10
p6eval rakudo 29834: OUTPUT[No applicable methods.␤current instr.: '_block11' pc 24 (EVAL_13:13)␤]
moritz_ another rakudo bug, and another bogus error message 17:11
bennymack-work eep
works on Int though
moritz_ yes 17:12
moritz_ types based MMD with Str somehow fails. 17:12
bennymack-work For lesson 6 I'm getting: You passed an initialization parameter that does not have a matching attribute. 17:17
moritz_ which code snippet? 17:18
bennymack-work In the SYNOPSIS
moritz_ d'oh, I have two lesson 6 ;)
bennymack-work i thought it was odd that it skipped lesson 05 17:19
perlgeek.de/blog-en/perl-5-to-6/05-...asses.html
lambdabot Title: blog | Perlgeek.de Blog, tinyurl.com/5ccj9l
moritz_ in case it's not clear, I didn't limit my examples to stuff that works at the moment ;/ 17:20
bennymack-work ok. but there's no way to make this particular example work? 17:21
moritz_ I guess it complains about the inherited attribute 17:22
so by removing the colour from the initialization at least the rest should work
bennymack-work ok, yes that does work. 17:23
I also had to comment out the "method area" in Shape otherwise I get this: Missing '}' at line 4, near "... } #"
moritz_ rakudo doesn't do '...', and it sometimes requires a ; after a } although that's not officially needed 17:27
pmurias ruoso: i would rather have s1p emit m0ld instead of C as gcc is sluggish and we can load mold directly or as bytecode in the future 18:06
pmurias ruoso: we shouldn't use ¢ in places where the root namespace is avalible 18:37
krzys pugs: my @array = (3,7,9,11); say @array[0..1,2,3] 20:59
p6eval pugs: OUTPUT[37911␤]
krzys pugs: { my @array = ((11,12,13,14,15)(21,22,23,24,25)); my @slice = ((0,4)(0,4); say @array(@slice); } 21:02
p6eval pugs: OUTPUT[*** ␤ Unexpected "@slice"␤ expecting "=", term postfix, operator, ")", context, ":" or "("␤ at /tmp/julJz4IK84 line 1, column 54␤]
moritz_ )( is illegal 21:03
dr_df0 pugs: { my @array = ((11,12,13,14,15),(21,22,23,24,25)); my @slice = ((0,4),(0,4)); say @array(@slice); } 21:04
p6eval pugs: OUTPUT[*** Cannot cast from VList [VInt 11,VInt 12,VInt 13,VInt 14,VInt 15,VInt 21,VInt 22,VInt 23,VInt 24,VInt 25] to Pugs.AST.Types.VCode (VCode)␤ at /tmp/Zj4woO5Uxx line 1, column 83-97␤]
dr_df0 pugs: { my @array = ((11,12,13,14,15),(21,22,23,24,25)); my @slice = ((0,4),(0,4)); say @array[@slice] } 21:05
p6eval pugs: OUTPUT[11151115␤]
dr_df0 moritz_ )( ... i missed that 21:06
Lichtkind does "file.name".:X work? 22:07
TimToady STD thinks (1,2,3)(1,2,3) is just fine 22:20
moritz_ TimToady: yes, but it's not what dr_df0 would expect (most likely) 22:21
TimToady it just doesn't mean multiple dimensions :)
moritz_ it's a postcircumfix:<( )> on a list, right?
Lichtkind TimToady: does :r also match do literals? 22:25
moritz_ imagines TimToady waking up, opening irssi with sleepy eyes, and being DDOSed by #perl6 ;) 22:26
TimToady suspects moritz_ is correct on both points 22:33
Lichtkind: I think :r would be pretty useless if it didn't work on literals 22:35
Lichtkind TimToady: true but i ask because all the examples in spec where with var 22:36
TimToady it's a little odd that it means that the methods are defined in Str 22:37
assuming they are true methods, which is the latest thinking
or maybe they live in Any and coerce to Str 22:38
moritz_ it seems weird to have a method with the name of pair
TimToady it does kinda reserve a little namespace for such methods
moritz_ how would you define such a thing? method postfix:':r' { ... } ? 22:39
TimToady which we have also made use of with .:delete and whatever .:exists ends up being called
moritz_ is that a trigil? ;-) 22:40
TimToady we could do worse than the postfix declaration 22:41
originally I was thinking just method :r, but making trait notation a variant of pair notation mades it a problem
moritz_ but do we really need a grammatical construct for every file test operator? 22:42
is $fn ~~ :r such a bad huffman coding, as compared to $fn.:r?
TimToady yeah, and I'm not sure that $file:r is a good construct 22:43
originally we had all the :r things interpreted by another method
well, I don't make decisions when I've just woken up and been DDOS'd :) 22:44
moritz_ ;)
Lichtkind TimToady: thanks i really didnt say goodbye in kobenhavn 22:45
moritz_ slowly wanders off towards bed... 22:45
TimToady Lichtkind: if everyone at a conference said goodbye to everyone else we'd have to schedule an extra week
fullermd But with an extra week, you'd run into people you'd already said goodbye to, so you'd need extra time to re-bye... 22:46
moritz_ O(n²)
Lichtkind true 22:46
Lichtkind www.youtube.com/watch?v=bLiX5d3rC6o 23:14
lambdabot Title: YouTube - Meet "Emily" - Image Metrics Tech Demo
Lichtkind ups wrong channel

Powered by Cro and the Raku® Programming Language.

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

Thank you!