The topic for #perl6 is: pugscode.org/ planetsix.perl.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 TimToady on 25 November 2008.
00:50 alc joined 00:51 cspencer joined 00:53 [particle] left
Caelum should I read "perl6 and parrot essentials" or is it completely out of date? 01:02
01:10 Larry joined, Larry left 01:11 Larry joined 01:12 Larry is now known as Hell_Kaiser, [particle] joined, Hell_Kaiser left 01:15 Hell_Kaiser joined 01:16 cspencer_ joined 01:17 cspencer left
pmichaud Caelum: it's very much out of date. 01:22
01:23 Limbic_Region joined 01:29 hercynium left 01:34 [bjoern] joined
pasteling "[bjoern]" at 88.75.226.0 pasted "Latest svn, win32 build failure (In file included from perl5\p5embed.c:1:0: ...)" (32 lines, 1.2K) at sial.org/pbot/34137 01:35
[bjoern] perhaps it's about the " in '/Perl/lib/CORE"/EXTERN.h'? 01:37
not sure where that is coming from...
Caelum pmichaud: I see, thank you. What should I read then to learn perl6? "beginning perl6" is not going to be out until october 01:38
pmichaud Caelum: that's a good question at the moment. You might look and see if there's anything useful on the perl6 wiki, or else there are the synopses. 01:39
perl6 wiki: www.perlfoundation.org/wiki/perl6
Caelum will do, thank you 01:40
pmichaud oops, wrong url.
just a sec
[bjoern] ( www.perlfoundation.org/perl6 )
pmichaud www.perlfoundation.org/perl6/index.cgi?perl_6 01:41
bjoern's url works also (and is probably more canonically correct)
Limbic_Region hrm 01:46
suggestion if anyone here has maintenance rights to www.perlfoundation.org/perl6/index.cgi?fud 01:47
it is unclear that some of the FUD has no rebuttal
www.perlfoundation.org/perl6/index....an_modules
for instance
oh, and some rebuttals really should be updated 01:49
01:49 icwiener left, jferrero joined
Limbic_Region "If you do not want to have to keep Perl 5 around, the Ponie project aims to port Perl 5 on Parrot....." 01:49
Ponie has been dead for years 01:50
[bjoern] So I take it it's Configure.PL that tries to get the path to the perl5 headers... 01:55
and it seems it tries to get the path out of the libpath flag... 01:56
hmm looks like a non-greedy match gone wrong there
that does not appear to be the problem 02:03
hmm it may be caching the configure.pl ... 02:05
rakudo_svn r34682 | pmichaud++ | [rakudo]: eval() returns Nil for void expressions (RT #61868, lathos++) 02:10
02:13 hercynium joined
pugs_svn r24698 | lwall++ | [S02] clarification of meaning of HyperWhatever 02:14
[bjoern] Does `$flags =~ s{([\\"'])}{\\$1}g;` do anything useful?
Oh i see... 02:15
Caelum my perl6 segfaults :( 02:17
oh right, I forgot to make clean first 02:18
pugs_svn r24699 | lwall++ | [S03] clean up some awkward phrasing 02:23
02:27 kisu left 02:28 DemoFreak left
pasteling "[bjoern]" at 88.75.226.0 pasted "New build failures, ActivePerl 5.10.x's /Perl/lib/CORE/win32iop.h probably fails to define intptr_t in a manner compatible with mingw" (13 lines, 637B) at sial.org/pbot/34138 02:29
[bjoern] I only got there by stripping the quote marks from Pugs.buildinfo manually and setting the file to read only
somehow changes to Configure.PL didn't propagate quite as they should have 02:30
So let's patch the perl header file and see... 02:31
Great, now it fails because it cannot delete Pugs.buildinfo 02:32
02:33 kisu joined
[bjoern] ah I take it I have to modify the configure.pl in some Pugs-6.2.13.14.tar.gz file... 02:34
02:46 cspencer_ left
rakudo_svn r34683 | pmichaud++ | [rakudo]: Enable whatever star meaning "full slice" (e.g. @a[*]) 02:50
02:55 Whiteknight left 03:03 cataska left 03:05 jhorwitz left 03:17 azawawi joined
azawawi hi 03:17
meppl good night 03:28
03:30 meppl left 03:43 Limbic_Region left 04:00 elmex left 04:02 elmex joined, c9s__ joined
azawawi # last day morning in 2008 ;-) 04:11
04:21 s1n left 04:47 hercynium left 04:50 kisu left 04:54 kisu joined 05:37 simcop2387 left 05:38 simcop2387 joined 05:47 alc left 06:09 brunoV left 06:11 jferrero left 06:18 justatheory joined 07:22 ChrisDavaz joined 07:33 samlh joined
samlh perl6: (0..1,2)[1].perl.say 07:39
p6eval pugs: OUTPUTĀ«\1ā¤Ā»
..rakudo 34688: OUTPUTĀ«2ā¤Ā»
..elf 24699: OUTPUTĀ«Can't use string ("2") as an ARRAY ref while "strict refs" in use at (eval 117) line 3.ā¤ at ./elf_f line 3861ā¤Ā»
samlh I beleive rakudo is wrong; should submit a rakudobug? 07:41
07:49 Hell_Kaiser left 07:57 samlh_ joined 08:00 samlh left, samlh_ is now known as samlh 08:08 samlh left 08:09 vixey joined 08:15 [bjoern] left 08:27 azawawi left 08:28 DemoFreak joined 08:33 iblechbot joined 08:34 azawawi joined 09:09 barney joined 09:12 pdcawley joined 09:17 DemoPhreak joined 09:33 DemoFreak left 09:40 vixey left, duke_leto joined, vixey joined
rakudo_svn r34694 | bernhard++ | [codingstd] Satisfy trailing_space.t and pdd_format.t 10:00
10:09 pmurias joined 10:14 ejs joined 10:23 kisu_ joined 10:24 pmurias_ joined, pmurias_ left 10:26 tomyan joined 10:28 kisu left
pmurias ruoso: i'm going to the runloop integration with Coro now, it's going to be slow but i don't know how to do it otherwise 10:32
10:37 yantom joined 10:48 tomyan left 10:49 justatheory left 10:51 maerzhase joined 10:57 kidd left 11:02 kidd joined 11:07 pdcawley left 11:10 ejs left 11:17 kisu__ joined
moritz_ @tell masak I'm working on adding isa, does, can, clone to S29, but atm I can't commit here due to technical limitations 11:20
lambdabot Consider it noted.
11:25 kisu_ left 11:46 clintongormley joined
pmurias moritz_: what stopes you from commiting, lack of a subversion client or a pesky proxy? 11:50
12:05 tacocat joined 12:15 icwiener joined 12:17 kisu_ joined 12:18 slavik joined 12:19 rindolf joined 12:23 kisu__ left 12:32 masak joined
ruoso TimToady, Foo::.bar() really parses in a weird way... it considers "bar" a sym, instead of a dotty 12:35
12:38 tacocat left 12:39 tacocat joined 12:40 tacocat left 12:43 azawawi is now known as azawawi2009 12:57 meppl joined 13:51 justatheory joined 14:04 vixey left 14:05 vixey joined
ruoso TimToady, forget it... it just works now... ;) 14:07
masak just like that? 14:08
lambdabot masak: You have 1 new message. '/msg lambdabot @messages' to read it.
masak @messages
lambdabot moritz_ said 2h 47m 57s ago: I'm working on adding isa, does, can, clone to S29, but atm I can't commit here due to technical limitations
masak moritz_++
moritz_: I'm dangerously close to working on S29 too today. thanks for the update.
slavik fix eval? 14:09
:P
perl6: my $a = 5; eval 'say $a';
p6eval elf 24699, pugs: OUTPUTĀ«5ā¤Ā»
..rakudo 34701: RESULTĀ«undefĀ»
slavik ooh, one of them works :D
masak hey, something's changed :)
slavik my $a = 5; eval 'say $b';
masak rakudo: my $a = 5; eval 'say $b'; 14:10
p6eval rakudo 34701: RESULTĀ«undefĀ»
masak that's correct, at least.
rakudo: my $a = 5; eval 'say $b'; say $!
p6eval rakudo 34701: OUTPUTĀ«Scope not found for PAST::Var '$b'ā¤Ā»
masak rakudo: my $a = 5; eval 'say $a'; say $! 14:11
p6eval rakudo 34701: OUTPUTĀ«Scope not found for PAST::Var '$a'ā¤Ā»
masak hah!
that's wrong.
masak reports this new development
slavik: I'd fix it, but the bug is likely somewhere in lexicals land, which scares me a bit. 14:14
pugs_svn r24700 | ruoso++ | [mildew] Foo::.method works now
slavik :(
ruoso mildew: module Foo { }; $OUT.print(Foo::.name.FETCH) 14:15
p6eval mildew: OUTPUTĀ«Could not find variable $?PACKAGE in the lexical scope.ā¤Ā» 14:16
ruoso hmm... mildew evalbot is a bit outdated
masak slavik: all right, I'll look. but I might not get very far.
slavik :D 14:17
masak I've always wanted to know how the eval function works anyway. 14:19
14:19 kisu_ is now known as kisu
masak rakudo: eval "use Foo" or die $! 14:20
p6eval rakudo 34703: OUTPUTĀ«Can't find ./Foo in @INCā¤current instr.: 'die' pc 14778 (src/builtins/control.pir:188)ā¤Ā»
masak hm, that's different from the output on my bow.
s/w/x/
updating.
pmurias ruoso: does make test in mildew work for you? 14:21
ruoso pmurias, I had only one or two tests failing 14:22
ruoso re-running
pmurias, the p5 test has two sub-tests failing 14:24
pmurias fixed
pugs_svn r24701 | pmurias++ | [mildew] fixed test count in p5.t
ruoso and classhow_add_method_real.t needs to be implemented with the new signature
slavik is there support for .perl on matchers yet? 14:27
rakudo: say 2**32
p6eval rakudo 34703: OUTPUTĀ«4294967296ā¤Ā»
slavik rakudo: say 2**64
p6eval rakudo 34703: OUTPUTĀ«1.84467440737096e+19ā¤Ā»
slavik :(
masak slavik: I'm getting increasingly curious as to what it is you're trying to build. 14:28
slavik masak: I would like to be able to match a string against a grammar and get a dump of the structure of the matcher object that is returned 14:29
masak slavik: aye, that'd be neat. is it specced? 14:30
slavik masak: not sure :(
masak: is it possible, looking at a var to tell what the type is, like if it's a plain scalar or a reference to an array and such
masak slavik: the var or the value it contains? 14:31
slavik the var
masak slavik: $x.VAR.WHAT, perhaps?
slavik like $var =5; $var2 = 1,2,3,4; to be able to tell that $var holds a single value and that $var2 is a reference to an array
masak rakudo: 'foobar1bar2' ~~ / [ bar \d ]+ /; say $/.perl 14:32
p6eval rakudo 34703: OUTPUTĀ«Method 'perl' not found for invocant of class 'Match'ā¤current instr.: '_block14' pc 78 (EVAL_15:43)ā¤Ā»
slavik that's interesting ...
masak slavik: well, $var.WHAT is enough for that.
slavik I see 14:33
masak slavik: becuase you're only interested in the value inside the variable.
slavik rakudo: my $var =5; say $var.WHAT;
p6eval rakudo 34703: OUTPUTĀ«Intā¤Ā»
slavik rakudo: my $var =1..6; say $var.WHAT;
p6eval rakudo 34703: OUTPUTĀ«Rangeā¤Ā»
slavik rakudo: my @list = 1..5; my $var = @list; say $var.WHAT;
p6eval rakudo 34703: OUTPUTĀ«Arrayā¤Ā»
slavik awesome
rakudo: my @list = 1..5; my $var = @list; say $var.perl;
p6eval rakudo 34703: OUTPUTĀ«[1, 2, 3, 4, 5]ā¤Ā»
slavik rakudo: my %list = 1..5; my $var = @list; say $var.perl; 14:34
p6eval rakudo 34703: OUTPUTĀ«Scope not found for PAST::Var '@list'ā¤current instr.: 'parrot;PCT;HLLCompiler;panic' pc 146 (src/PCT/HLLCompiler.pir:102)ā¤Ā»
slavik rakudo: my %list = 1..5; my $var = %list; say $var.what, $var.perl;
p6eval rakudo 34703: OUTPUTĀ«Odd number of elements found where hash expectedā¤current instr.: 'parrot;Perl6Hash;!STORE' pc 5748 (src/classes/Hash.pir:154)ā¤Ā»
slavik rakudo: my %list = 1..6; my $var = %list; say $var.what, $var.perl;
p6eval rakudo 34703: OUTPUTĀ«Method 'what' not found for invocant of class 'Perl6Hash'ā¤current instr.: '_block14' pc 122 (EVAL_14:55)ā¤Ā»
slavik rakudo: my %list = 1..6; my $var = %list; say $var.perl;
p6eval rakudo 34703: OUTPUTĀ«{"1" => 2, "3" => 4, "5" => 6}ā¤Ā»
slavik rakudo: my %list = 1..6; my $var = %list; say $list.what; 14:36
p6eval rakudo 34703: OUTPUTĀ«Scope not found for PAST::Var '$list'ā¤current instr.: 'parrot;PCT;HLLCompiler;panic' pc 146 (src/PCT/HLLCompiler.pir:102)ā¤Ā»
slavik rakudo: my %list = 1..6; my $var = %list; say %list.what;
p6eval rakudo 34703: OUTPUTĀ«Method 'what' not found for invocant of class 'Perl6Hash'ā¤current instr.: '_block14' pc 122 (EVAL_14:55)ā¤Ā»
slavik :(
masak .WHAT, not .what 14:37
slavik oh
rakudo: my %list = 1..6; my $var = %list; say %list.WHAT;
p6eval rakudo 34703: OUTPUTĀ«Hashā¤Ā»
slavik masak: that's a bit inconsistant ...
.perl, but .WHAT
is .perl a function?
masak slavik: yes. and .WHAT is (ideally) a macro.
slavik I see
masak well, .perl is a method. :)
slavik masak: method/function ... 14:38
masak aye.
slavik masak: same **** different name :P
masak except when it isn't.
slavik is it possible to dereference a var as a hash to see what functions/macros and such it has?
14:38 yantom left
masak it will be. 14:38
slavik awesome 14:39
masak %hash.^methods
slavik I mean to do it for $var
masak or %hash.WHAT.methods (I think)
slavik: exemplify.
slavik in perl5, you can take a package, treat it as a hash and print out all the functions/variable/whatever it has 14:40
masak you can? neat!
slavik err, you didn't know that? 14:41
masak I know at least classes will have such introspection capabilities in Perl 6. don't know about modules and packages and such.
slavik: no, I'm relatively inexperienced as a Perl 5 programmer. 14:42
slavik I see
masak slavik: I use it a lot, but I'm no expert.
slavik it would be nice, because then you can literally check if the module/whatever has the functionality you need
no need to rely on version and such, unless the output from stuff changes 14:43
14:43 jhorwitz joined
slavik masak: the package introspection is how I was able to get a list of the Purple and Pidgin API that is exposed to the Perl interpreter. :) then it was a matter of looking up the appropriate C function to see what it does 14:44
masak slavik: I suggest you make it your mission to have this corner of Perl 6 specced and eventually implemented. 14:45
slavik: I also predict that this will make you very happy.
slavik yes
I love that Perl let's you do everything ...
masak slavik: seriously. go put your suggestions in the nearest Spec document. 14:46
wolverian including misusing apostrophes! graa!
wolverian runs outside screaming
masak wolverian++ # 哈哈哈
wolverian are those houses? :) 14:47
masak wolverian: I thought you were outside. 14:48
wolverian: no, they're sinographs for laughter.
wolverian: I've found that's a very logical way to depict laughter.
wolverian ah. I thought it was the scenery. 14:49
I have a laptop; I can run outside _and_ IRC.
masak wolverian: I can picture it very clearly.
slavik masak, will look at it ...
but been busy 14:50
wolverian yet another normal day in my life; irc.fi/wolverian might help.
masak ah, .fi
hello, neighbour.
slavik masak: here's another one ... since it is possible for evaled code to inherit the current state (defined vars and such), what if it was possible for the oposite? define a var in an eval and then access it after eval 14:51
so that basically, the eval simply copies/pastes the codes like a #define
wolverian maybe I will run to sweden!
14:52 tomyan joined
masak wolverian: screaming all the way! 14:52
wolverian absolutely. though I suppose if I want to run over land, the detour is a bit long.. 14:53
masak wolverian: or run right up to the border, throwing the extra apostrophe as long as you can over it.
14:54 tomyan left
wolverian haha. 14:55
masak++
masak if I can keep this up, I'll acheive my life-long dream: being able to write like _why. 14:56
poignantguide.net/ruby/chapter-5.html 14:58
slavik: see 'macros'.
slavik will do 14:59
in that chapter?
masak S06, methinks. 15:01
somewhere near the bottom.
slavik: got ack?
slavik huh? 15:02
masak search.cpan.org/dist/ack/
slavik what's the new url for the synopsis?
never used it 15:03
masak slavik: I suggest you try it. specifically, download Pugs and docs/Perl6/Spec, and use ack on it. you'll never have to search manually again. 15:04
slavik I see
masak
.oO( maybe now would be a good time to implement that spec-ack-bot we've been mumbling about )
slavik: perlcabal.org/syn/S06.html
bbiab
slavik bbiab? 15:05
15:05 masak left
ruoso pmurias, I was thinking we could refactor mildew, creating a src/VAST directory with each vast type in a separated file 15:05
slavik, Be Back In A Bit
slavik oh, ok
brb is shorter
ruoso pmurias, mildew is starting to be too big in loc 15:06
pmurias ruoso: ok 15:09
you mean like moving VAST::comp_unit::emit_m0ld to src/VAST/comp_unit.pm 15:11
?
ruoso yes...
pugs_svn r24702 | ruoso++ | [mildew] initial support for Symbolic_unary 15:12
pmurias seperating the traversal routines from the main code would be a good start
ruoso then we do a "for <src/VAST/*.pm> { require $_ }"
pmurias i'm fighting with Coro now feel free to refactor mildew 15:15
15:18 tomyan joined 15:26 kisu left 15:27 masak joined, tomyan left
masak slavik: 'brb' is shorter, but it also IMO conveys a shorter time span 15:30
15:30 kisu joined
slavik oh 15:30
SamB how about biab? that's shorter than bbiab but conveys the same idea, no? 15:31
masak SamB: aye.
offhand, I don't see a difference between the two. 15:32
Matt-W aieee
my parrot won't compile
masak Matt-W: 'make realclean; perl Makefile.PL; make'? 15:33
Matt-W masak: I shall try 15:36
slavik make realclean; perl Makefile.PL; make; make perl6; sudo make reallyinstall
that's what I do :)
and an svn update after make realclean :) 15:37
Matt-W I don't install my parrot yet
which means my perl 6 scripts have #!/home/matthew/Compiling/parrot/languages/perl6/perl6 at the top
pmurias ruoso: have you ever tried using gobby?
Matt-W but I can live with that :)
slavik k 15:38
Matt-W how's rakudo doing for rule/grammar support at the moment?
slavik Matt-W: really good :)
Matt-W: bcacm.org/~slavik/
15:39 alester joined, ft joined
Matt-W slavik: thanks. I'm planning to try something with them and thought I should check if it's a pointless undertaking yet 15:40
ruoso pmurias, gobby? 15:41
slavik Matt-W: afaik there are no assertions yet
Matt-W slavik: I don't know what assertions are yet, so perhaps I won't miss them too much 15:42
I really need to play with rules to understand more about how they work
slavik Matt-W: after you read the post, you will start missing them ;)
Matt-W: rules, tokens, etc are just regex with some preapplied flags.
pmurias ruoso: gobby is a "multiplayer" editor 15:43
slavik grammar = hierarchy of said rules, tokens, regex
pmurias: I believe that is IRC :P
pmurias slavik: do you consider cat(1) an editor? 15:45
slavik hmm 15:46
no, but cat | sed is :)
pmurias doesn't sed make only one pass? 15:47
i guess if we all used ed we could have an editorbot
Matt-W I'm afraid I'm very fussy and am only truly happy when I'm using vim 15:48
ruoso pmurias, is there a debian package for gobby?
pmurias yes gobby
ruoso pmurias, I cant host a gobby session, because I'm behind asquerade 15:52
pmurias you mean masquerade? 15:53
ruoso yes 15:54
Matt-W woo, I have a perl6 again
I shall have to remember make realclean for next time, as make clean didn't sort it out
pmurias ruoso: opened one on feather.perl6.nl:6522 15:56
ruoso pmurias, it doesn't seem to have syntax highlight 15:57
ah it does
15:58 kisu left 15:59 kisu joined
pmurias the editor doesn't support vim binding unfortunatly... 15:59
Matt-W do grammars, rules and objects share the same namespace? 16:00
slavik Matt-W: grammars and classes, probably
pmurias rules share the namespace with methods 16:01
ruoso pmurias, hmm... I think emacs+irc seems more productive
Matt-W okay
16:02 kisu left 16:15 lichtkind joined, kisu joined 16:16 eric2561 joined
eric2561 @tell unobe i added you to the git-hub repo as a collaborator 16:17
lambdabot Consider it noted.
pugs_svn r24703 | pmurias++ | [smop] trying out using Coro for runloop integration - proof of concept pulp 16:22
masak eric2561: how's the repo going? 16:28
16:29 duke_leto left
ruoso pmurias, I'm doing that refactoring... so if you plan to change mildew, please wait a little bit... 16:29
masak Matt-W: instead of providing the whole path to perl6 in your perl scripts, consider creating an alias.
Matt-W: `alias perl6=/.../parrot/languages/perl6` 16:30
Matt-W masak: at this stage, I'm quite content with the current arrangement
masak and putting it in your ~/.bash_profile
Matt-W: ok. you do what you prefer.
Matt-W I am having some trouble with a rule though
slavik how about PERL6LIB that is predefined?!
:P
Matt-W: see my example, if you use rule, the spaces between braces matter 16:31
masak slavik: doesn't solve the above problem, but yes, it'd be nice.
slavik hence the use of tokens in my article
Matt-W I thought whitespace was always metasyntactic in perl 6 rules
slavik Matt-W: not in rules, tokens yes, rules no 16:32
regex is also /x I think
masak all of them are /x in Perl6 16:33
16:33 DemoPhreak left
Matt-W it appears that the :sigspace modifier (default on rules) makes *some* whitespace significant 16:34
eric2561 masak just stalled while i'm vacationing ;) the few days i've had to work on it i've been to busy with real work...and now its back to cleaning the house, but i'll be back soon
masak eric2561: sounds good. looking forward to seeing it grow and prosper.
slavik Matt-W: I prefer to use token :) 16:35
ruoso pmurias, it wasn't that hard... I'm about to commit the refactoring.. 16:36
szabgab rakudo: $year = 2008; $year++; $year.say 16:38
p6eval rakudo 34710: OUTPUTĀ«Scope not found for PAST::Var '$year'ā¤current instr.: 'parrot;PCT;HLLCompiler;panic' pc 146 (src/PCT/HLLCompiler.pir:102)ā¤Ā»
szabgab rakudo $year = 2008;
rakudo: $year = 2008; 16:39
p6eval rakudo 34710: OUTPUTĀ«Scope not found for PAST::Var '$year'ā¤current instr.: 'parrot;PCT;HLLCompiler;panic' pc 146 (src/PCT/HLLCompiler.pir:102)ā¤Ā»
szabgab rakudo: my $year = 2008;
p6eval rakudo 34710: RESULTĀ«2008Ā»
szabgab rakudo: my $year = 2008; $year++;
p6eval rakudo 34710: RESULTĀ«2008Ā»
szabgab rakudo: my $year = 2008; $year++; $year.say
p6eval rakudo 34710: OUTPUTĀ«2009ā¤Ā»
szabgab rakudo: my $year; $year //= 2009;
p6eval rakudo 34710: OUTPUTĀ«get_bool() not implemented in class 'Timer'ā¤current instr.: 'parrot;PCT;HLLCompiler;command_line' pc 1442 (src/PCT/HLLCompiler.pir:771)ā¤Ā»
masak reports rakudobug 16:40
rindolf szabgab: hi. 16:41
16:41 vixey left
rindolf Happy new year 16:41
16:43 cspencer joined, ejs joined
pmurias rindolf: happy new year 16:44
rindolf pmurias: thanks. 16:45
pugs_svn r24704 | ruoso++ | [mildew] refactor VAST::* to separated files, should be easier to maintain...; 16:46
ruoso still counts a few hours (and a second) to new year
masak yes, can't they leave those seconds alone? 16:47
now we have to go '3 .. 2 .. 1 .. hold it .. happy new year!' this year.
ruoso and you never know if the clock was already updated or not 16:48
because the second is added 00:00:00 UTC
slavik rakudo: say 6..1;
p6eval rakudo 34710: OUTPUTĀ«ā¤Ā»
slavik rakudo: say -6..1;
p6eval rakudo 34710: OUTPUTĀ«-6-5-4-3-2-101ā¤Ā»
ruoso I always thought 6..1 would return 6,5,4,3,2,1
slavik ruoso: I thought so, too 16:49
perl6: say 6..1;
p6eval elf 24703, pugs, rakudo 34710: OUTPUTĀ«ā¤Ā»
slavik oh, wow
icwiener say reverse 1..6 16:50
rindolf say 6..1:-1
rakudo: say 6..1:-1
p6eval rakudo 34710: OUTPUTĀ«Statement not terminated properly at line 1, near ":-1"ā¤ā¤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)ā¤Ā»
rindolf rakudo: say 6..1,-1
p6eval rakudo 34710: OUTPUTĀ«-1ā¤Ā»
ruoso rakudo say 6..1 :by -1 16:51
pmurias how was the env variable to tell where the perl5 modules live called?
ruoso perldoc perlrun should tell you
icwiener rakudo: say reverse 1..6 16:52
p6eval rakudo 34710: OUTPUTĀ«654321ā¤Ā»
ruoso rakudo: say 6..1 :by -1
p6eval rakudo 34710: OUTPUTĀ«Statement not terminated properly at line 1, near ":by -1"ā¤ā¤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)ā¤Ā»
masak needs parens. 16:53
icwiener rakudo: say 6..1:by(-1)
p6eval rakudo 34710: OUTPUTĀ«Statement not terminated properly at line 1, near ":by(-1)"ā¤ā¤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)ā¤Ā»
masak rakudo: say (6,,1):by(-1) 16:54
p6eval rakudo 34710: OUTPUTĀ«Statement not terminated properly at line 1, near ":by(-1)"ā¤ā¤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)ā¤Ā»
masak rakudo: say (6..1):by(-1)
p6eval rakudo 34710: OUTPUTĀ«Statement not terminated properly at line 1, near ":by(-1)"ā¤ā¤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)ā¤Ā»
masak ok, Rakudo simply doesn't implement that yet.
icwiener S03 says: 2..1:by(-1) but reverse 1..2 is preferred because it works for alphabetic ranges as well. 16:57
masak icwiener: not true anymore. :) 16:58
at least, I've written tests for ('C'..'A'):by(-1)
see recent discussion on this channel
(as in, a few days back)
pugs_svn r24705 | ruoso++ | [mildew] support :(\$capture) in signatures, all tests pass now ;) 16:59
icwiener Ok. Is S03 going to be updated accordingly? I mean, is it standard procedure?
masak icwiener: the spec and the tests will just have to fight it out, I guess, 17:00
icwiener What does that mean? 17:01
17:01 vixey joined
ruoso now mildew supports receiving a capture as-is and passing it forward to another method without enforcing any context on it... :) :) :) 17:02
this means that you can call foo(bar(),baz(),bla()) and foo might delay enforcing the context as long as it wants
17:10 Morpheus[PL] joined
Morpheus[PL] Hi 17:10
pmurias Morpheus[PL]: hi
Morpheus[PL] Hi pmurias (: 17:11
17:11 ejs left
Morpheus[PL] What I need to try perl6? 17:13
Only pugs or pugs+parrot..?
slavik Morpheus[PL]: pugs or rakudo :)
17:13 hercynium joined
slavik I'd suggest rakudo since it is more feature complete atm 17:13
rakudo = perl6 on parrot 17:14
Morpheus[PL] Hm
So what is pugs for ?
It's different perl6 implementation?
pmurias yes, pugs is written in haskell and dead atm 17:15
17:15 mberends joined
Morpheus[PL] So which one will be official perl distribution in future? 17:15
slavik Morpheus[PL]: anything passing the spec test :) 17:16
pugs_svn r24706 | lwall++ | [undef.t] there is no my() function...declarators require space 17:17
17:21 Morpheus[PL] left
ruoso TimToady, "is export" only means that it is exportable, right... you still need to "use Foo <symbol>" to import it, right?, does the :DEFAULT tagset means that it is imported even without explicit request? 17:25
17:25 cspencer left, smtms left
TimToady that was how it was designed, but I'm thinking multis should always be imported by default 17:26
most languages that do multiple dispatch just make them all global 17:28
we don't do that, but importing all the multis we find anywhere seems like the next best thing
where, in this case, "next best" actually means "better" :) 17:29
pmurias if do different modules declare a multi with proto should we just mix them together? 17:31
as multis without a proto declaration declare additional meaning and ones with a proto declaration override it 17:32
TimToady we could just merge them into a multisig proto
17:33 jrockway joined
TimToady then we complain only on a multi that doesn't match either of the protos 17:33
on the other hand, it might pessimize rewriting named args to positionals, and make it much harder to do multiple dispatch 17:34
17:34 araujo left
literal hm, are private object attributes read only by default? 17:35
slavik rakudo: my $a = 1..10; for =$a { say $_; }
p6eval rakudo 34712: OUTPUTĀ«1ā¤2ā¤3ā¤4ā¤5ā¤6ā¤7ā¤8ā¤9ā¤10ā¤Ā»
TimToady readonly to whom?
slavik rakudo: my $a = 1..10; for =$a { say; }
p6eval rakudo 34712: OUTPUTĀ«ā¤ā¤ā¤ā¤ā¤ā¤ā¤ā¤ā¤ā¤Ā»
literal to the object itself 17:36
TimToady no, they're rw
literal oh, ok, so Ovid was just being overly verbose in his hangman source then
use.perl.org/~Ovid/journal/38191
TimToady yes, "is rw" is applied only to the accessor, and since there aren't any accessors there, it's a Useless Use of "is rw" 17:38
literal will Rakudo eventually warn of such things? 17:39
pmurias if someone implement it
* implements
literal yeah, I'm just asking if that's something to be desired or not :) 17:40
TimToady we could require it in the spec, but that seems like overkill
in general, it's fine to warn about anything the compiler knows is completely bogus
it's when it starts guessing that it gets into trouble 17:41
there are some instances of .= that STD currently guesses about wrongly, for instance
literal that's temporary, right? I thought getting rid of guessing was one of the objectives with Perl 6 :) 17:42
TimToady well, guessing on warnings has less danger than guessing on the actual p6 parse... 17:44
17:44 xinming joined
literal eric2561: so how does this work, what do I do if I want to push a patch (e.g. euler solution) to the perl6-examples repo? 17:48
TimToady and in the case of .= it continues parsing as if it's a mutating method call, but it doesn't like my Fooz $f .=(:x(1)) from t/oo/construction.t
ruoso TimToady, it feelds awkward that multis get to the default tagset where regular subs doesn't 17:51
azawawi2009 starts highlighting hangman ;-) 17:52
TimToady it's because multis are really more like methods than like functions
17:53 barney left
TimToady you don't have to worry about exporting methods either 17:53
they Just Happen
the only reason we don't just shove all multis into GLOBAL is that we want different lexical scopes to be able to support different languages
and we don't want magical action at a distance, esp at run time 17:54
so we want to know at least the set of protos by compile time
we could go as far as to disallow run-time multis in packages unless there's a compile-time proto in that package 17:56
because it would be really nice for the optimizer to know exactly what set of multis it's dealing with
azawawi2009 TimToady: p5 '~=' => p6 '~~', right? 17:57
TimToady =~ rather
in p5 ~= would be the xor assignment operators
azawawi2009 so in line 15: subset Letter of Str where { $_ =~ /^ <[a..z]> $/ }; 17:58
TimToady or maybe that's ^=
azawawi2009 i get obsolete use of =~, please use ~~
17:58 azawawi2009 is now known as azawawi
TimToady indeed 17:58
the warning is entirely correct 17:59
masak I was going to write that in a comment.
just haven't gotten around to it.
17:59 xinming_ left
TimToady and p6 will parse that as $_ = ~/.../ 17:59
17:59 s1n joined
TimToady so it's Simply Wrong 17:59
18:00 apeiron left
ruoso TimToady, ok... mildew/smop doesn't support multis yet, so I'm more concerned about the exportation of regular subs 18:01
s1n jerry gay is responsible for the S19 draft, correct?
18:03 pdcawley joined
[particle] yes, i am. 18:04
ruoso TimToady, anyway... have you thought about the 'capture in item context' issue?
s1n [particle]: i have a question about one part, it says single letter options cannot be clustered 18:05
[particle]: why? that not only breaks from perl5 but also most unices
[particle] you need to get the updated version
s1n lemme check the cabal...
[particle] clustering is now allowed again, see...
perlcabal.org/syn/S19.html 18:06
s1n ahh yeah, okay, thanks heh
sorry, i was sifting through _a lot_ of spec commit emails heh
[particle] no problem, i committed a bunch of changes yesterday which i made on holiday travel flights 18:07
s1n ahh okay, well i clearly have a few weeks worth of ml emails to sift though, thanks for saving me from making an embarassing email :) 18:08
[particle] i'd be very happy to hear any other comments you have
ruoso TimToady, smop currently implements the changed behavior, you can see how that looks like in pugs/v6/mildew/t/return_function.t
s1n there's a bunch to digest, looks like the arguments will get very complex very quick but i'm not really that much of a perl expert :) 18:09
the open/close delimiters look confusing, not sure how you're going to use that 18:10
[particle] basically, we need to be able to pass values through to subsystems without perl parsing them 18:11
s1n is that the best way to do it? ...
how is --argsproc options,values,options,values any different than what you have?
in other words, why is the closing delimiter needed? 18:12
[particle] because the subsystem might take an option that is a valid perl 6 option
need to make sure perl doesn't get confused
s1n pardon my ignorance, but couldn't you assume the next token shouldn't be parsed? 18:13
azawawi syntax highlighted hangman -> feather.perl6.nl/~azawawi/hangman/
s1n btw, autoloop-print is missing the '--' in front
[particle] ++parser -o -- -foobar-file- ++/parser -o foo -O pbc
s1n what about: --parser '-o -- -foobar-file' -o foo -O pbc 18:14
[particle] then you need to deal with shell quoting 18:15
TimToady doesn't allow file globbing in args very well
[particle] and what if you want to ... right
s1n oh, well just asking
[particle] these are good questions
s1n it's rather unique, so i wanted to know why
TimToady most users will only see the mundane options 18:16
even if they desugar to something with ++foo ... ++/foo
s1n TimToady: isn't that true with perl5 too :)
TimToady in this sense p5 only has mundane options :)
s1n i'm not saying it's bad, just unique, that is most users will not be familiar with command-line syntax that has delimiters like that 18:17
[particle] well, perl 6 is a language spec, not an implementation spec. it needs to deal with multiple implementations.
TimToady I do think I'd capitialize all the magical ++FOO thingies though because it's hard to see them otherwise
[particle] imagine how nice life would be for c programmers if all c compilers took the same options 18:18
TimToady what there's other compilers than gcc?
s1n that's absurd 18:19
There Can Be Only One
TimToady it must be 2008 or something
slavik [particle]: that would make sense, and we can't have none of that
[particle] ok, so it looks like i'm on the right track :) 18:20
slavik :D
s1n [particle]: have to checked to see how well ++ and ++/ play with unix shells?
or other shells for that matter
[particle] TimToady: i can capitalize them, it just looked like shouting to me. would titlecase be loud enough?
s1n IMPORTANT THINGS ARE ALWAYS SHOUTED!
TimToady I think they want to be shouted
slavik is there a C compiler for parrot?
[particle] ok, then. 18:21
slavik: there's a c header parser
slavik doesn't count :P
TimToady otherwise the proper subsystem might not hear :)
[particle] TimToady: i wonder how subsystems will return errors to perl 6, and does that need to be specced?
s1n [particle]: another thing, something like autoloop-split should probably be ++... 18:22
TimToady p6 knows how to deal with failure :)
s1n that is, ++F use Text::CSV ++/F
rather than the {}?
18:23 buu left
[particle] well, i see the compiler (and other) subsystems removing the -- before things like '--CHECK{...}' 18:23
then executing the code contained within 18:24
18:24 icwiener left
s1n why is {} different than option closures though? 18:24
[particle] so the {...} syntax allows arbitrary perl 6 code to run
s1n that's providing 2 different kinds of option delimiters
pmurias what's the runtime equivalent of -O 18:25
?
[particle] pmurias: for rakudo, it's --target=[parse|past|post|pir]
pmurias what i really meant if with -O i can specify what to compile to, how to i specify how to execute the code when not compiling 18:26
[particle] i forget what pugs did
that will depend on your runtime
pmurias pugs has -B and -C like mildew and kp6 had? 18:27
[particle] i don't know if perl will take for example foo.past as input and compile it the rest of the way
TimToady --CHECK seems unnecessary when you can just say -e 'CHECK {...}'
18:27 ovid joined
TimToady bbl & 18:28
[particle] i'm not certain i can just say 'CHECK{...}' as it may be interpreted as a filename
18:29 Psyche^ joined
ovid If I submit a bug report for Rakudo, I would like to include patch with a test, but I don't want a failing test suite. Are there guidelines on submitting TODO, or "must not run" tests? How do they get converted to passing tests later? 18:29
[particle] ovid: can you make it into a spectest, and fudge it for rakudo? 18:30
ruoso ovid, you can add a fudged test
ovid Spectest? Does that mean commit it to the old pugs repository? I don't even know my user/pass for that any more :)
[particle] #?rakudo todo 'RT #nnnnn' 18:31
ovid, i can give you a new commit bit
ovid @particle: OK.
lambdabot Unknown command, try @list
[particle] ovid your current email is: publiustemp-perl6pugs@yahoo 18:32
ovid Ooh, let me check to see if that's still active. brb
Looks like that's long gone. I'll recreate it. 18:33
[particle] you'd still have to remember the password :)
apparently, you're an administrator, and i can't delete administrator commit bit accounts 18:34
oh, i see, i can only delete pending bits anyway
ovid What's the checkout URL? I'll see if I can remember it.
[particle] i can resend the email
svn.pugscode.org/pugs
ovid Email address added. 18:35
[particle] ovid: message sent 18:36
ovid Thanks
rindolf Hi ovid 18:37
ovid: Happy New Year.
ovid HI Shlomi. Thanks.
Er, rindolf. Sorry. :)
Hope you (and everyone else) has a happy new year, too.
rindolf ovid: what's up?
ovid Just trying to figure out how to submit test patches for bugs. 18:38
rindolf ovid: which bugs?
ovid Any I find. There are several in RT.
rindolf ovid: ah. 18:39
[particle] ovid: it's a little annoying, since you need to deal with multiple repositories, but that's the way of the future, i hear.
ovid It would be a little less annoying if it were at least git instead of Subversion :)
rindolf ovid: have you released that module you blogged about a lot on use.perl.org?
[particle] i hate git on windows.
ovid Which module?
I've blogged about many. 18:40
rindolf ovid: don't remember.
ovid: but you've blogged about it a lot.
And then stopped.
Let me find it on your journal.
It had a Japanese-sounding name.
ovid Bermuda? (Not Japanese sounding, but that might be what you mean)
s1n [particle]: you use windows? 18:41
18:41 Patterner left, Psyche^ is now known as Patterner
rindolf ovid: maybe. Let me see. 18:41
[particle] ovid: nice work on the hangman program. i noticed at least one magic number there though
18:41 araujo joined
[particle] s1n: yes, i've been developing on windows for many, many years 18:41
s1n [particle]: i'm sorry :) 18:42
ovid Number of tries on finding a valid word or appropriate word length? It's a bit hackish and embarrassing, but I struggled to make it work with what I knew and what Rakudo offers. Fun, though.
rindolf ovid: yes - use.perl.org/~Ovid/journal/35655 18:43
[particle] 6..* is what i remember
ovid And all of those "self!private_method" calls are just ugly!
[rindolf]: never released it. It's nowhere close to good to go. I could toss it on github at some point, but it's abandonware right now. :/
rindolf ovid: ah. 18:44
18:44 apeiron joined 18:46 rindolf left
pugs_svn r24707 | particle++ | [S19] fix typo, s1n++ 18:50
r24708 | particle++ | [S19] metasyntactic options should shout, to stand out. TimToady++ 18:54
18:54 rindolf joined 19:00 abebae joined
ovid [particle] Are you sure that email went out? I just sent a test one from gmail and received it a few seconds later. [email@hidden.address] 19:00
NM. Spam folder :)
pugs_svn r24709 | Ovid++ | Add test to check autovivification of accessing a top-level hash element. [perl #61882] 19:03
19:10 pdcawley left
masak azawawi: feather.perl6.nl/~azawawi/hangman/h...ippet.html is very beautiful. 19:13
s1n sigh, man getting 5.10 into gentoo has been a real struggle 19:14
[particle] s1n: it's easy on windows ;)
s1n [particle]: they decided making separate ebuilds for each module would be a good idea. it wasn't (imho) 19:15
ovid Hmm, took me a moment to realize why this wasn't working: ack -i 'pigs skip' t/spec 19:16
[particle] azawawi: seems coloring is off for "You guessed '$letter'"; in hangman, due to embedded quoting
s1n ovid: they skip now? i thought it was cool that they could fly 19:17
ovid Well, you know, genetic engineering and all.
TimToady [particle]: you missed the -e
masak [particle]: what's off about it?
ovid So with 'fudge', you can't skip a test in the middle of a test block? It must be the whole block or nothing? 19:18
[particle] method guess_letter ($letter) has blue '$', red 'letter'
'$letter' has purple "'" and blue '$letter'
TimToady [particle]: as in -e 'CHECK...'
[particle] TimToady: ah, gotcha 19:19
masak ovid: someone might have said this already, but parens isn't required around the list in an array assignment.
ovid: nice script, btw.
TimToady only thing is, -e would throw off the line count
ovid masak: post the 'parens' comment to use.perl or the mailing list. Others should see that so they can get the new idioms drilled in.
masak ovid: gotcha. 19:20
[particle] actually, method guess_letter has blue '$l' and red 'etter'
ovid [masak]: thanks :)
[particle] ovid, masak: thoughts about a script repo with liberal commit policy? 19:21
ovid Tempting, but you might get a lot of junk in there (might be better than a lot of nothing, though) 19:22
[particle] like, say, pugs? examples/ maybe?
masak [particle]: is eric2561's Perl 6 Examples repo liberal enough for you?
[particle] never heard of it
masak [particle]: github.com/eric256/perl6-examples/
[particle] please don't say it's git.
eh, well.
s1n hah
masak it's git.
ovid git++
s1n [particle]: come to the dark side :) 19:23
we have cookies
ovid And git.
s1n yes, and git :)
masak (cookies and git)++
19:23 apeiron_ joined
[particle] git and tonic, maybe 19:23
masak :) 19:24
s1n [particle]: at least install a vm already
TimToady all the self!foo should be replacable with $!foo
19:24 apeiron left, apeiron_ is now known as apeiron, justatheory left
TimToady and as pointed out earlier, 'is rw' is useless on private vars 19:25
[particle] s1n: i have 14 vms
s1n [particle]: i assume there's one with git
19:26 GitHub148 joined
GitHub148 perl6-examples: 3David Romano 07master0 SHA1-f8b6172 19:26
Added wsg (Winter Scripting Games) directory and an example
is.gd/eiYh
19:26 GitHub148 left
pugs_svn r24710 | Ovid++ | Trying to make this suitable for "make spectest" with Rakudo. 19:26
r24710 | Ovid++ |
r24710 | Ovid++ | The skip tests work, but the first TODO test doesn't TODO.
r24710 | Ovid++ | Also, tests die part way through, but I can't figure out how to make part of a
r24710 | Ovid++ | test skip instead of all of it.
ovid If anyone wants to add that test to ls t/spectest.data in languages/perl6 (just locally, not as a commit), I'd be grateful if you could explain the issues I raised in that commit. Sorry to be a pain :/ 19:27
[particle] i wonder if pmichaud++'s rval branch will fix these hash autovivification errors
masak TimToady: by "useless", you mean "the default, so you're wasting keystrokes", right? 19:28
ovid [TimToady] Replace self!foo with $!foo? What if I have "has $!foo" and "my method foo"? That's a bug I had to work around.
[particle] ovid: did you know you can create t/localtest.data and run 'make localtest' to test only those files you want?
masak ovid: first time I've seen the =open($file) idion. neat! 19:29
TimToady ovid: Doctor, it hurts when I do this...
ovid [particle]: nope. I wrote vim bindings that let me run perl6 tests from my editor. That's just the way I'm used to doing things. Maybe not the best.
19:30 GitHub2 joined
GitHub2 perl6-examples: 3David Romano 07master0 SHA1-0de80d0 19:30
[wsg] Added README
is.gd/eiYY
19:30 GitHub2 left
[particle] turns off enter/leave messages 19:31
ovid [TimToady]: are you saying that people shouldn't have methods with the same names as variables? I routinely have 'sub foo { shift->{foo} }'. It's a common idiom. Or have I misunderstood?
[particle] then have the accessors auto-generated
TimToady if it's a public method, you should just use $.foo
if it's private and you know it's a no-op, why do you have it? 19:32
ovid [particle]: but that doesn't work if I need to do some any work before returning the variable (such as precalculate it on demand).
Or fetch it from a Web service, or check the user's permissions, etc. 19:33
[TimToady]: I had get_word() because it could be read publicly, but is set privately. I'll have to read through the docs some more.
TimToady then give the backing store a different name, since it obviously doesn't mean the same thing :) 19:34
ovid scratches his head :)
pugs_svn r24711 | particle++ | [S06] update to shortname syntax in command-line examples 19:37
Matt-W Ow my head 19:38
I clearly don't understand rules as much as I thought I did
pugs_svn r24712 | particle++ | [S19] update C<--CHECK{...}> syntax with more standard C<-e 'CHECK{...}'> 19:39
Matt-W Oh
masak Matt-W: I know the feeling. 19:40
Matt-W There seems to be some automatic chomping going on
ovid [TimToady]: so if naming a private method the same name as a variable is discouraged and a no-op, should it be illegal? It was a hard bug for me to track down.
masak Matt-W: that's likely not the rules, but the readline.
Matt-W masak: yes, is it supposed to chomp by default? I didn't tell it to 19:41
masak Matt-W: =$*IN chomps by default, yes.
Matt-W Right
Okay, so that's all behaving properly then
nnunley How lazy is START? Could ovid use that for his lazy initialization pattern? (ie has $foo = START { ... } )
Matt-W Now I have some tokens that match entire lines in my input. Can I write a rule or token or something which can match a specified sequence of them in a multiline input string? 19:42
19:43 apeiron left
masak Matt-W: exemplify. 19:43
[particle] rule some_lines { <one_line> <another_line> <a_different_line> } 19:44
masak right. where each rule for a line looks like { ^^ ... $$ }
s1n [particle]: what does CHECK do?
Matt-W gargh 19:45
curse my Perl 5 brain
19:45 apeiron joined
Matt-W oh also, how do I say 'a line that doesn't begin with a .'? 19:46
masak s1n: (from S04) 'at compile time, ALAP, only ever runs once'
TimToady ^^ <!before '.'>
Matt-W TimToady: thanks 19:47
masak smiles at the ^^
TimToady smiles at $$ :) 19:48
masak 哈哈
TimToady funny that that also means pugs 19:49
Matt-W It's like I can see that underneath it all makes sense... I'm just not tuned in yet
masak is there someone here who runs a bot on feather? I'd like to ask someone who's set one up a few questions. 19:50
Matt-W oh dear now none of the tokens match
masak Matt-W: nopaste?
TimToady: you lost me. what also means pugs?
TimToady 哈 19:51
masak oh! I didn't know that.
TimToady 哈 哈 [ha1] /laughter/yawn/
哈 哈 [ha3] /a Pekinese/a pug/
masak how exceedingly appropriate.
TimToady "the mouth is fitting" :) 19:52
masak aye.
TimToady: I'm sometimes toying with the idea of creating a database of hanzi, with focus on character components. ISTR you have already made something like this. 19:54
ovid Are there any docs on how to "fudge" tests. I appear to be failing at it, badly.
pasteling "Matt-W" at 81.101.136.8 pasted "Silly problem with rules" (48 lines, 791B) at sial.org/pbot/34147
TimToady masak: 54c8 09哈mouth.l fit.r 19:55
masak TimToady: nice.
what's the 09?
I think I got the rest.
TimToady 9 strokes
masak aah.
TimToady can search on that too
masak convenient. 19:56
[particle] for golfing? ;)
TimToady you'd have to ask putter about that...
masak groans
19:56 clintongormley left
[particle] #perl6 rejects 48-hour cease-pun request 19:58
Matt-W It wouldn't be IRC without puns
masak what's the latest word on grammar matching syntax? is the syntax at the bottom of sial.org/pbot/34147 ok?
TimToady it's just a method call
19:58 clintongormley joined
masak so it should be StoryGrammar.story instead? 19:59
TimToady and Grammar.parse($string) should work
masak is that specced?
[particle] ovid: i'm looking at autovivification.t now, but will be distracted by a phone call
ovid No worries. I'm not as concerned about that feature, per se, as I am about being able to safely add tests for Perl 6. 20:00
TimToady or StoryGrammar.new($string).TOP
[particle] yes, trying to help you Get It Right.
masak TimToady: I gotta write this down somewhere.
pugs_svn r24713 | Ovid++ | Trying to make this fudgeable test work. Need to understand fudge better.
r24713 | Ovid++ | Mmm ... fudge.
20:01 Whiteknight joined
ovid OK, I think my housemate and I need to run off and look at pretty fireworks soon. Happy New Year everyone! 20:01
masak ovid: and to you, sir.
20:02 ovid left
Matt-W .parse($string) on a grammar doesn't seem to be implemented in rakudo atm 20:02
azawawi masak: thx 20:03
masak Matt-W: would you like to submit a TODO rakudobug about it?
Matt-W masak: I might do that
masak Matt-W++
azawawi masak: Padre + Padre::Plugin::Perl6 + p5 PPI rocks ;-) 20:04
20:11 c9s_ joined
TimToady also .parsefile($filename) 20:11
mberends Matt-W, could your =$ifh { ... } loop simplify to: $contents=slurp($input_file); ?
Matt-W umm 20:12
perhaps
ruoso I think I'll declare EXPORT::ALL and EXPORT::DEFAULT on every package from the start, it should be easier than checking if that package is there...
Matt-W yes, yes it can
mberends: thanks
mberends :) 20:13
Matt-W &
ruoso TimToady, I've been naming packages with the "::" in the end, like Foo::Bar::baz means looking for "Foo::" in the current scope, than looking for Bar:: inside that package, and for &baz inside the last 20:14
20:14 c9s left 20:16 eric2561 left
mberends $contents = slurp($input_file).chomp; #may be even better to suppress a superfluous final \n 20:16
ruoso that means the type and the package are stored with different names
pugs_svn r24714 | particle++ | [S19] s/bundl/cluster/g; update date/version 20:17
20:18 clintongormley left
literal can I do something like "use lib '.'" in Perl 6? 20:21
20:22 eric2561 joined 20:27 eric2562 joined
mberends probably not yet, I looked in vain a few weeks ago 20:34
PERL6LIB seems to be the only facility so far 20:35
20:44 iblechbot left 20:45 eric2561 left 20:55 Limbic_Region joined 20:57 vixey left 20:59 slavik left
masak literal: it's not implemented in Rakudo yet. 21:04
mberends masak, apart from run(), do you know of a qx{} or `backtick` facility in Rakudo? 21:07
masak mberends: nope. 21:08
mberends :(
masak I don't do shell games a lot.
Matt-W: did you send rakudobug a message about .parse? otherwise I'll do it now. 21:09
21:10 buubot left 21:17 vixey joined
masak mberends: why, what do you need? 21:17
mberends: why not create a sub that mimics the behaviour of infix:<` `>?
21:22 Limbic_Region left
rindolf Did someone implement line numbers for run-time errors in Rakudo yet? 21:24
[particle] no 21:27
21:36 alester left
rindolf [particle]: I seeee. 21:38
21:41 dukeleto joined, Whiteknight left
mberends masak, I did hack something together with run() and output redirection to a temp 21:46
file, but it's ugly. The application is a web server, currently using netcat for TCP. 21:48
21:48 kisu left, eric2562 left
pugs_svn r24715 | ruoso++ | [mildew] "is export" implemented, "is export(:DEFAULT)" still not implemented 21:49
ruoso the (:DEFAULT) part is parsed in a more complicated way than I expected
[particle] ruoso: what pugs dir do i need to check out to see all the mildew code? 21:51
21:52 cspencer joined 21:54 kisu joined
literal are block/loop labels unimplemented in Rakudo? 21:56
[particle] like loop: for @x -> $y { next loop unless $y > 5 } 21:57
?
literal yeah
[particle] i don't think so
rakudo: my @x = 1,5,10; loop: for @x -> $y { next loop unless $y > 5 }; 21:58
p6eval rakudo 34724: OUTPUTĀ«Statement not terminated properly at line 1, near ": for @x -"ā¤ā¤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)ā¤Ā»
literal that's the same thing I got
statement not terminated properly
[particle] yep, it's not parsing the label
mberends nopaste? 21:59
[particle] pasteling? 22:00
use rafb.net iirc
azawawi 2009 here finally ;-) 22:02
azawawi prepares world domination plans for 2009 ;-)
[particle] can't wait till christmas. 22:03
mberends masak, crude Rakudo qx{} implementation: rafb.net/p/em5i8592.html
22:04 rindolf left 22:07 Exodist left, Exodist joined 22:08 cspencer left 22:14 eric2561 joined
eric2561 hates his ISP 22:14
arg how did i get a 1 on my name! ;)
22:15 DemoFreak joined
mberends eric256(1), an IRC client often renames you if it thinks your nick is already online, to preserve uniqueness 22:16
22:25 maerzhase left
eric2561 evil ;) 22:27
22:34 Ehtyar joined 22:37 hanekomu_ joined
azawawi @tell moritz_ Moose 0.64 is released. Can you install it on feather1 please? 22:38
lambdabot Consider it noted.
azawawi Not bad, Moose 0.64 on the last day of 2008 22:39
# this fixes win32 installation problem for Class::MOP
22:41 hanekomu_ left
masak mberends: nice. 22:45
mberends: but I agree getting rid of the tempfile would be an improvement. 22:46
azawawi masak: hi
masak azawawi: greetings, earthling. 22:47
azawawi masak: 2009!
azawawi 2009++
masak azawawi: we're still waiting for it here :)
ten minutes left. 22:48
azawawi it is the same ;-)
we had it 48 mins ago
masak rakudo: 2009++
p6eval rakudo 34725: OUTPUTĀ«Unable to set lvalue on PAST::Val nodeā¤current instr.: 'parrot;PAST;Val;lvalue' pc 556 (src/PAST/Node.pir:161)ā¤Ā»
masak azawawi: I think you have to put 2009 in a scalar, and then ++ the scalar :P 22:49
azawawi lol
rakudo: year2009++; 22:50
p6eval rakudo 34725: OUTPUTĀ«Could not find non-existent sub year2009ā¤current instr.: '_block14' pc 48 (EVAL_12:36)ā¤Ā»
azawawi rakudo: $year2009++;
p6eval rakudo 34725: OUTPUTĀ«Scope not found for PAST::Var '$year2009'ā¤current instr.: 'parrot;PCT;HLLCompiler;panic' pc 146 (src/PCT/HLLCompiler.pir:102)ā¤Ā»
azawawi rakudo: my $year = 2008; $year++;
p6eval rakudo 34725: RESULTĀ«2008Ā»
azawawi rakudo: my $year = 2008; ++$year;
p6eval rakudo 34725: RESULTĀ«2009Ā»
azawawi reads about upcoming netbooks chipsets... 22:51
22:54 eric2562 joined
masak akh. if I had thought of it earlier, I would have made a Rakudo script that animated fireworks... 22:56
Matt-W masak: I can hear lots going off here, i don't think you need to simulate any more 23:05
eric2562 masak: you still can i've got 8 more hours to go till 2009 ;) 23:09
azawawi # 2008 is yesterday's news here ;-) 23:10
23:12 eric2561 left
masak eric2562: IƤll keep that in mind :) 23:13
23:16 Exodist left
literal eric2562: what do I do if I want to push to perl6-examples? 23:18
eric2562 hey literal
did you fork it on git? or just download the public clone?
23:18 lichtkind left 23:20 smtms joined, jhorwitz left 23:21 justatheory joined
eric2562 btw i've got a perl script to autoamticaly add colatborators..just need somewhere to host it...might host it on feather, but i'm not sure how to do that without making my password just a weee bit too public ;) 23:21
but i can add you as a collabotor and then you can push directly to the repo, or you can fork it and submit a pull request
Juerd shutdown -r now "HAPPY NEW YEAR" ;) 23:22
23:23 smtms left
vixey man 23:24
23:28 justatheory left
literal eric2562: haven't done either 23:28
eric2562 literal: then the choice is up to you, if you wanna register i'd be happy to add you as a collaborator (i.e. get you a commit bit) 23:29
23:31 eric2561 joined 23:38 armagad joined
TimToady ruoso: storing packages under "Foo::" key seems okay to me, though presumably one could also store it magically within the Foo protoobject somehow 23:41
ruoso TimToady, well, it is... Foo.WHO 23:45
TimToady presumably the long name has all the version numbers too... 23:46
ruoso [particle], pugs/v6/mildew
TimToady either that, or Foo::Bar::baz can mean different things in different scopes
ruoso that's how it's implemented in mildew now
the package name is always a local alias 23:47
TimToady sounds good
23:49 eric2562 left
TimToady for some reason it's really easy for me to fall into P5-Think now and then... :) 23:49
ruoso ;)
the parsing of "is export(:DEFAULT)" surprised me 23:50
TimToady equiv to :export(:DEFAULT) really
ruoso the part between the parenthesis is a statement
TimToady indeed, and :DEFAULT is just a term
ruoso can it be simplified? 23:51
or at least can I treat it as if it could be simplified?
TimToady do we want to disallow :export($?FOO) is the question...
is export<foo bar> is possible
ruoso are those equivalent: "is export(:DEFAULT)" and "is export<DEFAULT>" 23:52
TimToady some value in preserving the pair "look" for consistency though
no, not semantically 23:53
would be more like is export('DEFAULT')
ruoso I see... but...
what would "is export('DEFAULT')" mean then 23:54
?
TimToady presumably the trait still has to poke an alias into ::EXPORT::DEFAULT 23:55
but traits are active, not passive, so doesn't seem like a big problem
ruoso so, if the later parse easier, can I use it 23:56
?
TimToady I need to glare at S11 a bit and think about it, but maybe
might need to reserve bare names for something else like aliases though 23:57
ruoso btw... this export tagsets need some clarification
TimToady doesn't everything?
ruoso heh
"is export<DEFAULT>" has a much cleaner parsing 23:59

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!