»ö« | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, alpha:, pugs:, std:, or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by moritz_ on 25 June 2010.
00:01 Psyche^ joined 00:05 Patterner left, Psyche^ is now known as Patterner
dalek odel: 4c8ee84 | jnthn++ | dotnet/runtime/Metamodel/Representations/P6 (3 files):
Eliminate use of Nullable<T>, which should make life a bit easier for mberends++ when porting.
 00:11
00:13 lichtkind left 00:15 flatwhatson left 00:16 flatwhatson joined
dalek odel: b23a318 | jnthn++ | dotnet/runtime/Metamodel/SharedTable.cs:
Flesh out STable a little more; this adds find_method and postcircumfix:<{ }> caching.
 00:24
00:34 hudnix left 00:37 drbean left 00:45 meppl left 00:46 lest_away is now known as lestrrat
jnthn -> sleep 00:54
00:59 whiteknight left 01:05 jferrero joined
TimToady phenny: tell masak yes, checking scope constraints on FIRST is in my todo file 01:07
phenny TimToady: I'll pass that on when masak is around.
01:17 jferrero left 01:23 Trashlord joined 01:39 cggoebel left 01:47 Italian_Plumber joined 01:49 molaf_ joined 01:52 molaf left
[Coke] git question: How do I kill all the changes I've made locally since my last 'push' ? 02:09
(I'm ahead by N commits, and I don't want any of them.)
02:10 hudnix joined
colomon some form of git reset, I think 02:10
02:10 Italian_Plumber left 02:12 Italian_Plumber joined 02:14 Italian_Plumber left 02:19 gfx joined
diakopter jnthn: interesting: bugsplat.info/2010-08-22-perl-with-a-lisp.html 02:19
02:32 agentzh joined
colomon must remember he has a lot of stuff to do, and not just spend the next week translating that blog post's code to p6.... 02:32
02:33 meppl joined
sorear [Coke]: git reset --hard origin/master 02:36
02:36 cheeps left
[Coke] sorear++ 02:39
02:46 azert0x left 02:47 meppl left 02:53 xinming joined
tylercurtis Is there a named subrule form of \w? 03:27
03:40 [Coke] left, [Coke] joined
ingy tylercurtis: hi 03:49
tylercurtis ingy: Hi. 03:50
ingy :) 03:51
you wanted something?
tylercurtis ingy: Remember how I was supposed to add a LICENSE to yaml-pm6? I finally got around to doing so, but I wanted to know what you want the COPYRIGHT section of the README to say before I commit it. 03:52
ingy tylercurtis: how about something like any of my cpan modules 03:54
03:55 [Coke] left
ingy tylercurtis: gist.github.com/544745 03:55
tylercurtis ingy: well, same terms as Perl itself seems kinda inappropriate for a Perl 6 module, since that could be anything from AL 1.0/GPL if you interpret it as Perl 5, AL 2.0, MIT, or who knows what else. In addition, the "Ingy döt Net." part will become only partially accurate once I start working on it. 03:57
ingy: gist.github.com/544748 here's the standard-ish (in that everything but the Copyright 2010. <SOMETHING> part came from TPF's webpage) notice for the AL 2.0 (IIRC, that's what you said to put as the license, right?). 03:59
ingy tylercurtis: I think you should add your name to authors but the copyright should just be in one name 04:01
04:01 [Coke] joined
tylercurtis ingy: Works for me. 04:01
ingy: Create a seperate AUTHORS or CREDITS file or just add an AUTHORS section to the README? 04:06
tylercurtis has some slight changes to push to the loader branch that don't make it work but at least let it compile.
04:17 sftp left 04:21 redicaps joined 04:22 redicaps left, shelling_lab left, shelling_lab joined 04:39 shelling_lab left 04:49 shelling_lab joined 04:50 molaf_ left 05:07 Eevee left 05:08 gfx left 05:26 drake1 joined
drake1 hello, how do you code something like pastebin.ca/1923137 in perl6 / rakudo ? 05:27
05:29 kaare joined 05:30 kaare is now known as Guest81071
drake1 things like dynamic method construction 05:33
tylercurtis drake1: which bit are you asking about?
drake1 the ``sub'' and the my @bin initialization 05:34
tylercurtis You could do them exactly the same way. Perl 6 does have some nicer ways to do some of it, though. 05:35
drake1 oh, and is it possible to export a perl6 program in VM-code only ie. without the source code itself 05:36
tylercurtis For example, in Perl 6, list assignment is looser than comma, so you don't need the parentheses around the list in the my @bin assignment.
drake1 tylercurtis: good
nice
tylercurtis Well, technically, you would have to use my or our or such to declare the @bin variable, since Perl 6 is strict by default(and for now, at least, is strict always).
title => "Winamp" can be written perhaps a little more nicely as :title<Winamp>. 05:37
drake1 that's what im used to, only forgot to in this test script. first run with Tk;
tylercurtis And so on for the other items in the hashes.
drake1 not .title = "Winamp" 05:38
tylercurtis In Perl 6, "foreach my $bin (@bin) {...}" becomes "for @bin -> $bin {...}".
drake1 fine 05:39
tylercurtis You also don't need the -> in your hash dereferences in Perl 6.
drake1 so foreach is dropped?
TimToady yes, always for now
tylercurtis drake1: for is always foreach. If you want C-style for, you want loop.
drake1 tylercurtis: like with multi-dimensional hashes in perl5 ?
TimToady yes, but you can drop it on the first one now too :) 05:40
drake1 TimToady: OK
TimToady because it's always a ref/deref at that boundary
drake1 naturally
good
tylercurtis $bin->{title} would have to become either $bin<title> or $bin{'title'} since postcircumfix:<{ }> doesn't autoquote anymore. 05:41
TimToady only => autoquotes now
drake1 im also tired of the -> thingy
TimToady and it doesn't do -foo like p5
but => indicates real pairs, so we don't need the - to keep tk in sink
sync
assuming we had Tk bindings... 05:42
tylercurtis -> becomes . for methods, as well.
TimToady not sure what blizkost wants to cross over to p5 there...
drake1 then it's getting closer to ``sh'' perhaps
TimToady if so, it's an accident :) 05:43
tylercurtis drake1: you can now drop the sub from "sub { ... }", also.
drake1 perl is alot more friendly with data structures etc
tylercurtis: to it's just the curlies?
05:43 snearch joined
TimToady rakudo: my $closure = { 'say hi drake1' }; $closure() 05:44
drake1 nice
p6eval rakudo b958a1: ( no output )
TimToady er...only written right
drake1 it's a lot of nice improvements
TimToady rakudo: my $closure = { say 'hi drake1' }; $closure()
p6eval rakudo b958a1: OUTPUT«hi drake1␤»
drake1 you're continuing to support the ``fork'' ? 05:45
TimToady those are just the surface improvements
drake1 sweet
tylercurtis I'm not sure Perl 6 supports "new MainWindow;" syntax anymore. I though I had seen it in the spec, but I can't seem to find it at the moment.
TimToady though the 'for @array -> $elem {...}' syntax actually points to the fundamental notion that all blocks are really just lambdas
drake1 just copied from another website. I never use objects like that, normally 05:46
TimToady no, you'd either have to put a : after MainWindow, or flip it around to MainWindow.new
no more guessing about indirect objects
drake1 fine enough
more like ``ld'' import, then 05:47
TimToady so the colon is now required after such an object
but lets you add more args
tylercurtis What's the name for that syntax?
TimToady we still call it indirect object
it's just not ambiguous anymore
tylercurtis thought it had some fancy linguistics-inspired name that he can never remember. 05:48
tylercurtis appears to be wrong.
TimToady it's really more like the direct object in English 05:49
drake1 with adjective constructors
TimToady rakudo: class Dog { has $.ears; has $.tail }; my $fido = new Dog: tail => 'long', ears => 'short'; say $fido.perl 05:50
p6eval rakudo b958a1: OUTPUT«===SORRY!===␤Confused at line 22, near "my $fido ="␤»
TimToady maybe rakudo doesn't parse that yet 05:51
std: class Dog { has $.ears; has $.tail }; my $fido = new Dog: tail => 'long', ears => 'short'; say $fido.perl
p6eval std 32087: OUTPUT«ok 00:01 119m␤»
TimToady the STD parser doesn't mind it
drake1 cool
cool syntax/
TimToady esp when formatted one initializer per line 05:52
which is hard to show in irc
drake1 a bit more active than bind configuration 05:53
you know with wildcard_name { parameter1=blah\nparameter2 ... 05:55
TimToady tylercurtis: \w === <alnum>
we have to preserve the clocking of the code to better report syntax errors 05:56
so it takes a bit more punctuation to make sure the parser stays in sync, since it's a richer language 05:57
many syntax errors show up by noticing two terms in a row
so that's always an error in Perl 6
std: 42 42
p6eval std 32087: OUTPUT«===SORRY!===␤Two terms in a row at /tmp/LcB0P4wdGb line 1:␤------> 42 ⏏42␤ expecting any of:␤ bracketed infix␤ infix or meta-infix␤ statement modifier loop␤Parse failed␤FAILED 00:01 116m␤»
tylercurtis TimToady: if I modify alnum, will that change the behavior of \w, then? I don't think I'm going to end up doing that, since the grammar in question turned out to be simple enough to not need word boundary checks, but it's nice to know in case I do need to. 05:58
TimToady but it can show things that are more subtle, like a missing semicolon
std: my $a = 42␤my $b = 43; 05:59
p6eval std : ( no output )
TimToady heh
tylercurtis rakudo: my $a = 42␤my $b = 43
p6eval rakudo b958a1: OUTPUT«===SORRY!===␤Confused at line 22, near "my $a = 42"␤»
05:59 cono left
TimToady tylercurtis: uh, I think changing alnum would probably change \w in STD, but not sure about rakudo 06:00
std: my $a = 42␤my $b = 43;
p6eval std : ( no output )
TimToady std: my $a = 42;␤my $b = 43;
p6eval std : ( no output )
TimToady hmm, evalbot hosed?
std: my $a = 42;
p6eval std : ( no output )
drake1 it's really good to see it becomes clear 06:01
TimToady odd, it was just there...
drake1 perl5's a bit messy
TimToady yes, just a bit
drake1 fine enough to the hardcore scripts
TimToady std: 42
p6eval std : ( no output )
06:02 uniejo joined
TimToady it shouldn't be stuck mid compile, because it is supposedly using snapshot semantics 06:02
perhaps the server is overloaded or something 06:03
rakudo: 42
p6eval rakudo b958a1: ( no output )
TimToady rakudo: say 42
p6eval rakudo b958a1: OUTPUT«42␤»
TimToady std: 42 06:04
p6eval std 32087: OUTPUT«ok 00:01 114m␤»
TimToady std: my $a = 42␤my $b = 43;
p6eval std 32087: OUTPUT«===SORRY!===␤Two terms in a row (previous line missing its semicolon?) at /tmp/S0YtT21PCl line 2:␤------> <BOL>⏏my $b = 43;␤ expecting any of:␤ POST␤ bracketed infix␤ infix or meta-infix␤ postfix␤
..postfix_prefix_meta_operator␤ statement modif…
TimToady here it figures out a semicolon seems to be missing, since there's a line break 06:05
tylercurtis As I write tests for a toy LISP interpreter, I find myself thinking about how convenient it is to use functions as list ops, without parenthesizing the arguments. That seems a little ironic. 06:12
drake1 well, printf "%s\n" $F |egrep '\.l|\.y' for now. thanks for the introduction 06:15
bye
06:15 drake1 left 06:21 wtw joined 06:27 moritz__ is now known as moritz_ 06:28 snearch left 06:40 mberends left 06:45 barika joined
moritz_ good morning 06:55
and good evening to sorear :-) 06:56
tylercurtis Good night, #perl6. 06:57
06:58 tylercurtis left 07:35 tadzik joined
tadzik o/ 07:35
07:37 wamba joined 07:39 rgrau joined 07:44 azawawi joined
azawawi hi 07:44
07:45 rokoteko_ joined 07:49 barika left, nadim left, huf left, phenny left, renormalist left, aesop left, rokoteko left
azawawi sorear: ping 07:50
phenny: tell sorear I tried irclog.perlgeek.de/perl6/2010-08-19#i_2722179 and it worked. Thanks. How about documenting that somewhere? 07:51
phenny down? 07:53
07:54 lestrrat is now known as lest_away
sorear no, you're just on the wrong channel 07:55
phenny is in the other #perl6 now
azawawi other #perl6? 07:56
07:56 drbean joined
sorear do you know what a netsplit is 07:57
azawawi nop but i just read about it 07:58
08:00 barika joined, nadim joined, huf joined, phenny joined, renormalist joined, aesop joined, card.freenode.net sets mode: +v phenny
azawawi sorear: on a unsuccessful parse, i get a lot of uninitialized variable errors. It that normal? 08:00
sorear not for me 08:03
it might be for you; disable the warnings if you have to 08:04
azawawi std: say 1; 08:05
p6eval std 32087: OUTPUT«ok 00:01 115m␤»
azawawi std: say;
p6eval std 32087: OUTPUT«Potential difficulties:␤ Unsupported use of bare 'say'; in Perl 6 please use .say if you meant $_, or use an explicit invocant or argument at /tmp/cuPE3P6cui line 1:␤------> say⏏;␤ok 00:01 114m␤»
08:06 patrickas joined 08:07 patrickas left
azawawi sorear: here is what im getting when trying it on Strawberry Perl 5.12.0 08:08
azawawi checks it again on his ubuntu vm 08:09
sorear I am sorry but I cannot help with any Strawberry-specific problem 08:13
it works fine on a 5.12.0 built from the git sources 08:14
azawawi sorear: yeah i know. i couldnt even get blizkost on run on that strawberry...
08:16 fod left 08:18 thebird joined 08:24 tadzik left 08:27 dakkar joined 08:29 daxim joined, drake1 joined
azawawi sorear: ok i tried all the steps again on ubuntu and I still get the same warnings. Any clue to what is happening? Am i compiling CORE.setting in a wrong way or should i get a compatible CORE.setting with the same STD version? 08:29
drake1 another question. something like pastebin.ca/1923194 with pipe-line shell setup, how's that in perl6 ? 08:30
azawawi sorear: this is what i use to compile the setting if it is not there pastebin.ca/1923196
sorear azawawi: you can't use STD->parse* twice in one process without a huge amount of locals, and even then it's not reliable 08:33
see sys_compile_module for the current list
08:33 zulon joined
azawawi sorear: i know but it is executed once after syml/CORE.syml is created 08:34
moritz_ drake1: a lot of IO stuff isn't specified yet. So I'm not sure how it will end up looking
azawawi sorear: second runs yield the same warnings. Am i missing something here?
drake1 moritz_: ok
sorear azawawi: are you changing directories at all anywhere? 08:35
azawawi sorear: nop, just that simple example 08:36
sorear: and i made sure that S:H:P6 STD stuff is deleted before installing STD
sorear well, there's something wrong, probably related to STD's lack of reusability 08:38
azawawi tries older STD versions 08:39
sorear: maybe the change to STD:: namespace made these errors 08:41
drake1 use Tk; won't allow perl to fork so had to use stdio for extern solutions
azawawi sorear: older STD behaves the same
drake1: why dont you try using blizkost? Write p5 in perl6 08:42
drake1 blizkost?
azawawi drake1: github.com/jnthn/blizkost
drake1: embeds perl5 in Rakudo perl 6
sorear azawawi: you're on your own 08:43
drake1 azawawi: nice. ill try it
azawawi sorear: Thanks. I thought so :) 08:44
sorear: im rolling back S:H:P6 to the older STD copying script.
08:44 shade\ left
drake1 no perl6 module to associate a file descriptor with an io buffer? 08:45
08:46 azawawi left 08:47 shade\ joined, shade\ left 08:54 lest_away is now known as lestrrat, lestrrat is now known as lest_away 08:55 lest_away is now known as lestrrat 08:58 zulon left 08:59 drake1 left
jnthn morning 09:05
dalek ecza: 36e966d | sorear++ | src/ (3 files):
*very* hacky first take on $/ and $¢

They're only available inside regex embedded blocks.
ecza: 862b2be | sorear++ | / (2 files):
Parse $/ and $¢
ecza: eb16e02 | sorear++ | src/Niecza/Actions.pm:
Implement parsing of $<foo> form
moritz_ next step: $0
niecza: 'abc' ~~ / <alpha> { say $<alpha> } / 09:09
p6eval niecza eb16e02: OUTPUT«Unhandled exception: Unable to resolve method alpha in class Cursor␤ at /tmp/VHk_61qKRO line 0␤ at lib/SAFE.setting line 976␤ at /tmp/VHk_61qKRO line 0␤ at /tmp/VHk_61qKRO line 0␤ at lib/SAFE.setting line 988␤ at lib/SAFE.setting line 346␤ at /tmp/VHk_61qKRO line 1␤
..at lib/…
moritz_ niecza: 'abc' ~~ / $<alpha>=(.) { say $<alpha> } / 09:10
p6eval niecza eb16e02: OUTPUT«===SORRY!===␤Can't locate object method "new" via package "RxOp::Any" (perhaps you forgot to load "RxOp::Any"?) at /home/p6eval/niecza/src/Niecza/Actions.pm line 546.␤Attribute (zyg) does not pass the type constraint because: Validation failed for
..'ArrayRef[RxOp]'…
moritz_ sorear: what built-in named rules are available?
sorear ws. 09:12
also, while $<foo> parsing may be implemented... 09:13
niecza: ' ' ~~ / <ws> { say $<ws> } /
p6eval niecza eb16e02: OUTPUT«Unhandled exception: Unable to resolve method at-key in class Cursor␤ at /tmp/da70PO5Syg line 1␤ at /tmp/da70PO5Syg line 1␤ at lib/SAFE.setting line 975␤ at /tmp/da70PO5Syg line 0␤ at lib/SAFE.setting line 933␤ at /tmp/da70PO5Syg line 0␤ at lib/SAFE.setting line 976␤
..at /tm…
moritz_ niecza: ' ' ~~ / <ws> { say ':' ~ $/{'ws'} ~ ':' } / 09:14
p6eval niecza eb16e02: OUTPUT«Unhandled exception: Unable to resolve method at-key in class Cursor␤ at lib/SAFE.setting line 836␤ at /tmp/U4fSrSdzxx line 1␤ at /tmp/U4fSrSdzxx line 1␤ at lib/SAFE.setting line 975␤ at /tmp/U4fSrSdzxx line 0␤ at lib/SAFE.setting line 933␤ at /tmp/U4fSrSdzxx line 0␤
..at lib…
09:20 shade\ joined, shade\ left, Mowah joined 09:25 drbean left, mulander left 09:27 shade\ joined, shade\ left
sorear is now trying to get niecza's parse->past-ish stage to accept STD 09:28
I don't think jnthn is going to try to compete with me in parsing STD.pm6 in the next two weeks 09:29
moritz_ I don't expect that either
he's working on much more low-level stuff
sorear moritz_: one flaw in niecza which affects the evalbot that I'm not sure how to fix 09:31
moritz_: niecza keeps a cache of precompiled modules
moritz_: sometimes I like to make binary-incompatible changes to the code generation, but I don't want to invalidate the cache on any compiler tweak 09:32
is there a more elegant way out than just a Parrot-style version number?
09:33 shade\ joined
sorear looks like the next big error generator is adverbs. hmm. 09:33
rakudo: my %foo; say %foo<x> :exists;
moritz_ sorear: have a dummy revision file, and make all modules depend on it
p6eval rakudo b958a1: OUTPUT«===SORRY!===␤Confused at line 22, near "say %foo<x"␤» 09:34
moritz_ sorear: and change that revision file whenever you want them to recompile
sorear would that work across git?
moritz_ sure; if you commit these changes to the revision file, its timestamp will be updated too
09:35 ruoso left
jnthn sorear: No, I generally stay well away from the grammar engine. :-) 09:40
moritz_ sorear: I'm curious about your mid to long-term goals for niecza. Will you try to attract users and contributors? what will you do after you have STD bootstrapped? 09:41
09:42 ppant joined
moritz_ github.com/moritz/perlgeek.de/blob/...ro-sum.txt # anybody want to comment on my next blog post before it goes live? 09:44
sorear moritz_: after STD is bootstrapped, I'm going to try the metacircular aspects of Perl 6 - BEGIN and eval 09:47
I'm going to try to set up a real REPL 09:48
there may be an attempt at macros, or maybe not; I would need to figure out how to resolve meta type issues first 09:49
TimToady has said that the compiler and the compilee have logically distinct interpreters, which raises questions on how they can share AST nodes and strings
github-- # presenting a .*\.txt URL as text/html 09:50
moritz_: 1, how does worldwide population growth figure into this 2. is it just me, or has the US gov't become a disproportionate exporter of demographics? 09:53
moritz_ sorear: if you have better links, please share them :-)
mathw o/
jnthn o/ mathw 09:54
moritz_ I'll write a bit about population growth 09:55
09:59 tadzik joined, Mowah left 10:00 Mowah joined, Mowah left 10:03 Mowah joined 10:04 Gothmog_ joined 10:07 rokoteko_ is now known as rokoteko, rokoteko left, rokoteko joined 10:08 ggoebel joined
sorear ggoebel: are you reachable? 10:09
10:09 agentzh left
moritz_ perlgeek.de/blog-en/perl-6/not-zero-sum.html loliblogged 10:10
sorear perl 6 hackers have their blogs on paper sticks? 10:11
moritz_ paper stick?
jnthn lol
moritz_ doesn't get it
jnthn moritz_: Think it's a place on "lolly"
*play
szbalint moritz_++ 10:12
moritz_ oh my
sorear they're called lolipops in the states
10:12 Mowah left
jnthn Yes, in the UK too :-) 10:12
Or "lolly" as slang. :-)
moritz_ we have "lolly" too
jnthn Sweet. :-)
10:12 Mowah joined
moritz_ but I didn't make the jump to paper sticks 10:13
szbalint The connection was a bit *thin*.
moritz_ that's why I didn't *thin*k of it 10:14
10:14 Mowah left 10:15 Mowah joined
moritz_ I've also hackernews'ed it - let's see what comes out of it :-) news.ycombinator.com/item?id=1626335 10:17
10:19 Mowah left, Mowah joined 10:23 ppant left 10:24 Mowah left 10:25 Mowah joined, Mowah left 10:26 Mowah joined 10:27 Mowah left, pnu left, Mowah joined, Mowah left 10:29 lestrrat is now known as lest_away
dalek ecza: f61a6db | sorear++ | / (3 files):
Implement operator adverb processing
 10:32
10:32 pnu joined, Mowah joined 10:34 azert0x joined, azert0x left 10:35 Mowah left, Mowah joined
sorear Mowah: What's the problem? 10:36
10:36 azert0x joined
moritz_ ignores joins and leaves 10:37
sorear what about quits?
moritz_ too
12:36 3 #perl6: JOINS PARTS QUITS NICKS
azert0x hello
sorear down to 205 sorry()s in STD parse
moritz_ from my ignorance list
hi azert0x
tadzik I miss an irc client which will keep those events in some separate windows, visible, but not polluting the tal
the fact is these are almost useless on irc, where everyone idles like hell anyway 10:38
hi azert0x
moritz_ tadzik: I ignore them in irssi, and if they interest me, I look into the web based IRC logs
sorear tadzik: that's what I was planning when I got sidetracked.
moritz_ though they aren't quite complete
tadzik sorear: sidetracked? By Perl 6?
there's an old irssi theme having that, but it segfaults 10:39
sorear by *implementing* perl6 :|
tadzik :)
sorear I pretty thoroughly hate irssi
tadzik on weechat
but not a love-hate relationships, these are just tools
sorear I was writing a replacement in perl6 when I realized a couple features I needed were missing
who says those are mutually exclusive? 10:40
tadzik hm?
isn't there some nice irc client in Perl which we can extend all way long? 10:41
sorear next... metaoperators 10:47
this will be a little bit interesting. 10:48
(I'm mainly concerned with infix_postfix_meta_operator:sym<=>)
sorear wonders if the user should be able to define metaops with names like that
dalek ecza: 5b403a4 | sorear++ | src/Niecza/Actions.pm:
Implement \$\d+ syntax
 10:50
sorear out 10:51
10:58 meppl joined 11:07 scottp joined, scottp left 11:26 Mowah left 11:34 envi^home joined 11:36 envi^home left 11:40 envi^home joined 11:45 M_o_C joined 11:47 araujo left
takadonet morning all 12:02
12:07 [particle] left 12:08 [particle] joined 12:10 zulon joined 12:15 ruoso joined 12:16 zulon left, M_o_C left 12:17 zulon joined 12:22 sftp joined 12:23 araujo joined 12:30 zulon left
[Coke] RT: 673 active tickets; 152 new; 12 need tests. 12:40
moritz_ :/ 12:41
12:42 rgrau_ joined
TiMBuS mo' users, mo' tickets 12:42
[Coke] moritz_: by which I mean "are assigned to moritz". =-) 12:45
12:47 drbean joined 12:52 Italian_Plumber joined 12:56 timbunce joined 12:58 karb joined 13:04 karb left 13:07 drake1 joined
drake1 it's my perl day today heh 13:07
just wrote this pastebin.ca/1923290 13:08
and started to think that maybe perl6 is more nice to call functions from returned object pointers
moritz_ what are these pointers you're talking about? 13:09
drake1 object methods
sry
the ->pack on like 13 eg 13:10
s,like,line
jnthn Well, you save a character :-)
moritz_ .pack instead of ->pack 13:11
jnthn Other things would likely get neater too :-)
moritz_ like, no more - in front of named options
drake1 with dots instead of arrows
jnthn "sub { my $widget = shift;" could become maybe -> $widget { ... }
tadzik looks like the Canvas methods is using some funny named parameters paring
drake1 the dot colors of that Canvas grid drove me close to insane on MSN and cpan to find the property
13:12 foodoo joined, orafu left
drake1 not => { ... } instead of ``sub''? 13:13
13:14 orafu joined
drake1 instead of => sub { ... } sry 13:15
13:15 zulon joined
jnthn $mw.bind('<ButtonPress-1>' => -> $widget { ... }) # pointy block way 13:15
$mw.bind('<ButtonPress-1>' => sub ($widget) { ... }) # anonymous sub way
drake1 and `pairs' by the arrow instead
jnthn The => -> looks kinda funny though :-)
drake1 yes 13:16
13:16 Guest81071 left
drake1 but => is still to associate a pair? 13:17
moritz_ not "still"
perl 5 has no idea of pairs
drake1 seems like -> int argc, int argv alike prototyping 13:18
moritz_ in perl 5, => is just like a comma, except that it quotes the LHS
drake1 yes I know with first one being the even positions
moritz_ buubot: eval: [my $a = 5 => 'b' => a => 4]
buubot moritz_: [5,"b","a",4]
moritz_ whereas in Perl 6, the => actually constructs a Pair object 13:19
PerlJam moritz_: well, it quotes barewords on the LHS, not just anything :)
moritz_ bareword identifiers, right
drake1 does anyone know how to get a direct pixmap to X instead of all those objects? 13:23
moritz_ aren't there Tk widges for displaying pixel graphics?
drake1 I don't know
moritz_ I'd be surprised if not 13:24
drake1 something like an X pixmap string
whatever
it's fine enough with perl objects
moritz_ search.cpan.org/~srezic/Tk-804.029/...Pixmap.pod 13:25
drake1 XPM then
13:26 bbkr joined
drake1 then $mw->bind would then be $mw.bind or just mw.bind ? 13:29
in perl6
tadzik $mw.bind, variable is a variable 13:30
drake1 and it's still called a reference variable?
tadzik well, it's just a scalar 13:31
drake1 no special type
moritz_ since (nearly) everything in Perl 6 is a reference, it doesn't make sense to call something explcilitly a reference 13:32
drake1 nice
but perl6 know we're asking for an object method by the dot? 13:33
knows*
moritz_ with the dot you *always* ask for an object method
drake1 alright
jnthn ...what other sort of method is there in Perl 6? :-)
moritz_ well, .() is probably not an object method 13:34
jnthn oh, I see what you mean.
drake1 jnthn: only playing smart with OO terms
moritz_ other languages distinguish "object methods" and "class methods" 13:35
jnthn Yeah
moritz_ in Perl 6, a "class method" just means calling a method on the type object
drake1 class methods are still package::method ?
tadzik package.method
moritz_ TypeObject.method()
like
drake1 nice 13:36
moritz_ Gymnast.new().walk-to-bar()
jnthn ;-)
moritz_ there .new() is a method on the type object
drake1 and perl6 also has arbitrary constructors with blessings ?
moritz_ yes
drake1 cool
moritz_ though you get one free for default 13:37
13:37 Helios- left 13:38 Helios- joined
drake1 then perl is close to perfect 13:38
are there any pattern modifiers for locale compatible BREs and EREs? 13:45
moritz_ what are BREs and EREs? 13:46
PerlJam basic regex and extended regex
they are POSIXy I think. 13:47
drake1 \(basic\|extended\) regular expression
sometimes perl is very rich in meta characters
moritz_ I don't know; Perl 6 doesn't use POSIX regexes :-)
drake1 read a bit about the angle bracket expressions for pattern meta control 13:48
seems it'
s
gonna be possible
13:50 wooden left
drake1 even in a `lex' alike fashion by default 13:53
with implicit /.../e's actions and stuff 13:54
13:55 uniejo left
drake1 perfect for ASM generators 13:55
maybe a bit slower than lex+yacc+cc 13:59
PerlJam slower for now, but still more expressive. 14:01
moritz_ well, we found that Perl 6 is a pretty good language to write a compiler in
drake1 very friendly, usually
moritz_ which is why big parts of Rakudo are written in Perl 6
drake1 with a proper build system, the compilation time isn't really an issue 14:02
PerlJam whenever the code is optimized for speed, I'd be willing to be the total time of development using Perl6 will be much less than lex+yacc
drake1 probably
also a matter of taste
14:03 lamstyle joined 14:06 Trashlord left
bbkr lex+yacc are very complicated when compared to P6 grammar with :actions attached. check for example masak's HTML::Template - grammar is on one screen. i use Eyapp at work and most jobs can be done with P6 grammars in 50% less time. 14:10
14:10 amkrankruleuen joined 14:11 drbean left
drake1 I love to code C 14:11
tadzik I need to lear about this grammars Actions
drake1 sometimes Perl is a lot more complicated
moritz_ and sometimes a lot easier :-) 14:12
drake1 when things are getting tiny and specific
moritz_: depends on many things
bbkr drake1: me too. and I think it will be possible soon (or it is possible now) to have very clear P6 grammar and C libs to interpret AST.
drake1 hmm 14:13
moritz_ drake1: sure; for low level stuff C is easier; but string manipulations are just a PITA in C
more so if they can contain binary data
drake1 depends how you wrap it
14:13 Italian_Plumber left
moritz_ well, the fact that you have to wrap it speaks for itself 14:14
in Perl, it is already wrapped
bbkr tadzik: the best way to see how grammar actions work is to analyze JSON::Tiny (available on CPAN)
drake1 C99 macro code is typically far enough string magic
but I like perl to
depends on my personal mood 14:15
sometimes the one sometimes the other 14:16
moritz_ github.com/moritz/json/blob/master/...Grammar.pm
github.com/moritz/json/blob/master/...Actions.pm
14:17 Trashlord joined
drake1 a bit paradoxical, how perl5 is completely utf-8 for a trillion different scripts while trying to be the a single language for everything at the same time 14:17
moritz_ doesn't understand that statement 14:18
you can use perl5 in all-latin1 if you want
14:20 alester joined
drake1 I mean with `sh' you always switch scripting language, in perl you stay or can stay in perl 14:20
tadzik moritz_: make $<pair>>>.ast.flat; is actually make $<pair>».ast.flat?
>>> looks not pretty
moritz_ tadzik: yes; I planned to change that
now that the applicable rakudobug is fixed 14:21
tadzik so Grammar is Parsing, and Actions is generating the actual data structure? 14:22
14:22 perlygatekeeper joined
tadzik duh, I need to finally read this grammar chapter 14:22
drake1 the actions stack the compilation
in fine binary strings
moritz_ tadzik: yes, you should 14:23
tadzik: and report back whether it answers your questions
amkrankruleuen Hello, i find documentation for Sockets in Perl6 14:31
moritz_ where?
amkrankruleuen generally 14:32
moritz_ are you searching for documentation? 14:33
perlcabal.org/syn/S32/IO.html#IO%3A...%3A%3AINET seems all we have right now :( 14:34
amkrankruleuen Ups, sorry, i sleep -.-, i s/find/search/
moritz_ there are some working examples that you could use as inspiration, though
amkrankruleuen Thanks
tadzik moritz_: so far the book looks nice, I even noted down some bugs and typos 14:35
moritz_ in particular github.com/mberends/http-server-simple
tadzik++
tadzik I sat and started reading it from the very beginning
let me translate it to english
moritz_: can I fix them myself?
moritz_ tadzik: sure
hugme: add tadzik to book
hugme hugs tadzik. Welcome to book!
tadzik that'd be the second Warsaw Perl Monger fixing the Perl 6 Book :) 14:36
after Piotr Fusik
moritz_: oh thanks :)
suprise hugme :)
moritz_ thinks we could use a dozen Warsaw Perl Mongers :-) 14:37
14:37 x3nU is now known as mbialy
drake1 IO::Pipe, that's exactly what I needed. Thank you moritz_. 14:37
tadzik :)
speaking of Warsaw Perl Mongers, how is require doing in Rakudo?
14:38 pnate joined
moritz_ it blocks on me not getting it to parse, and nobody else is working on it :/ 14:38
rakudo: use Test <is ok plan>;
p6eval rakudo b958a1: OUTPUT«===SORRY!===␤Confused at line 22, near "use Test <"␤»
tadzik looks like pdf conversion is eating letters 14:39
moritz_ the parsing code is in src/Perl6/Grammar.pm line 458
tadzik: are you looking at the latest PDF?
dalek ok: 8746dbc | moritz++ | src/ (2 files):
[subs] rip out section about multi methods; it belongs into the OO chapter, and probably with quite different wording
tadzik moritz_: yep 14:40
moritz_ and I think that the <arglist> *should* be parsing <is ok plan>
tadzik imgur.com/xeGFO
moritz_: ↑
moritz_ ouch
tadzik also, I don't know how to fix that one: imgur.com/Sle6T 14:41
besides another lowercased c
drake1 sed '/^\.PL/,/^\.LP/!d' */* |perl . Gotta run. Bye 14:42
14:42 drake1 left
tadzik also I have no tools to build the book :/ 14:43
moritz_ tadzik: the README tells you what you need... it comes with every modern Linux distributions
s/s$//
tadzik speak for yourself, CRUX has no inkscape in official ports ;) 14:44
should I add myself to contributors?
moritz_ s/modern/modern and mostly complete/
:-)
tadzik: yes, please
tadzik CRUX is complete. It's just frugal in terms of packages :) 14:45
moritz_ I can't build it currently, because at $work I don't have the tools either
will try some fixes at home later on
14:46 zulon left
dalek ok: 54b6b80 | tadzik++ | src/regexes.pod:
Fixed another broken C<> in regexes.pod
 14:46
ok: d28f1b3 | tadzik++ | src/basics.pod:
Fixed C<> in basics.pod
tadzik I think the first one fixed that second bug
moritz_ aye
and I think the first bug is just missing indention
it seems the building code uncondtionally cuts off the first 4 characters of code blocks 14:47
tadzik so indenting it will fix it? This appears in many places 14:48
moritz_ I'll test it when I get home
if that's the fix, easy karma :-)
tadzik :D 14:49
speaking of easy karma...
jnthn tadzik++ 14:50
;-)
14:50 sftp left
tadzik oh noes, you have spoiled the effect! :) 14:50
speaking of easy karma, **dalek**...
14:50 sftp joined
dalek ok: 426a900 | moritz++ | src/basics.pod:
markup fix, noticed by tadzik++
 14:52
ok: 081b5ef | tadzik++ | lib/book.sty:
Added myself to contributors
tadzik ...there we go :)
I mean the second one :)
14:52 mbialy is now known as x3nU
tadzik so these are just spaces, yep? I remember seeing another few bugs like this one 14:53
or maybe I didn't
14:54 wtw left
moritz_ there are some more 14:55
src/basics.pod lines 491 to 502 14:56
tadzik www.reddit.com/r/programming/commen...bsets_and/ I see a room for some nice propaganda response to that comment :)(
moritz_ and src/subs-n-sigs.pod lines 76 and later
tadzik aye
are you fixing it now?
moritz_ no; I want to test if my last commit actually fixed anythinig 14:57
but I can't do that right now
so I have to wait for tonight
tadzik building inkscape atm, I may be able to test it in a few minuts 14:58
moritz_ (reddit) replied, though my reply seems a bit lame
tadzik not bad 14:59
lue hello o/ 15:05
bbkr rakudo: sub foo(Int $x) {say "called with $x"}; foo(any(1,2,3)) # it that correct method dispatch or a bug? looks like a bug because Int !~~ Junction
tadzik hi lue \o
p6eval rakudo b958a1: OUTPUT«(timeout)»
moritz_ timeout looks bad
rakudo: sub foo(Int $x) {say "called with $x"};
p6eval rakudo b958a1: ( no output ) 15:06
bbkr it returns "called with 1" "called with 2" "called with 3" locally on Pisa release
moritz_ and that's spec
that's what we call "autothreading"
jnthn That's what auto-threading does.
bbkr ah, awesome! 15:07
tadzik autothreading alredy works? yay. I thought threads in Parrot are NYI
15:08 justatheory joined
mathw it doesn't necessarily do it in parallel 15:08
conceptually yes
moritz_ authoreading != threading
mathw but at the moment it doesn't use threads to do it
tadzik oh, I see
bbkr rakudo: multi sub foo(Int $x) {say "called with number $x"}; multi sub foo(Str $y) {say "called with string $y"}; foo(one(1,"a")) # this dispatch is purely AWESOME! 15:11
p6eval rakudo b958a1: OUTPUT«(timeout)called with number 1␤» 15:12
bbkr eval has some hickup today :( 15:13
jnthn bbkr: That took...effort...to get right. :-)
[Coke] partcl-nqp: puts [namespace children ::]
p6eval partcl-nqp: OUTPUT«␤»
[Coke] ww!
15:13 isBEKaml joined
moritz_ bbkr: what's odd is that the p6eval has a pretty load at the moment 15:14
15:15 scoon joined
isBEKaml good evening, #perl6! 15:16
moritz_ o/ isBEKaml 15:17
isBEKaml hugme: hug moritz # for his response to the reddit thread. esp "Steal" xD 15:18
hugme hugs moritz
15:18 scoon left, justatheory left
tadzik hehe 15:19
bbkr hmm, i found some inconsistency. %($x) calls method $x.hash but @($x) does not call $x.array. 15:27
i'm not sure if it's anywhere in the spec
moritz_ what does @() call? .flat?
TimToady probably .list 15:28
bbkr rakudo: class A { method list {say 1}}; %(A.new) # checking
jnthn rakudo: class A { method flat { say "yes" } }; say @(A.new)
p6eval rakudo b958a1: OUTPUT«Method 'hash' not found for invocant of class 'A'␤ in main program body at line 22:/tmp/rhvL2pb_Pp␤»
rakudo b958a1: OUTPUT«A()<0x5fbf5e0>␤»
bbkr rakudo: class A { method list {say 1}}; @(A.new) # checking
p6eval rakudo b958a1: OUTPUT«1␤» 15:29
jnthn rakudo: class A { method list { say "yes" } }; say @(A.new)
p6eval rakudo b958a1: OUTPUT«yes␤1␤»
jnthn ah, bbkr beat me to it :-)
bbkr .list :) thanks, everything clear now
TimToady we try to keep the flat/slice distinction lazy
15:42 pyrimidine joined
x3nU where can i find list of date&time related functions in perl6 (if there's any)? 15:42
tadzik S32 is it? 15:44
PerlJam x3nU: perlcabal.org/syn/S32/Temporal.html
tadzik yep
TimToady you can find these things via perl6.org specifications linking to perlcabal.org/syn/ 15:45
x3nU thanks :) 15:46
15:48 zulon joined 15:50 Mowah joined 15:55 ruoso left
pmichaud good morning, #perl6 15:57
TimToady back atchya
jnthn afternoon, pmichaud :-) 15:58
tadzik pre-evening, pmichaud :)
TimToady hugme should randomly hug people who use :( or :/ 16:00
pmichaud it doesn't do that yet? :(
16:00 risou joined
tadzik :( 16:00
that'd be a nice feature
TimToady if not overdone 16:01
jnthn But what about all the extra traffic? :/
tadzik hugme: hug jnthn
hugme hugs jnthn
TimToady it should track who it has not hugged recently
jnthn
.oO( hugma )
TimToady it should especially hug someone new to the channel 16:02
[Coke] thinks that's a little overboard.
TimToady I mean if they use a negative emoticon
[Coke] oh, I thought you meant on join. 16:04
Þ
bbkr rakudo: say "mmmmmmmmmany" ~~ /m ** 1..0/ # a bug? inverted ranges (correct according to STD) matches infinite amount in regexp repetition ** op 16:05
p6eval rakudo b958a1: OUTPUT«mmmmmmmmm␤»
TimToady it could also detect when someone asks a question and there's no activity for the next N minutes
jnthn Oh wow, :Þ is a really cool Icelandic smiley.
TimToady bbkr: yes, that should match 0 times 16:06
should probably be a compile error, really
PerlJam suddenly wonders what / x ** 'a'..'z' / would mean 16:07
sjn jnthn: :Ð
pmichaud bbkr: yes, it's a bug.
PerlJam: I suspect it would parse as / [ x ** 'a'] .. 'z' / 16:09
bbkr reports 16:10
pmichaud std: / <[9..0]> / # just checking 16:11
p6eval std 32087: OUTPUT«ok 00:03 117m␤»
16:11 tylercurtis joined 16:13 molaf joined 16:14 foodoo left, tadzik left
pugssvn r32088 | lwall++ | [STD] detect empty range on ** quantifier 16:14
gfldex i like hugme as it is right now because the human interaction that is required to trigger the hug gives that hug a meaning
bbkr inverted ranges should be legal. there are many examples when silent-skip of inverted range is most DWIM behavior. 16:16
pmichaud inverted ranges are okay, but probably not if the range is a constant. 16:17
jnthn bbkr: I suspect this is only looking for literals that are inverted.
TimToady jnthn: that is correct 16:18
16:18 pmurias joined
pmurias hi 16:18
bbkr jnthn: yes
TimToady empty character class ranges are a bit harder to detect, since we'd have to deal with \x and such
that should probably be done in the semantic layer 16:19
pmichaud agreed
afk, lunch and errands 16:22
bbkr rakudo: say "" ~~ /m ** 1..-1 / # tiny bug - error message should say "positive integers" because -1 IS an integer. 16:25
x3nU i know that i'm asking stupid questions but it's because there's no good perl6 tutorial 16:26
p6eval rakudo b958a1: OUTPUT«===SORRY!===␤Only integers or '*' allowed as range quantifier endpoint at line 22, near "-1 / # tin"␤»
x3nU how do i read Str from socket?
16:26 Mowah left
moritz_ x3nU: there is indeed no good documentation for that - see github.com/mberends/http-server-simple for an example though 16:26
x3nU i heard that there was get() method, but well... it *Was*
TimToady std: say "" ~~ /m ** 1..-1 / 16:28
p6eval std : ( no output )
TimToady std: 42
p6eval std : ( no output )
TimToady has p6eval abandoned using 'make snap' for std? 16:29
16:29 zulon left
moritz_ no 16:29
16:31 spq1 joined 16:32 ambs joined 16:34 ambs left 16:35 snearch joined
TimToady std: say "" ~~ /m ** 1..-1 / 16:35
p6eval std 32088: OUTPUT«===SORRY!===␤Malformed range at /tmp/NlvQf0HNFI line 1:␤------> say "" ~~ /m ** 1..⏏-1 /␤ expecting quantifier␤Parse failed␤FAILED 00:01 119m␤»
TimToady std: say "" ~~ /m ** 1..0 / 16:36
p6eval std 32088: OUTPUT«===SORRY!===␤Empty range at /tmp/CL9etXG2gw line 1:␤------> say "" ~~ /m ** 1..0⏏ /␤Parse failed␤FAILED 00:01 119m␤»
bbkr TimToady: I'll add this output to both tickets. 16:38
std: "" ~~ /m ** 1..Inf / 16:40
p6eval std 32088: OUTPUT«===SORRY!===␤Malformed range at /tmp/S0dgnvVdrx line 1:␤------> "" ~~ /m ** 1..⏏Inf /␤ expecting quantifier␤Parse failed␤FAILED 00:03 119m␤»
bbkr well, 1..Inf is not malformed range :) 16:41
std: 1..Inf
p6eval std 32088: OUTPUT«ok 00:04 116m␤» 16:42
bbkr inconsistency here
16:47 ruoso joined, isBEKaml left
TimToady it can't match Inf of them, only * of them :) 16:47
matching is harder than counting :) 16:49
dalek ok: 917d907 | moritz++ | src/ (2 files):
more markup fixes; turns out we need four leading spaces in each code block
 16:50
x3nU ok, i know how to read str from socket 16:55
16:55 justatheory joined
x3nU but what if i want read bytes or integer 16:56
bbkr TimToady: you're right. and that's why it shouldn't complain about malformed range, because range is perfectly legal, but not in this context. STD error is misleading.
moritz_ x3nU: oha has submitted a patch that allows reading bytes 16:57
x3nU: it's being reviewed
16:59 dakkar left
x3nU so in best case i will be able to do that in a month? ;f 17:00
moritz_ in a week, if you're lucky
(and if you build the compiler from source)
x3nU using compiler builded from repo is bad idea imho ;/ 17:01
bbkr TimToady: most P5 converts won't see difference between 1..* and 1..Inf meaning. and if they get "malformed range" error they will report bug because the range is valid.
TimToady tylercurtis: in your subset article, a 16-year-old is neither a Child nor an Adult 17:02
mathw TimToady: that's because they're a Hassle
tylercurtis TimToady: thanks. 17:03
tylercurtis fixes.
TimToady bbkr: I don't think that's going to be a major problem
normal ranges also allow negative numbers and non-numerics, and we don't allow those here either 17:07
it's much more likely for a p5 programmer to try to leave out the right side of .. entirely in this case 17:08
std: /x ** 1../
p6eval std 32088: OUTPUT«===SORRY!===␤Malformed range at /tmp/jJajhI08ZT line 1:␤------> /x ** 1..⏏/␤ expecting quantifier␤Parse failed␤FAILED 00:01 117m␤»
tylercurtis suspects that, for now at least, Rakudo's error messages are much more relevant for (new) P5 converts. 17:09
moritz_ tylercurtis: yes, but rakudo gets its parse error messages from STD
17:12 thebird left 17:13 pmurias left
bbkr TimToady: I'm just saying that "expecting quantifier" part of this error message is not helpful at all because there are some "hidden" limitations about this quantifier (positive Int or *). But I assume not muany implementations will use directly STD errors, so maybe you're right that it won't be major problem 17:13
bbkr is back to bughunting 17:14
17:15 pmurias joined
TimToady I just don't want the grammar to end up ten times larger to catch marginal mistakes that almost anyone can figure out rapidly; there has to be some balance between that and an error message that just says '?' 17:16
ingy hi TimToady. Got a sec? 17:17
moritz_ and the error message does indicate the position pretty well
TimToady had one, but it expired 17 seconds ago...
moritz_ rakudo: / ; /
p6eval rakudo b958a1: OUTPUT«===SORRY!===␤Confused at line 22, near "/ ; /"␤»
TimToady std: / ; /
p6eval std 32088: OUTPUT«===SORRY!===␤Semicolon must be quoted at /tmp/BkGopLO1bQ line 1:␤------> / ;⏏ /␤Parse failed␤FAILED 00:01 115m␤»
moritz_ std: / - / 17:18
p6eval std 32088: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter (must be quoted to match literally) at /tmp/AnzpibclQL line 1:␤------> / -⏏ /␤Can't call method "from" on unblessed reference at STD.pm line 53782.␤FAILED 00:01 115m␤»
ingy I need a word/term... general term for a piece of data like a string or num
TimToady oopsie
moritz_ ingy: "atom"?
TimToady scalar? item?
datum?
ingy "object" feels correct but too overloaded
TimToady thingie?
ingy the value could be a list 17:19
TimToady Moo
moritz_ "value"?
ingy value and object are two
except $value->value feels yucky
bbkr "item"
PerlJam bbkr: I don't think the limitations will be hidden if everywhere we talk about ranges in regexp, we use normal ranges as an analogue rather than begin the same. 17:20
s/begin/being/
ingy the thingy has type and value properties
variable seems wrong
moritz_ "instance"?
tylercurtis Thing?
PerlJam ingy: "object" :) Just because it's too overloaded in general doesn't mean it can't work. 17:21
ingy thing is too wishy washy
x3nU case insensitive regex is still unimplemented?
moritz_ x3nU: no
tylercurtis rakudo: say 'AbC' ~~ /:i abc/
p6eval rakudo b958a1: OUTPUT«AbC␤»
ingy I am on PerlJam's camp, but thought I'd ask here 17:22
moritz_ rakudo: say 'AbC' ~~ m:i/abc/
p6eval rakudo b958a1: OUTPUT«AbC␤»
x3nU wow
moritz_ now with modifiers on the outside
bbkr ingy: maybe "chunk" ?
tylercurtis Noun.
ingy I like wow ;)
17:22 rhr_ left
ingy x3nU++ 17:22
datum 17:23
TimToady dat
17:23 snearch left
ingy nice 17:24
TimToady++ for inventing new words
but I think I want a common one 17:25
17:25 BlueT_ joined
ingy more acmeist 17:25
acmeism is inclusive of boring languages
17:25 envi^home left
ingy :P 17:25
tylercurtis thesaurus.com/browse/thing?r=76&...amp;ch=the
ingy tylercurtis: I been surfing too :) 17:26
portion 17:28
sjn wonders where to find documentation for the batch method
17:29 rhr joined
colomon sjn: don't know if it exists yet. 17:30
sjn S07 mentions itr
colomon and come to think of it, I don't think it's supposed to exist at all any more -- the batch method, I mean, not the documentation
sjn oh?
TimToady S07 is quite out of date 17:31
colomon I believe it's been replaced (more or less) by munch.
which probably also isn't documented.
that area of the spec and implementation has been under heavy construction recently. 17:32
sjn colomon: so you'd write (1,2,4...*).munch(10) ? 17:33
colomon any objections to my removing batch from the spectests (it's only in two tests at the moment) and Rakudo?
sjn: yes
sjn ok
why the name change?
colomon star: say (1,2,4...*).munch(10)
p6eval star 2010.07: OUTPUT«1248163264128256512␤»
colomon star: say (1,2,4...*).munch(10).perl
p6eval star 2010.07: OUTPUT«(1, 2, 4, 8, 16, 32, 64, 128, 256, 512)␤» 17:34
17:34 pmurias left
colomon It behaves by taking the first N items off of the list (like using shift N times in a row), so munch was considered more accurate than batch, because it conveys the notion that the items are now gone from the list. 17:34
sjn right 17:35
why not just call it first?
TimToady first does something else 17:36
colomon again, first doesn't convey the notion that the elements are gone from the list
sjn ah
right
colomon star: my $a := (1, 2, 4 ... *); say $a.munch(3).perl; say $a.munch(3).perl 17:37
sjn and .shift(10) would be kinda odd
p6eval star 2010.07: OUTPUT«(1, 2, 4)␤(8, 16, 32)␤»
colomon I think shift(10) has been seriously considered, but people felt it was better to have a separate method name for now.
TimToady .munch might default to .munch(*), which would let the list decide how big to make the batch 17:38
17:39 risou left
sjn fwiw, I like term reuse, when possible :) 17:39
TimToady shift was also felt to be more specific to arrays than to lists
we're trying to keep the APIs somewhat distinct 17:40
sjn hm.
bbkr rakudo: %() # Parrot VM: PANIC: Out of mem! 17:45
p6eval rakudo b958a1: ( no output )
bbkr hmm
does it go OOM locally for you on Pisa build?
sjn shift and munch(1) are the same type of actions, but still different... 17:46
is it possible to have lists that begin with * ? 17:47
(or Inf)
17:47 justatheory left
moritz_ sure 17:48
rakudo: say (Inf, 1, 2, Inf).perl
p6eval rakudo b958a1: OUTPUT«(Inf, 1, 2, Inf)␤»
TimToady you're confusing lists with ranges
sjn e.g. (*...*+*,1,1)
moritz_ arithmetics with Inf is pretty boring 17:49
Inf +-*/ finite number == +- Inf
all other operations are either +-Inf or NaN
bbkr rakudo: %(); say "alive"; 17:50
p6eval rakudo b958a1: ( no output )
bbkr reports
sjn right, so munch() works on lists _and_ ranges, but not arrays 17:51
TimToady rakudo: say (* R... 1,1,*+*).munch(10).perl 17:52
p6eval rakudo b958a1: OUTPUT«(1, 1, 2, 3, 5, 8, 13, 21, 34, 55)␤»
moritz_ scary. 17:53
sjn rakudo: (1..*).munch(4).perl.say 17:54
p6eval rakudo b958a1: OUTPUT«Method 'munch' not found for invocant of class 'Range'␤ in main program body at line 22:/tmp/PsFUmJmtTg␤»
sjn right
sjn is even more confused
maybe I should play with something safer :-P 17:55
moritz_ .[^10] is pretty safe
sjn rakudo: (1..*)[^4].perl.say
moritz_ rakudo: say (1..*).[^10].join(', ')
p6eval rakudo b958a1: OUTPUT«(1, 2, 3, 4)␤»
rakudo b958a1: OUTPUT«1, 2, 3, 4, 5, 6, 7, 8, 9, 10␤»
sjn likes the munch() thing mostly because of the possibilities for making Edvard Munch jokes :) 17:56
moritz_ rakudo: say %(a => 1, b => 2).perl 17:57
p6eval rakudo b958a1: ( no output )
17:58 tylercurtis left
sjn on the down side, I'm getting a bit worried about the keyword inflation in Perl6 :-( 17:59
17:59 TimToady left
moritz_ .munch is not a keyword 18:00
sjn well
ok, "term inflation"?
moritz_ "method inflation"?
dalek kudo: 689bb75 | moritz++ | src/builtins/Parcel.pir:
fix %() segfault. Patch courtesy of TiMBuS++
 18:01
sjn "concept inflation"
or as we n00bs like to call it, "ETOOMUCHTOLEARN" 18:02
moritz_ I kinda share your concerns 18:03
OTOH you don't need to learn them all
sjn proposed to jnthn to give a talk in Oslo about "patterns" in Perl6 (where that term means "concepts that are common all over perl 6, that you can put together to create really cool tings") 18:04
multimethods is one such thing
another is metaoperators 18:05
moritz_ that's actually a nice idea
sjn I'm sure there are others</perl6 n00b>
my problem with that thought, plus the munch/shift thing, is that they don't really go together 18:06
18:06 araujo left
pugssvn r32089 | moritz++ | [t/spec] tests for %() hash context forcer; also tests RT #75502 18:06
sjn in my brane at least - munch and shift are so similar, it seems weird they're named differently 18:07
18:07 araujo joined
sjn the concept is really "take n items of the beginning of this sequence", right? 18:08
where shift just takes 1, and munch takes any
but they do it differently according to what kind of sequence they operate on
(at which point, my thoughts go "this smells awefully multimethod-like") 18:10
18:13 macroron joined 18:14 macroron left, macroron joined
pugssvn r32090 | moritz++ | [t/spec] get rid of wrong plan 18:16
18:17 cotto_work joined, plainhao joined 18:20 am0c^ joined 18:25 jakk_ joined, rlb3 joined
moritz_ std: / - / 18:26
p6eval std 32088: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter (must be quoted to match literally) at /tmp/iBU9UZsd8y line 1:␤------> / -⏏ /␤Can't call method "from" on unblessed reference at STD.pm line 53782.␤FAILED 00:01 115m␤»
moritz_ rakudo: / - /
p6eval rakudo b958a1: OUTPUT«===SORRY!===␤Confused at line 22, near "/ - /"␤»
18:26 jakk_ left
moritz_ nqp: / - / 18:27
p6eval nqp: OUTPUT«Confused at line 1, near "/ - /"␤current instr.: 'parrot;HLL;Grammar;panic' pc 552 (src/cheats/hll-compiler.pir:205)␤»
18:27 Phiend joined 18:28 alanhaggai joined
Phiend hello, I have a simple question on roles : I know it is possible to add a role to an object or class, but I never hear of the role being removed. Possible/how ? 18:28
PerlJam Phiend: sounds painful 18:29
Phiend laughs
moritz_ Phiend: don't think it's supported
Phiend if not, that would be kinda a glaring design issue.
moritz_ removing a role screams for trouble 18:30
Phiend well, it is reasonable. happens all the time in real life, for example. an it would be orthogonal behavior, darn it :)
18:31 tylercurtis joined
moritz_ phenny: tell mberends if you replace 'plan *;' with a specific number, please spectest before committing 18:32
phenny moritz_: I'll pass that on when mberends is around.
colomon Phiend: would you expect every existing object of that class to be modified to lose the attributes and methods from that role? 18:33
pugssvn r32091 | moritz++ | [t/spec] remove a wrong plan; avoid STDOUT output that confuses TAP 18:34
moritz_ std: $
p6eval std 32090: OUTPUT«===SORRY!===␤Non-declarative sigil is missing its name at /tmp/GlPSJbdGbL line 1:␤------> <BOL>⏏$␤Bogus statement at /tmp/GlPSJbdGbL line 1:␤------> $⏏<EOL>␤ expecting twigil␤Parse failed␤FAILED 00:03 116m␤»
moritz_ std: @
p6eval std 32090: OUTPUT«===SORRY!===␤Non-declarative sigil is missing its name at /tmp/vckluP58Zp line 1:␤------> <BOL>⏏@␤Bogus statement at /tmp/vckluP58Zp line 1:␤------> @⏏<EOL>␤ expecting twigil␤Parse failed␤FAILED 00:03 116m␤»
jdv79 wrong chan - sorry
Phiend colomon: for truly dynamic roles, yes. 18:35
pugssvn r32092 | moritz++ | run more tests 18:36
18:36 chromatic joined
chromatic Allowing role removal would violate the type system. 18:36
colomon chromatic++ 18:37
Phiend perhaps it is a design issue, then
moritz_ aye; it would make every operation unsafe
gfldex what does not mean that you can't implement a prototype based object system with perl 6
chromatic You're welcome to write a custom MOP which allows that, but you get to respond to bug reports when that happens.
Phiend chuckles
moritz_ it sounds comparably safe to allow ripping out integers from a language 18:39
or any other built-in types
chromatic Now I can imagine that it's useful to pretend that an entity does not perform a role in certain cases, but I expect it's more useful to override the DOES() equivalent, whatever that's called in Perl 6 today. (Haven't tried lately; don't remember.)
moritz_ well, method does() in the metaclass 18:40
chromatic $entity.^does() ?
Phiend from a design standpoint, it makes sense though that some roles be removable. For instance if you were writing code that modelled admin privileges, you would want to be able to revoke those privileges. In the real world, this often happens with 'roles' - once a cop, not always a cop for instance. It is fine if you want to keep perl6 simple, however I would assume roles would be able to
reflect transient behavior occasionally.
chromatic That's really not a role as much as it is instance data. 18:41
moritz_ Phiend: just because something is called a "role" doesn't mean you have to implement it with actual rules
s/rules/roles/
colomon "perl6 simple" !!!!!!!!!!!!!!!
18:41 sorear left
moritz_ in fact it's probably a pretty bad idea 18:41
18:41 rlb3 left
Phiend yes, maybe roles shouldn't actually be roles ? 18:41
moritz_ because you'd have to generate code for each roles
Phiend well, maybe this is something to think on for a future version of p6. maybe it is a bad idea, but then maybe not. 18:42
chromatic It's a very bad idea to allow decomposition of roles from classes. 18:43
That said, if you implement your admin system in terms of runtime decoration of instances with roles, you could make that work.
Phiend but doing so would make it orthogonal. mmmmmmmm...orthogonal.
chromatic Orthogonal is useless if the system doesn't work.
Phiend chrome, you mean runtime subtraction of roles is possible ? 18:44
colomon you keep saying "orthogonal", but what I keep hearing is "insane"
Phiend I get that a lot :)
chromatic Runtime manipulation of the class of an object is (in theory) possible.
moritz_ you could just have a "reset" role that just overrides the privilege methods that the other roles might have mixed in
Phiend well, how many calculators have you seen with '+' but no '-' ?
m: yes, but why would you need to do that hack in a well-designed language ? 18:45
tylercurtis Peano arithmetic.
chromatic How many calculators have you seen where you can say 10.02 looks like a rational number, but it's an integer. I promise.
tylercurtis always had fixed point calculators, as far as he could tell. 18:46
gfldex Phiend: if the Perl 6 role would not have been called role, would you still want it to be removeable?
moritz_ Phiend: why would you not need that in a well-designed language?
colomon Phiend: what exactly do you think the opposite of removing a role at run time?
Phiend what would we call it then : a biscuit ?
moritz_ a trait
chromatic an aspect 18:47
moritz_ a mixin
an interface
whatever
Phiend I think the name role is good enough.
chromatic I hope not a mixin or an interface.
tylercurtis A class.
Phiend an interface, not quite.
chromatic a partial class
Phiend gfldex: no matter what you call it, my answer is yes. 18:48
chromatic Okay, so when you decompose a role out of a class at runtime, do you run the constraint solver? 18:49
Do you reoptimize all of the code that could possible use that class?
Phiend I don't know - not a design expert.
chromatic Do you modify all of the instances of that class?
moritz_ Phiend: what do you think should happen if you remove the Numeric role from Int at runtime, and then somebody calls 3.sin?
chromatic What if they have user data that goes away, because those are attributes of the role?
Do you save them?
Phiend kaboom!
chromatic Do you discard them?
moritz_ right
gfldex do you remove all methods that come from classes that use attributes that come from a role? 18:50
chromatic What if it's a parametric role>
Phiend I didnot say every role should have this ability, but you should be able to, if it makes sense.
moritz_ ok
chromatic It never makes sense.
moritz_ that#'s a good idea
Phiend how this is done is left up to those crazy enough to go for it.
moritz_ as long as it's only allowed on the empty set of roles
gfldex :-> 18:51
Phiend chrome: sure it does. You have a role as a licensed driver. If you should do a few illegal DUI operations, you could lose that role
chromatic That's instance data.
That's not a role.
Phiend roles can be granted on instance objects, no ?
chromatic Sure. 18:52
gfldex the name role put an idea into your head that works very well with the reality you live in, but quite bad with programming languages
moritz_ and I must admit that some of us gave examples likes this, which might not be the best OO design 18:53
gfldex well, it does work quite well with javascript. But you trade languages niceness with speed and unpredictable runtime behaviour
Phiend I remain unconvinced; it would seem natural for a language that seeks to go along with how you think to actually do it.
gfldex: it does ?
colomon I dunno, I'm not seeing any obvious reason why removing a role *from an instance object* would necessarily be a terrible thing, though it would probably be a pain to implement cleanly.
gfldex at lest for mozilla you can fool around with the prototype chain 18:54
moritz_ Phiend: it's ok. You can drop a topic even if you're not convinced that your idea is bad :-)
18:54 sorear joined
chromatic Why would you decompose the LicensedDriver role from the Driver class when one instance of the Driver loses his license? 18:54
Phiend mor: I take that as an insult, sir
moritz_ Phiend: you shouldn't
colomon it's generalizing that to removing a role from an entire class that's completely insane.
Phiend chrome: you would not - you would decompose the instance/object, obviously.
I didn't even suggest that...although now that I think about it... 18:55
moritz_ colomon: it's interface sanity. sin(5 bot not Numeric) # BOOOM
chromatic Runtime application and deapplication of roles from instances makes sense.
Phiend I think so :)
moritz_ if you're unlucky, such cases blow up in your own code
chromatic Runtime decomposition of roles from classes does not. You cannot do it safely in any but the most trivial circumstances.
colomon moritz_: yeah, but you can always do stuff that blows up code. 18:56
moritz_ and people will write bug reports because your code blows up, even though it's not your problem
Phiend <nods>
moritz_ colomon: if you have good type constraints and precondtions, you can avoid that
colomon I mean, why shouldn't you be able to say my $a = 5 but not Numeric?
yes, that would mean Numeric functions wouldn't work on it.
moritz_ because code all around assumes that Int ~~ Numeric 18:57
and if you have instances where that's not true, your type system is screwed
gfldex role foo { has $.a; }; class bar does foo { method bumm () { say $.a; } } }; my $o1 = bar.new(); remove role foo from bar; $o1.bumm(); # what would happen?
18:57 masak joined
colomon moritz_: on an instance -- ie, this particular $a -- not on every Int. 18:57
masak ahoy, #perl6!
phenny masak: 01:07Z <TimToady> tell masak yes, checking scope constraints on FIRST is in my todo file
moritz_ colomon: code we have depends on 1 ~~ Numeric
masak no rush.
Phiend it should throw a runtime exception, I would expect. 18:58
gfldex :->
colomon moritz_: 1 is not an instance of an object, it's a literal.
moritz_ Phiend: and disable all optimizations that are based on nominal types?
Phiend m: I dunno !
moritz_ well, that would be a consequence
masak: but is it FIRST on the todo list? :-) 18:59
Phiend I'm not saying that decomposition on every role is the right thing - I would think not. Once an int, always a filthy int.
chromatic That's what we're saying. You can't decompose roles from classes at runtime.
Phiend true dat
chromatic The system does not have sufficient information to do so, even if it were safe. 19:00
Phiend but I think such an operation on instances does make sense, in some situations.
chromatic Sure; no one's disputing that.
moritz_ well, I do
Phiend ok, we are then in agreement.
except for moritz.
chromatic Sure; no good-hearted person could possibly dispute that!
moritz_ :-)
Phiend much is explained :)
well, thats all I desired !
19:01 TimToady joined
chromatic There are probably better ways to accomplish most goals (in your Driver case, a subset type) but it's a possible operation. 19:01
jnthn chromatic is right - you *lose* information when you compose roles into a class. When you compose roles into an instance, it actually creates an anonymous subclass and puts the role into that. So in that case you could re-bless to the superclass, I guess.
Phiend how would you code that then ?
masak moritz_: maybe it's NEXT on the todo list :P
chromatic You *lose* that information on purpose, as a design goal of roles. 19:02
19:02 mberends joined
jnthn chromatic: Right. :-) 19:02
Phiend thats part of what makes roles great
it increases the KISS-itude
jnthn I've mostly tended to think of roles as re-usable bits of functionality that you use to build stuff, rather than pluggable things to add and remove. 19:03
chromatic Phiend, something like subset LicensedDriver of Driver where *.points < 3 && age > 17;
Phiend interesting...although it is a hack methinks.
chromatic How so? It's a parametric type, like you might find in ML.
jnthn chromatic: I'd tended to see them as refinement types rather than parametric types. 19:04
TimToady imagines running some subset of DESTROY submethods if you rebless yourself to a parent class
Phiend well, I think so because I don't know subsets so good, and when I think of removing roles, I want to say stuff like 'bob isnt aLicensedDriver' - I'm a lazy coder ;)
jnthn TimToady: I wasn't offering to implement it, just saying its possible. :P
TimToady: Though yes, running the DESTROYs could make sense. 19:05
TimToady I've been wanting to talk about mixins for the last 1/2 hour, but appflux was unusable
chromatic You don't have to write code to specify *how* to make Bob an unlicensed driver. You can write declarative code to specify the rules as to why Bob has just become an unlicensed driver.
TimToady took you guys about 20 minutes to get there...
Phiend knock yourself out, TT !
chromatic You can ignore the mechanism and describe the policy.
TimToady yes and no 19:06
a language designer must simultaneously know everything and nothing
chromatic You need a Leelu wig. 19:07
PerlJam If you already know nothing, how do you get to know everything? :)
moritz_ that's the problem.
rokoteko PerlJam: ?
TimToady the basic problem is that performance and exceptions leak through every abstraction
Phiend chrome: but what if the decomposition is a decision made at runtime, ad hoc ? 19:08
PerlJam and knowing things (or no things) isn't enough either, you have to *decide* which is important. And *that's* the hard part if you ask me.
rokoteko You must be humble enough to understand you know nothing, as I see it. So basically he willingness always to learn something new. :) Not saying "I know everything, Im not interested.." But OTOH, Im not a language designed. :)
Phiend ya gotta have Vision. Unlike Microsoft. 19:09
TimToady is now contemplating a form of mixin that is really delegation
chromatic If you have subset types (whether you call them parametric or refinement), than any type dispatcher can destructure to examine specific attribute values.
19:09 tylercurtis left
chromatic That's a runtime decision, but a fair amount of type-based dispatch can be runtime dispatch anyway. 19:10
PerlJam TimToady: temporized?
TimToady no
if you mixin a role, and you make not only an anon class but a new object that delegates to the parent object
then you can just switch back to the parent, and the child will be GC'd 19:11
the main problem with that is identity 19:12
moritz_ rakudo: / | /
p6eval rakudo 689bb7: ( no output ) 19:13
moritz_ rakudo: / | /; say "alive"
p6eval rakudo 689bb7: OUTPUT«alive␤»
chromatic What does that offer over swapping to the anon class?
TimToady so you want the delegating child to pretend to have the same identity as its parent
automatic GC
jnthn TimToady: I'd hope that we have enough flexibility in the meta-model for that to be done in a module.
moritz_ std: / | /
p6eval std 32092: OUTPUT«===SORRY!===␤Null pattern not allowed at /tmp/seCzwubd7Z line 1:␤------> / | ⏏/␤ expecting quantifier␤Parse failed␤FAILED 00:01 115m␤»
TimToady when you unmixin
jnthn: sure, I'm just speculatin'
jnthn TimToady: Oh, I'm not saying it's a bad idea. :-) 19:14
chromatic Suppose the anon class remembers its parent (because it has to).
TimToady just sayin' that making it two different objects makes it really clear which memory belongs to each 19:15
jnthn Aye. 19:16
chromatic True. You don't have to morph.
TimToady morphing is also possible, of course
19:16 jferrero joined
jnthn Wrapping existing stuff rather than mutating it always makes it easy to "go back". 19:16
gfldex i'm trying to find a real world example where somebody is revoked a role and you dont want to keep any record that he/she ever had this role. I can't find any. On top of that drivers licenses are a really bad example here.
TimToady but I don't think calling the appropriate DESTROYs would do it completely
unless you can GC within an object's storage :)
Phiend would a Power Rangers example help ? 19:17
chromatic DESTROY on morph is a hard problem. Trust me.
TimToady and that's likely to depend on the repr
jnthn TimToady: Yeah, that's the other thing that bothers me about re-blessing to a superclass.
TimToady we're only discussing morph to a parent class here
jnthn Yes, immediate parent makes it less bad.
chromatic If all of these anon classes are singletons though, instance storage and class storage could be the same thing.
TimToady so calling the right DESTROY is just a smop, but reclaiming memory is a different issue
jnthn But it'd still need the repr to co-operate I guess. 19:18
chromatic Identity seems like the least worst problem to solve.
TimToady one wonders if the mixed object can serve as its own class
the inside-out way of repeating what chromatic said 19:19
chromatic Yes, but then wouldn't $mixed.^does( 'Class' ) be true suddenly?
jnthn TimToady: It may want to wrap the meta-object too.
TimToady huh? we don't have Class in that sense in Perl 6
jnthn It took me a while longer to realize that www2.parc.com/csl/groups/sda/public...or-web.pdf is basically a lot of words for "let's do delegation" :-) 19:20
TimToady I guess the problem is that :U would be false
jnthn *while longer than it should have done
TimToady: When you say "serve as its own class" I'm not sure what meaning of class you're meaning. :-) 19:21
TimToady me too :)
moritz_ std: / \p /
jnthn ;-)
p6eval std 32092: OUTPUT«===SORRY!===␤Unrecognized regex backslash sequence at /tmp/lZEfIRsa8p line 1:␤------> / \⏏p /␤Check failed␤FAILED 00:03 116m␤»
TimToady that should probably be a .obs() 19:22
moritz_ ie "Obsolete use of..." ?
sorear good * #perl6
jnthn o/ sorear
moritz_ \o 19:23
rakudo: /\p/
p6eval rakudo 689bb7: OUTPUT«===SORRY!===␤Confused at line 22, near "/\\p/"␤»
masak o/ sorear
jnthn masak! \o/
jnthn didn't see masak sneak in :-)
masak jnthn: only half here... :)
jnthn Ah 19:24
Clearly just your invizible half. :-)
TimToady the problem is we should really parse the argument to \p as well, and you example wouldn't trigger that
Alias waves 19:25
masak Alias! \o/
moritz_ particles
Alias I'm heading off to TechEd today for the sekrit Microsoft Open Source track
Anyone got anything they want me to talk to MS about
Phiend Visual COBOL 19:26
chromatic Not DDOSsing CPAN Testers.
Alias :)
PerlJam Alias: providing resources for Perl 6 development.
Alias PerlJam: What kind
Phiend yeah: $$$$$$$$$$$$, preferrably in gold.
PerlJam Alias: whatever jnthn needs :)
chromatic Build servers for Parrot and Rakudo would be lovely.
PerlJam Alias: what chromatic said 19:27
jnthn PerlJam: We can get Microsoft to give me beer? ;-)
Alias chromatic: I'll add that to the list for the testing lab 2.0
chromatic Thank you.
Phiend don't they have a micro-brewery ? <ducks>
jnthn lol :-D
PerlJam Alias++
jnthn Phiend: Trouble is it only makes soft drinks...
19:27 pnate left
Alias jnthn: Anything you want? MSDN license? (I have one spare) 19:28
Phiend bites his tongue...
jnthn Alias: I have a lot of Microsoft stuff already 'cus I do .Net dev in my $other-job.
PerlJam jnthn: couldn't you use a really nice machine for dev work? 19:29
:)
jnthn ;-)
Alias Not entirely sure I can get hardware out of them
Phiend that would be cute - ask them for MacBooks
jnthn Alias: I don't know how much interest they have in knowing that folks are looking at Perl 6 on .Net (multiple, in fact).
moritz_ rakudo: /<[\]]>/ 19:30
p6eval rakudo 689bb7: ( no output )
Alias jnthn: I've got a 20 minute talk tomorrow
jnthn Alias: It may be worth mentioning that there are such efforts afoot. They can't take an interest if they don't know about them. :-)
Alias I'll include it
PerlJam Does microsoft have any real interest in perl these days?
Phiend well, this is a new record for me - I never been on here this long without being kicked... 19:31
PerlJam: that is a GOOD question !
They've got IronRuby and IronPython.... but missing some perl loving.
PerlJam Phiend: well, the important part would be, if they do ... for what?
Phiend Yes. 19:32
Alias PerlJam: They care about InterOp in general
Phiend they do ???
rokoteko PerlJam: I dunno. I think perl people should have interested on making it work on windows systems.
Phiend pardon my skepticism.
I think it is good for Perl, and actually good for Microsoft, although I fear they don't have the wisdom to see it that way. 19:33
rokoteko PerlJam: it's anyways the most popular platform. That's just a sad fact. :/
Alias Phiend: The more code works on Windows, the more code gets run on Windows, the more licenses people will buy
Phiend true dat.
well, roko. things are changing. laptops are rising, and OS X is getting bigger. 19:34
Alias Plus, I'm sure there's a soft spot for Perl in MS
Tene There's a perception I've sen before that the more open-source code windows-using people run, the easier they can migrate away from MS.
Phiend MUHAHAHAHAHA
Alias They built Windows using Perl for years and years
Phiend Tene: yes, but lets not mention that.
moritz_ rakudo: /a[b}/ 19:35
p6eval rakudo 689bb7: OUTPUT«===SORRY!===␤Confused at line 22, near "/a[b}/"␤»
Phiend yes, in the past Microsoft has been good to Perl. Even though part of the reason I got uninstalled at my last job was that it was a "Microsoft shop", which made perl unpopular.
moritz_ std: /a[b}/
p6eval std 32092: OUTPUT«===SORRY!===␤Unable to parse bracketed regex at /tmp/cYo4scQBFP line 1:␤------> /a⏏[b}/␤Couldn't find final ']'; gave up at /tmp/cYo4scQBFP line 1:␤------> /a[b⏏}/␤ expecting any of:␤ metachar␤ quantifier␤
..regex atom␤Parse …
PerlJam "uninstalled"? That's an interesting metaphor. 19:36
Alias Phiend: I'm hoping that Strawberry Professional will help with that. Make's Perl look more like a "normal" Windows language
Phiend why do that to Perl though ?
19:36 molaf left
Phiend or is this a visual studio add in ? 19:36
Alias So it will suck less?
It's a Visual Studio equivalent
Phiend Perl doesn't suck. windows does.
Alias Plus no command line
Phiend well, it sounds good.
Alias Unless you want it
Phiend command line : old farts love it.
however you get perl in there, is good. We must infiltrate, integrate and destroy ! 19:37
hopefully at some point during all that, there will be beer involved. 19:38
moritz_ phenny: ask pmichaud it seems that the p6regex tests in nqp-rx assumes that nqp-rx should support the :x($count) modifier -- is that sitll a valid assumption?
phenny moritz_: I'll pass that on when pmichaud is around.
Phiend well, I've exceeded my luck quotient for today; everybody weng chung ! 19:39
19:39 chromatic left, Phiend left 19:42 amkrankruleuen left
dalek p-rx: 705f966 | moritz++ | t/p6regex/ (8 files):
[p6regex-test] fudge it to be clean; fix a few tests along the way
 19:43
19:47 masak left
dalek p-rx: dfbe437 | moritz++ | build/Makefile.in:
include p6regex-test in "test" target
 19:55
20:02 timbunce left 20:03 Alias left
ingy re: the datum/object/thing naming discussion, I am settling on "value" for now. This is mainly due to the fact that they are always thought about in the context of things returned by functions. 20:03
and everyone knows that a function returns a value.
QED
BBQ 20:04
ETC
FTW
PerlJam I thought they returned PArcels
ingy meh
TimToady not are values are valuable
[Coke] not everything is a perl6.
ingy [Coke]: but perl6 is certainly a everything 20:05
x3nU how do i convert Match to Str? 20:07
ingy x3nU: tmtowtdi
moritz_ $/.Str
or ~$/
TimToady or just use it as a string
20:07 timbunce joined
x3nU also imho it sucks that i must conver it 20:07
convert it*
moritz_ TimToady: rakudo currently requires a conversion to Str for hash keys 20:08
because $/ ~~ Positional, so the .{}(Positional) multi is called
which iterates over the argument
and since $/ doesn't flatten, it recurses infinitely
jnthn Maybe we need a Match candidate... 20:09
(that knows to strinfigy it)
er, stringify
TimToady I guess the question is what will DWTM
jnthn It kinda feels like we're adding a (allbeit useful) special case there. 20:10
It makes me feel slightly dirty.
As if there should be a better way for this to work. 20:11
moritz_ rakudo: 'abc' ~~ /(.)(.)(.)/; my @a := $/; .say for @a; 20:13
p6eval rakudo 689bb7: OUTPUT«abc␤»
moritz_ is that right?
I would kinda expect something @-sigiled to be flat
rakudo: 'abc' ~~ /(.)(.)(.)/; my @a := $/.flat; .say for @a;
p6eval rakudo 689bb7: OUTPUT«a␤b␤c␤» 20:14
x3nU also, how do i convert str to uppercase? there is .tou but it don't work with Str
PerlJam rakudo: say "foo".uc
p6eval rakudo 689bb7: OUTPUT«FOO␤» 20:15
20:15 daxim left
PerlJam rakudo: say uc("foo") 20:15
p6eval rakudo 689bb7: OUTPUT«FOO␤»
x3nU lol, i don't know why i was thinking about tou 20:16
but thanks :)
pugssvn r32093 | lwall++ | [STD] handle \p redirection 20:20
20:20 Mowah joined
moritz_ rakudo: my @a := 1; say @a.perl 20:20
p6eval rakudo 689bb7: OUTPUT«1␤» 20:21
moritz_ rakudo: my @a := 1; say @a ~~ Positional
p6eval rakudo 689bb7: OUTPUT«0␤»
moritz_ rakudo: my Str $a := 1
p6eval rakudo 689bb7: OUTPUT«You cannot bind a variable of type Int() to a variable of type Str().␤ in 'infix:<:=>' at line 685:CORE.setting␤ in main program body at line 22:/tmp/22REpULr3l␤»
20:22 Helios- left, Helios joined 20:23 Helios left
jnthn moritz_: Some type of problem there... 20:23
moritz_ some type of problem ticketed there :-) 20:24
jnthn moritz_: I think we get away with it for assignment because list assignment means you can't fail to assign something positional.
20:24 Helios joined
moritz_ jnthn: is the problem that @a doesn't have a Positional type constraint? 20:25
jnthn Yeah, I think so. 20:26
mberends phenny, tell masak oha++'s Socket.pm and INET.pm patch in gist.github.com/543580 tests out OK on all the webservers unmodified in github.com/mberends/http-server-sim.../examples/ so it looks good to commit it. This does not prove binary file I/O yet however, that requires some rewriting. 20:28
phenny mberends: I'll pass that on when masak is around.
mberends: 18:32Z <moritz_> tell mberends if you replace 'plan *;' with a specific number, please spectest before committing
20:28 guidj0s_ joined
moritz_ phenny++ 20:29
jnthn o/ mberends 20:30
mberends moritz_: I did spectest, that's what delayed the commits. Have I let some errors through? Sorry if that is the case.
moritz_ mberends: yes, two mis-plans; I fixed them, so no problem
mberends thanks moritz_++ :)
20:31 tylerni7 joined 20:32 kirillm left
moritz_ std: sub design-ice-cream-mixture(:$name!, $base = 'Vanilla') { 20:33
p6eval std 32093: OUTPUT«===SORRY!===␤Can't put optional positional parameter after variadic parameters at /tmp/PCimRuuR2z line 1:␤------> cream-mixture(:$name!, $base = 'Vanilla'⏏) {␤Unable to parse block at /tmp/PCimRuuR2z line 1:␤------> eam-mixture(:$name!,
..$ba…
mberends o/ jnthn: I'm going to have to apply a dependency injection circularity saw to the 6model java hierarchy. Still wondering where to cut...
moritz_ std: sub design-ice-cream-mixture(:$name!, $base = 'Vanilla') { }
p6eval std 32093: OUTPUT«===SORRY!===␤Can't put optional positional parameter after variadic parameters at /tmp/hmxUfXl2CG line 1:␤------> cream-mixture(:$name!, $base = 'Vanilla'⏏) { }␤Potential difficulties:␤ $base is declared but not used at /tmp/hmxUfXl2CG line
..1:␤…
moritz_ TimToady: that error message is mighty confusing
:$name! doesn't look rather variadic to me
jnthn I think it's Just Wrong.
moritz_ std: sub design-ice-cream-mixture($base = 'Vanilla', :$name!) { } 20:34
jnthn Probably just bug somewhere.
p6eval std 32093: OUTPUT«Potential difficulties:␤ $base is declared but not used at /tmp/wR4NDCkoBN line 1:␤------> sub design-ice-cream-mixture(⏏$base = 'Vanilla', :$name!) { }␤ok 00:01 119m␤»
jnthn mberends: Oh, curious.
mberends: Why, ooc?
mberends jnthn: it's a long loop. SharedTable used to be at the end, now it's also at the beginning. 20:35
jnthn :S
mberends jnthn: via ThreadContext and Context 20:36
jnthn I'm curious why I didn't hit the problem in the .Net version.
20:36 PhatEddy joined
jnthn mberends: Is this some "can't have circular references between classes" thingy? 20:36
dalek ok: 677ac11 | moritz++ | src/subs-n-sigs.pod:
Fix order of named and positional parameters.

Thanks to David H Adler and James E Keenan in RT #77220
jnthn mberends: If so, I guess I just hadn't appreciated that Java had such a restriction. 20:37
mberends jnthn: yes, hence dependency injection as a solution.
jnthn Ah
That sucks.
.oO( Surprise! It's Java! )
mberends jnthn: nvm, first nom, then hacking 20:38
jnthn mberends: Oh well. At least you didn't offer to port it to VBScript... ;-)
moritz_ rakudo: sub a(:$a!, $b = 2) { }
20:38 slavik left
p6eval rakudo 689bb7: OUTPUT«===SORRY!===␤Can't put required parameter after variadic parameters at line 22, near ") { }"␤» 20:38
mberends jnthn: lol
jnthn moritz_: I blame STD! ;-)
PhatEddy rakudo: my $ = 2; say 'hello'
p6eval rakudo 689bb7: OUTPUT«hello␤»
20:38 timbunce_ joined
moritz_ jnthn: sure; I'll just retitle the ticket 20:38
PhatEddy std: my $ = 2; say 'hello'
20:38 timbunce left
p6eval std 32093: OUTPUT«ok 00:01 118m␤» 20:38
20:39 timbunce_ is now known as timbunce
jnthn moritz_: I'll probably be lazy and let STD fix it first. ;-) 20:39
pugssvn r32094 | colomon++ | [t/spec] Add some new series operator tests.
moritz_ wfm.
PhatEddy rakudo: my $ = 2; say $ 20:40
jnthn loves having blog stats
p6eval rakudo 689bb7: OUTPUT«===SORRY!===␤Non-declarative sigil is missing its name at line 22, near "$"␤»
jnthn std: my $ = 2; say $
p6eval std 32093: OUTPUT«===SORRY!===␤Non-declarative sigil is missing its name at /tmp/bl2JeNupzH line 1:␤------> my $ = 2; say ⏏$␤Confused at /tmp/bl2JeNupzH line 1:␤------> my $ = 2; say $⏏<EOL>␤ expecting twigil␤Other potential difficulties:␤
..…
moritz_ jnthn: so, how many visitors did your 6model post attract?
jnthn Just 192 so far.
moritz_ probably too young to have number of RSS followers 20:41
jnthn moritz_: But after 2 days its the second most popular post.
moritz_: Yes, it's only the 4th post on the blog.
It's quite young.
moritz_: I haven't got any historical stats to compare it to, of course.
dalek kudo: 819e2ce | colomon++ | src/core/List.pm:
Remove List.batch method, as it is obsolete.
 20:42
20:42 molaf joined
moritz_ jnthn: my p6 blog posts usually attract between a few hundred and 2k visitors 20:42
many of them through planetsix and perlphere.net 20:43
PhatEddy jnthn: you noticed the earlier $ = 2 example that std and rakudo took. Is that the one you're going wait for std on? Is it ok?
moritz_ perlsphere.net actually
jnthn PhatEddy: No, I was talking about the one moritz_ found in signature parsing.
PhatEddy: The my $ = 2; is silly but OK. 20:44
moritz_ std: sub f(:$x, :$y!) { }
p6eval std 32093: OUTPUT«ok 00:02 117m␤»
jnthn PhatEddy: It's a degenerate case of things like my ($a, $, $b) = foo(); where you don't want the second thing stored anywhere. 20:45
moritz_ std: sub f($x?, $y) { } 20:46
p6eval std 32093: OUTPUT«===SORRY!===␤Can't put required parameter after optional parameters at /tmp/FIXs6hnsIk line 1:␤------> sub f($x?, $y⏏) { }␤Potential difficulties:␤ $x is declared but not used at /tmp/FIXs6hnsIk line 1:␤------> sub f(⏏$x?, $y) {
..…
PhatEddy rakudo: say $.perl 20:47
p6eval rakudo 689bb7: OUTPUT«Lexical 'self' not found␤ in main program body at line 2:/tmp/LIXuBOBNea␤»
PhatEddy rakudo: say Any.perl
p6eval rakudo 689bb7: OUTPUT«Any␤»
jnthn std: say $.perl 20:48
p6eval std 32093: OUTPUT«ok 00:03 116m␤»
jnthn We could whine about that one I guess.
20:50 macroron left
moritz_ we do :-) 20:51
jnthn moritz_: :P 20:52
moritz_: I was thinking more, like, compile time-y, and helpfully. :-)
PhatEddy Um ... what should say $.perl have done? 20:53
jnthn PhatEddy: It's equivalent to self.perl
PhatEddy: But outside of a method, there is no self. 20:54
PhatEddy rakudo: say self.perl
p6eval rakudo 689bb7: OUTPUT«(timeout)»
jnthn rakudo: say self.perl 20:55
PhatEddy rakudo: say 'hello'
p6eval rakudo 689bb7: OUTPUT«===SORRY!===␤Symbol 'self' not predeclared in <anonymous> (/tmp/YyC4Me9q8E:22)␤»
rakudo 689bb7: OUTPUT«hello␤»
dalek ok: 98d51fe | moritz++ | src/subs-n-sigs.pod:
[subs] positional parameters come before named ones
20:56 molaf left
PhatEddy If it's ok I'll try to research a bit further and file the critter ?? 20:56
if needed
20:57 drake1 joined
jnthn PhatEddy: I'm all for a ticket that proposes making those two fail more consistently/helpfully. 20:57
I'm even happier with a patch that does so. ;-)
drake1 now I wrote some more perl pastebin.ca/1923627 20:58
how would that be perl6?
moritz_ drake1: are you sure that does what you want it to do? 20:59
drake1 yes
moritz_: but only with mode=0
moritz_ ah, unpack has a ($$) prototype 21:00
I thought it would slurp up the whole file in line 04
drake1 in perl6?
no I set the record size with $/
21:00 guidj0s_ left 21:01 guidj0s_ joined
moritz_ no, perl 5 21:01
oh right
well
still, if it were in list context, it would slurp the whole file
which was my point of confusion 21:02
drake1 is it possible to compare two arrays with implicit iteration ?
for exact identical values
21:02 guidj0s_ left
jnthn Maybe try @a >>===<< @b 21:02
drake1 unpack only read in a single record
moritz_ or @a == @b && @a >>===<< @b 21:03
because @a >>===<< @b dies if they have different lengths
jnthn oh, true 21:04
21:04 sekimura joined
moritz_ -> sleep 21:05
drake1 output snip pastebin.ca/1923644
PhatEddy rakudo: my (Any, @x) = (3, 10); say "@x" 21:06
p6eval rakudo 689bb7: OUTPUT«@x␤»
21:06 dju left
PhatEddy rakudo: my (Any, @x) = (3, 10); say "{@x}" 21:06
p6eval rakudo 689bb7: OUTPUT«3 10␤» 21:07
PhatEddy rakudo: my ($, @x) = (3, 10); say "{@x}"
p6eval rakudo 689bb7: OUTPUT«10␤»
drake1 PhatEddy: I thought about comparing @magic with ( 0x32,0x04) 21:08
Tene moritz_: use the angles the other way on something you want autoextended 21:10
21:10 pyrimidine left
drake1 if (@magic == (0x32,0x04) ) { # PSF1 format should work? 21:12
they're always the same length
Tene rakudo: say "foo" if (1,2) == (3,4); 21:13
p6eval rakudo 689bb7: OUTPUT«foo␤»
drake1 hmm
Tene that casts both sides to a scalar and does a numeric comparison on the result.
rakudo: say "foo" if (1,2) >>==<< (3,4);
p6eval rakudo 689bb7: OUTPUT«foo␤»
Tene Oh... 21:14
drake1 so you still have to iterate manually?
Tene rakudo: say "foo" if all((1,2) >>==<< (3,4));
drake1 explicitly*
p6eval rakudo 689bb7: ( no output )
Tene rakudo: say "foo" if all((1,2) >>==<< (1,2));
p6eval rakudo 689bb7: OUTPUT«foo␤»
Tene rakudo: say "foo" if all((1,2) >>==<< (1,3));
p6eval rakudo 689bb7: ( no output )
Tene So, that works fine for you.
drake1 nice
21:14 perlygatekeeper left, PhatEddy left
Tene the >> << make the operation run on individual elements from each list, and produces a list of the results. 21:15
gfldex std: my @a = 1..3; say [?] @a; 21:16
p6eval std 32094: OUTPUT«===SORRY!===␤Prefix requires an argument at /tmp/wJwz1P3CBl line 1:␤------> my @a = 1..3; say [?⏏] @a;␤Parse failed␤FAILED 00:01 119m␤»
jnthn [op] needs an infix
drake1 I could always say: if("@magic" eq "@psf1") I guess 21:18
21:18 justatheory joined 21:19 ruoso left
drake1 or ultimately pack a binary string 21:19
21:21 eternaleye_ joined
drake1 if pack ("CC",(0x32,0x04)) eq <FONT> 21:21
21:22 eternaleye_ left
drake1 better get the X11 pixmap toggle I wrote a bit earlier combined with this PSF reader. then it's gonna be a nice 'r' and a nice EUR for ISO8859-1[56] 21:28
21:30 timbunce left
x3nU i have written simple module 21:31
but
because lack of tutorials it's really ugly 21:32
and i don't know how to fix it
so i hope someone helps
github.com/xenu/Net-Daytime
21:34 eternaleye left, eternaleye joined 21:35 timbunce joined 21:37 dju joined 21:40 dju left, dju joined 21:41 Trashlord left 21:49 drake1 left
gfldex x3nU: timeZoneToTimeZone would want to be a hash if i would have wrote it :) 21:51
x3nU gfldex: makes sense 21:52
21:54 tylercurtis joined
gfldex x3nU: And that regex don't got any names in it. You might want to have a grammer instead and export it for other folk to use. 21:56
x3nU could show me example? :) 21:57
could you*
tylercurtis wishes a bit that he didn't have somewhere to go earlier during the removing roles discussion. 21:58
x3nU seriously, perl6 *really* needs a good tutorial, it would bring a lot of supporters 21:59
22:01 Mowah left 22:02 tadzik joined
gfldex x3nU: another note to that hash thingy. If some country comes up with the brilliant idea to introduce another timezone. Only 10 years or so later the ISO will have a name for it and your package would have to updated or subclassed. Updating that hash could be done even at runtime, without any hassle. 22:02
tadzik oh hai
22:02 Mowah joined, Mowah left
gfldex and there are tutorials :) 22:02
x3nU btw timezones
gfldex rosettacode.org/wiki/Category:Perl_6
22:02 Mowah joined
x3nU thing that pisses me off is duplicated timezone acronyms 22:02
gfldex github.com/perl6/perl6-examples 22:03
x3nU anyway i'm going to fix my module tommorow
22:03 alester left
x3nU thanks for gfldex 22:03
thanks for help gfldex *
22:03 Mowah left
TimToady std: /\pL/ 22:03
x3nU good night :)
p6eval std 32094: OUTPUT«===SORRY!===␤Unsupported use of \pL; in Perl 6 please use <+isL> at /tmp/eK_kFp8ncS line 1:␤------> /\pL⏏/␤Parse failed␤FAILED 00:02 115m␤» 22:04
22:06 colomon left, Eevee joined 22:10 colomon joined
TimToady rakudo: say "foo" if (1,2) eqv (1,2) 22:14
p6eval rakudo 689bb7: OUTPUT«foo␤»
TimToady rakudo: say "foo" if (1,2) eqv (1,3)
p6eval rakudo 689bb7: ( no output )
TimToady I dunno why you guys are futzing with >>===<< 22:15
22:16 fod joined
sbp original question was: 22:16
22:01 <drake1> is it possible to compare two arrays with implicit iteration ?
22:02 <drake1> for exact identical values
TimToady and in what sense does eqv not do implicit iteration?
rakudo: say "foo" if (1,2) ~~ (1,2) # also works, unless RHS contains wildcard 22:18
p6eval rakudo 689bb7: OUTPUT«foo␤»
22:18 am0c joined 22:21 wamba left
Tene TimToady: i didn't remember eqv. 22:24
22:27 timbunce left 22:32 whiteknight joined 22:35 tylercurtis left 22:40 xinming left, xinming joined 22:45 shade\ left 22:46 lest_away is now known as lestrrat 22:48 shade\ joined 22:49 risou joined
cognominal this may be of interest to people here : code.google.com/p/es-lab/wiki/Traits 22:55
23:01 lue left 23:07 shade\ left
cognominal voilà qui m'avait échappé car je ne lis guère linuxfr mais que je trouve via une recherche google : linuxfr.org//2010/07/21/27151.html 23:08
23:11 shade\ joined 23:14 xinming left
cognominal oops, wrong window again :( 23:14
23:15 xinming joined
mberends cognominal: nice all the same :) 23:16
"une affaire de mugs" :-) 23:17
cognominal masak++ 23:23
23:26 Pavlz joined
Pavlz hello 23:26
do you know which license is used by perl6 ? 23:27
thanks
iw0fzw paolo
gfldex Pavlz: your question is not that easy to answer 23:28
jnthn Pavlz: Depends what you mean by "Perl 6", since it's a spec, and implementations of it are free to use whatever license they wish.
mberends Pavlz: Rakudo has the same as Perl 5, Artistic License 2.0. Other implementations vary.
jnthn Pavlz: One popular implementation is Rakudo, and its license is github.com/rakudo/rakudo/blob/master/LICENSE
Pavlz why is used a camel as logo ? 23:29
is funny, but i am curios to know 23:30
mberends Pavlz: O'Reilly trademarked it as part of their series of book cover pictures.
gfldex that camel is a butterfly
for perl 6 that is
Pavlz camel is not a butterfly 23:31
gfldex indeed
Pavlz camel is an animal which walk in the desert
gfldex i
23:32 crythias joined
gfldex i'm not entirely sure you understood me as i wanted to 23:32
mberends Pavlz: camel => O'Reilly book, onion => Perl, butterfly => Perl 6
Pavlz comes out on Artistic License 2.0 23:34
This license is a free software license, compatible with the GPL thanks to the relicensing option in section 4(c)(ii).Artistic License 2.0
and this is described on Artistic License 2.0 23:35
www.gnu.org/licenses/licenses-list....reLicenses 23:36
so Artistic License 2.0 is compatible with the GPL 23:37
mberends Pavlz: yes
Pavlz and i am happy for this
in which contest can be used PERL ? 23:38
wath is its own point of force ? for what is really good ?
mberends Pavlz: mainly text processing, "glue" between applications, and thousands of other uses 23:39
Pavlz one example
mberends Pavlz: most examples are in search.cpan.org/ 23:40
gfldex the BBC uses perl for most of their websites 23:41
mberends Pavlz: if you're interested in learning, see www.perl101.org/
Pavlz and how you know to do this about BBC ?
23:41 Trashlord joined
gfldex Pavlz: are you using an online translation thingy? 23:42
Pavlz no
i am typing myself 23:43
and i made any error
i know this
but it's normal, cause i am english
gfldex i'm asking because i can't find meaning in your last question 23:44
23:45 drake1 joined
Pavlz i am italian, and is just good that i type in English, because there are many italian people, which don't know Italian, and you think that to know English is more important when people, don't speak the language of their own country ? 23:45
23:46 draxil left
drake1 in perl6, can you subclass without blessing the initial class like this pastebin.ca/1923798 ? 23:47
Pavlz in italy there are 6.000.000 ofilliterate,and you think that English is the primary focus ? 23:48
23:48 ruoso joined
mberends Pavlz: your English is very much better than my Italian. And many of us enjoyed a very well run conference in Pisa recently: conferences.yapceurope.org/ye2010/ 23:50
Pavlz when there are 6.000.000 people of illeterate, you must ask to yourself if Berlusconi won for its own program, or for the fact that people don't know him in the IRL = In the Real Life
gfldex Pavlz: i did not intend to offend you and am very sad that my english does not allow me to be as gentle as mberends 23:51
Pavlz i am happy for your conference, but i don't know if you have seen what this the real situation here in Italy
you are not offending me 23:52
TimToady however, this seems somewhat OT for #perl6
Pavlz yes, is OT for #perl6 23:53
TimToady it's okay, as long as it's not interfering with people trying to work
mberends
.oO( there's that TimToady++ sending us back to work again ;)
 23:54
TimToady but some of us do feel like we have to read everything here
23:55 draxil joined
drake1 if you actually called the constructor through a sub class, how would perl6 manage to associate the new tier in the chain? 23:56
TimToady drake1: in what sense are you subclassing? I just see functional wrappers around an anonymous
hash
drake1 typical
23:56 lestrrat is now known as lest_away
jnthn class Cat { }; class Lolcat is Cat { } # this is how to subclass in Perl 6 23:57
And the default inherited new constuctor will do the right thing.
So will your own new method and self.bless(...)
drake1 jnthn: I mean how the constructor in a parent class would associate the sob class with the object 23:58
jnthn: the class name of the sub class still has to be blessed?
jnthn drake1: No, you don't have to do that in Perl 6. The self in the constructor in the parent class is a type object of the subclass, and that's how it knows what to bless into. 23:59

Powered by Cro and the Raku® Programming Language.

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

Thank you!