pugscode.org/ planetsix.perl.org/ | nopaste: sial.org/pbot/perl6 | evalbot: perl6: say 3; (or rakudo:, pugs:, elf:, etc) | irclog: irc.pugscode.org/
Set by mncharity on 5 January 2009.
00:10 aindilis left 00:11 wknight8111 joined
ovid OK. Time to hit the sack. Night all! 00:12
00:12 ovid left 00:18 pbuetow left 00:29 Limbic_Region left 00:30 c9s_ is now known as c9s 00:31 Caelum left, alc joined 00:34 Caelum joined, kane_ left 00:35 wknight8111 left, Limbic_Region joined, wknight8111 joined 00:36 |jedai| left, |jedai| joined 00:37 hudnix left 00:39 |jedai| left, hercynium_ joined, hudnix joined 00:55 meppl joined
meppl good morning 00:56
00:58 nipotan joined 01:21 [particle] joined 01:22 lambdabot left 01:25 lambdabot joined 01:40 [particle]1 left 01:49 stephens left 01:53 DemoFreak left 02:10 eternaleye joined 02:21 mj41_ joined 02:28 Alias_ joined 02:37 mj41 left 02:43 Exodist left 02:59 wknight8111 left 03:06 meppl left 03:10 GeJ joined 03:11 hercynium_ left 03:23 cspencer joined 03:28 preflex left 03:30 preflex joined 03:32 [particle]1 joined 03:44 GeJ left 03:46 Limbic_Region left 03:50 [particle] left 04:01 elmex_ joined 04:12 stephens joined 04:17 elmex left, elmex_ is now known as elmex 04:18 unobe joined 04:19 sahadev joined
sahadev ?eval grep {$_ % 2 == 0}. 1..10 04:25
04:31 alc left, sahadev left 04:33 |jedai| joined 04:41 alester left 04:43 sahadev joined
sahadev ?eval 42.say 04:44
hello, anybody here?
TimToady perl6: 42.say 04:48
p6eval elf 24892, pugs, rakudo 35476: OUTPUT«42␤»
TimToady perl6: grep {$_ % 2 == 0}. 1..10 04:49
p6eval rakudo 35476: OUTPUT«Statement not terminated properly at line 1, near ". 1..10"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
..pugs: OUTPUT«*** ␤ Unexpected " 1"␤ expecting ".", "\187", ">>", "=", "^", operator name, qualified identifier, variable name, "...", "--", "++", "i", array subscript, hash subscript or code subscript␤ at /tmp/vfo483sjPN line 1, column 20␤»
..elf 24892: OUTPUT«Parse error in: /tmp/KaAiHB3qza␤panic at line 1 column 18 (pos 18): Statement not terminated properly␤WHERE: grep {$_ % 2 == 0}. 1..10␤WHERE: /\<-- HERE␤ STD_red/prelude.rb:99:in `panic'␤ STD_red/std.rb:355:in `eat_terminator'␤ STD_red/std.rb:269:in `block in
.....
TimToady perl6: grep {$_ % 2 == 0}, 1..10
p6eval rakudo 35476: RESULT«[2, 4, 6, 8, 10]»
..elf 24892: OUTPUT«Undefined subroutine &GLOBAL::infix__37 called at (eval 121) line 3.␤ at ./elf_f line 3861␤»
..pugs: RESULT«(2, 4, 6, 8, 10)»
cspencer TimToday: i've got a question regarding one of the things you said in the last minutes posted on rakudo.org
TimToady whazzat? 04:50
cspencer you'd said "the character class should not pay attention to magical increment" - just to be clear, what do you mean by that?
is that as in the ++ magical increment?
ie) $x = 'A'; $x++
TimToady for example, / <[ \0 .. \xff ]> / should match any chr from 0..255 04:51
cspencer right, ok
sahadev perl6: grep {$_ % 2 == 0}, 1..10
cspencer am i missing the magical increment part? :)
p6eval rakudo 35476: RESULT«[2, 4, 6, 8, 10]»
..elf 24892: OUTPUT«Undefined subroutine &GLOBAL::infix__37 called at (eval 121) line 3.␤ at ./elf_f line 3861␤»
..pugs: RESULT«(2, 4, 6, 8, 10)»
cspencer (or perhaps i'm unclear as to where the magic part lies) :) 04:52
TimToady perl6: my $x = 'Z'; say ++$x
p6eval elf 24892: OUTPUT«No viable candidate for call to multimethod prefix__43_43($) at (eval 119) line 4␤ at ./elf_f line 3861␤»
..pugs, rakudo 35476: OUTPUT«AA␤»
TimToady note AA, not [
cspencer right, ok 04:53
TimToady that's magical
cspencer ah, ok, i understand now
thank you :)
sahadev thanks TimToady, for pointing out the evalbot syntax. 04:54
cspencer perl6: my $x = "\xff"; say ++$x
sahadev (i should have read the topic, of course)
p6eval elf 24892: OUTPUT«No viable candidate for call to multimethod prefix__43_43($) at (eval 119) line 4␤ at ./elf_f line 3861␤» 04:55
..pugs: OUTPUT«Ā␤»
..rakudo 35476: OUTPUT«ÿ␤»
TimToady ooh, rakudo bug
04:55 meppl joined
cspencer that does look a little different than pugs, doesn't it! 04:55
TimToady it incremented 255 and got 255 04:56
cspencer perl6: my $x = "\xff"; say $x
p6eval elf 24892: OUTPUT«\xff␤»
..pugs, rakudo 35476: OUTPUT«ÿ␤»
cspencer indeed, rakudo was lacking in the magic department on that one
TimToady but I suppose ++ is undefined outside of official ranges
cspencer there are defined cases in which the ++ makes sense, correct? which synopsis is that defined in? 04:57
perl6: my $x="\x0"; say --$x 04:58
p6eval elf 24892: OUTPUT«No viable candidate for call to multimethod prefix__45_45($) at (eval 119) line 4␤ at ./elf_f line 3861␤»
..rakudo 35476: OUTPUT«
..pugs: OUTPUT«-1␤»
cspencer is \x used to defined a character? 04:59
05:00 elmex left
cspencer as indexed by a hex number 05:00
TimToady S03:396
cspencer thanks 05:01
s1n are ranges not allowed to be reversed?
TimToady ranges never auto-reverse in Perl 05:02
s1n rakudo: (1..10).reverse.perl.say
p6eval rakudo 35476: OUTPUT«[10, 9, 8, 7, 6, 5, 4, 3, 2, 1]␤»
s1n hmm, that worked there...
not auto-reverse, just reversable via .reverse
TimToady \x works in double quotes and such
cspencer s1n: i think that worked because it was converted to a list by the range's reverse method
is that how it should display in perl6 though? 05:03
or should it output as: 10..1:by(-1) ?
is there a criteria which specifies when it should flatten? 05:04
TimToady the subject of context of .perl is still a bit problematic in various ways
cspencer is that for lack of specification at the moment? 05:05
TimToady more like not knowing what the eventual context will be 05:06
cspencer ah 05:07
TimToady arguably, that means all .perl should be returned in the form of a Capture, when in doubt
05:11 alc joined
cspencer are Captures specced out in a synopsis? 05:12
i've seen them appear more frequently as of late 05:13
but haven't found a detailed discussion of them
TimToady spack: Capture 05:17
hmm, probably got that wrong... 05:18
cspencer spack is a bot of some sort, i take it?
05:18 aindilis` left
cspencer though not one with timely responses, perhaps :) 05:19
TimToady anyway, S02, S03, S04, S05, S06, S09, S12, and S29
to begin with... :)
cspencer hah, ok, i'll review those then for my enlightenment :) 05:20
05:23 cspencer left
s1n how do i convert a Num to a Str? 05:28
TimToady perl6: ~42.000 05:30
05:31 stephens left
p6eval elf 24892: RESULT«42␤» 05:31
..pugs, rakudo 35476: RESULT«"42"»
s1n thanks
TimToady perl6: Str(42.000)
p6eval rakudo 35476: OUTPUT«invoke() not implemented in class 'Perl6Str'␤current instr.: '_block14' pc 66 (EVAL_17:41)␤»
..pugs: OUTPUT«*** No such subroutine: "&Str"␤ at /tmp/OihDiC6bXO line 1, column 1 - line 2, column 1␤»
..elf 24892: OUTPUT«Undefined subroutine &GLOBAL::circumfix_S_32_41 called at (eval 117) line 3.␤ at ./elf_f line 3861␤»
s1n is doing that supposed to be valid? 05:32
TimToady *nod*
05:32 dukeleto joined
s1n TimToady: where can i read about more about Str? 05:33
TimToady any type may be used that way
s1n it seems split on s02 and s29
yeah, but i'm wanting to look at what i can do with Str though 05:34
TimToady Typename($x) is simply a coercion 05:36
s1n TimToady: okay, i've tried that before and i saw it didnt work, but i want to know what methods Str has and what I can call against it 05:37
ouch
TimToady most methods are defined in S29 currently
s1n oops heh, wrong window
okay, thanks
TimToady though it's kinda mislabelled Functions
It's more like Methods
hard to know how to organize it all though 05:38
s1n well, it'd be nice if i say "gee, i'm trying to do something with Num, here's all the things Num supports" with outlinks to other sections 05:39
like how Num can be used with Regex, IO, etc
i'm having a hard time with some things not quite dwim. it may be the impl but it's every harder searching through the spec 05:42
rakudo: my $var = 10; for (1..$var).reverse -> $t { say ~$t.split(/\d/).perl }
p6eval rakudo 35476: OUTPUT«["", "", ""]␤["", ""]␤["", ""]␤["", ""]␤["", ""]␤["", ""]␤["", ""]␤["", ""]␤["", ""]␤["", ""]␤» 05:43
05:43 [particle] joined 05:47 sahadev left
TimToady rakudo: my $var = 10; for (1..$var).reverse -> $t { say (~$t).comb(/\d/).perl } 05:48
p6eval rakudo 35477: OUTPUT«["1", "0"]␤["9"]␤["8"]␤["7"]␤["6"]␤["5"]␤["4"]␤["3"]␤["2"]␤["1"]␤» 05:49
TimToady rakudo: my $var = 10; for (1..$var).reverse -> $t { say $t.comb(/\d/).perl } 05:50
p6eval rakudo 35477: OUTPUT«["1", "0"]␤["9"]␤["8"]␤["7"]␤["6"]␤["5"]␤["4"]␤["3"]␤["2"]␤["1"]␤»
TimToady rakudo: my $var = 10; for reverse 1..$var -> $t { say $t.comb(/\d/).perl } 05:51
p6eval rakudo 35477: OUTPUT«["1", "0"]␤["9"]␤["8"]␤["7"]␤["6"]␤["5"]␤["4"]␤["3"]␤["2"]␤["1"]␤»
05:57 Trey left, Trey joined 05:59 [particle]1 left 06:10 mberends left 06:31 justatheory left 06:37 maerzhase joined 06:52 ashizawa joined 06:54 maerzhase left 06:58 namenlos joined 06:59 pdcawley joined 07:28 maerzhase joined 07:30 DemoFreak joined 07:34 meppl left 07:39 namenlos left 07:40 namenlos joined 07:42 maerzhase left, maerzhase joined, maerzhase left, maerzhase joined 07:43 maerzhase left 07:47 vixey joined 07:51 iblechbot joined 08:02 masak joined 08:04 [particle]1 joined 08:17 Cantacuzenus joined 08:22 [particle] left
masak rakudo: class A { my $.x = 100; my $.y = $!x * 1.05 }; say A.y 08:37
lambdabot masak: You have 1 new message. '/msg lambdabot @messages' to read it.
p6eval rakudo 35480: OUTPUT«Lexical 'self' not found␤current instr.: 'parrot;A;_block20' pc 137 (EVAL_20:62)␤»
masak @massage
lambdabot azawawi said 2d 12h 12m 39s ago: use.perl.org/~azawawi/journal/38255
masak @clear
lambdabot Messages cleared.
moritz_ @clear 08:38
lambdabot Messages cleared.
moritz_ that's what I was missing so far :-)
masak don't know if that helps, though.
ok, today's question: can a module set $_ of the 'use'-ing module? 08:39
08:39 pmurias joined 08:41 maerzhase joined 08:45 _jedai_ joined, |jedai| left
Matt-W Oh that's how lambdabot's messages work 08:46
masak: my instinct says 'it should be possible to do that' 08:47
masak if I knew you were all wondering, I'd have told you sooner... :)
Matt-W: how?
Matt-W umm
Matt-W waves his hands around
masak I'm sorry, that's not enough. :) 08:48
I'm writing real code, and this is a real question.
Matt-W Oh
Well I don't know how it would have been implemented if it's been implemented
masak I'm more interested in how it could be done at all.
does $OUTER::_ cover this case?
if not, what does? 08:49
Matt-W It would be some sort of special name for 'the module that used me'
moritz_ masak: not $OUTER::_, but $CALLER::_
masak moritz_: great, thanks.
Matt-W Yes, something like that :P
moritz_ but I doubt that rakudo supports this notation 08:50
masak it doesn't.
I can always submit a TODO ticket.
masak does that 08:51
I wish I could tell you what wonderful new thing I need this for.
moritz_ he :-)
masak but I think you'll just have to wait like everybody else.
Matt-W Damn
I keep trying to write Perl 6
But this code's in Perl 5... 08:52
moritz_ masak: I know that too well :-)
erm, Matt-W :)
Matt-W oh how much nicer this would be if I could just say @.refreshes.end
moritz_ damn, ma<tab> is ambigious, I have to pay more attention
Matt-W that's okay, I'm doing it too :)
I keep hitting m<tab> 08:53
and there are quite a few people with m in here
masak it's a popular letter. 08:54
Matt-W Well I suppose it could be worse. At school when I was 6, there were three other Matthews in my class
masak Matt-W: what do you mean? you _can_ say '@.refreshes.end'
Matt-W masak: not in Perl 5
moritz_ but not [*-1], I suppose
masak Matt-W: ah, now ISWYM.
Matt-W moritz_: that wouldn't be the same thing, in this case I need the last index 08:55
moritz_ is still ignorant
why doesn't @.refreshes.end work?
Matt-W Because I'm coding in Perl 5
in Rakudo it's fine
moritz_ ah
Matt-W I've used it, it's lovely
Unfortunately I'm currently at work, and using Perl 5 08:56
masak Matt-W: if you're not using Perl 6, you're off topic :P
Matt-W And admiring the beauty of Perl 6 in comparison
It really is a better language
moritz_ Matt-W: maybe Moose and Autobox can get you there
masak Moose++
09:01 _jedai_ left
masak is defined a method on Object or Any? i.e. do junctions autothread over .defined or not? 09:01
rakudo: say (undef&2).defined
p6eval rakudo 35480: OUTPUT«./parrot: error while loading shared libraries: /home/evalenv/parrot/blib/lib/libparrot.so.0.8.2: invalid ELF header␤»
masak oh, it's that time of the hour. 09:02
09:02 |jedai| joined
Matt-W it's rebuilding? 09:03
masak aye.
Matt-W In a few years I'm going to spend every coding moment appreciating <> on hashes... 09:04
masak rakudo: say (undef & 2).defined
p6eval rakudo 35480: OUTPUT«1␤»
masak so, on Object, then.
if Rakudo is anything to go by.
moritz_ that looks... wrong 09:05
oh noes
masak moritz_: :)
moritz_ it can't be defined in Any
Matt-W isn't it just saying the junction object is defined?
moritz_ yes
and I was expecting it to autothread
Matt-W so you were expecting 0&1
masak it would if it were defined in Any.
Matt-W has flashbacks to quantum computing 09:06
moritz_ masak: but then undef.defined wouldn't work
masak: because undef is an Object
masak moritz_: guess not. it would have to have a separate definition.
or something. I dunno.
Matt-W by the way, is there any movement towards being able to have junction types for object attributes? 09:07
masak Matt-W: in the Perl 6 spec? isn't it allowed already? 09:08
std: class A { has Int|Str $.foo }
p6eval std 24892: OUTPUT«00:05 83m␤» 09:09
09:19 cott0 left 09:27 donaldh joined 09:32 pasteling left 09:33 pasteling joined
Matt-W masak: I mean as implemented in Rakudo 09:33
masak: currently it can't handle them
masak Matt-W: ah.
Matt-W I guess jnthn's the person for that
masak an #parrot's the channel.
s/an/and/
09:36 |jedai| left 09:37 |jedai| joined, meppl joined 09:41 riffraff joined 09:42 pmurias left, jimmy_ joined, namenlos left 09:43 namenlos joined
jimmy_ hello masak, a question 09:43
masak hi jimmy_.
let's hear it.
jimmy_ how can I use jimmy both #parrot and #perl6? 09:44
masak jimmy_: sometimes names are registered on freenode. then you can't use them.
your IRC client will usually tell you.
09:48 Southen_ joined
masak jimmy_: did that help you? 09:49
jimmy_: it seems to me that the nick 'jimmy' is already taken on freenode.
jimmy_ I don't know. but I can use jimmy once I left #parrot 09:52
09:52 jimmy_ left 09:53 jimmy_ joined
masak jimmy_: no, whether you're on #parrot or not shouldn't affect the situation. 09:53
but if the real 'jimmy' (on freenode) is logged in, you can't take his nick.
09:53 Southen left, jimmy_ left
moritz_ just do a '/whois jimmy', and you'll see that the nick is taken on freenode. You're out of luck. 09:54
what do you think why I have a _ at the end of my nick?
masak moritz_: he's gone for the moment.
09:54 jimmy_ joined
masak jimmy_: and if he's not, you can only hold it temporarily. 09:54
jimmy_: so I suggest you either register 'jimmy_' or choose another nick here on freenode. 09:55
jimmy_ ah, I had used jimmy ever. but I can't use it now.
masak jimmy_: (you can't use 'ever' like that -- bu keyi. you need to use 'once') 09:56
jimmy_ once
masak (not that it matters a lot. both are understandable.)
jimmy_ I used jimmy ever. is it right? 09:57
masak jimmy_: no.
jimmy_: "I haven't ever used jimmy".
jimmy_: "I have used jimmy once."
silly, I know. but that's the way it works. 09:58
jimmy_ i know now, ever usually used to be question
masak jimmy_: oh, right!
jimmy_: as in "Do you ever help around here?" 09:59
jimmy_ She was happier than she had ever been.
so I can say 'I had ever used it'?
masak yes, there it means 'than all previous moments'. 'ever' sort of means 'always' sometimes. 10:00
jimmy_: not standalone, no.
jimmy_: but you can say 'I wondered if I had ever used it'.
jimmy_ 'I had ever used it', it is wrong?
masak yes, it sounds funny. 10:01
moritz_ it's easy for me, because we have the same word in German ("jemals"), with the same oddities
you'd say "I have used it once" instead
jimmy_ yes, sometimes it is funny, and so does chinese.
masak it Swedish, it's "någonsin"
jimmy_ s/does/is/ 10:02
moritz_ that also sounds funny :-)
masak jimmy_: Chinese is much less worse in this regards, if you ask me.
jimmy_ yep, Chiense is hieroglyph. 10:03
lumi_ jimmy_: Do you mean that Chinese is funny, or that it also has this distinction?
jimmy_ nope, on the other hand.
masak jimmy_: hieroglyph is not the term. that's for ancient Egyptian. 10:04
lumi_ Ideogram, I think?
masak lumi_: well, some of the characters are. not all.
lumi_ masak: Some aren't ideograms?
masak lumi_: an ideogram is a character that conveys an idea.
lumi_: some are pictograms, essentially painting the thing. 10:05
lumi_: and some are sound-borrowings from unrelated characters.
jimmy_ it is Ideogram at one time
masak lumi_: many are combinations of other characters.
jimmy_ or pictograph
masak jimmy_: right, sorry. that's the term.
I've seen the term 'sinograph' for chinese characters. I kinda like it. 10:06
of course, you can say 'hanzi' if you expect to be understood.
moritz_ what I thought about yesterday (and now we're 100% off-topic): since in Mandarin the tone (or the change of tone) carries information, can you sing in Mandarin to arbitrary meldodies, and still preserve the meaning? 10:07
jimmy_ yes.
masak moritz_: the dependence on subtitles or text sheafs is greater, to be sure.
moritz_: however, if you keep within the classical themes such as life, love and death, I don't think that it's a problem. 10:08
jimmy_ tone carries feeling.
but not meaning.
masak moritz_: it's also a nice trick to avoid learning the spoken tones perfectly, because the sung tones take precedence. I've done it. :)
lumi_ masak: So it can sometimes be that song lyrics are ambiguous? Or that you can't use unexpected words in songs? 10:09
masak lumi_: I don't know about unexpected words, but it seems to me that context is very important in Mandarin Chinese in general, not just in songs.
moritz_ jimmy_: that's interesting, nobody ever told me that before
masak lumi_: ambiguity happens almost however you do it, but is most often saved by context. 10:10
lumi_ There's the Lion-Eating Poet in the Stone Den, (iirc) as one extreme...
jimmy_ moritz: tone also helps to catch the meaning. It is important too.
masak lumi_: yes, but that's not current Mandarin.
10:11 barney joined
masak moritz_: I don't really know what jimmy_ is trying to say, but tones definitely matter for parsing a word in Mandarin. 10:11
it would be so much easier if they didn't. :/
jimmy_ shoulde say reach, not catch.
hey barney.
a question
masak: tone is helped to reach the meaning. 10:12
lumi_ Actually: What's a good way to catch up with p6 progress? I've been out of the loop for the last year or more 10:13
masak think of it as three dimensions. one for consonants, one for vowels and one for tones. English only reserves the first two dimensions for the parsing of a word, but even in English, if someone says "attintion", you probably guess that "attention" was what was intended.
on a scale, the consonants are most important for catching the meaning, then the vowels, then the tones.
lumi_ (Sorry to ask on-topic questions :P)
masak lumi_: take a look at November the wiki.
github.com/viklund/november 10:14
moritz_ lumi_: maybe my "tidings" postings in perlgeek.de/blog-en/perl-6/ can be of some help
the cover only the last few months, and don't give a really high-level overview, but still...
lumi_ masak & moritz_: Thanks :) 10:15
masak lumi_: also, make sure to play Druid. it's fun. :)
github.com/masak/druid/
lumi_ moritz_: And in a word or two, where is most p6 work going on nowadays? Not pugs, right?
moritz_ lumi_: pugs is sleeping/dead. Rakudo (Perl 6 on parrot) makes the most progress
lumi_: rakudo.de/ shows its progress in the test suite 10:16
jimmy_ in English, If someone says 'attintion', then I don't know what he said. even he says 'attention', I still don't know what he said.
masak Pugs isn't dead, it's just sleeping.
10:16 vixey left
masak jimmy_: :) 10:16
10:16 ludan joined
jimmy_ It's hard to me to listen to someone in English. 10:17
barney: ping 10:19
masak: will perl6 support documentation i18n officially, or is there any schedule? 10:20
masak jimmy_: good question. what, specifically, are you interested in having? 10:21
i18n is a very broad subject.
moritz_ schedule? you're kidding :-)
masak our schedule is, and has always been, to deliver Perl 6.0.0 by Christmas.
moritz_ man, we don't even have English documentation, except from incomplete design documents and unmaintained, incomplete other stuff
masak we have stuck to that schedule for years now. 10:22
I don't see what people are complaining about.
moritz_ heh, nice point of view
10:22 cotto joined 10:24 [particle] joined
jimmy_ a old topic is that perl is not widely used in china, for lack of native documentation. 10:25
Matt-W masak: conveniently failing to specify which Christmas is a masterstroke 10:26
masak we should employ the 100 or so students who translated the last Harry Potter book in a matter of hours, and set them on translating perldoc.
Matt-W: it's not new, though.
Matt-W masak: true
moritz_ and probably just as old is the problem that most non-chinese don't speak it well enough, or care, to translate the documentation
masak Matt-W: but yes, it's a nice way of giving the right answer: "when it's done". 10:27
Matt-W moritz_: also, they say that for good translation you need a native speaker of the target language
jimmy_ and be misunderstood by a few news with prejudice
moritz_ usually the main developers can contribute the i18n infrastructure, but the actual translations need to come from however is interested in that language
masak aye.
moritz_ that why I proposed to work on an exception hierarchy/role model, because that's something that an English-Only speaker can do, but it enables i18n of error messages 10:28
which IMHO is a good foundation
when we force @implementors to think about i18n, we strengthen i18n culturally 10:29
10:29 cotto left
jimmy_ yes, perldoc is not native, that is the problem. 10:29
moritz_ masak: re 100 students, I know that wasn't entirely serious, but I'd still like to point out that it's not gonna work 10:30
I had troubles translating parts of S05 into German, and I'm relatively deep in the matter
so having 100 perl illiterate translators is going to buy us exactly nothing 10:31
masak moritz_: you're right. of course.
Matt-W That's the hard part
masak I'm not sure the HP translation turned out that well either.
Matt-W You need people who understand Perl 6 and have a very good grasp of their target language, and there just aren't that many such people
masak oh, and I'd say the synopses would be hard for anyone to translate into any language.
jimmy_ synopses and perldoc 10:32
Matt-W synopses aren't documentation though, really 10:33
masak first, we need something closer to perldoc than S29 is at present.
Matt-W we're going to need reference docs
masak I think that'll be one of my goals for 2009.
user documentation so crisp, that people will want to read it just for its literary qualities. 10:34
Matt-W cool
donaldh masak++
Matt-W I can get behind that
And help, maybe
jimmy_ perl5, translating synopses had been done in china, the problem is perldoc.
masak Matt-W: excellent.
donaldh I might be able to help there too. 10:35
10:35 alc left
masak Matt-W, donaldh: I'll make sure to put a README somewhere with plans, so that such work can be coordinated. I see no reason not to use the Pugs repo for this. 10:35
Matt-W I do quite enjoy writing explanations of things 10:36
moritz_ perldoc2.sourceforge.net/ already has the infrastructure for perl5
Matt-W And it'll ensure I have an excellent knowledge of Perl 6 :)
donaldh It's a good way of learning.
masak moritz_: I'll have a look at that. thanks.
moritz_ don't re-invent the weel unless you see a good reason
masak moritz_: what do you mean?
jimmy_ moritz_: it's not official 10:37
masak finds that implementing features in Rakudo is a good way to learn Perl 6
Matt-W well if your wheel has fallen off and is in several pieces, you might want to consider a different one...
10:37 cotto joined
moritz_ jimmy_: and it won't become official unless it improves very much 10:37
Matt-W masak: there is that, but one also has to understand Rakudo
jimmy_ moritz_: then it never become official
moritz_ masak: I think this site has tools that help with translatings (keeping track of paragraphs, or whatever), so I wanted to say "don't reinvent them because you didn't know they exist" 10:38
masak moritz_: my goals are more humble than perldoc2. I want to provide excellent Perl 6 documentation in English. but I'll definitely check perldoc2 out first.
moritz_ jimmy_: that's bad luck. But as an organization you can't declare a 50% done and partly out-of-date documentation "official" 10:39
jimmy_: that said, "official" is overrated. One of perl's greatest strengths is CPAN, and that's also not "offical" perl 10:40
and thinking that something will get more voluntueers because somebody declares it official isn't realistic either
Matt-W No, but it'd be nice to have something to point to with the 6.0.0 announcement and say 'here is the documentation' 10:41
jimmy_ moritz_: the core
not a third party.
10:41 [particle]1 left
masak Matt-W: that's what we're aiming for. let's make that the goal. 10:41
moritz_ jimmy_: what about it?
jimmy_ at least, the core lib documentation shoud be.
masak 'Everybody wants the core.'
moritz_ jimmy_: but again, how can you declare something official that doesn't exist? 10:42
jimmy_ as php.net
Matt-W masak: sounds like a good goal to me
moritz_ perldoc.perl.org ist just as official as php.net, I think
masak Matt-W: I'll have some time later today to write something about it, I hope.
jimmy_ php.net support it officially. 10:43
Matt-W masak: I look forward to reading it
barney jimmy: pong
masak Matt-W: will keep you posted.
jimmy_ and so does ubuntu.
Matt-W perldoc.perl.org is the documentation that comes with the Perl distribution, so surely it's as official as one can get?
moritz_ aye
masak yes. 10:44
jimmy_ barney: In PBC_COMPAT, there were two reduplicate line from you. I don't if there is a mistake.
Matt-W:but just english. 10:45
Matt-W oh you mean officially supported documentation in multiple languages
jimmy_ yes.
moritz_ jimmy_: but non-english docs can't become offical unless they exist.
Matt-W all that would need is for somebody to write them
moritz_ jimmy_: so first somebody would have to find contributors for them
Matt-W and maintain them 10:46
moritz_ and they aren't easy to find
masak i18n documentation is hard! let's go shopping.
jimmy_ actually, they wouldn't be translated completely . 10:47
Matt-W Impossible for me, I speak some German but nowhere near as much as would be necessary to write docs in German
barney jimmy: I'll add it to the Parrot 0.9.0 ticket, I don't want to invalidate PBC for that change 10:48
masak Matt-W: documentation translation should definitely be done by native speakers of the target language, I think.
jimmy_ like wiki, somesone tranlated one line and another one translated another line.
barney++ 10:49
jimmy_ is gonna home now.
10:49 ludan left, jimmy_ left
Matt-W masak: absolutely - you need that fluency to write comprehensible, idiomatic text 10:49
masak aye. 10:50
Matt-W So I can only write in en_GB :) 10:51
10:51 ludan joined, Alias_ left
masak and I'm not technically qualified to write in English at all, by that standard. but I think I've read enough manuals in my day to be able to put some sort of prose together, that can at least be polished by native English speakers. 10:52
10:53 Alias_ joined
Matt-W Well you're doing fine in here :) 10:54
And I'm not at all averse to polishing other people's text
10:54 elmex joined
masak Matt-W: thank you. looking forward to writing documentation. 10:55
Matt-W: have you seen use.perl.org/~masak/journal/38170 ?
the omissions list is quite high on my priority list. I hope to have it done by February. 10:56
help appreciated.
10:56 Alias_ is now known as AdamKennedy 10:59 AdamKennedy is now known as Alias_
Matt-W masak: I'll take a look 11:02
masak it's often things that need a question or clarification here at #perl6, or at p6l, and then they can simply be added to S29. 11:03
Matt-W my browser's currently stuck waiting on images.use.perl.org :( 11:04
masak quite pleasant work, and the result will be a good start for user docs.
Matt-W: well, the blog post contains no images, so you're not missing out. :)
Matt-W well unfortunately I have no text yet either
masak huh. 11:05
Matt-W it's probably just the work proxy being a pain
it does this sometimes
ahah, got it 11:07
masak: well some of those should be fairly easy, others will need more research, but it's nice to have a list of things to attend to 11:09
masak Matt-W: exactly.
Matt-W: if you decide to do one, do drop me a note. that way, there will be less double work. 11:10
and less S29 conflicts.
fewer* :)
Matt-W I will 11:11
I'll see if I can do one or two this evening
masak cool.
me too.
lunch & 11:18
11:18 masak left 11:33 brunoV joined 11:44 riffraff left 11:51 |jedai| left 11:52 |jedai| joined 11:57 ashizawa left 12:18 icwiener joined 12:30 |jedai| left, |jedai| joined 12:34 [particle]1 joined, iblechbot left 12:39 OuLouFu left, orafu joined
rakudo_svn r35485 | jonathan++ | [rakudo] For now, when we add_type we will put it not in the current blocks's symbol table, but instead we will search outwards to find the first one representing a package and put it there. This deals with the final regression test failure that enabling the type registration 12:40
..brings up. It does mean lexical classes etc would be too visible - aside from the fact that we don't actually implement them yet, so we'll work out a solution to this when we get there.
r35486 | jonathan++ | [rakudo] Enable the type registry. Causes no regressions in the spectests or sanity tests.
12:44 maerzhase1 joined 12:47 agentzh left 12:54 [particle] left 12:56 |jedai| left, |jedai| joined 12:59 maerzhase left
rakudo_svn r35487 | jonathan++ | [rakudo] Parse parametric role declarations. (Note - trying to write one will just cause a runtime explosion; this is just the parsing!) 13:10
13:12 alc joined 13:19 DemoFreak left 13:27 icwiener left 13:31 iblechbot joined 13:38 aindilis joined 13:41 Lorn joined 13:47 mberends joined
moritz_ rakudo: class a { method B { "works" } }; say a.new.B 13:50
p6eval rakudo 35487: OUTPUT«works␤»
moritz_ jnthn++
jnthn :-) 13:53
Matt-W Is that new?
jnthn moritz_: Do we have much in the way of spectests to unfudge?
Matt-W: Before you could only really write classes that started with an uppercase letter. 13:54
Matt-W aaah
jnthn And subs would only behave properly if you started them with a lowercase letter.
Matt-W nice
so we're no longer Java then :)
moritz_ jnthn: I'm running autounfudge right now
jnthn: but don't expect too many new tests, most tests are written to reflect the standard naming scheme 13:56
jnthn Aye.
moritz_: Maybe would be good to write some new ones. 14:07
moritz_ jnthn: I'll do, if autounfudge doesn't find anything 14:09
14:10 athenot left 14:11 athenot joined
jnthn moritz_: Excellent, thanks! 14:13
14:18 pmurias joined
Matt-W Always good to write more tests... 14:25
14:27 ejs joined 14:28 ejs left, ejs joined 14:35 aindilis left, Alias_ left, Alias__ joined, preflex left, aindilis joined 14:37 preflex joined
moritz_ Matt-W: feel free :-) 14:40
14:45 unobe left, alc left 14:47 mj41 joined 14:54 [particle] joined 14:59 eric256 joined, pmurias_ joined, pmurias_ left 15:01 mj41_ left 15:10 ruoso joined 15:12 [particle]1 left 15:17 Exodist joined, pmurias left
rakudo_svn r35489 | pmichaud++ | [rakudo]: spectest-progress.csv update: 282 files, 6233 passing, 0 failing 15:20
Matt-W 6,233, very nice 15:22
how many fudged though 15:23
moritz_ Matt-W: you can see that in the chart on rakudo.de 15:27
eric256 hey...who managed November?
moritz_ eric256: masak does
(and viklund)
Matt-W moritz_: doh, silly me 15:28
eric256 ah okay. i wanted to use their CGI.pm and didn't know the best way to do it, like fork theirs or...dunno
15:31 justatheory joined 15:34 aindilis left 15:43 mberends left 15:52 eternaleye left 15:54 rhr joined 16:07 vixey joined, rhr_ left 16:12 aindilis joined, ejs left, DemoFreak joined
diakopter moritz_: count on TimToady to use untranslatable english in the synopses :) 16:16
eric256 has a hard enough time translating the synopses into english ;) 16:17
16:18 ejs joined 16:21 Alias__ left 16:24 duke_leto joined 16:29 alanhaggai joined 16:45 maerzhase1 left 16:46 |jedai| left, |jedai| joined 17:10 |jedai| left 17:11 |jedai| joined 17:14 [particle]1 joined 17:18 tomyan left 17:20 donaldh left 17:22 ejs1 joined, ejs left 17:25 Whiteknight joined 17:29 jhorwitz joined 17:30 azawawi joined
azawawi hi 17:30
17:32 pmurias joined, [particle] left
pmurias ruoso: hi 17:32
17:36 andrew_ joined, cognominal left 17:40 ruoso left, andrew_ is now known as wknight8111 17:41 namenlos left 17:45 ludan left 17:46 cognominal joined 17:52 pdcawley left 17:53 pmurias left 17:56 Whiteknight left 18:01 masak joined 18:04 pdcawley joined 18:09 Psyche^ joined 18:17 hercynium left 18:19 hercynium joined 18:23 riffraff joined
riffraff hi 18:24
18:25 Patterner left, Psyche^ is now known as Patterner, |jedai| left 18:26 pbuetow joined, |jedai| joined 18:27 pdcawley left
pugs_svn r24893 | azawawi++ | [S:H:P6] Fixed dependency on Test::Exception, jq++ (0.033 -> CPAN) 18:40
18:40 pmurias joined
pugs_svn r24894 | masak++ | [t/spec] fixed spelling of "instantiate" 18:40
18:42 brunoV left
masak hi, I've been Warnocked on a question I care about. what should I do? 18:48
[particle]1 stop caring? :P 18:49
jnthn masak: Sob loudly.
18:49 [particle]1 is now known as [particle]
jnthn masak: What question? 18:49
masak jnthn: the thing about read-accessing class vars from within the class declaration.
like 'my $.other = $.first + 2' 18:50
jnthn Oh, I read that and thought...hmm.
masak :)
jnthn $.x desugars to self.x
masak yes, I know.
jnthn So under that definition we have issues.
masak it's such a natural thing to do.
jnthn Yeah, I can see the temptation to write that. 18:51
I didn't yet think of a way to allow it without making for confusion.
masak ok, that's a goodish kind of Warnock.
jnthn My problem is that 18:52
has $.x = 42; has $.y = $.x * 2;
Will people expect those to work on the current instance?
If so, self has to be the current instance.
masak that would be awesome, too. 18:53
but not as important.
jnthn But then self has to mean something else (maybe the proto) to make it work for class ones.
masak yes, I see that.
jnthn And I haven't through through exactly the implications of that yet.
It may well be able to be made to work just like that. 18:54
masak sounds promising.
jnthn: by the way, resulting from a discussion here earlier today, tonight I'm going to initiate a project in the Pugs repository with the aim of providing Perl 6 implementations with a full set of user docs for Christmas. 18:55
any comments on that? 18:56
[particle] a merry christmas it will be, then. 18:57
jnthn masak: Not beyond, "that's awesome, thank you!"
18:57 meppl left
masak :) 18:57
goodie.
jnthn masak: Will you try to write them in pod6?
masak jnthn: good idea.
I think.
jnthn I know there's the early beginnings of a perldoc in Rakudo.
Matt-W Evening
jnthn I think [particle]++ started that.
Matt-W Guess what I was just thinking about 18:58
masak Matt-W: Christmas? :)
Matt-W No, documentation
And I'm already confused
masak Matt-W: great!
that's just what we need, an itch to scratch.
Matt-W S29 as it stands talks about exists on Array
masak Matt-W: yes...? 18:59
Matt-W but there are mentions in S03's history that exists was removed, its role filled instead by contains
masak Matt-W: no...
contains() compares two containers.
it has different semantics from exists() 19:00
what was removed was the method form of exists, to be replaced by an adverb.
Matt-W That's one of the things that confused me, yes :)
ahah
masak Matt-W: think of contains as the subset relation between sets.
Matt-W so if you say @a.contains(@b), you get True back if (and only if) all the elements of @b are present in @a 19:02
i.e. @b considered as a set is a subset of @a considered as a set 19:03
masak aye.
Matt-W excellent
masak that's my understanding based on reading the smartmatch tables in S03.
Matt-W writes something to that effect 19:04
19:04 pdcawley joined 19:06 elmex left, elmex joined
Matt-W Now, would @a.contains(@b) be False if @a and @b are both empty? 19:07
masak Matt-W: no, True. 19:08
Matt-W because it's subset
and there are no elements in @b for @a to contain, so by definition it takes all of them
masak Matt-W: by the logic that it doesn't need to be a _proper_ subset.
Matt-W err, contains all of them
and so in fact it's True whenever @b is empty 19:09
masak i.e. it's enough that all of @b's elems are in @a. and they are, vacuously.
19:10 duke_leto left
TimToady I think you're misreading S03 19:10
masak oh :/
TimToady .contains only ever has a Hash as the invocant, I think
masak TimToady: O RLY?
masak checks 19:11
19:11 |jedai| left
TimToady so it's probably synonymous with .{X}:exists 19:11
19:11 elmex left
masak Any Hash superset relation X.contains($_) 19:11
TimToady X is the Hash 19:12
masak but "superset" means...
19:12 |jedai| joined
TimToady cut-n-paste error? 19:12
masak note that it doesn't say _containment_ relation.
TimToady: but yes, you seem to be right about the Hash thing. 19:13
Matt-W So what does .contains mean on a Hash?
masak apparently, it means :exists 19:14
TimToady in any case, superset would probably need an all($_) in the subscript
Matt-W so it's completely redundant? 19:15
masak I think I must have misread S03 when concluding that Array.contains existed.
Matt-W Maybe I should just stick to grammar pedantry. 19:18
I missed why :exists has a : on the front
ah, adverb 19:21
masak aye.
TimToady since adverbs naturally modify the preceding operator, and the preceding operator is a subscript, it works, and the notation to access the element is the same 19:23
regardless of whether you want to modify the final operator to test for existence or to delete the element from its container 19:24
Matt-W It's making sense now
TimToady anyway, .contains probably dates from when we still weren't sure what we'd replace exists with
masak I really like that we're having this discussion. 19:25
Matt-W So we don't actually need a .contains on Hash then?
pmurias masak: user docs = reference documentation? 19:26
masak pmurias: I dunno, is it?
Matt-W well users will need reference documentation
masak all I want is an executable somewhat like perldoc, where users type in their Perl 6 questions, and magically get their answers. 19:27
that's the official goal.
Matt-W 'all I want'
masak oh, and also that they will be so happy when they read the docs that they almost start crying out of gratitude. 19:28
almost, but not quite.
Matt-W No I thought the crying was because Perl 6 is so amazing
masak that, too.
the combination of the two might actually put the over the edge.
s/the/them/
Matt-W But you're right, the people who write the docs will be heroes wherever Perl programmers gather 19:29
masak it's not that hard a task. it just requires a lot of iterations.
that's why we start today. :)
Matt-W and clarifications
so .contains is out then
masak Matt-W: clarifications are a kind of iterations. :)
Matt-W yes
pugs_svn r24895 | lwall++ | [S03] remove .contains fossil 19:30
Matt-W yup, it's gone
masak :)
except from Jurassic Park. 19:31
Matt-W so instead of what we were talking about of %h.contains(@a) we now say %h<<@a.all>>:exists, I think 19:33
masak looks right to me. 19:34
std: my %h; my @a; %h<<@a.all>>:exists
p6eval std 24894: OUTPUT«00:07 87m␤»
19:34 |jedai| left
Matt-W and most certainly not %h<@a.all>:exists... unless you really mean it :) 19:34
masak: looks encouraging
masak aye. 19:35
Matt-W There's a ticket in RT for implementing .contains on Hash
#60234
19:35 |jedai| joined
masak Matt-W: thanks, closing it. 19:35
Matt-W Just trying to make sure there are no remnants hanging about to confuse people 19:36
19:36 [particle]1 joined
Matt-W I'm confused enough as it is 19:36
TimToady %h<<@a.all>> is wrong a couple ways 19:38
first, it'd have to be %h<<@a.all()>> 19:39
masak oh no, him again :)
TimToady second, the words will be divided on spacs
*spaces
masak oh, right. 19:40
Matt-W doh
TimToady you want %h{@a.all}
Matt-W of course
masak TimToady++
Matt-W I get so hung up on the <<>> and <> that I keep forgetting we still have (and need) {}
19:40 [particle] left 19:46 alester joined
Matt-W Okay 19:47
next thing
masak: "Code has a .sig" - is this different to .signature which is mentioned elsewhere, or is it indeed meaning .signature? Or do elsewheres mean .sig? 19:49
masak Matt-W: hm.
I don't think I abbreviated it when I wrote it.
buubot: spack \b\.sig\b
buubot masak: S03-operators.pod:1
Matt-W there's a .sig mention in S03
masak yes.
buubot: spack \b\.signature\b 19:50
buubot masak: S12-objects.pod:1 S29-functions.pod:1
masak could be that .sig is a typo.
Matt-W Code Signature sig compatibility $_.sig is a subset of X ???
pmichaud I call typo. 19:51
TimToady more like, can't make up my mind :) 19:52
but the long one is fine
19:54 pdcawley left
Matt-W long one it is then 20:06
20:14 donaldh joined 20:19 ejs1 left 20:20 Lorn left 20:21 justatheory left 20:28 jgoulah joined 20:29 pmurias left
jgoulah moritz_: thanks for your ilbot... probably wouldn't hurt to add Bot::Basicbot to your deps list -- or even better have a Makefile.PL that lists deps so people can just 'make installdeps' 20:29
also I have to ask why is this channel on freenode and not irc.perl.org :)
20:30 barney left 20:33 plu joined 20:45 riffraff left 20:46 |jedai| left 20:47 |jedai| joined 20:49 kisu_ left 20:50 kisu joined
pugs_svn r24896 | masak++ | [u4x] initial commit -- just a README file so far 20:54
masak use.perl.org/~masak/journal/38279 20:56
masak needs to sleep now
see you tomorrow, people. 20:57
20:57 masak left 20:59 wknight8111 left 21:02 Cantacuzenus left 21:05 elmex joined
jgoulah moritz_: Date::Simple too 21:05
21:13 elmex left
moritz_ jgoulah: I'll look into it 21:15
21:16 pdcawley joined 21:17 kisu left 21:21 kisu joined
jgoulah moritz_: cool, yeah for some reason I cant get the channels links to work though 21:23
21:23 lisppaste3 joined
moritz_ jgoulah: did you copy the .htaccess file? 21:24
jgoulah ah
actually yeah it exists
i can try to trace through it 21:26
moritz_ did you look into the error.log?
21:28 azawawi left
eric256 is there going to be a way for pod to be directly linked to the code it is in? i.e. class and method documentation? 21:33
moritz_ eric256: yes.
avar moritz_: how? I thought you could access the POD document through a hash, but there wasn't a way of really telling what function or thing the pod documented programatically 21:36
in the sense of lisp or python that is which have docstrings
moritz_ avar: there was a lenghty discussion on p6l about it, TheDamian vs. Mark Overmeer iirc 21:37
eric256 well that should make the technical parts of masaks idea pretty easy, then just right modules for everything with the documentation
moritz_ avar: there was agreement that such a thing would be possible, and Damian proposed a method with markers and some such 21:38
eric256 if we can extend classes then they could even extend the classes builtin in Parrot with functionality that is implemented in perl6 directly ;)
moritz_ avar: but in the end we have to wait for Damian's next iteration
jgoulah moritz_: yeah it doesn't seem to be rewriting properly 21:39
moritz_: i'll debug and get back with you
rakudo_svn r35506 | jonathan++ | [rakudo] Get the signature in Perl6MultiSub straight from the property, rathre than via a method. Bit faster, and solves an ordering issue in the forthcoming role changes. 21:40
eric256 ahh, well rubys way of just using the last comment before the method seems pretty straight forward ;) 21:41
moritz_ jgoulah: the rewrite rules probably assume that you've installed it into the root dir of a vhost
eric256 has no idea about implementation of such though ;) 21:42
moritz_ eric256: problem is, Perl 6 is pretty hard to parse... 21:44
eric256: and to found out where the method declarations ends, you have to parse it 21:45
eric256: and one of the goals was that a Pod parser shouldn't have to parse Perl 6
21:46 Tene left, Tene joined
pugs_svn r24897 | moritz++ | [t/spec] test for the presence of a type registry 21:49
rakudo_svn r35507 | jonathan++ | [rakudo] Initial refactor of roles to work towards parametric role support. Now we have a Perl6Role object installed in the namespace. It knows how to produce a Parrot-level role when given parameters, by doing a multi-dispatch on them. There's a bunch of nasty things to make 21:50
..enums keep working (that code is over-ripe for a big refactor soon), plus many comments of things that remain to be done. Happily, discounting enum stuff which will shrink a lot soon, actions.pm grows little.
eric256 moritz_: ahh, i would have though the pod and perl parser would be one and the same. but making them seperate would make this particular task more difficult 21:51
i'll have to see if i can find the discussion on p6l
21:56 [particle] joined 22:05 wknight8111 joined 22:09 justatheory joined 22:14 [particle]1 left 22:20 alanhaggai left 22:21 ryanc left 22:23 Tene left
pugs_svn r24898 | moritz++ | [t/spec] use $*OUT instead of $*DEFOUT in basic print tests 22:27
22:28 ryanc joined
moritz_ perl6: say $*EXECUTABLE_NAME 22:33
p6eval elf 24898: OUTPUT«Can't call method "Str" on an undefined value at ./elf_f line 649.␤ at ./elf_f line 3861␤» 22:34
..rakudo 35508: No output (you need to produce output to STDOUT)
..pugs: OUTPUT«␤»
moritz_ perl6: say $*EXECUTABLE_NAME
p6eval elf 24898: OUTPUT«Can't call method "Str" on an undefined value at ./elf_f line 649.␤ at ./elf_f line 3861␤»
..rakudo 35508: No output (you need to produce output to STDOUT)
..pugs: OUTPUT«␤»
moritz_ my local rakudo says ../../parrot for this 22:35
wknight8111 rakudo: say $*EXECUTABLE_NAME 22:38
p6eval rakudo 35508: No output (you need to produce output to STDOUT)
moritz_ perl6: say $*PID 22:39
p6eval rakudo 35508: No output (you need to produce output to STDOUT)
..pugs: OUTPUT«␤»
..elf 24898: OUTPUT«32152␤»
moritz_ rakudo seems borked atm
sh: line 1: 32168 Segmentation fault ./parrot languages/perl6/perl6.pbc /tmp/VjxhW91kJ1 >> /tmp/hwYHEmW3BF 2>&1
jnthn Ouch.
moritz_: Tried a clean build? Or is it one? 22:40
moritz_ trying a realclean...
wknight8111 rakudo isn't borked, it's become sentient and decided not to interact with us because we are peons
22:40 iblechbot left
moritz_ you mean it's too 31337? ;-) 22:41
wknight8111 far too 31337 22:42
22:42 funktio left 22:45 zev left, donaldh left, tokstolle joined
moritz_ does 'parrot perl6.pbc file > out' work the same on windows and unix? 22:50
(neglecting any path issues)
jnthn moritz_: Yes, should do. 22:51
I often do --target=pir file.p6 > dump 22:52
So I can look through output if needed.
moritz_ is trying to write some tests for things that need another rakudo instance
like 'say [1, 2, 3]'
jnthn moritz_: Why does that need another instance? 22:53
Oh, to check it gives output?
moritz_ yes
jnthn Aha. :-)
moritz_ rakudo: module A { say "alive" }; 22:55
p6eval rakudo 35511: OUTPUT«alive␤»
moritz_ rakudo: module A; say "Alive";
p6eval rakudo 35511: OUTPUT«Alive␤»
moritz_ jnthn: how does redirecting of STDERR work on windows? 22:56
jnthn 2> IIRC
moritz_ can I do 2>&1 like on Linux 22:57
jnthn I'd have to check... :-)
moritz_ ie both FDs into the same file?
please do
rakudo: say $*PID
p6eval rakudo 35511: OUTPUT«Use of uninitialized value␤␤»
eric256 regarding pod from above: it would be realy cool for module makers to be able to actualy output documentiation as part of an error. i.e. "$method used wrong: proper usages is $method.pod.usage" (lots of arm waving is used around any of those variable names, its just psuedo code/ideas) /me goes back to reading relevant p6l on the matter 22:58
moritz_ eric256: since exceptions and error messages aren't specced yet, you can influence the matter by writing your own specs ;-)
jnthn 2>&1 doesn't work, no :-(
oh, hang on 22:59
eric256 hmm a debuging exception handler the gives good output. kinda like use diagnostics.
that would be fancy
23:00 |jedai| left, |jedai| joined
jnthn moritz_: perl -w -e "print $x . 42" 1> out 2>&1 23:01
That puts both into out
moritz_ eric256: that could use the same mechanism as i18n'ing error messages
jnthn: thanks
eric256 which brings around i18n 23:02
moritz_ now what's missing is a defintion of what execption to throw in which case 23:03
23:04 pbuetow left 23:07 _SamB_ joined 23:08 SamB left
moritz_ rakudo: module A { sub b { return 'bar' }}; say A::b() 23:11
p6eval rakudo 35511: OUTPUT«b␤»
moritz_ that's wrong, isn't it?
rakudo: module A { sub b($x) { return 'bar' }}; say A::b("blubb") 23:12
p6eval rakudo 35511: OUTPUT«bblubb␤»
jnthn hmm.
rakudo: module A { sub b { return 'bar' }}; A::b
p6eval rakudo 35511: RESULT«[{ ... }]»
moritz_ it would explain a lot of the worreis of the November folks
jnthn rakudo: module A { sub b { return 'bar' }}; say A::b
p6eval rakudo 35511: OUTPUT«bar␤»
jnthn rakudo: module A { sub b { return 'bar' }}; say A::b() 23:13
p6eval rakudo 35511: OUTPUT«b␤»
jnthn rakudo: module A { sub b { return 'bar' }}; A::b()
p6eval rakudo 35511: RESULT«[{ ... }]»
jnthn ...huh?!
moritz_ when you don't say() something, p6eval executes it again, with .perl attached
so in this case a COde object is returned 23:14
jnthn I know, I'm just besumed by the bug!
moritz_ ah
should I rakudbug it?
jnthn Please.
moritz_ rakudo: module A { sub b($x) { return 'bar' }}; say (A::b("blubb")).perl 23:16
p6eval rakudo 35513: OUTPUT«[{ ... }, "blubb"]␤»
jnthn ah, feck 23:17
I bet I know what's to blame for that.
moritz_ so basically A::b() returns a Code object instead of executing it?
jnthn sucks
TimToady functions, who needs 'em...
jnthn I adjusted term the other day to make it not ignore namespaces.
23:18 _SamB_ is now known as SamB
moritz_ rakudo: module A { sub b($x) { return 'bar' }}; say A::b("blubb").() 23:18
jnthn We had a test before that passed for epicly the wrong reason. ;-)
p6eval rakudo 35513: OUTPUT«invoke() not implemented in class 'ResizablePMCArray'␤current instr.: '_block14' pc 78 (EVAL_21:42)␤»
moritz_ rakudo: module A { sub b($x) { return 'bar' }}; say A::b.("blubb")
jnthn Anyway, I musta hashed it up.
p6eval rakudo 35513: OUTPUT«bblubb␤»
moritz_ doesnt' even want to know where teh ResizablePMCArray comes from
jnthn :-/ 23:19
moritz_ so why the bloddy hell does S11-modules/export.t still pass? 23:20
ah, because it uses sub name eq return value 23:21
how convenient.
jnthn lol
*that's* the test I did the change to pass ;-)
pugs_svn r24899 | moritz++ | [t/spec] test subs should return something different than their name, because 23:24
r24899 | moritz++ | the underlying Code object might stringify to its name (Rakudo regression)
23:25 |jedai| left, |jedai| joined 23:28 kisu left
pugs_svn r24900 | moritz++ | [t/spec] more sub call tests 23:29
23:30 kisu joined
moritz_ so much for my project on testing say() by invoking another copy of rakudo 23:30
bedtime now 23:31
23:33 tokstolle left 23:35 meppl joined 23:45 hercynium left 23:49 eric256 left 23:56 Limbic_Region joined 23:57 vixey left

Powered by Cro and the Raku® Programming Language.

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

Thank you!