»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.
tadzik hmm 00:00
rakudo: "\e\e".split('\x[1b]').perl.say
p6eval rakudo 00848d: OUTPUT«("\x[1b]\x[1b]")␤»
masak tadzik: you should patch in a special case for "\e" in the pretty-printer
tadzik $what.trans('\e' => 'E').split('E'); looks awful
but well, original Term::ANSIColor's way is even worse :) 00:01
masak that's not what I was suggesting, though :)
rakudo: "\e\e".split("\x[1b]").perl.say
tadzik that's what I worked out
p6eval rakudo 00848d: OUTPUT«("", "", "")␤»
tadzik oh
I used '', not "", right
:)
masak that's not what I was suggesting either. :) but no matter. 00:02
00:02 cbk left
tylercurtis rakudo: say pir::fdiv__III(-3, 4); say pir::div__III(-3, 4); say pir::fdiv__NNN(-3, 4); say pir::div__NNN(-3, 4); 00:02
p6eval rakudo 00848d: OUTPUT«0␤0␤-1␤-0.75␤»
tylercurtis rakudo: say pir::fdiv__III(-3, -4); say pir::div__III(-3, -4); say pir::fdiv__NNN(-3, -4); say pir::div__NNN(-3, -4); 00:03
p6eval rakudo 00848d: OUTPUT«0␤0␤0␤0.75␤»
00:07 rlb3_ joined 00:08 kfo joined
[Coke] I am still here. waiting for teh kerfuffle to settle. 00:09
00:09 rlb3_ is now known as rlb3_home
masak [Coke]: I'm all settled. don't know about the others :) 00:10
[Coke] jnthn: we can always do a point release of parrot. we've done it before.
jnthn [Coke]: I don't think this would be serious enough to warrant it.
[Coke] but if you can hack around it, it will save me great headache. I don't wish to do /3/ big releases this week. =-) 00:11
jnthn++ ;)
jnthn [Coke]: Well if we desparately want such an op that's right, we could just make a dynop in perl6.ops, and then switch to the Parrot one after it's fixed in the future. 00:12
[Coke] jnthn: [:facepalm:] 00:13
tylercurtis is building a parrot with a working fdiv_i_i_i to see how it would affect rakudo's div performance.
[Coke] that makes a helluva lot more sense! =-)
00:13 Schwern left
tadzik is there something like wantarray in Perl 6? 00:13
jnthn tadzik: No
tadzik thanks jnthn 00:14
[Coke] (there is only zuul)
jnthn tadzik: Context works the opposite way in Perl 6 to Perl 5.
tadzik jnthn: what do you mean, the opposite way?
jnthn tadzik: moritz_++ wrote a really great explanation on his blog at some point
tadzik in 5-6? 00:15
jnthn tadzik: seems so, yes. perlgeek.de/en/article/5-to-6#post_06
[Coke] do we care about renaming the old announcement dirs?
tadzik reads
jnthn I wasn't sure if it was in 5-6 offhand, but yes, it is.
[Coke] er, files
jnthn tadzik: Oh, I'm not sure that was the one I wanted actually
tadzik: aha! 00:16
mberends zavolaj-users: try the new unix-fork.p6 example -Ofun :)
jnthn tadzik: perlgeek.de/blog-en/perl-6/immutabl...ntext.html
mberends: huh, it fails here
:P
mberends jnthn: you could try it on pivo ;) 00:17
tadzik sounds sane
colomon here's the thing. the reason pmichaud
jnthn "A 1GB netbook froze with 19 children."
colomon 's version was so fast is that it switched to doing Nums instead of Rats.
jnthn Poor childern. My parry are with the father. I'm truly sorry for their lots. 00:18
[Coke] boy, y'all are much better about updating the news.
cxreg heh
jnthn mberends: Does this...actually work too? :-)
colomon getting working fdiv would be nice, but I don't think it will get a crazy speed improvement for us.
mberends jnthn: yes, up to 18 children. at 100MB per Rakudo process, it was heavy. 00:19
tadzik rakudo: sub foo (Str *@a) {}; foo("foo", "bar", "asd", 7)
p6eval rakudo 00848d: ( no output )
cxreg mberends: a m&s gc is decidedly anti-cow
tadzik huh. Doesn't it check the types?
jnthn 100MB memory use is pro-pig though.
tadzik: I think type constraints on slurpies are kinda bogus. 00:20
pmichaud colomon: ping -- any word on getting some speed back?
jnthn tadzik: I mean, what if it's an infinite list?
pmichaud oh, we can use fdiv
just use the num version
tadzik oh, right
colomon pmichaud: I'm spectesting the 22s version right now, it just hit the trig functions
s1n can anyone explain why DateTime went back to immutable again?
pmichaud pir::fdiv__NNN($a, $b).Int
cxreg mberends: wait, what? 00:21
sub fork() returns Int is native('libz') { ... }
libz?
mberends cxreg: wfm, libc didn't ;)
pmichaud or, even better, acall "upgrade_to_num_*
[Coke] looks like the news section is all set. anyone want to bump something up from the changelog into teh announcement?
cxreg mberends: that might only be working by accident 00:22
[Coke] multi-whatever-currying seems big to me.
jnthn [Coke]: Reviewing both
mberends cxreg: I think on Linux fork() is system and may not even be in a lib, but you have to name *some* lib.
jnthn [Coke]: Yes, that would be a worthy candidate.
pmichaud looking 00:23
jnthn ooh, feeds are missing
colomon closure and gather / take fixes, if they're not already in the announcement?
cxreg mberends: that's what im saying. in #plparrot and #parrot, we talked once about how to NCI from your resident symbol table
mberends: parrot has support for it, not sure about zavolaj
pmichaud I thought I put feeds in the announcement
tadzik s:g/foo/bar/ is NYI?
cxreg jnthn++ implemented it last week, i thought
sorear do try the other two syntaxes 00:24
jnthn pmichaud: maybe you did
mberends cxreg: afaik zavolaj still requires a lib, but it's a good idea to make the lib optional
jnthn pmichaud: I'm looking at ChangeLog though :-)
pmichaud I didn't update the ChangeLog
jnthn pmichaud: I've added it, I'm just pondering what else may be missing
pmichaud: Is it worth mentiong that Int ~~ Num is gone in this release? 00:25
pmichaud Yes.
jnthn Adding the ChangeLog
cxreg so are you guys doing both a monthly release and star?
pmichaud star is a distribution release
today is a compiler release
[Coke] feeds are in there.
cxreg righto
[Coke] pmichaud: you said the script for generate percentages is broken? 00:27
(er, of passing spectests)
pmichaud [Coke]: yes. 00:28
dalek kudo: 245b4fe | Coke++ | docs/announce/2010.07:
Slight mod to verbage, add another entry from Changelog.
mberends yawn. snore. \o
kudo: afd65e7 | jonathan++ | docs/ChangeLog:
Couple more ChangeLog items.
[Coke] pmichaud: hokay. leaving off that step...
00:28 PZt joined
colomon spectesting now in integration... 00:29
[Coke] ingy: ping 00:30
ingy apparently needs a CREDITS entry. 00:31
jnthn pmichaud: Shoudl the announce have a mention that this is not R*?
pmichaud jnthn: yes.
I'm not at a point where I can make such a mention, though.
[Coke] I will take a first stab at it. 00:32
pmichaud (so if someone else can draft, I'll edit)
jnthn [Coke]++
Announce so far looks good, anyway.
00:32 elmex left
dalek kudo: 937177e | jonathan++ | docs/ChangeLog:
Also a note in the ChangeLog about enum improvements.
 00:33
00:34 lest_away is now known as lestrrat
[Coke] jnthn: I think Util wrote most of it. =-) 00:37
00:37 elmex joined
[Coke] star includes compiler and modules and... a pony? 00:37
tadzik and a book
00:38 mikehh joined
tadzik and a Cake 00:38
tylercurtis How about a kitty?
[Coke] book's separate, but timed with, I think.
tadzik Acme::Meow?
tylercurtis tadzik: Exactly!
tadzik tylercurtis: I'll have to write some tests then :)
00:39 eternaleye joined
tylercurtis [Coke]: I think Rakudo * is supposed to include the PDF of the book. 00:39
masak Star includes an airbag, for when the tutorials go too fast.
rakudo: enum A <a b c>; say A.enums.WHAT
p6eval rakudo 00848d: OUTPUT«Method 'enums' not found for invocant of class 'EnumMap'␤ in main program body at line 22:/tmp/RiJhCH9ygw␤»
[Coke] an enum says WHAT?
masak ESTILLTOOSOON
[Coke] what is, the release? 00:40
are we waiting on something from you, masak?
00:40 felliott left
tylercurtis Also perl6.vim and Blizkost according to wiki.github.com/rakudo/rakudo/whats...nto-rakudo 00:41
[Coke] ah, that's what I need, danke 00:42
00:42 Limbic_Region joined
masak [Coke]: nope, waiting on nothing from me. I made it! :) 00:43
see what you think about the new "Operator overloading" section of the blog post draft gist.github.com/485816
if there are no comments, I'll go ahead and publish it now. 00:44
[Coke] msgbot?
colomon [Coke]: I just pushed my changes a minute ago.
tadzik masak: multi sub infix:<±>($number, $fuzz) is before Operator Overloading section 00:45
[Coke] masak - there is a +/- builtin?
dalek kudo: 3528a10 | (Solomon Foster)++ | src/core/Int.pm:
Optimize use of .sign in infix:<div>.
masak rakudo: say "still on 00848d?"
p6eval rakudo 00848d: OUTPUT«still on 00848d?␤»
masak tadzik: thanks, fixing.
[Coke] colomon: so, back to original speed and passing spectest?
00:45 eternaleye left
colomon [Coke]: not sure what you mean by original speed. 00:46
[Coke] rakudo: say 15 ± 5;
p6eval rakudo 00848d: OUTPUT«===SORRY!===␤Confused at line 22, near "say 15 \x{b1} 5"␤»
colomon it's the fastest it was today when it passed all the tests. :)
[Coke] colomon: apparently last release was 2x as slow as the previous, or something. you eliminated the factor-of-two slowdown, yes? 00:47
colomon I shaved about 55% off the execution time of that script.
[Coke] masak: you use that example but it fails.
colomon so yes, it should be about where it was last release.
[Coke] colomon++ to you, sir.
tadzik [Coke]: the definition is above the section start
masak [Coke]: what tadzik said. now fixed.
00:47 rlb3_home left
masak the section is about operator overloading. :) 00:48
hm, I don't need a multi on that sub...
tylercurtis colomon: a working fdiv_i_i_i appears to reduce the time taken by that script by about 66% compared to before your changes.
00:48 isBEKaml left
tadzik nice examples, masak++ 00:49
masak \o/
tadzik bah, apparently Term::ANSIColor is alredy in proto 00:50
[Coke] tylercurtis: how about compared to nwo?
tylercurtis colomon: probably rather minor difference from as is now that you've made it so much faster.
00:50 azert0x left
tylercurtis [Coke]: Haven't tried that yet. 00:50
tadzik Jakob Ilves, is he on #perl6 sometimes maybe?
[Coke] ok. compare that, if it's another big win, I'll consider holding the release for someone to make a dynop. 00:51
masak rakudo: say "?"
p6eval rakudo 937177: OUTPUT«?␤»
dalek kudo: 87e77d0 | Coke++ | docs/announce/2010.07:
add a "we are not star" note to the compiler release announcement
kudo: f2aa8d1 | Coke++ | docs/announce/2010.07:
rerun the contributer script.
masak whistles "we are not star" 00:52
rakudo: enum A <a b c>; say A.enums.WHAT
p6eval rakudo 937177: OUTPUT«EnumMap()␤»
masak rakudo: enum A <a b c>; say A..WHAT
p6eval rakudo 937177: OUTPUT«Not enough positional parameters passed; got 0 but expected 1␤ in 'WHAT' at line 710:CORE.setting␤ in main program body at line 22:/tmp/gA1pJtBwqX␤»
masak oops. 00:53
rakudo: enum A <a b c>; say A.WHAT
p6eval rakudo 937177: OUTPUT«A␤»
masak \o/
masak closes ticket
tylercurtis colomon: can you point me to any quick spectests to run that incorrect infix:<div> behavior would mess up?
00:54 ashleydev left
colomon t/spec/S03-operators/arith.t 00:54
tylercurtis: you actually changed parrot for this fix? 00:55
jnthn masak: Nice post!
masak thankz.
tylercurtis colomon: my local checkout, yeah.
masak it's going into print now.
just have to... um... make yet another small Perl 6 script to convert to use.perl.org's idiosyncratic idea of HTML. 00:56
dalek kudo: 5c33c5a | Coke++ | docs/release_guide.pod:
This release is heading out the door soon.
 00:57
kudo: 8a9027a | Coke++ | docs/announce/2010.07:
I am fairly certain that Rakudo is not yet committing code to her own repository
[Coke] ... though it would be super awesome if she wasn't so lazy!
colomon 66% better is a decent bit faster than my fix. but seems like it is very late in the day to try to squeeze it into the release if it needs parrot changes as well.
[Coke] colomon: no, you'd just make his parrot builtin op fix a dynop for rakudo.
jnthn [Coke]: no no, Rakudo became self-aware this month too 00:58
[Coke] and yes, I really am not doing a parrot release today, again. =-)
jnthn: let her fix the damn announcement, then!
;)
colomon ah, [Coke] I have not the faintest idea how to do a dynop
[Coke] jnthn and I have some experience in that area.
and perl6 is already using dynops, so it's not a heavy lift. 00:59
uh, why does perl6.ops have an $Id $ string? =-)
colomon you guys be the judge, then? if it's easy to do the dynop thang, I'm pretty sure the rakudo part of things is trivial. then it's just a matter of another spectest....
sorear [Coke]: because it used to be in svn? 01:00
jnthn [Coke]: Laughing dried grapes.
[Coke] someone show me the op.
colomon tylercurtis: can you add say $n.WHAT; to the end of the test script to make sure it is outputting a Rat?
[Coke] sorear: it was kind of rhetorical. =-)
tylercurtis colomon: meh, apparently, I didn't correctly fix fdiv_i_i_i.
colomon ah. 01:01
I vote we leave this for the next release.
jnthn [Coke]: tbh I'd not worry about it.
+1
colomon always nice to have another 10% speedup in your pocket. ;)
[Coke] WFM. 01:02
Util [Coke]: Your rerun of the contributer script removed "döt net" from Ingy's name.
[Coke] tylercurtis: if you get it working as an op, it's basically a CNP into src/ops/perl6.ops with a new op name.
cxreg mberends: patch for zavolaj ready, reads from resident symbol table
[Coke] Util: not my fault. he's not in CREDITS.
cxreg (no zlib required :) )
[Coke] I will gladly rerun if someone adds him. =-)
cxreg: resident evil table? 01:03
Util [Coke]: Fair enough.
01:03 eternaleye joined
cxreg [Coke]: is native('burial ground') 01:03
sjohnson rakudo: for (^5) { my %hash; say "dub-tee-eff" if defined(%hash); %hash<something> = 1; }
p6eval rakudo 937177: OUTPUT«dub-tee-eff␤dub-tee-eff␤dub-tee-eff␤dub-tee-eff␤dub-tee-eff␤»
01:03 timbunce left
sjohnson anyone here able to explain why %hash stays defined, after it is redeclared with a my? 01:04
tylercurtis rakudo: my %hash; say %hash.defined
p6eval rakudo 937177: OUTPUT«1␤»
tylercurtis sjohnson: because apparently empty hashes are defined. Is that wrong? 01:05
[Coke] there's yer problem.
sjohnson rakudo: my $string; say $string.defined
p6eval rakudo 937177: OUTPUT«0␤»
sjohnson ... that threw me for a good loop at work 01:06
tylercurtis rakudo: my @array; @array.defined.say;
p6eval rakudo 937177: OUTPUT«1␤»
sjohnson maybe TimToady can help me
*puppy dog eyes*
help me with explanation, i mean. just so i get the idea.
pmichaud back again... anything from the release need anything from me? 01:07
(in the next 8 mins, before the soto?)
[Coke] pmichaud: double check the Star note in the announcement.
pmichaud I'm thinking it might want to be higher
01:07 pjcj left
[Coke] sure. 01:08
pmichaud make it the 2nd paragraph, maybe?
tylercurtis sjohnson: I think composite containers are always defined.
01:08 masak left
dalek kudo: 3d67b9d | Coke++ | src/ops/perl6.ops:
remove svn fossil
 01:08
tylercurtis sjohnson: based on the implication of this from S02 "Assigning or binding Nil to any composite container (such as an Array or Hash) empties the container, resetting it back to an uninitialized state. The container object itself remains defined."
01:08 masak joined
sjohnson thank you, tylercurtis 01:09
sorry to highlight you, TimToadie
sowwie
pmichaud "July 29[2]" feels weird.
01:09 eternaleye_ joined
pmichaud maybe "July 29, 2010 [2]." 01:09
or "July 29 [2]"
[Coke] pmichaud: done. 01:10
01:10 eternaleye left
[Coke] .. except for that. 01:10
01:10 pjcj joined
[Coke] ooh, the push failed. bundling.. 01:10
pmichaud anything else that needs my input/review?
au|irc sjohnson: ().defined is also true; generally, ".defined" is true for allocated containers, even if it has a size of 0.
pmichaud [Coke]: does anyone else have a "hold request" on the release? (more) 01:11
at this point, I consider the release to be largely in your hands -- you can decide when to cut it. :)
sjohnson you learn something everyday!
[Coke] pmichaud: colomon was the last. not waiting on anyone else.
pmichaud okay, great.
[Coke] pmichaud: I'm doing the build/test now. 01:12
masak rakudo: say "foo".substr(4).WHAT
01:12 achromic left
p6eval rakudo 937177: OUTPUT«Mu()␤» 01:12
[Coke] oh, and i can move up the family room to do this and stop being a hermit.
masak submits rakudobug
pmichaud [Coke]: if you run into any difficulties (you shouldn't, but...) and it's getting late where you are, I don't see any major difficulty with holding until tomorrow
anyway, it's in your ballpark. 01:13
for everyone else ... feature freeze until release :)
tylercurtis sjohnson: I expect the reasoning is that if you made @ and % variables default to undefined objects like scalars do, you either have to manually initialize them or or be unable to distinguish an empty array/hash from lack of an array/hash.
pmichaud (unless Coke++ requests or approves :)
okay, time for SOTO here...bbiaw
tylercurtis The former if you initialize them to the type object, the latter if you say that empty arrays/hashes are undefined. 01:14
dalek kudo: 5b5c8f5 | Coke++ | docs/announce/2010.07:
I said, WE ARE NOT STAR.
01:14 achromic joined 01:17 timbunce joined, snarkyboojum joined 01:20 timbunce left 01:24 fridim joined
masak loliblogged: use.perl.org/~masak/journal/40459 01:25
[Coke]: happy releasing!
masak seeks the bed 01:26
01:26 masak left
tylercurtis Oh. That's what I did wrong! I only changed fdiv_i_i, not fdiv_i_i_i. 01:26
01:26 kjeldahl joined
tadzik phenny: tell masak you forgot to bold the Operator Overloading section 01:26
phenny tadzik: I'll pass that on when masak is around.
tylercurtis facepalms. 01:27
01:27 plobsing joined 01:28 justatheory joined
au|irc phenny: tell masak great blogpost :) looks like use.perl ate the <...> in infix:<...> as markup, though 01:28
phenny au|irc: I'll pass that on when masak is around.
01:28 felliott joined 01:29 felliott_ joined, felliott left, felliott_ is now known as felliott
[Coke] is it normal for there to be a small # of spectest failures? 01:30
colomon no 01:31
what are you seeing?
sjohnson rakudo: my @list; @list.push(undef); @list.defined.say
p6eval rakudo 937177: OUTPUT«===SORRY!===␤Unsupported use of undef as a value; in Perl 6 please use something more specific:␤ Mu (the "most undefined" type object),␤ an undefined type object such as Int,␤ Nil as an empty list,␤ *.notdef as a matcher or method,␤ Any:U as a type constraint␤
.. or fail() as a failur…
sjohnson ... what have I DONE
rakudo: my $buffer; my @list; $buffer .= 'bad' if @list; @list.push(0); $buffer .= 'good' if @list; say $buffer 01:32
p6eval rakudo 937177: OUTPUT«===SORRY!===␤Quoted method name requires parenthesized arguments at line 22, near " if @list;"␤»
sjohnson rakudo: my $buffer; my @list; $buffer ~~ 'bad' if @list; @list.push(0); $buffer ~~ 'good' if @list; say $buffer 01:33
p6eval rakudo 937177: OUTPUT«Any()␤»
sjohnson （´ー｀　）
sjohnson gives up
01:33 gbacon joined
sjohnson rakudo: say 'happy'.' juice' 01:33
p6eval rakudo 937177: OUTPUT«===SORRY!===␤Quoted method name requires parenthesized arguments at line 22, near ""␤»
sjohnson rakudo: say 'happy'~' juice' 01:34
tylercurtis colomon: While figuring out what I did wrong with fdiv_i_i_i, I tried fdiv_n_n_n, which gives the correct result for the example script and isn't noticeably slower than the broken _i_i_i. It also passed arith.t(well, arith.rakudo; I assume arith.t should have some failures itself?). Haven't compared to yours yet. About to try your change then try with working fdiv_i_i_i.
p6eval rakudo 937177: OUTPUT«happy juice␤»
sjohnson rakudo: my $buffer; my @list; $buffer ~= 'bad' if @list; @list.push(0); $buffer ~= 'good' if @list; say $buffer
p6eval rakudo 937177: OUTPUT«good␤»
sjohnson sorry to clog the channel tylercurtis, i'm done.
tylercurtis sjohnson: "undef" isn't valid in Perl 6.
colomon tylercurtis: default parrot fdiv_n_n_n works?
sjohnson thanks for the help
tylercurtis colomon: if you still use box__PI, yes. 01:35
[Coke] ... I haven't released yet. 01:36
colomon trying here. 01:37
[Coke]: please do not wait on us!
tylercurtis colomon: Surprisingly, it appeared to be slightly faster than broken fdiv_i_i_i. But that's probably not a significant difference. I didn't fully spectest.
[Coke] I'm not, I'm waiting on the (&#$ spectests. =-)
colomon [Coke]: \o/
arith.t passes 01:39
wow, timing script works and is down to 10s 01:40
tylercurtis++ # another big win
I used upgrade_to_num_if_needed instead of boxing -- that's how most of the other Int operators do things. 01:42
though I don't think it will ever need to upgrade, so there's probably a way to shave a few more cycles off there.
01:43 masonkramer joined, s1n left
sorear TimToady: ping 01:43
[Coke] spec. test. so. slow. w. t. f. 01:47
ah. don't run the spec tests with TEST_JOBS, because they eat a TON of memory. 01:48
ah well, I'm already in S32.
tylercurtis colomon: yeah. Your previous version is significantly slower than fdiv_n_n_n for me. 01:49
colomon [Coke]: S32 takes a substantial portion of the time, alas. 01:50
Util [Coke]: Shall I commit gist.github.com/486902 ? 01:51
(Added Ingy to CREDITS; made contributors.pl work with UTF-8 and ignore fake "Rakudo Perl" name)
[Coke] nopaste.snit.ch/22220
spectest failures. 01:52
is that evil?
colomon yes
what the heck?
[Coke] Util: Go for it, that should be safe.
Util++ 01:53
colomon [Coke]: try
make t/spec/S05-mass/properties-derived.t
I'm wondering if you just ran out of memory or something.
[Coke] Util: also fix the list in teh release.
Util [Coke]: Will do.
[Coke] colomon: no, it fails even when run like that. 01:54
colomon huh.
works fine here.
01:54 jaldhar left
colomon you're on linux? 01:55
[Coke] no, OS X 10.6
colomon double huh
[Coke] darwin, intel.
colomon 10.5 here
[Coke] that's with a --gen-parrot 01:56
colomon same here
tylercurtis gets the failures with a non gen-parrot on OS X 10.5. 01:57
[Coke] I am guessing this is not worth killing the release over.
can someone do a spec test on linux for me?
colomon [Coke]: I'll pull a fresh rakudo and try it on my big 64-bit linux box 01:58
[Coke] ARGH.
does stresstest really also run spectest?
colomon I dunno, they did all that while I was on vacation. 01:59
02:00 s1n joined, Limbic_Region left
dalek kudo: 93ec069 | util++ | (3 files):
Added Ingy to CREDITS; made contributors.pl work with UTF-8 and ignore fake
02:00
tylercurtis [Coke]: forgot that, it's only the s05 tests I'm getting failures for. 02:01
02:02 paul1234 left
snarkyboojum I've had failure in t/spec/S05-mass/properties-derived.t on OS X for ages 02:05
failures even
for several months at least 02:06
tylercurtis snarkyboojum: the Ideograph stuff?
snarkyboojum unicode weirdness
yeah
colomon do you guys have ICU? 02:07
snarkyboojum tylercurtis: always related to \x[9FC4]
yep
02:07 timbunce joined
[Coke] pretty sure i"m icu enabled, yes. 02:08
02:08 justatheory left
snarkyboojum definitely have ICU here 02:08
tylercurtis also does. 02:09
snarkyboojum last time I brought it up moritz_ mentioned that it was a known issue on some platforms
colomon huh, t/spec/S05-mass/properties-derived.t doesn't work at all on my Linux box. perhaps no ICU is the problem?
(for me, I mean, not for you) 02:11
tylercurtis [Coke]: which test does the s10 test fail on?
dalek kudo: 40bf707 | Coke++ | docs/release_guide.pod:
stresstest runs spectest also - don't need it 2x.
 02:12
kudo: d1cd713 | Coke++ | (3 files):
Merge branch 'master' of github.com:rakudo/rakudo
02:14 jaldhar joined
lue ohai 02:16
02:16 timbunce left
sorear TimToady: Can you explain 'my module CORE;' ? I thought that the setting, and YOU_ARE_HERE code, was supposed to go into GLOBAL 02:17
(the part that isn't 'my'-scoped)
snarkyboojum I seem to get intermittent failures in S10-packages/basic.t - just ran it a few times.. a couple of runs where everything passes and a few with failed subtests which seems to abort and fail with "Bad plan." 02:18
[Coke] tylercurtis: it's in the nopaste.
(or else I don't know. =-)
running stresstest now, will post those... 02:19
ugh, only on s03
02:19 achromic left 02:20 skangas joined, paul1234 joined
sorear rakudo: $Foo::x = 42; say "alive" 02:23
p6eval rakudo 937177: OUTPUT«Null PMC access in getprop()␤ in '&infix:<=>' at line 1␤ in main program body at line 22:/tmp/1dH4ftG5W5␤»
sorear std: $Foo::x = 42; say "alive"
p6eval std 31790: OUTPUT«ok 00:01 117m␤»
sorear Is that valid (no runtime errors) Perl 6?
au|irc sorear: it is 02:27
as per S10/Package nesting
02:28 au|irc is now known as au|afk|lunch 02:30 reg7 joined
sorear au|afk|lunch: I'm speaking of the fact that there is no declarator for Foo::x 02:30
snarkyboojum tylercurtis: I think [Coke] might be seeing the same kinds of errors as I've described for S10-packages/basic.t - i.e the test is crashing after test 13 and test 36 for me, but intermittently 02:31
02:31 au|afk|lunch is now known as au|irc
sorear wow, you eat almost as fast as I do 02:31
au|irc it's as fast as pushing lunch on thunks queue 02:32
I read that as "Foo::<$x> = 42" and would think the package obj would be vivified, as per S10 02:33
might be wrong, though. 02:34
02:37 au|irc is now known as au|afkForReal 02:42 colomon_phone joined
sorear let me rephrase that. 02:43
rakudo: package Foo { }; $Foo::x = 42; say "alive"
p6eval rakudo 937177: OUTPUT«Null PMC access in getprop()␤ in '&infix:<=>' at line 1␤ in main program body at line 22:/tmp/FMfdkLTHlI␤»
sorear should that be legal?
no *package* vivification here
tylercurtis Is it cheating on my chemistry placement test if I write a Perl 6 module for strongly-typed unit-correct calculations and use that as my calculator? :) 02:44
snarkyboojum tylercurtis: I'd be interested if you managed to run a Perl 6 module *on* your calculator :P 02:46
tylercurtis snarkyboojum: not on my calculator. :) As my calculator. The test is online. 02:47
tadzik hah :) 02:48
[Coke] holy crap, this stresstest isnt going to finish until friday. 02:55
snarkyboojum [Coke]: what's it up to? :) 02:56
02:57 colomon_phone left
tylercurtis ugh... I think I'm going to have to do .Num.Int to get .kilo to actually return an Int when called on an Int. 02:57
03:03 meppel joined 03:04 reg7 left 03:07 meppl left
cxreg rakudo: my $foo = pir::null__P() 03:08
p6eval rakudo d1cd71: OUTPUT«Null PMC access in isa_pmc()␤ in '&infix:<=>' at line 1␤ in main program body at line 22:/tmp/8u0MgbDSLX␤»
cxreg why do you get that when trying to store a null?
03:09 kid51 left 03:11 gbacon left, masonkramer left
sorear snarkyboojum: Do you have a .NET calculator? 03:13
I know they make Java cellphones
[Coke] advent day 17 03:14
03:14 colomon_phone joined
snarkyboojum [Coke]: oops - that's a memory intensive test 03:16
sorear: nope :P
03:18 nimiezko left 03:19 kkrev joined
[Coke] guess it was a bad night to have 150 tabs open in Chrome. 03:22
done. nopaste.snit.ch/22223 03:26
same properties-derived failrus and s10-packages.
S10-packages working for folks on linux ? 03:27
03:33 alester left 03:35 shade_ joined
cxreg ok, i think i need a commit bit on zavolaj 03:39
sorear hugme: show zavolaj 03:41
hugme sorear: sorry, I don't know anything about 'zavolaj'
cxreg crud
sorear ping jnthn 03:42
cxreg jnthn: ping :)
03:42 sorear sets mode: +v hugme
cxreg forks 03:42
03:44 shade_ is now known as SHADE\, SHADE\ is now known as shade\
cxreg there we go: github.com/cxreg/zavolaj/commit/37c4b6112 03:44
03:48 DemoFreak left 03:50 DemoFreak joined
[Coke] ok, those tests do pass on feather. 03:51
03:55 tedv left
sorear std: augment class Foo {} 03:58
p6eval std 31790: OUTPUT«===SORRY!===␤Can't augment class Foo without MONKEY_TYPING at /tmp/aG9wyrJklb line 1:␤------> augment class Foo ⏏{}␤Potential difficulties:␤ Can't augment class Foo because it doesn't exist at /tmp/aG9wyrJklb line 1:␤------> augment class
..…
03:59 ashleydev joined
sorear std: class Foo::Baz { }; class Foo { class Bar { augment class Baz {}; }; } 03:59
p6eval std 31790: OUTPUT«===SORRY!===␤Can't augment class Baz without MONKEY_TYPING at /tmp/kpi_NJM45B line 1:␤------> ass Foo { class Bar { augment class Baz ⏏{}; }; }␤Check failed␤FAILED 00:01 115m␤»
sorear How does that work?
tadzik use MONKEY_TYPING;
sorear std: use MONKEY_TYPING; class Foo::Baz { }; class Foo { class Bar { augment class Baz {}; }; } 04:00
p6eval std 31790: OUTPUT«Compiled ./lib/MONKEY_TYPING.pm6␤ok 00:01 115m␤»
sorear std: use MONKEY_TYPING; augment class Foo {}
p6eval std 31790: OUTPUT«Potential difficulties:␤ Can't augment class Foo because it doesn't exist at /tmp/uY2COpb4mH line 1:␤------> use MONKEY_TYPING; augment class Foo ⏏{}␤ok 00:01 114m␤»
sorear How does that work? 04:01
tadzik I think you need to have a class first
sorear How does Perl 6 find baz
tadzik: no. I want to know why the Baz version *works*
tadzik no idea
sorear If I ask "How does X work", you should assume that I'm talking about something that works.
If I want to know about something that doesn't work, I will ask "Why doesn't X work" 04:02
tadzik my bad
04:02 paul1234 left 04:05 kkrev left 04:07 kkrev joined
sorear ah, I see, I think 04:09
STD seems to be aliasing $*CURLEX to $*CURPKG
04:09 xinming left
sorear 23:09 <@Alias_> I continue to think that until Perl 6 is real, until you can have it fully installed and experimentable in less than a minute, nobody will really care very much 04:10
I think this sounds like a request for binary packages of Rakudo
Is that doable?
04:12 colomon_phone left
snarkyboojum sorear: only on a .NET calculator 04:12
sorear well, Alias maintains the official windows packages of Perl5 04:13
cxreg sorear: this from the guy who's still scared of Git
sorear so .NET works
cxreg: huh? 04:14
cxreg i think debian is already shipping rakudo
yeah, it is
Get:4 ftp.us.debian.org experimental/main rakudo 0.1~2010.01-1 [1,357kB]
sorear: alias has git-phobia :) 04:15
or something approximating it
[Coke] trying to post to use.perl.org. article submission . preview .takes me to search.pl?op=journals, not a preview. wthf. 04:16
04:16 plobsing left 04:18 shade\ left
[Coke] righhht. can't even click on the login input box on firefox. 04:18
cross use.perl.org off my list of places to publish.
crap. can I haz a login on rakudo.org? 04:19
04:19 shade_ joined 04:21 paul1234 joined, kkrev left 04:22 mtnviewmark_ joined
[Coke] I will not be offended if someone just posts the announcement on rakudo.org for me. I will also take a login and do it myself. or take a login after someone has done that. or a beer. 04:22
04:22 mtnviewmark_ left
[Coke] release is done, go about your business. =-) 04:22
04:23 kkrev joined
cxreg hacks SPI support into plparrot 04:24
snarkyboojum [Coke]++ # release-fu 04:25
04:34 Raynes joined
[Coke] is tiiiiiired. 04:36
szabgab cxreg: debian is also shipping padre but people still would like to have an easy, on click installer for it. Even on linux 04:37
04:37 kkrev left
[Coke] is also trying to get Albany.pm to meet next week. 04:37
anyone going to be in the area on thursday? =-)
cxreg szabgab: so use ubuntu :D
[Coke] szabgab: you would be proud!
szabgab I think easy installation is critical for acceptance
snarkyboojum szabgab: enjoyed your screencasts - keep em coming :)
cxreg szabgab: why should rakudo be easier to install than perl 5? 04:38
szabgab cxreg: I am using Ubuntu, but I am not sure why does that make a difference?
snarkyboojum cxreg: why shouldn't it be? :D
szabgab cxreg: only if we want people to start using it :)
cxreg heh
04:38 kkrev joined
[Coke] for most users, installing perl is a noop. =-) 04:38
"oh, look, it came with the system." 04:39
cxreg well whatever. ubuntu has those fiddly things for adding your own sources now
we should set one of those up maybe
snarkyboojum what's happening with try.perl6.org?
szabgab [Coke]: right but that's an old perl
not the one freshly baked by p5p 04:40
04:40 molaf joined
tylercurtis rakudo: enum Foo <a b c> ; my Foo $f = Foo::a; say $f; my Foo $g = a; say $g; 04:41
p6eval rakudo d1cd71: OUTPUT«Type check failed for assignment␤ in '&infix:<=>' at line 1␤ in main program body at line 22:/tmp/GsJnM6Geob␤»
tylercurtis rakudo: enum Foo <a b c> ; my $f = Foo::a; say $f; my $g = a; say $g;
p6eval rakudo d1cd71: OUTPUT«0␤0␤»
tadzik snarkyboojum: moritz says the parts are pretty much done, and someone needs to stick it all together
it's try.rakudo.org by the way
tylercurtis rakudo: enum Foo <a b c> ; my $f = Foo::b; say $f; my $g = c; say $g; 04:42
snarkyboojum tadzik: oops ok - but cool
p6eval rakudo d1cd71: OUTPUT«1␤2␤»
szabgab case: I was just in a talk about garlicsim.org/ a simulator, I had my windows netbook with me. I looked for the windows installer, installed it and it was not the right one (it was without a gui) so I uninstalled and went home
tylercurtis rakudo: enum Foo <a b c> ; my $f = Foo::b; say $f; my $g = c; say $g; say Foo::a ~~ Foo;
p6eval rakudo d1cd71: OUTPUT«1␤2␤0␤»
tylercurtis Enumbug =/
rakudo: say Bool::True ~~ Bool; 04:43
p6eval rakudo d1cd71: OUTPUT«1␤»
tylercurtis rakudo: enum Foo <a b c> ; my $f = Foo::b; say $f; my $g = c; say $g; say Foo::a ~~ Foo; say Foo::a ~~ Int;
p6eval rakudo d1cd71: OUTPUT«1␤2␤0␤1␤»
snarkyboojum tadzik: not sure if I can help, but happy to if there's anything obvious to be done 04:45
:)
tadzik snarkyboojum: don't ask me, I have no idea :)
szabgab snarkyboojum: you are welcome (re screencasts) but I am also trying to figure out how could they be improved? 04:50
I see a sharp drop in the view stats from episode 1 to 4 04:51
04:51 kkrev left
szabgab viewers: 864 - 203 - 107 - 39 04:51
04:52 kkrev joined
tylercurtis szabgab: @viewers[2,3]>>++ 04:54
tylercurtis is now getting around to watching the 3rd and 4th.
sorear szabgab: what youtube doesn't tell you is the number of people who watched the end of #4 04:55
I bet it's around 50
er
the end of #1
szabgab maybe, but I'd still like to understand why people don't watch till the end? 04:56
is it as Alias said, to slow, archaic editor, not easy to try it? 04:57
sorear szabgab: because you didn't explain what the screencasts are in your blog posts 05:00
people had to start watching to know what you meant 05:01
after doing that, they didn't need to visit the rest
at least that's why I only visited the first
05:03 kkrev left, shade_ left
szabgab rakudo: my %h = ("Foo" => 2, "Bar" => 3); %h.fmt("%s %s"); 05:03
p6eval rakudo d1cd71: ( no output )
szabgab rakudo: my %h = ("Foo" => 2, "Bar" => 3); say %h.fmt("%s %s");
p6eval rakudo d1cd71: OUTPUT«Bar 3␤Foo 2␤»
szabgab rakudo: my %h = ("Foo" => 2, "Bar" => 3); say %h.fmt("%s<%s>"); 05:04
p6eval rakudo d1cd71: OUTPUT«===SORRY!===␤Symbol '%s' not predeclared in <anonymous> (/tmp/Fq9euLarDL:22)␤»
szabgab rakudo: my %h = ("Foo" => 2, "Bar" => 3); say %h.fmt("%s <%s>");
p6eval rakudo d1cd71: OUTPUT«Foo <2>␤Bar <3>␤»
05:04 kkrev joined
szabgab rakudo: my %h = ("Foo" => 2, "Bar" => 3); say %h.fmt("%s\<%s>"); 05:04
p6eval rakudo d1cd71: OUTPUT«===SORRY!===␤Symbol '%s' not predeclared in <anonymous> (/tmp/o77_iCYm4B:22)␤»
szabgab rakudo: my %h = ("Foo" => 2, "Bar" => 3); say %h.fmt("%s\<%s\>");
p6eval rakudo d1cd71: OUTPUT«Foo<2>␤Bar<3>␤»
szabgab rakudo: my %h = ("Foo" => 2, "Bar" => 3); say %h.fmt("%s\{\"%s\"\}"); 05:05
p6eval rakudo d1cd71: OUTPUT«===SORRY!===␤Unable to parse postcircumfix:sym<{ }>, couldn't find final '}' at line 22␤»
szabgab I wonder if there is an explanation for these errors or if they are bugs? 05:06
sorear try std:
szabgab std: my %h = ("Foo" => 2, "Bar" => 3); say %h.fmt("%s\{\"%s\"\}");
p6eval std 31790: OUTPUT«ok 00:01 120m␤»
szabgab std: my %h = ("Foo" => 2, "Bar" => 3); say %h.fmt("%s<%s>"); 05:07
p6eval std 31790: OUTPUT«===SORRY!===␤Variable %s is not predeclared at /tmp/FzktIaYAJf line 1:␤------> ("Foo" => 2, "Bar" => 3); say %h.fmt("%s⏏<%s>");␤Check failed␤FAILED 00:01 120m␤»
sorear '' is probably better if you don't want interpolatin 05:10
05:11 buubot left
szabgab rakudo: my %h = ("Foo" => 2, "Bar" => 3); say %h.fmt('%s<%s>'); 05:12
p6eval rakudo d1cd71: OUTPUT«Bar<3>␤Foo<2>␤»
szabgab right, I did not think of that
thanks
rakudo: my %h = ("Foo" => 2, "Bar" => 3); say %h.fmt('%s{"%s"}'); 05:13
p6eval rakudo d1cd71: OUTPUT«Bar{"3"}␤Foo{"2"}␤»
05:14 buubot joined
tylercurtis Is list assignment lazy? 05:15
sorear no
05:15 jaldhar left 05:16 jaldhar joined, kkrev left
sorear the exact status of list assignment wrt laziness hasn't been decided yet 05:16
to be safe, use :=
huh. STD just made a lexpad where OUTER:: points to a *package* 05:17
how does that make any sense?
05:17 hercynium left, hercynium joined 05:18 kkrev joined 05:20 lucs joined
sorear std: package Foo { my $x = 2; }; package Foo::Bar { say $x } 05:25
p6eval std 31790: OUTPUT«ok 00:01 118m␤»
sorear phenny: tell TimToady It's possibly to confuse STD.pm6 into setting up a package as the OUTER:: of a lexical scope. Try package Foo { my $x = 2; }; package Foo::Bar { say $x } 05:26
phenny sorear: I'll pass that on when TimToady is around.
sorear phenny: tell TimToady STD.pm6 line 5547 appears to be involved, but I'm not very knowledgable on the $*CURLEX handling 05:27
phenny sorear: I'll pass that on when TimToady is around.
snarkyboojum szabgab: It's most speculation I suppose. Not sure re the editor comment - could be. I suspect most people are comfortable using whatever editor they're most familiar with. I think it could also be related to people just dipping their toes in and having a peek - I know I often treat other "new" languages like that until I'm ready to give them a good look. Good luck with the rest.. I'll be sending links to them around and generally
advertising them on the web :)
s/most/mostly/
sorear What editor was szabgab using?
snarkyboojum vim I think
tadzik yep, vim 05:28
[Coke] ->> zzz
05:34 kkrev left 05:35 kkrev joined
szabgab one comment is sure important, that people need to be able to try out that stuff in order to want to go on watching, especially as I am not show the funky features of perl 6 but trying to teach it 05:40
so for the quick start I might need to direct them to the irc channel to use the bot
is there a web interface for the evalbot? 05:41
oh I remember I've already asked if I could talk to the evalbot in my privacy :)
but I don't remember how
tylercurtis szabgab: just send private messages to p6eval. 05:42
szabgab yep this worked: /msg p6eval rakudo: say 42
so let me try it from the web 05:43
tylercurtis rakudo: class Q_ { }; Q_.new.perl.say; 05:44
p6eval rakudo d1cd71: OUTPUT«Q_.new()␤»
tylercurtis rakudo: class Q { }; Q.new.perl.say;
p6eval rakudo d1cd71: OUTPUT«===SORRY!===␤Confused at line 22, near "Q.new.perl"␤»
tylercurtis rakudo: class Foo::Q { }; Foo::Q.new.perl.say;
p6eval rakudo d1cd71: OUTPUT«Foo;Q.new()␤» 05:45
snarkyboojum szabgab: once try.rakudo.org is up that might be worth mentioning 05:48
szabgab who is working on it? 05:49
I thought it would not be hard to connect a web pages to the current evalbot, would it ?
snarkyboojum szabgab: moritz_ put it up as a weekly rakudo challenge a while back.. not sure of status or who is working on it.. I was asking the same question earlier today
szabgab: maybe not terribly hard, but needs to be done safely etc (obviously) 05:50
szabgab: anyway, tadzik mentioned that perhaps most of the bits'n'pieces for it are complete, so it may not be far off 05:51
cxreg snarkyboojum: maybe you meant try.rakudo.org:8090 ? 05:54
seems pretty non-functional atm 05:55
tylercurtis ISTR someone having an at least somewhat working demo.
snarkyboojum cxreg: ah cool - seem the input field is currently disabled 05:56
seems*(
tylercurtis wonders if it's a mistake to attempt to parallel the naming conventions of the DTD in his HTML library. 06:01
tylercurtis wonders if it's even possible to use "naming conventions" to describe the HTML 4.01 Strict DTD without being a liar.
There's ALLCAPS, CamelCase, Firstlettercapitalized alllowercasescrunchedtogether, separated-by-hypens, separated.by.dots... Probably some more I'm forgetting. 06:04
06:10 uniejo joined
moritz_ in first approximation, everybody is a liar 06:16
szabgab hi moritz_ 06:19
moritz_ hi szabgab
szabgab so what about that web interface for the evalbot? 06:20
can I help with it to make it live?
moritz_ szabgab: I hope so :-)
szabgab what is its status?
moritz_ szabgab: cygx++ implemented all the backend logic, but so far nobody plugged it into the front end
szabgab oh I can try to do that 06:21
what is the backend?
snarkyboojum moritz_: am happy to lend a hand if I can be of assistance too :)
moritz_ snarkyboojum: look at github.com/moritz/try.rakudo.org/
and szabgab too
there's an architecture.svg that tries to explain the structure
and a DESIGN_BACKEND file, which is probably more accurate 06:22
huh
it seems that cygx didn't commit all his stuff :(
szabgab so is this listening on some socket that is not 80 right? 06:23
moritz_ ah, it's on feather3 in ~cygx/try-rakudo/
szabgab: yes 06:24
szabgab and the client would connect to that socket, send in some code, get a session id and then the connection is closed
moritz_ the client connects to the CGI script
which then connects to the socket
szabgab then later the client should poll the server?
06:25 chee left
moritz_ yes 06:25
szabgab ok so the client sends in standard GET or POST request?
moritz_ yes
szabgab and is that part running already?
moritz_ yes, but not with the code in the repo...
just a sec 06:26
try.rakudo.org/try-rakudo.tar.gz
that's the current backend code
feel free to dump it into the repo
szabgab I'd prefer to talk to the eral backend on the server :)
moritz_ 'eral'? 06:27
'real'?
szabgab real
try.rakudo.org:8090/ seems to give me an id
mberends cxreg: pulled your Zavolaj patch, verified it and pushed to jnthn/zavolaj :-) 06:28
moritz_ szabgab: do you want an account on the development machine?
szabgab I think I had an account on fether
but I don't even remember its address
cxreg mberends: sweet 06:30
szabgab moritz_: so which is the development machine?
moritz_ szabgab: feather3 06:31
szabgab full name?
moritz_ feather3.perl6.nl
szabgab ty
moritz_ try.rakudo.org :-)
cxreg mberends: turned out to be pretty simple :)
tadzik moritz_: might giving me some tips about File::Find? The P5 version passes no arguments to the callback function, and the file info can be accessed via some global variables. Can it be just passing a hash in Perl6 version? 06:32
cxreg mberends: the trickiest part was learning that you can't store pir::null__P() in a variable and then use it
moritz_ tadzik: sure. Or just the file name as first param
mberends cxreg: yes, surprisingly elegant. nice work!
tadzik moritz_: I currently pass the whole path, and I wonder whether it's important to give anything else. P5 version gives, via those variables, a path without a final name, a final name, and a full path with a filename. Weird, imho 06:33
moritz_ tadzik: a good solution would be to pass in an object which stringifies to the full name, and has accessors for the filename too 06:34
szabgab moritz_: you you pls check if there is a user called gabor or szabgab there?
oh I see it now
moritz_ szabgab: created a new one (feather != feather3), see /msg
tadzik moritz_: maybe… but wouldn't creating an object be an overkill for such a simple task? Or maybe not. How do you think? These objects won't live in memory for long anyway 06:35
06:35 au|afkForReal is now known as au|irc 06:37 molaf left, paul1234 left
moritz_ tadzik: the alternative would be to use contextual variables 06:38
tadzik contextual variables? 06:39
moritz_ rakudo: sub f { say $*x }; my $*x = 3; f() 06:40
p6eval rakudo d1cd71: OUTPUT«3␤»
moritz_ it looks up in the call chain
06:40 meppel is now known as meppl 06:41 eternaleye_ left
tadzik I think I'll go for objects, they're quite common in Perl 6 as I see. But out of curiosity, in which synopse can I read about those contextual vars? 06:42
cxreg moritz_: is that much like a dynamic variable in perl 5? 06:43
au|irc it's now known as dynamic variables in perl6 too... 06:44
cxreg tadzik: to read about that in perl5, perldoc perlsub | less -p ' Temporary Values via' 06:45
au|irc tadzik: in S02, look for C<DYNAMIC> which talks about the * twigil :) 06:46
(and "temp", which is like perl5's "local") 06:47
tadzik intereseting 06:48
cxreg incidentally, (less -p)++
au|irc wow, learns something new every day. cxreg++ :) 06:49
szabgab it seems that ever call to try.rakudo.org:8090/ gives me the same session ID 06:50
cxreg szabgab: C is for Cookie
or not, i deleted mine and still got the same one 06:51
unless '2521d99fd2fb58f4' means 'empty session' 06:52
moritz_ [Coke]++ # release
06:52 mulander joined
moritz_ hugme: add szabgab to try.rakudo.org 06:53
hugme hugs szabgab. Welcome to try.rakudo.org!
cxreg moritz_: it's like wildlfe research; tag and release
szabgab ty moritz_ but I am not sure what to do with it:) 06:54
06:54 mulander left
sorear returns from walk with a much better understanding of p6 packages 06:54
szabgab first I wanted to try a command line client for that back-end
tylercurtis .u smiling face
phenny U+263A WHITE SMILING FACE (☺)
szabgab maybe wrapped in perl 5
so people could integrate it into their IDE... 06:55
cxreg sorear: a vision quest of sorts?
szabgab cxreg: so I tried POST try.rakudo.org:8090/send?2521d99fd2c158f4
sorear cxreg: of sorts.
szabgab and ten typed in 42.say and ctrl-d 06:56
and no feedback received
sorear pmichaud does them too, it really does help
cxreg interesting
szabgab cxreg: are you the same as cygx or just similar letter? 06:57
cxreg if that's a person, i'm not him 07:02
szabgab as I understand that's the person who wrote the backend and has it in his(?) home dir on feather 07:03
cxreg heh
there's not enough hours in the day for me to have an alter ego 07:04
szabgab with an alter ego you might be able to do twice as much work :) 07:06
cxreg then i'd be like 1/3 of au|irc 07:10
07:13 MichaelRWolf joined
tadzik wonders if File::Find could be as powerful as find(1) one day 07:14
au|irc feels rather hexagonal :) 07:16
07:17 ldd left
tylercurtis moritz_: are you around? 07:18
moritz_ tylercurtis: more or less :-) 07:20
tylercurtis moritz_: how does one submit a patch for Rakudo?
07:20 mberends left
moritz_ tylercurtis: mailto:rakudobug@perl.org 07:20
cxreg tylercurtis: 'Submitting patches' in the README describes the process 07:22
07:22 zulon joined
moritz_ if it's something short-and-quick you can also nopaste patches here 07:24
zulon hi!
moritz_ (with the risk of them getting lost in the noise (or signal) here)
hi zulon
cxreg i'm fond of the github fork method :)
zulon moritz_: I think I found a typo in your blog post
$x max= 0; 07:25
$x min= 10;
tylercurtis moritz_: I may have cut the runtime of the script colomon and pmichaud were working on optimizing by about 50%(very unscientific measurement) on my box, at least. I'll run a spectest in the morning and submit the patch assuming it doesn't break and colomon doesn't do the same thing before then.
zulon shouldn't it be the opposite?
tylercurtis zulon: no. $x max= 0; means that $x becomes whichever is greater of $x and 0.
zulon oh, right
moritz_ zulon: no. When I want to cllip to the range of 0..10, I want -1 to result in 0
snarkyboojum which blog post? :D
moritz_ so -1 max 0
zulon yes, I'm stupid 07:26
sorry :)
moritz_ snarkyboojum: perlgeek.de/blog-en/perl-5-to-6/27-...dioms.html
zulon: I've made the same mistake too
just a matter of experience
tylercurtis zulon: That doesn't make you stupid. Or at least, I hope not.
snarkyboojum moritz_: sweet - thanks
tylercurtis didn't get it immediately either.
moritz_ tylercurtis: I'm eager to see that patch :-) 07:27
mathw Morning
snarkyboojum waits to hear moritz_'s accent doing a screencast of sorts for perl5 to 6 ;)
moritz_ tylercurtis: colomon++ did some tweaks shortly before the release; not sure if they are the same as yours
snarkyboojum: which of mine? the German? or the false Scottish? 07:28
snarkyboojum moritz_: hehe 07:29
when is TimToady doing his talk at OSCON?
is he doing a talk at OSCON? 07:30
frettled is there an OSCON?
:)
moritz_: if you do a passable Connery impression, I'd love to hear it.
snarkyboojum frettled: I thought OurGreatLeader was speaking there :P 07:31
tylercurtis moritz_: no. Mine uses fdiv_n_n_n(which does work, unlike fdiv_i_i_i). On the plus side, it greatly simplifies infix:<div> and makes it look like most of the other (Int, Int --> Int) infixes. I mentioned it to colomon before, so he might end up commiting it, but in case he for some reason doesn't, I thought I should submit a patch. Besides, it's an excuse to get my name in the release announcement for next month. ;)
tylercurtis ==> sleep # will spectest and submit patch in slightly later morning. 07:33
07:34 tylercurtis left 07:43 baest joined
snarkyboojum sorry to belabour a point, but where will the state of the onion address for 2010 be accessible? 07:45
tadzik how about adding Term::ANSIColor to wiki.github.com/rakudo/rakudo/whats...nto-rakudo ?
07:49 mberends joined
moritz_ +1 07:52
07:56 d4rt joined, d4rt left 07:57 dumdum joined, d4rt joined
dumdum rakudo: say 42 07:58
p6eval rakudo d1cd71: OUTPUT«42␤»
dumdum rakudo: say 42 "more"
p6eval rakudo d1cd71: OUTPUT«===SORRY!===␤Confused at line 22, near "say 42 \"mo"␤»
08:01 dumdum left 08:02 au|irc is now known as au|afk
moritz_ rakudo has less-than-awesome error messages for two terms in a row 08:05
08:06 zulon left
szabgab I think I'll try to make a screencast to show people how to use the IRC channel ad the evalbot 08:06
mathw Like, not flooding the channel with evalbot experiments? 08:07
moritz_ or like, flooding the channel with evalbot experiments? :-)
08:07 MichaelRWolf left
szabgab I'll also show how to do it in a private channel 08:07
moritz_ on the channel you don't see that people also /msg the evalbot
ight
r 08:08
szabgab but the question will the server manage it?
moritz_ we will see :-)
it's worth a trial
szabgab let's try it
tadzik szabgab: what was your audience last time (views)?
szabgab 876 - 207 - 111 - 60 08:09
for the 4 parts
moritz_ rakudo: say [+] 876, 207, 111, 60 08:10
szabgab the last one was uploaded 12 hours ago
tadzik so a few dozens more users of the evalbot, where the new users will most probably issue loads of requests
p6eval rakudo d1cd71: OUTPUT«1254␤»
szabgab rakudo: say [*] 876, 207, 111, 60
p6eval rakudo d1cd71: OUTPUT«1207671120␤»
szabgab rakudo: say [**] 876, 207, 111, 60
moritz_ runs an upgrade on feather3
p6eval rakudo d1cd71: OUTPUT«Inf␤»
moritz_ don't be surprised if things work funnily for a while
Get:1 ftp.us.debian.org sid/main libc-bin 2.11.2-2 [704kB] 08:11
Get:2 ftp.us.debian.org sid/main libc6 2.11.2-2 [3,877kB]
tadzik yay, every day is a new-perl6-post-on-reddit day
moritz_ that's a good start :-)
08:12 charsbar_ joined, xabbu42 joined
tadzik moritz_: is there a perfectly POD-documented module to be inspired with? 08:13
moritz_ tadzik: I have no idea. masak++ might know
tadzik masak is absent, sadly :)
08:13 justatheory joined
moritz_ I know. He usually shows up during the day, though 08:14
tadzik in what timezone does he live?
snarkyboojum tadzik: nice alliteration :p 'perfectly POD-documented'
08:14 pjcj left, dimid_ left
moritz_ UTC+2 (central europe) 08:14
tadzik triple P
08:14 Tene left
tadzik could be a badge in proto :) 08:15
08:15 charsbar left, Tene joined, Tene left, Tene joined, dimid joined
tadzik so, what nifty modules would be nice to have in R*? 08:15
08:15 pjcj joined 08:16 justatheory left
moritz_ all nifty modules, please :-) 08:16
moritz_ wants LWP-Simple
and MiniDBI
and a ponie!
tadzik they're alredy on a list
hmm, Acme::Ponie would be a nice Cake for R* 08:17
it should be able to do everything
and more
moritz_ a libjustdoit wrapper, via zavolaj!
tadzik Net::FTP would be nice. Do we have binary stuff working yet? 08:21
moritz_ ask masak :-)
tadzik file writing and stuff
:)
08:21 daxim joined
moritz_ mostly, yes 08:21
tadzik maybe some Term::Progressbar?
moritz_ I think it would benefit from a real-world application (the binary stuff)
08:22 kst left
tadzik or rather a real-world application would benefit from binary stuff 08:22
moritz_ every thing that's implemented is usually fragile at first, but the tests only cover some usage patterns 08:23
so only real-world application and bug reports lead to it working nicely
08:23 kst joined
tadzik hmm, that's right 08:23
sorear isn't the whole point of R* to attract real world applications to stuff that they would benefit?
moritz_ sorear: that's right 08:24
08:24 Eevee left
moritz_ sorear: but if we have more stuff working, we'll get error reports on a wider front 08:24
08:24 aesop joined
moritz_ so, more benefit here 08:24
08:24 Eevee joined
tadzik benefits lead to another benefits \o/ 08:24
Raynes I read that as "benefits lead to friends with benefits \o/" 08:25
Had to double-take.
08:26 tadzik left
sorear mberends: How would you feel about vill being moved from src/perl6 to src/ ?\ 08:28
mberends: for niecza, and any future project seeking to track STD, I'm wanting src/perl6 subcheckouts to be as lightweight as possible 08:29
08:30 pugssvn left, pugssvn joined 08:31 sorear sets mode: +v pugssvn, lestrrat is now known as lest_away 08:33 xinming joined
mberends sorear: vill should probably move to a git repo anyway. Will do that soon. 08:36
daxim ⟫Ö⟪ giant camelia 08:37
08:38 dakkar joined
patrickas good morning perl6! 08:46
mathw hi patrickas 08:48
patrickas mathw o/
08:49 pmurias joined
pmurias sorear: why not use the STD on CPAN? 08:50
sorear: i'll add what you need and there will be no need to checkout STD
sorear pmurias: a stable extension API and the ability to do automatic installs tracking changes to a STD_REVISION file 08:52
sorear wonders if local::lib will help 08:53
08:53 kst left
pmurias automatic installs tracking changes to a STD_REVISION file? explain 08:53
sorear TimToady++ regularly makes breaking changes to STD 08:54
having a STD_REVISION file allows me to track STD at my own pace
I have Makefile rules to automatically download and build STD whenever _REVISION changes
pmurias you want to avoid changes that break STD or change the AST format a bit? 08:55
sorear the latter 08:56
08:56 mac joined
mac hi 08:56
sorear Hello, mac
Welcome
pmurias: also, occasionally the language spec gets changed, like when YOU_ARE_HERE became {YOU_ARE_HERE}
pmurias had to fix that one recently 08:57
sorear that required modifications to the niecza setting and compiler
having explicit revision tracking means that I can update and fix in the same commit
and the evalbot never breaks
pmurias if you want to make niecza installable you will either have to prefix your STD like Niecza::STD or use the CPAN one 08:58
sorear I don't plan to make niecza installable from CPAN 08:59
pmurias why not?
sorear using Perl 5 is strictly a bootstrapping stunt for me
my #1 goal is to be able to compile STD.pm6
pmurias is not very keen on bootstrapping 09:00
sorear well, the entire compiler exists because I became frustrated with how inefficient viv/Cursor is 09:01
pmichaud good morning, #perl6 09:07
moritz_ good morning, pmichaud
mac morning? huh ;)
sorear good * pmichaud
mac Some of us have to wake up a "little" earlier;)
09:07 DemoFreak left
frettled or just good pmichaud :) 09:07
sorear mac: did we scare you away?
moritz_ frettled: like in China, where you say "ni hao", "you good" :-)
frettled moritz_: yes, exactly!
09:07 Mowah joined
pmurias TimToady: what can i rename to/prefix with LazyMap, Cursor, Actions so i don't get an UNAUTHORIZED release marker on CPAN 09:07
09:07 jnthn left, jnthn joined
mac No, sorear, in spite of I feel fine 09:08
09:08 DemoFreak joined
moritz_ it's always morning *somewhere* 09:08
09:09 _sri joined
pmichaud I've posted the release announcement to rakudo.org. Coke++ 09:11
09:12 broquaint joined
daxim what's the included book? 09:12
moritz_ daxim: "Using Perl 6" 09:13
daxim url?
moritz_ not quite finished yet :/
github.com/perl6/book/
daxim thanks
pmichaud hugme tweet rakudoperl Rakudo #31 "Atlanta" released: rakudo.org/node/74 09:14
hugme hugs pmichaud; tweet delivered
pmichaud hmm, I probably should've included "#perl6 #rakudo" in that tweet. 09:15
pmichaud is still trying to get the hang of twitter
moritz_ it's the first release where I'm truely convinced that it's better than alpha, for nearly all applications
sorear is somewhat suprised there wasn't a fork 09:16
09:16 patrickas left 09:17 patrickas joined
moritz_ what kind of fork? that continued developing alpha? 09:17
sorear yes
09:17 a3r0 joined
pmichaud alpha had lots of issues -- in many ways it had reached the end of being a useful development line 09:17
also, it was based on PGE, which didn't have protoregexes 09:18
(still doesn't)
sorear PGE is still maintained? 09:19
pmichaud not really
moritz_ it's mostly unbroken when a parrot change breaks it
09:19 _macdaddy left
moritz_ I wouldn't exactly call that maintenance :-) 09:20
pmichaud time for a bit more sleep here -- bbl 09:21
sorear ok, so it's as maintained as pugs 09:22
09:24 patrickas left
daxim .[ ^10 ]; # 10 most common words - where is the circumflex piece of syntax explained 09:26
moritz_ daxim: S03. Search for "upto"
pmurias sorear: who would want to hack on a dead-end branch of rakudo?
avar are there pre-built pdfs of the p6 book available somewhere?
moritz_ avar: yes, in the download section
they don't have the latest markup improvements, though 09:27
avar ah, from -06, seems a bit old :)
moritz_ avar: pmichaud++ volunteered for making a release this month
pmurias moritz_: would it be difficult to make automatic pdf builds of the book? 09:29
sorear pmurias: I was under the impression that point was highly contested
moritz_ pmurias: currently it's difficult to build, because it has quite some requirements on installed latex packages, tools and fonts 09:30
sorear What is the full set of psuedo packages that are legal after CALLER::? 09:32
just OUTER, UNIT, SETTING? 09:33
pmurias CALLER::CALLER?
moritz_ CALLER:: too
sorear only those 4? 09:34
ok
moritz_ not sure
there are always surprises in Perl 6
I wonder if there should be a syntax for declaring pseudo packages
mathw wonders why he's in the list of contributers to the Rakudo release 09:36
09:36 wamba joined
Juerd Of course. Colon, star, or backtick? 09:36
a3r0 I think Rakudo logo had better be modified,Japanese Shrine's gate is too religious and give a reluctant feeling to those who live in the country which was one of the japanese colony in the past..
like korea, china ,tiwan... 09:37
moritz_ a3r0: are you talking of personal experience, or general sentiment?
a3r0 i think it's general sentiment.
if we make t-shirt which contains shrine's gate.
moritz_ mathw: if you were ++'ed in one of the commit messages, you are mentioned
a3r0 I can't wear it in korea.
09:38 patrickas joined
moritz_ mathw: like, for mentioning that something's missing in a commit, spelling mistakes, LTA error messages or so :-) 09:38
a3r0 It's like nazi mark in Europe.
mathw moritz_: maybe it was a bug I spotted or something
moritz_ a3r0: it's new to me that it's that bad... do you have a link to something that describes the cultural resentments more deeply? 09:41
a3r0 en.wikipedia.org/wiki/Yasukuni_Shrine 09:42
www1.voanews.com/korean/news/a-35-2...2005-10-17 09:44
pugssvn r31791 | sorear++ | [S02] Move CALLER:: to the "reserved anywhere in a name" section, because CALLER::CALLER:: is explicitly described. 09:45
moritz_ buu: seen JimmyZ 09:46
sorry
buubot: seen JimmyZ
buubot moritz_: I don't think I've seen JimmyZ.
moritz_ useless bot
frettled a3r0: I fail to see the connection. The links you gave are regarding a specific disputed shrine, not about Japanese logotypes representing (or not representing) a shrine being a general problem. Or have I missed your point?
09:47 lest_away is now known as lestrrat 09:48 lestrrat is now known as lest_away
a3r0 Japan forced to go and worship the shirine to japanese colony countries. 09:49
moritz_ it's also curious that so far nobody from said countries complained 09:50
09:50 pat__ joined
frettled moritz_: it might be a generational thing 09:50
moritz_ neither JimmyZ from China, nor au|afk from Taiwan (iirc) 09:51
sorear I'd be suprised if JimmyZ was the only one 09:52
moritz_ right; but it's the one I explicitly remember as being from China :-) 09:53
09:53 patrickas left 09:56 envi^home joined 09:58 JimmyZ joined
frettled a3r0: What I don't understand is the connection between the Rakudo logo and what you're talking about. There's something missing, something that can explain to me how that logo is offensive. I understand that Japan's actions before, during, and after WW II have been reprehensible to many people in greater Asia, which is what your links also essentially tell a part of the story of. But what's the direct connection? You say it's like "the nazi mark" ... 09:59
... (I suppose you mean the black swastika on a white circle on a red flag), but I want to know _how_ it is like that.
JimmyZ Hi~~
moritz_ hi JimmyZ; if you backlog a few minutes, you might find an intersting discussion 10:00
frettled feels mostly clueless regarding that part of cultural history.
10:01 zulon joined
frettled goes for chocolate cake. 10:03
JimmyZ xingming has been being here since 2005, he was from china 10:04
moritz_ had some hazelnut cake already. Very tasty.
JimmyZ: do you have any bad feelings towards the Rakudo logo? 10:05
JimmyZ Nope
乐土 is chinese :) 10:06
a3r0 korean think shrine == militarism.
10:07 gbacon joined
JimmyZ and I like the door, hehe 10:07
10:07 hanekomu_9 joined
sorear here in the USA, which never was a Nazi colony, you don't even need the white circle and the red flag 10:08
the swastika alone is enough to get you spat on in the streets
moritz_ same here (obviously) 10:09
it's weird for Germans to travel to India, where it's still used in its original context :-)
sorear no similar situation exists with temple iconography, despite the fact that the Japanese came much closer to occupying US territory than the Germans ever did
a3r0 rki.kbs.co.kr/english/korea/korea_h..._con81.htm 10:10
rki.kbs.co.kr/images/img_history/8-...ligion.jpg
In order to obiliterate religion among Koreans, Japan built a shrine on Mt. Namsan in Seoul in 1920. All Koreans were forced to visit the shrine to worship the Japanese emperor and the Gods idolized by the Japanese people.
10:10 Sanitoeter left 10:11 zulon left 10:12 hanekomu_9 left
a3r0 i hope this article would be a good reason to modify. 10:12
moritz_ in the end it's pmichaud's call 10:14
10:14 Velvet joined 10:15 hanekomu_9 joined, hanekomu_9 left
sorear does $MY::x do a full lexical scope, hitting outers too? 10:15
10:17 a3r0 left 10:20 Sanitoeter joined
pmurias sorear: is it ok if i rename LazyMap,Action,Cursor to STD::LazyMap... 10:20
?
10:24 gbacon left 10:27 tadzik joined
sorear pmurias: you're asking the person who originally proposed it. of course it's ok with me 10:29
it's TimToady who might (already has) object(ed)
10:31 azert0x joined, JimmyZ left 10:32 masak joined
masak oh hai, #perl6! 10:32
phenny masak: 01:26Z <tadzik> tell masak you forgot to bold the Operator Overloading section
masak: 01:28Z <au|irc> tell masak great blogpost :) looks like use.perl ate the <...> in infix:<...> as markup, though
10:32 au|afk is now known as au|irc
masak tadzik, au: thanks. fixed. 10:32
<JimmyZ> and I like the door, hehe
LOL
interesting question about cultural associations and WWII. 10:33
au|irc .u 100DC 10:34
phenny au|irc: Sorry, no results for '100DC'.
au|irc hm. it's the closest I can get to a shrine in unicode :)
well, the rakudo.org artwork is specifically a Shinto shrine, so I can see the WWII connection 10:36
otoh, it's a nice diversity from the general Judeo-Christian vocabulary that perl uses :) 10:37
masak I've always felt Rakudo's logo to be a bit austere, especially next to the colorsplosion that is Camelia.
au|irc I like it, fwiw; it has that minimal Zen feel
masak huh -- release announcement was sent to parrot-dev, p6l and p6u -- but not to p6c. :) 10:39
szabgab I thought it was a gate but now I see it is a shrine, a gate to heaven?
10:39 zulon joined
masak oh well, [Coke]++ 10:39
au|irc szabgab: a gateway to the Perfect Language
which, as per S01, is defined as Perl 7 :) 10:40
10:40 |Velvet| joined
jnthn o/ 10:42
au|irc o ̷
jnthn [Coke]++ # release...I fell asleep before it happened :-) 10:43
szabgab o _
10:43 Velvet left
jnthn cxreg: What's your github id? 10:43
szabgab another gate: hu.wikipedia.org/wiki/Székelykapu 10:44
au|irc nice geometric constrast 10:46
*contrast
colomon can we check in patches yet?
errr, push commits? ;)
10:46 daxim left
jnthn colomon: Would seem so. Our only extra constraint is that we can't bump PARROT_REVISION until the 30th. :-) 10:47
colomon will we cut another release of Rakudo for R*?
masak that's an odd constraint :)
mathw because R* has to be based on a released version of Parrot 10:48
10:48 |Velvet| left
jnthn colomon: I somewhat suspect so. For example, we really want a fix for the roles + settings issue in R*. 10:48
colomon: And I didn't manage to get that into this release. :( 10:49
(er, yesterday's relesae)
10:49 Velvet joined
colomon I've got a patch based on tylercurtis's work last night which should make current Rakudo twice as fast as Kiev for that script we were working on optimizing yesterday. 10:50
jnthn colomon: Nice 10:52
colomon: So long as it doesn't need a Parrot revision bump, it's fine (and welcome :-)).
colomon tylercurtis++ did all the heavy lifting, I just put it together nicely.
pushed 10:53
10:53 Velvet left, Velvet joined
masak the trolls have joined the party. www.reddit.com/r/programming/commen..._knew_you/ -- give it a day or so, and the good guys'll show up, too. 10:53
colomon there are good guys on reddit? ;) 10:54
masak moritz_, chromatic...
dalek kudo: bc3e181 | (Solomon Foster)++ | src/core/Int.pm:
Reimplement infix:<div> to use pir::fdiv. tylercurtis++.
masak realises he shouldn't keep a list of The Good Guys in one easy-to-find place :) 10:55
au|irc masak: "map { $_ * $_ }, (1 ... *)" needs the parens :)
masak oh dear. yes. 10:56
masak fixes
frettled moritz_: Ah, well, at least nobody elected to use the Norwegian flag, which would offend a lot of moslems. :)
colomon au|irc: it's great to have you back with us here. :)
au|irc colomon: it's great to be back with y'all here :) 10:57
frettled moritz_: or the Swedish, in which case the Norwegians would be offended, right, masak? ;)
masak frettled: I have no ideas what offends you wusses over there. :P
colomon au|irc: you gave me my commit bit back in 2005 or so. I tweaked a thing or two, then ignored it for four years. 10:58
sorear it's great to have au here *period*
we're dangerously close to a critical mass of optimism
TiMBuS aw masak, you used the 'map' version instead of my genius solution 10:59
au|irc colomon: ah. good things that commit bits, unlike cookies, don't expire...
masak frettled: by the way, I spent a week in Värmland. it's so close to the Norwegian border, that one feels the influence of the language and the culture. I picked up a few Norwegian words.
colomon au|irc: :)
sorear au|irc: be careful what you wish for, I've had cookies without expiration dates before, they aren't all that great
masak TiMBuS: simplicity. your version was nice, though.
TiMBuS lol
frettled masak: oh noes! 11:01
masak I also feel like a "did nothing for four years" person. my biggest contribution in the Pugs era was trying, and failing, to port CPAN's Getopt::Long.
frettled :) 11:02
sorear I first payed attention to Perl 6 in Mar or Apr 2010
colomon my biggest contribution was noting that array reference via [ ] was O(N^2). (That was the very early days of Pugs!)
frettled my biggest contribution so far is joking in good humor on this channel ;)
au|irc colomon: ah right, via sieve.p6 11:03
jnthn I'm trying to remember what my first Pugs patch was. I think it was some copy-pastery that got the ASCII version of hyper-ops to parse as well as the unicode-y ones, which I was too lazy to work out how to type. :-)
au|irc very early indeed
colomon au|irc: just so.
masak told you the D&D-ers would like it :P -- twitter.com/dotandimet/status/19329500141 11:04
in other news, more than one name for anything confuses people: twitter.com/hiro_asari/status/19300163352 11:05
sorear colomon: N^2 !?
that's worse than a theoretically optimal linked list
colomon sorear: might have just be O(N) on each access. it was a long time ago. 11:06
frettled masak: all of those look like good PR to me
colomon that would have made seemingly O(N) code O(N^2)
sorear wonders what his biggest contribution was
colomon at any rate, it was very slow
au|irc indeed. seeing it was my second haskell program, lack of proficiency in managing data structures can be assumed :) 11:07
jnthn masak: They're all names for different things too. ;-)
masak sorear: I have a question. what's the best point in viv to go in and pull out a hierarchical data structure, rather than output it to STDOUT? or is there a better script than viv for this? 11:08
jnthn: tell me about it! if they were at least different names for the SAME thing, it's be slightly easier! :P
this looks like good PR, as well: www.pcworld.com/article/201743/perl...k=rss_news 11:09
frettled it almost looks like you're about to imminentize the escathon.
masak isn't that the whole idea? 11:10
"Apocalypse"... :)
sorear masak: if you just want a tree, rather than deparsed output, viv is entirely dead weight
use STD.pm6 directly
au|irc frettled: I see. so it's an Erisian Shrine and nobody would be offended. it all works out!
masak sorear: oki. any example code somewhere for that kind of thing?
colomon I like the "hints" there.
sorear perl -MSTD -E 'say YAML::XS::Dump(STD->parsefile("foo"))'
masak sorear++ 11:11
anyblag; I'm going offline now. going on a road trip with two friends. might not come back online today.
o/
szabgab finally a reddit conversation I could add my wealth of education
11:11 masak left
au|irc \o 11:11
frettled au|irc: or everybody should be offended. As you say, it all works out. :D
au|irc :) 11:12
frettled szabgab: oh, nice :) 11:13
11:21 screencast joined, xinming left
screencast hi , I am new to perl 6, I'll need some help ! 11:21
can someone tell me how can I add two numbers and print to the screen 11:22
cono rakudo: say 2+2
p6eval rakudo d1cd71: OUTPUT«4␤»
11:22 screencast left, xinming joined
szabgab crap 11:22
I had to stop the recording :( 11:23
au|irc nice irc performance art :)
szabgab let me start again :)
11:24 screencast joined
screencast hi I am new to perl 6, I'll need some help from you 11:25
how can I add two numbers and print them to the screen?
au|irc rakudo: 2+2
p6eval rakudo d1cd71: ( no output )
jnthn rakudo: say 39 + 3
cono :)
p6eval rakudo d1cd71: OUTPUT«42␤»
au|irc evidently, the "say" here is critical 11:26
jnthn I'd say so.
au|irc screencast: does that answer your question? :)
11:26 zulon left
screencast yes thank you 11:26
Juerd Also, existing Perl 5 code like "print 2 + 2" still works :) 11:27
cono rakudo: (2 + 2).say 11:28
p6eval rakudo d1cd71: OUTPUT«4␤»
au|irc rakudo: [+](2,2) 11:29
p6eval rakudo d1cd71: ( no output )
11:29 screencast left
au|irc rakudo: say [+](2,2) # sigh :) 11:29
p6eval rakudo d1cd71: OUTPUT«4␤»
jnthn or just
rakudo: say [+] 2,2
frettled rakudo: say :13<6*9>
p6eval rakudo d1cd71: OUTPUT«4␤»
rakudo d1cd71: OUTPUT«===SORRY!===␤Malformed radix number at line 22, near "<6*9>"␤»
jnthn prefers it without the parens :-) 11:30
frettled hmm
rakudo: say :13(6*9)
p6eval rakudo d1cd71: OUTPUT«69␤»
szabgab man , I have stage fright even when off-line recording :)
frettled nooo :(
rakudo: :13(6*9).say 11:31
szabgab let's see if it worked out in the recording well :)
p6eval rakudo d1cd71: OUTPUT«69␤»
au|irc rakudo: say 2 [[[+]]] 2
p6eval rakudo d1cd71: OUTPUT«4␤»
au|irc jnthn: cool, it works without the parens either way :)
pat__ au|irc : \o/
cono rakudo: eval("2+2").say
p6eval rakudo d1cd71: OUTPUT«4␤»
jnthn au|irc: It does occur to me that with the [[[+]]] nesting, we actually do have an infinite number of ways to add two numbers together in Perl 6. :-) 11:32
cono TMTOWDI
jnthn TAINOWTDI
au|irc TA∞WTDI 11:33
felliott rakudo: "hello".chomp.trans.say
moritz_ just started reading the pcworld article, and noted that it links to perl6.org, not the old dev.perl.org/perl6/ thing. \o/
p6eval rakudo d1cd71: OUTPUT«hello␤»
felliott rakudo: "hello\n".chomp.trans.say
p6eval rakudo d1cd71: OUTPUT«too few positional arguments: 1 passed, 2 (or more) expected␤ in main program body at line 22:/tmp/shV3pbU1mO␤»
frettled rakudo: $x=6*9; :13($x).say
p6eval rakudo d1cd71: OUTPUT«===SORRY!===␤Symbol '$x' not predeclared in <anonymous> (/tmp/VGEJIahW0b:22)␤»
jnthn moritz_: Yes, I was happy to see it was in general linking to good stuff.
frettled oh, drat, my
rakudo: my $x=6*9; :13($x).say
11:33 pat__ is now known as patrickas
moritz_ felliott: that's a bug 11:33
p6eval rakudo d1cd71: OUTPUT«69␤»
frettled copy+paste is so nice when you forget. hmm.
rakudo: my $x=6*9; :10($x).say
p6eval rakudo d1cd71: OUTPUT«54␤» 11:34
felliott moritz_: it seems to be a problem with Cool-str::substr and Any-list::join
cono jnthn: what means TAINOWTDI?
frettled Aha, I see what's going on, I confused the direction of the radix thingy.
jnthn cono: There's An Infiniate Number Of Ways To Do It
cono :D
jnthn *Infinite
moritz_ felliott: if you can figure out how to patch it, that would be awesome
felliott both return pir::<some string function>
11:34 tadzik left
moritz_ right 11:34
jnthn oh eww that bug 11:35
felliott I have a guess, but I'm not entirely sure...
moritz_ maybe add a ~ in front?
rakudo: say pir::substr__SSII('foobar', 1, 2)
p6eval rakudo d1cd71: OUTPUT«oo␤»
moritz_ rakudo: say pir::substr__SSII('foobar', 1, 2).trans
jnthn adds this one to his list of "problems that go away when Rakudo stops using various Parrot PMCs as base classes for it's builtin types"
p6eval rakudo d1cd71: OUTPUT«too few positional arguments: 1 passed, 2 (or more) expected␤ in main program body at line 22:/tmp/l8y45WPg8N␤» 11:36
felliott mortiz_: yes
moritz_ rakudo: say (~pir::substr__SSII('foobar', 1, 2)).trans
felliott I'm looking for a nopaste
p6eval rakudo d1cd71: OUTPUT«oo␤»
moritz_ felliott: nopaste.snit.ch
felliott If that's really the fix, we can close #76564 and #71088
nopaste.snit.ch/22234 11:37
I've also pushed it to github: github.com/felliott/rakudo/tree/par...rakudo_str 11:38
moritz_ felliott: spectesting now with these changes
felliott: if you want extra karma, you could make sure that we have tests for that in t/spec/integration/real-strings.t
felliott with it, I was able to unfudge 6 tests from spec/integration/real-strings.t
:)
moritz_ ah :-)
felliott good timing.
colomon \o/
moritz_ no need for the parenthesis btw.
frettled Am I missing something, or is there simply not a builtin for printing something in an arbitrary base?
felliott Oh, thanks! 11:39
moritz_ frettled: I know that the other way is implemented
frettled moritz_: automated conversion into base 10 is implemented, yes.
felliott How should I submit a patch to the spec? Via rakudobug? 11:40
frettled moritz_: for some reason, I thought the output would be in the same base as specified with the adverbial
moritz_ felliott: you mean for t/spec/?
felliott yes
moritz_ felliott: just give me your email address (via private message if you prefer), and I can give you a commit bit. These tests are in the pugs repository.
felliott okay, thanks! 11:41
moritz_ felliott: invitation sent (you might need to remove the port number from the confirmation URL though) 11:42
I haven't found a way to remove it from the email template :( 11:43
felliott moritz_: received! thank you 11:45
sorear <3 testsuites
moritz_ felliott: fix pushed, thank you very much. I haven't waited for the spectests to finish, but I'm pretty certain it doesn't break anything :-) 11:46
sorear hmm. does perl 6 have unfolds?
felliott I ran them overnight and didn't see any breakage
moritz_ is pretty proud of the spectests, even if he contributed only a tiny part of them
sorear even niecza's tiny internal testsuite has already saved me a couple months of labor 11:47
so much subtle breakage caught by 'make test; leave terminal for a minute'
dalek kudo: 4bf6c0f | moritz++ | src/core/ (2 files):
return Perl 6 strings from substr and join. Patch courtesy of felliott++
 11:48
ecza: a10c69a | sorear++ | (10 files):
Heavily refactor and improve the functionality of packages.
 11:49
frettled perl6: my $x = 54; say :13($x); 11:51
p6eval pugs, rakudo d1cd71: OUTPUT«69␤»
pugssvn r31792 | felliott++ | unfudge P6 strings tests, add tests for join & substr 11:54
szabgab not good, the recording, after I converted it to avi thinks that it is 17 min long 11:55
dalek ecza: dc56884 | sorear++ | (3 files):
Implement module as a synonym for package
sorear Is it mandatory that Foo/Bar.pm contain a definition of Foo::Bar or Foo::Bar::* ?
moritz_ don't think so 11:56
au|irc sorear: not from the spec, no. the EXPORTALL can then fail though if it's loaded via use 11:57
but "need Foo::Bar" would work it seems 11:58
11:58 wamba left 12:01 tadzik joined 12:02 snarkyboojum left 12:08 felliott__ joined, felliott left 12:11 cognominal left 12:13 cognominal joined, kkrev left 12:16 snarkyboojum joined 12:17 kkrev joined
szabgab the screencast did not come out well, 12:17
I hope you don't mind if I do it again
sorear why would we
12:18 tuxuday joined
au|irc . o O ("rakudo: sleep 30; say 2+2") 12:19
moritz_ rakudo: sleep 30; say 2+2
p6eval rakudo d1cd71: ( no output )
12:20 screencast joined
frettled attempted to phrase his stupidish radix/base question on p6l, maybe enlightenment is on the way. 12:20
screencast hi I am new to perl 6 I'll need some help
how can I add two numbers and print to the screen? 12:21
jnthn rakudo: say 37 + 5
p6eval rakudo d1cd71: OUTPUT«42␤»
au|irc hugme: hug screencast
hugme hugs screencast
screencast thank 12:22
au|irc np. in #perl6 you'll get the answer to life, universe & everything, plus a hug :)
12:24 pmurias left
moritz_ au|irc: in case it's not obvious, I wrote hugme after your "hug a troll" blog post :-) 12:24
au|irc yeah, I know. I'm honoured. 12:25
12:25 screencast left
moritz_ I was just looking for a funny name for the add-github-contributors thing 12:25
szabgab ahhh 12:26
au|irc moritz++ # hugtastic, as mst++ would say 12:27
moritz_ lol 12:28
frettled au|irc: perhaps with a few more swearwords sprinkled around, for good measure 12:29
12:29 takadonet joined
takadonet morning all 12:30
frettled hugme: hug takadonet
hugme hugs takadonet
au|irc frettled: very true, but that's what makes mst++ inimitable :) 12:31
takadonet .... thanks :) 12:32
frettled au|irc: yup, and I really like his style
…mostly for being refreshingly different
au|irc =) 12:33
12:37 snarkyboojum left 12:38 Raynes- joined, Raynes- left, Raynes left 12:39 Raynes joined
sorear has decided that {YOU_ARE_HERE} is just a really funny looking Whatever, and curries the entire module it's embedded in into a funny looking closure 12:41
moritz_ whatever. :-) 12:44
au|irc . o O ( {Total_Perspective_Vortex} ) 12:48
12:48 patrickas left
sorear translation: I now know *exactly* how to implement custom settings 12:49
takadonet rakudo: my @a = <1 2 3>; say @a>>+2 12:52
p6eval rakudo 4bf6c0: OUTPUT«===SORRY!===␤Missing << or >> at line 22, near "2"␤»
12:53 patrickas joined
moritz_ rakudo: my @a = <1 2 3>; say @a>>+<<2 12:56
p6eval rakudo 4bf6c0: OUTPUT«Sorry, sides are of uneven length and not dwimmy.␤ in 'hyper' at line 179:CORE.setting␤ in 'hyper' at line 211:CORE.setting␤ in main program body at line 1␤»
moritz_ rakudo: my @a = <1 2 3>; say @a>>+>>2
p6eval rakudo 4bf6c0: OUTPUT«345␤»
13:00 macroron joined
sorear keline? 13:05
takadonet moritz_: thanks :) 13:06
rakudo:my @a=<1 2 3>;say @a>>.say 13:07
rakudo: my @a=<1 2 3>;say @a>>.say
p6eval rakudo 4bf6c0: OUTPUT«1␤2␤3␤111␤»
takadonet is that a bug?
moritz_ doens't look like a bug to me 13:08
what would you expect instead?
szabgab as I can see there is a growing number of tweets marked with #perl6
au|irc the screencasts is very effective :) 13:09
takadonet was is it printing 111 at the end? 13:10
s/was/why
moritz_ takadonet: you have two calls to say() in there
takadonet ...
moritz_: I need a coffee
moritz_ takadonet: the .say method returns 1 for each array item
szabgab au|irc: you mean we see hundreds already joining the channel ? :)
takadonet moritz_: just ignore me please hehe 13:11
moritz_ 15:11 -!- Irssi: #perl6: Total of 191 nicks
not too bad, if you ask me :-)
au|irc szabgab: maybe "join us on #perl6" has a localized meaning on twitter :)
(j/k, mostly)
szabgab it has a certain self references 13:12
moritz_ corky.net/dotan/log/images/perl6-watchmen_a.jpg
you can always ilnk to perl6.org/community/irc 13:13
au|irc szabgab: I like the sound of your keyboard. it sounds like a real keyboard :)
szabgab :) 13:14
moritz_ twitter.com/handraiser/status/19099769071 # I like this one 13:15
13:18 zulon joined
[Coke] (ugh. don't port File::Find to perl6!_ 13:18
tadzik why not?
[particle] interface == suck
tadzik the original one?
[particle] try Class::Path or is it Path::Class
moritz_ isn't that something completely different? 13:19
tadzik sounds so
frettled But feel free to create something that a) works well and b) is a workable substitute for `find` ;)
[particle] no, it's based on File::Find, without the brain-dead interface
tadzik File::Find is like `find $where`
moritz_ wasn't that File::Find::Rule?
[particle]: I think File::Find is a very useful directory iterator. It's just not what many people want 13:20
[particle] there's also File::Find::Rule, which is somewhat better than F:F
frettled Path::Class is nothing like File::Find
moritz_ frettled: iirc it's more like File::Spec, right?
frettled It seems to be, yes, I've never really had a use for any of them. 13:21
[particle] oops, yes, File::Spec. sigh, someone wake me up.
tadzik github.com/tadzik/perl6-File-Find # now without the 'interface == suck' "get the data you want from some global variables"
moritz_ shakes some coffeine into [particle]
[particle] File::Find::Rule is the better interface to File::Find
au|irc actually, dir('...')->recurse() is comparable to File::Find :) 13:22
[Coke] (PGE) one version of partcl uses it, so it occasionally gets tested "in the real world", anyway.
au|irc but F::F::R has a better interface, yeah.
tadzik I was thinking about making File::Find something find(1) comparable
moritz_ tadzik: you shouldn't restrict $callback to Sub. Callable would be sufficient, I'd say
tadzik which, as far as I can see, is File::Find::Rule. Any reason why not to merge this two and have one good module for Perl 6? 13:23
moritz_: will fix, thank you moritz_++
moritz_ likewise there's no reason to restrict $dir to Str
[Coke] 16:56 < patrickas> oh ok ...
16:56 < Util> pmichaud, [Coke]: Thanks! Will push now
16:56 <@pmichaud> (mostly affects Rat arithmetic) yes, but a single / automatically puts you into the realm of Rat arithmetic. :-)
moritz_ dir() will stringify its argument anyway
tadzik: also a single multi sub find(&callback, *@dirs) could replace the one 13:25
frettled Does F::F::R chdir() or not?
moritz_ it's configurable in the p5 File::Find
to me it seems wrong to default to chdir()ing, because it's an application-wide process 13:26
frettled yep
moritz_ and since you're calling a callback, that itself could chdir, and screw up everything else
reentrance safety ftw!
frettled I use F::F with no_chdir => 1, I was just wondering whether it was set to 0 or 1 in F::F::R, the docs don't say.
tadzik moritz_: sounds sane, will apply 13:27
moritz_ calls it a day
PerlJam moritz++ IMHO we are in the "what can we add?" phase, while we should be in the "what can we take away?" design phase.
13:27 _macdaddy joined
frettled But as long as our F::F does the sane thing, I'm happy. 13:27
PerlJam good morning all
frettled tadzik++ for bothering
PerlJam: \o
tadzik frettled: hey, I'm sitting here all the time asking silly questions, it would be really impolite to give nothing in return :) 13:28
also, what do you think of adding functionalities to File::Find instead of porting File::Find::Rule?
frettled Since today is the first time I read about F::F::R, I'm not sure, hee-hee. 13:30
dalek kudo: 8c8b656 | Coke++ | docs/release_guide.pod:
add perl6-compiler, which masak++ noted was missing.
 13:31
kudo: b2af272 | Coke++ | src/core/ (3 files):
Merge branch 'master' of github.com:rakudo/rakudo
tadzik I was thinking that File::Find may eventually be something more powerful than just a tool to call a callback on all the files in a directory
frettled I must admit that my use of File::Find has never been more than a Perl-internal way of handling find(1), rather than using a system call and processing that. 13:32
[Coke] urk. mispaste there for a few rows.
tadzik now when I see File::Find::Rule I see that it's actually hell useful, and I think it won't hurt File::Find itself to have such an awesome stuff in it itself 13:33
[Coke] My comment on file::find was just that i don't want to promulgate that interface anymore.
frettled So, basically, my needs for a file finding module is to get a list of files (with full relative or absolute path spec), based on selection criteria. With File::Find I often ended up sending the test to a subroutine, with F::F::R I might actually use the builtins.
tadzik what interface?
13:33 M_o_C joined
PerlJam [Coke]++ amen to that 13:34
tadzik you mean the variables, [Coke]?
[Coke] I meant the whole thing. =-) but sure, the global variables are the worst. 13:35
frettled I guess it's the hash references and function references.
[Coke] ugh. I got a merge commit. how do I avoid that?
(maybe it'll look nicer in p6, Iunno) 13:36
tadzik [Coke]: the globals are now gone and withdrew with much prettier Perl6 way
[Coke] with withold judgement until he sees the module then! =-)
s/with/will/
tadzik so, any objections to extending the functionality of File::Find with the rules like in File::Find::Rule?
frettled [Coke]: github.com/tadzik/perl6-File-Find
tadzik: no objections from me, at least, I think the rules in F::F::R seem very useful. 13:37
tadzik now as I see F::F::R, F::F seems… poor
frettled yup
au|irc ☺ ☾ λ ӝ # g'night lambda butterflies... \o
13:37 au|irc is now known as au|zzz
frettled au|zzz: tee-hee, 'night! 13:38
13:38 M_o_C left 13:39 a3r0 joined
tadzik I just _love_ commits like this 13:39
1 files changed, 22 insertions(+), 30 deletions(-)
less code than before, and works better than before
frettled nice!
tadzik moritz_++ # nice ideas for sanity fixes 13:40
[Coke] it is amazing to me to contrast the positive energy in this group with the negative I see on p5p. 13:41
(not that all of p5p is like that, but the lows are very low.)
PerlJam hugme: hug [Coke]
[Coke] perl6++
hugme hugs [Coke]
PerlJam Coke: it's all the hugs :)
13:42 zulon left
[Coke] hugs-- 13:42
mac Can some1 check this out? wklej.to/JEBC Maybe I missunderstand public/protected attributes in classes, then please explain me how it goes. 13:43
[Coke] ah, crap, forgot to update wikipedia.
doing soo...
tadzik hugme: hugme
hugme tadzik:
13:43 a3r0 left
tadzik hugme: oh you 13:44
[Coke] ... or now.
frettled [Coke]: Which Wikipedia article is that?
[particle] it's amazing to me that you even read p5p and still have enough energy to be positive
13:44 umphy joined
tadzik is p5p that bad? 13:45
sorear [Coke]: p5-p or #p5p?
[particle] [Coke]++ for completing the dual-release this month. welcome to the club!
tadzik mac: is rw maybe?
mac: params are ro by default
[Coke] the mailing list. 13:46
PerlJam mac: careful calling "has $!bar" protected. It's not at all like C++'s protected.
[Coke] frettled: Rakudo. someone updated, but I found a typo, so that's ok. sadly, they linked to the rakudo.org release which isn't mine. ah well.
[particle]: (p5p) skim. occasional refeferences to Grants that come up that require attention. 13:47
mac Indeed, tadzik's right.
frettled [Coke]: aha! good!
[Coke] [particle]: just ``fun'' trying to avoid the sniping and vitriol to get at the info in those posts. =-)
mac PerlJam: so how should I understand it?
13:47 umphy left
frettled mac: This one may help: perl6advent.wordpress.com/2009/12/1...-and-more/ 13:48
tadzik www.nntp.perl.org/group/perl.perl5....59781.html 13:49
this stabbed me one day when I stumbled upon p5p
„its more a general observation that *any* new feature will break perl in unexpected ways, due to the complexity of perl and the weird ways in which all its
component parts interact”
(whoops)
frettled tadzik: Schwern and Jesse are far more positive in their responses, though :) 13:51
mac frettled: thx, I'll check this
frettled mac: but don't hesitate to ask more questions, I just wanted to point you to the awesome advent calendar. 13:52
13:52 uniejo left
tadzik PerlJam: still, aren't they more like protected than like private? They're visible in the inheriting classes, aren't they? 13:54
mac Ok, thanks, this site looks like fair knowledge source 13:55
13:56 bphillips joined
au|zzz phenny: tell masak there's an extra Mu() in the "Operator overloading" section, line 4, between the sub and say, #fyi :) 13:56
phenny au|zzz: I'll pass that on when masak is around.
13:56 au|zzz is now known as au|zzzForReal
PerlJam szabgab++ I don't pay attention for a day and he's posted 2 more screen casts. 13:56
tadzik These great-yet-unoficiall documentation should be gathered somewhere. Advent calendar, moritz_++ Perl 5 to 6, szabgab++'s screencasts... 13:58
wolverian huh, weird, I don't see S12 specifying what "private" means.
13:58 perlygatekeeper joined
moritz_ tadzik: sure. Start it somewhere, others will chime in 13:59
sorear wolverian: Nothing 14:00
Privacy in Perl is a cultural thing
tadzik moritz_: on the rakudo wiki maybe?
moritz_ tadzik: yes. Or in the 'star' repository 14:01
hugme: add tadzik to star
hugme hugs tadzik. Welcome to star!
wolverian sorear: are you saying $!foo attributes are seen by inheriting classes?
tadzik \o/
moritz_: so a pod in a Star repo? 14:02
moritz_ tadzik: yes
frettled tadzik: perl6.org/documentation/ has links to several thingies
moritz_ wolverian: they should not
tadzik will do
wolverian moritz_: that's some kind of privacy. as is the whole support for ! vs. .
tadzik frettled: yep, but no advent calendar, which is awesome. And even though, nice info included with R* is a great idea 14:03
PerlJam Is there a video of the state of the onion somewhere?
sorear wolverian: $! names don't automatically inherit, but you can access them qualified
class Foo { has $!x } ; class Bar { method k { say $!Foo::x } }
14:03 ashleydev left
wolverian ah, thanks. so it's kind of a mix between private and protected, then, as they are known in other languages? 14:04
moritz_ sorear: that requires and explict trust
s/and/an/
wolverian that changes things. :) 14:05
tadzik: does that answer your question?
frettled tadzik: it's in the pugs source, docs/feather/perl6.org/source/documentation/index.html — do you have commit access to that?
tadzik frettled: nah
moritz_ time to change that
tadzik: if you /msg me your email address, I'll give you a comimt bit 14:06
tadzik gee, I feel so important :]
moritz_ tadzik: invitation sent - please add yourself to AUTHORS in the first commit 14:07
tadzik so, is File::Find going into star? 14:08
moritz_: it's svn, right?
moritz_ (oh, I forgot to tell felliott__ the same)
tadzik: sure, why not?
tadzik moritz_: Just asking to be sure, nothing against it
I'll have to learn to use that :)
moritz_ let me re-iterate that everybody who has commit access to the pugs repo can also invite new committers at commitbit.pugscode.org/admin/project/Pugs/people 14:09
frettled tadzik: I'll cobble together a quickref for you
moritz_ can and should
tadzik frettled++ # thank you!
moritz_: oh, your "why not?" was about File::Find, not svn at pugs? 14:10
moritz_ tadzik: yes
tadzik multithreaded talk with the same person causes race conditions
moritz_ aye :-)
tadzik: the general opinion is that all modules that are useful to a somewhat wider audience (and that work with Rakudo) should go into R* 14:11
tadzik I see
just to be sure, docs with R* are to be written in Pod6? 14:12
14:12 ashleydev joined
moritz_ has no idea how docs will be managed 14:12
tadzik so far announcement is plaintext, I'll go for plaintext to, there will be time to unify it later 14:13
frettled tadzik: pastebin.com/FqkM1vm8
tadzik frettled++ 14:14
frettled It does not help you resolve conflicts and such, but it's okay for getting started.
You're welcome!
mac sorear: could you answer it, to be sure. Private fields can be accessed (like protected in other langs) by the class name and private accessor ($!), like: $!ClassName::field, am I right?
[particle] is feeling more awake 14:15
frettled [particle]: almost, ahem, excited? :) 14:16
Physics jokes are under-appreciated. :( 14:17
sorear mac: I think so 14:18
wolverian moritz_ said that only works if the class trusts the accessing class.
which I think is correct from reading S12, but I'm not sure. :(
jnthn mac: As far as I know, only if the class declares itself as "trusts OtherClass"
mac: Rakudo currently gives attributes protected visibility rather than the private they should have.
sorear I don't really see the point in implementing that 14:19
mac Uhmm... Where to declare "trusts"? In the header of class?
14:19 moritz_ sets mode: +oooo wolverian frettled jnthn [particle], moritz_ sets mode: +oo mac tadzik
mac class name trusts someOtherClass { 14:19
?
sorear monkey typing renders all "security" a joke
jnthn No, just
class Foo { trusts Blah; }
iirc
Oh, maybe it's a trait...I forget off hand. 14:20
mac woo, now I feel important too ;)
moritz_ sorear: in the end there will be some introspection mechanisms for getting foreign attributes anyway
frettled sorear: It's not quite a joke, because enabling monkey typing means that you're indicating that Here Be Monkeys.
jnthn Anyway, Rakudo will get this right when the meta-model foundations get re-done.
mac jnthn: kk thanks
frettled scratches its head, eats a banana, and makes noises.
jnthn And at the same time when reference to an attribute not declared inside the class you're currently working in will be a compile time error. 14:21
PerlJam besides ... without monkey typing, how are we going to port all that ruby code over to Perl 6 ;-)
moritz_ it's a bit like inside-out objects in p5 - it's not "secure", but it's quite a barrier
frettled PerlJam++
jnthn So we catch typos.
14:21 frettled sets mode: +o PerlJam, moritz_ sets mode: +o PerlJam, frettled sets mode: +o Juerd
jnthn yay, I have my cinameon bun back! 14:21
frettled jnthn: bun party!
jnthn omnomnom@@@ 14:22
14:22 cruftius_minimus joined
mac One more question, if can ;) I'm in doubt with handles. I`ve read example from advent and I'm thinking over if it's possible to "bind" whole methods from that class which I "handles", or just those which I "whitelist"? Because there is written "Renaming is just one option that is available; the delegation syntax offers many other alternatives." 14:23
my $bob = DogWalker.new(name => 'Bob', dog => $fido); - is there any other way to "bind" methods from other class into mine? 14:24
tadzik will the specs be included in R*?
moritz_ mac: 'handles' is more like forewarding, not binding 14:25
jnthn moritz_: handles works in two ways. In one, it generates forwarder methods. In the other, it says "if you search the class hierarchy and find no matching methods, check if you can forward to one of these instead"
er 14:26
mac
:-)
The m-space is ever crowded. :-)
frettled UTF-8 nicks FTW?
jnthn irc is too aulde to handle 'em. :-( 14:27
mac Ok, thanks, after dinner I'll check this ;)
14:27 Velvet left
[particle] auld lang sirc 14:27
jnthn 16:34 [freenode] -!- джнтн Erroneous Nickname
:-( 14:28
moritz_ mac: ou might want to check t/spec/S12-attributes/delegation.t
*you
PerlJam jnthn: Does rakudo's "handles" ... er, handle all of the same things as Moose yet?
jnthn PerlJam: I don't know all of the things Moose's one handles
PerlJam I guess I should read S12 to see where the intersections are anyway
jnthn PerlJam: It handles quite a lot of cases though.
PerlJam: See also to the test file moritz_ just mentioned. 14:29
For what Rakudo actually implements
PerlJam aye
jnthn iirc we the whole test file.
moritz_ is glad that the special cases for array and hash attributes in conjunction with 'handles' are gone
jnthn But not quite the whole spec
moritz_: ME TOO! \o/
moritz_: It made it a LOT less of a PITA to implement.
Well, not that I didn't once implement the other cases too...
moritz_ jnthn: and much nicer to use, too :-)
jnthn Yeah, quite 14:30
Was a bit too magic.
mathw special cases are usually bad
tadzik lolipushed: github.com/rakudo/star
mathw which makes me think, is there a method by which a user-defined class which does Positional can indicate how it can autovivify?
moritz_ mathw: WHENCE 14:31
tadzik You might want to review this: github.com/rakudo/star/commit/30100...f4e0381ccf
moritz_ tadzik++
mathw moritz_: excellent 14:32
tadzik fine :)
moritz_ I see you also updated download-stuff.pl. Very good.
tadzik yep
14:33 ashleydev left
tadzik moritz_: line #45 could also be removed now, no? 14:35
moritz_ tadzik: yes; I'm updating that file anyway
tadzik frettled: nice cheatsheet, but how do I initally clone the repo? svn co? 14:38
moritz_ svn co
tadzik moritz_: I got the invitation, but the link for setting the password seems dead 14:39
sorear oh man this is so surreal.
moritz_ tadzik: remove the port number from the link
tadzik alright, works 14:40
[Coke] bug handling... should we be using new to mean "tickets no one has vetted yet", and open to mean "yes, this is an issue of some kind" ?
at the moment, new+open is just one big pile, afaict.
moritz_ right
jnthn [Coke]: We haven't tended to do that.
[Coke]: If a ticket is not worth of attention, reviewer can reject it. 14:41
14:43 squeeky joined 14:44 zulon joined
[Coke] I may open a few a new-tickets-I-think-are-bugs. 14:44
nothing to see here.
(it will be helpful, I think, as we get more reports from non-devs)
rakudo: my %h; %h<one><two> = 1; %h.perl.say 14:45
p6eval rakudo 4bf6c0: OUTPUT«{"one" => {"two" => 1}}␤»
14:46 azert0x left, azert0x joined
[Coke] rakudo: (368 + 274 +5).say 14:49
p6eval rakudo 4bf6c0: OUTPUT«647␤» 14:50
moritz_ buubot: eval: $^O 14:51
buubot moritz_: linux
[Coke] pmichaud, jnthn: permission to close 49171? seems like the ROADMAP is getting more use than that ticket.
14:51 mberends left
[Coke] (I'd rather use the builtin priority field, anyways, if we're going to track that in RT) 14:51
moritz_ [Coke]: ask pmichaud, he seems to be the main user 14:52
ab5tract rakudo: my %h; %h<for> = "the"; %h<love> = "of"; %h<perl> = 6; %h.say
p6eval rakudo b2af27: OUTPUT«love ofperl 6for the␤»
ab5tract rakudo: my %h; %h<for> = "the"; %h<love> = "of"; %h<perl> = 6; %h.perl.say
p6eval rakudo b2af27: OUTPUT«{"love" => "of", "perl" => 6, "for" => "the"}␤»
[Coke] moritz_: I just did?
jnthn: can you change the summary of RT#50118 to match the current "left to do" ?
moritz_ [Coke]: oh, I only read the 'jnthn' part, sorry
14:53 gbacon joined
jnthn [Coke]: Opened it to glance at later. 14:54
(Got $dayjob bits going on ATM. :-))
[Coke] can someone with a checkout of the spectests add a marker for RT#58018 to the fudged __cmp tests? 14:58
14:58 ktne joined
pugssvn r31793 | moritz++ | [t/spec] add RT #58018 marker to tests in range.t 14:59
[particle] whenever did $dayjob become a priority around here?
14:59 shade_ joined
[Coke] moritz_++ 15:00
[particle] ...for me, it was over a year ago.
15:00 felliott joined
jnthn [particle]: I'm trying to keep the two in balance somewhat. :-) 15:03
[particle]: Even if the allure to just hack on Rakudo all the time is strong. :-)
15:03 traqy joined
[particle] i miss it greatly 15:03
15:06 baest left 15:08 shade_ is now known as shade\
[Coke] moritz_: I'm assigning some tickets to you, for great justice. 15:10
mac is afk
moritz_ [Coke]: yes, I've seen it :-)
15:10 mac is now known as mac_|afk
moritz_ I've closed the REPL thing without tests, because it's not in the spec, and we don't have a good way for non-spec tests in rakudo 15:11
[Coke] rakudo: say 10000000[1]
p6eval rakudo b2af27: ( no output )
[Coke] moritz_: 'sfine, I trust your judgement on those or i'd not have assigned them to you. =-)
15:12 zulon left
moritz_ :-) 15:12
15:12 levon joined 15:13 pyrimidine joined, levon is now known as Guest32833
TimToady sorear, pmurias: I'm fine with naming any XXX to STD::XXX except for STD itself :) 15:13
phenny TimToady: 05:26Z <sorear> tell TimToady It's possibly to confuse STD.pm6 into setting up a package as the OUTER:: of a lexical scope. Try package Foo { my $x = 2; }; package Foo::Bar { say $x }
TimToady: 05:27Z <sorear> tell TimToady STD.pm6 line 5547 appears to be involved, but I'm not very knowledgable on the $*CURLEX handling
15:14 Guest32833 left
TimToady ==> @oscon 15:14
frettled TimToady: and is it good?
moritz_ I've just pushed a build script to the 'star' repo 15:17
on my machine it successfully builds rakudo, puts it into the local $PATH, and runs a small sanity check
jnthn rakudo: sub TimToady { <awesomeness skillz> }; my @oscon; TimToady ==> @oscon; say @oscon.perl;
p6eval rakudo b2af27: OUTPUT«["awesomeness", "skillz"]␤»
moritz_ next up: installing ufo and modules 15:18
pugssvn r31794 | moritz++ | [t/spec] unfudge some range tests, [Coke]++
jnthn moritz_: For module installation
moritz_: It occured to me that if we hack up a way to copy the modules into the pls "cache" directory and scribble a project file that claims they are fetched
moritz_: Then we can just pls them the rest of the way. 15:19
moritz_ jnthn: oh, nice idea
anyway, I think I should by some food before continuing
15:19 sftp left 15:22 sftp joined
cxreg jnthn: shockingly enough, my github id is cxreg :) 15:23
15:23 jedai left
jnthn cxreg: And it was...Zavolaj you wanted to commit to? 15:23
cxreg well, I pushed to a fork and mberends++ merged it 15:24
so i'm good for now
moritz_ hugme: list
hugme: list projects
hugme moritz_: I know about Math-Model, Math-RungeKutta, MiniDBI, book, gge, hugme, ilbot, java2perl6, json, modules.perl6.org, november, nqp-rx, nqpbook, perl6-examples, perl6-wtop, process-cmd-args, proto, pugs, star, svg-matchdumper, svg-plot, tardis, temporal-flux-perl6syn, try.rakudo.org, tufte, ufo, web, yapsi
15:25 jedai joined 15:27 tuxuday left
[Coke] moritz_: just "Coke" is fine. (IHF!) 15:30
moritz_ phenny: ask masak if temporal-flux-perl6syn still needs to be in the hugme projects list 15:31
phenny moritz_: I'll pass that on when masak is around.
[Coke] offers [particle] some koohii 15:35
[Coke] is pretty sure no kana on this kb.
15:38 wagnerf joined
wagnerf hi, would anyone know how to use opendir in perl6 or where to look for documentation ? 15:39
i'm begining a little bit
dalek ecza: a7a25a2 | sorear++ | (12 files):
Refactor initialization procedure

handling of CORE.setting much closer to STD.		 15:40
jnthn wagnerf: There's no opendir/readdir/closedir - just use the "dir" function.
moritz_ wagnerf: dir() returns a list of files in the current directory
15:40 kjeldahl left
jnthn wagnerf: If you want to iterate over the directories, do for dir() { ... } 15:40
Or for dir() -> $filename { ... }
wagnerf thanks a lot
jnthn I think it also takes a regex too, moritz_?
moritz_ yes 15:41
wagnerf ha good, i also have some globs to do
moritz_ dir(/\.txt$/)
wagnerf thanks
15:41 ashleydev joined
jnthn moritz_: Nice :-) 15:42
15:42 zostay left 15:45 felliott left, zostay joined 15:50 _macdaddy left 15:54 felliott joined, grussell joined 15:56 am0c joined
[Coke] takes regex, or glbo? 15:57
[Coke] is used to glob from ... everywhere.
moritz_ [Coke]: anything that smartmatches, really
am0c .. 15:58
16:02 xinming left 16:03 xinming joined 16:04 timbunce joined, felliott left 16:08 Trashlord joined 16:09 TiMBuS left
moritz_ ouch. SVG::Plot segfaults. 16:13
16:14 eternaleye joined
moritz_ rakudo: BEGIN { @*INC.push: 'lib' }; use Test 16:15
p6eval rakudo b2af27: ( no output )
moritz_ rakudo: BEGIN { @*INC.push: 'lib' }; use Test; say "alive"
p6eval rakudo b2af27: OUTPUT«alive␤»
16:20 cognomore joined 16:24 justatheory joined 16:28 ashleydev left 16:29 patrickas left 16:30 d4rt left
moritz_ disaster 16:30
rt.perl.org/rt3/Ticket/Display.html?id=76692 16:31
takadonet not good 16:32
jnthn ffs 16:33
moritz_: Wehre's the script that crashes it? 16:34
Ah, I need to install SVG::Plot 16:35
moritz_ which in turn requires xml-writer and SVG
jnthn argh 16:36
OK
moritz_ I thought you had pls working? :-)
jnthn pls itself "works" 16:37
It then writes a Makefile that has no chance of working on nmake
The more I think about it, the less I want an installer doing makefile generation. 16:38
But I know @other feel different
moritz_ does it write the makefiles itself? or does it delegate to ufo?
moritz_ is all in favor of ditching makefiles
jnthn moritz_: It writes it itself. 16:40
16:40 PacoLinux left
jnthn Writing a file that we're going to struggle to make right on multiple systems, which in turn just runs commands, feels silly to me. We may as well just run the commands. 16:40
moritz_ jnthn: fwiw, precompilation is not necessary to reproduce the segfault
16:40 PacoLinux joined
jnthn ufo I can see the point of generating a Makefile 16:41
You want to just compile the bits you need and that have dependencies etc.
But for installing, we're just going to compile the whole thing once and install it.
moritz_ akshually if I recompile SVG::Plot, but not the dependencies, it segfaults 16:42
jnthn moritz_: OK. When I first saw the stacktrace I thought "ah, it blows up in the binder, so this'll be easy"
moritz_ oh
jnthn moritz_: Sadly though, the binder is in the stacktrace but it's unlikely it's relevant.
moritz_ even with removing all .pir files it segfaults from time to time
jnthn moritz_: The segfault happens in the GC proccess
[Coke] jnthn: the reason to use makefiles is so you don't have to rewrite make.
jnthn [Coke]: We don't *need* make!!! 16:43
(for this)
[Coke] if we don't need all of make... ok. but I don't want "Perl 6 Ant"
jnthn [Coke]: The Makefile we generate just runs a bunch of commands.
[Coke]: If projects want to ship a Makfeile, they can and pls runs it.
[Coke] no dependency resolution, etc?
jnthn [Coke]: Yes but to generate the Makefile we already had to do the dependency resolution. 16:44
[Coke]: That is, we determine what depends on what, and then write a Makefile with that inforamtion in it, and run it.
16:44 macroron left
jnthn So we're not getting Make to do much other than re-compute an ordering that was already computed. :-) 16:44
In general I agree totally - don't re-invent make.
[Coke] "make -j"
moritz_ [Coke]: will fail 16:45
[Coke] ... but I see your point.
moritz_: then you're not doing the deps right? =-)
but as long as in general jnthn agrees I'm going back tow ork. =_)
moritz_ [Coke]: no, so far only writing the production rules in the right order, or something
jnthn Anyway, I can try and fix the Makefile first.
It just feels like using a tool for the sake of it here though, that's all. 16:46
[Coke] jnthn: no, you write core code instead. :P
jnthn [Coke]: I'd love to, but nobody else has jumped in to hack on Win32 support for pls. :-( And I'd rather we didn't ship something broken on such a widely deployed OS. :-) 16:48
moritz_: Is the backtrace always the same? 16:49
16:49 cjk101010 left
Leonidas is it possible that the current release of rakudo lacks Configure? 16:50
moritz_ jnthn: not quite 16:51
Leonidas: there should be Configure.pl
jnthn moritz_: Got any more examples?
Leonidas moritz_: I am trying to install Web
moritz_ jnthn: nopaste.snit.ch/22239
jnthn moritz_: thanks
moritz_ Leonidas: oh, I don't think that works with current Rakudo
Leonidas moritz_: nopaste.snit.ch/paste 16:52
heh
moritz_ Leonidas: that URL doesn't contain a paste
Leonidas moritz_: sorry, I copied the url without looking, the proper url is nopaste.snit.ch/22240 16:53
jnthn moritz_: Hmm. The binder is in the trace again, but in a different place.
moritz_ jnthn: want more traces?
16:53 jaldhar left
jnthn moritz_: Yeah 16:53
sorear Leonidas: Configure is Perl 5, @*INC is Perl 6
jnthn moritz_: I'm still trying to get a feel for if the binder is certainly to blame or if it just happens to be in the stack traces.
moritz_ sorear: rakudo alpha + proto also had a Configure.pm 16:54
Leonidas sorear: so how to I install Web? I thought that Web was Perl 6 only.
moritz_ iirc you need the January release of Rakudo, and proto 16:55
jnthn: next backtrace infinite-loops in Parrot_gc_mark_PMC_alive_fun -> Parrot_FixedPMCArray_mark
jnthn: and the binder is in the BT, again
16:55 cdarroch joined, cdarroch left, cdarroch joined
Leonidas hmm, I think I'm going to wait for rakudo star 16:55
moritz_ jnthn: nopaste.snit.ch/22241 16:56
16:56 xinming_ joined
sorear moritz_: jnthn: Are you sure this isn't the same lazy list bug? 16:57
moritz_ and another one: nopaste.snit.ch/22242 16:58
16:58 justatheory left
moritz_ that seems to be at cleanup 16:58
16:58 xinming left
jnthn sorear: The one that results in deep recursive marks? 17:00
sorear yes
jnthn sorear: That eventually blow the C stack?
sorear yes
jnthn sorear: I'm not convinced because in what moritz_ has shown me so far, three of them were only 20 frames deep or so.
sorear oh.
moritz_ right, only one was larger 17:01
jnthn All of them seem to have hash destruction in common.
Apart from the longer one which has hash marking as the top frame.
17:01 timbunce left 17:02 japhb left
jnthn That they all go through the binder is somewhat suggestive though. 17:02
moritz_ just got another one with 1155 frames
jnthn moritz_: Is the top hash-y at all?
moritz_ jnthn: lots of marking in there, one of them hash 17:03
nopaste.snit.ch/22244
another one from the binder 17:04
jnthn moritz_: I don't see the binder in that last one? 17:05
17:05 sukria left
moritz_ jnthn: nopaste.snit.ch/22245 two more that involve the binder 17:06
jnthn moritz_: OK. I think we're probably beyond the point where it's just chance then.
17:07 timbunce joined
jnthn moritz_: Seems it's always about the binder and hash. 17:07
moritz_: Working on re-producing it here. 17:09
[Coke] std: / <[a-z]> / 17:10
p6eval std 31794: OUTPUT«===SORRY!===␤Unsupported use of - as character range; in Perl 6 please use .. at /tmp/lqLPfKUdfV line 1:␤------> / <[a-z⏏]> /␤Parse failed␤FAILED 00:01 116m␤»
moritz_ rakudo: /<[a-b]>/
p6eval rakudo b2af27: OUTPUT«===SORRY!===␤Obsolete use of hyphen in enumerated character class;in Perl 6 please use .. instead at line 22, near "b]>/"␤»
17:10 justatheory joined 17:11 ashleydev joined
[Coke] RT#69682 is probably very near closable. 17:11
17:14 mscha joined
mscha rakudo: say 0..9.pick(*); 17:14
p6eval rakudo b2af27: OUTPUT«0123456789␤»
mscha rakudo: say (0..9).pick(*);
p6eval rakudo b2af27: OUTPUT«3745680912␤» 17:15
mscha rakudo: say (0..*).pick(*);
p6eval rakudo b2af27: ( no output )
mscha :-(
rakudo: say (0..*).pick(2);
p6eval rakudo b2af27: ( no output ) 17:16
tadzik Unlike other languages, Perl assumes that novices will eventually learn how to read code.
I like this one
jnthn Wouldn't the probability of an random number between 0 and infinity being something that your computer could represent without running out of memory be infinitely small? :-) 17:17
PerlJam jnthn: 5 <-- there's my random number between 0 and infinity. I randomly choose it all the time :-) 17:18
TimToady jnthn: the probably of a random number between 0 and 1 being representable is infinitesimal
17:18 dakkar left
jnthn TimToady: I shoulda said integer. :P 17:18
sub rand { 4 #`picked by dice role, totally random!` } 17:19
17:20 cruftius_minimus left
mscha rakudo: for (1..100) { print [+] (1..6).pick(2, :replace), " "; } 17:20
p6eval rakudo b2af27: OUTPUT«3596127245796995476666848666102966881011745377107348737658692888861048788947127445867105»
mscha rakudo: for (1..100) { print [+] (1..6).pick(2, :replace); print " "; } 17:21
p6eval rakudo b2af27: OUTPUT«6 5 12 3 8 3 7 10 6 4 7 6 11 4 3 3 8 7 11 11 11 9 3 7 4 5 7 12 2 4 11 6 7 9 8 7 3 6 2 5 4 10 8 9 8 4 6 3 4 5 4 5 6 5 9 8 8 9 6 6 8 7 6 11 6 4 11 10 9 6 8 9 10 7 9 4 3 5 9 9 3 7 9 5 5 7 2 9 12 9 9 9 8 »
PerlJam rakudo: for 1..100 { [+] (1..6).pick(2, :replace); }.classify(*.Int).Str.say 17:22
p6eval rakudo b2af27: OUTPUT«===SORRY!===␤Confused at line 22, near "for 1..100"␤»
PerlJam rakudo: (for 1..100 { [+] (1..6).pick(2, :replace); }).classify(*.Int).Str.say 17:23
colomon rakudo: (1..100).map({ [+] (1..6).pick(2, :replace); } ).sort.Str.say
mscha rakudo: sub hello(Str $planet) { say "Hello, { $planet.ucfirst }!" } hello('world');
jnthn moritz_: Where is synopsis.pl from? I don't see it in github.com/moritz/svg-plot
p6eval rakudo b2af27: ( no output )
rakudo b2af27: OUTPUT«===SORRY!===␤Confused at line 22, near "sub hello("␤»
jnthn mscha: Need a ; after the } of the sub 17:24
mscha rakudo: sub hello(Str $planet) { say "Hello, { $planet.ucfirst }!" }; hello('world'); 17:25
PerlJam rakudo: (for 1..10 { $_ + 3 }).say
p6eval rakudo b2af27: OUTPUT«Hello, World!␤»
rakudo b2af27: OUTPUT«␤»
jnthn mscha: BTW, you could also write:
rakudo: sub hello(Str $planet) { say "Hello, $planet.ucfirst()!@ }; hello('world');
p6eval rakudo b2af27: OUTPUT«===SORRY!===␤Non-variable $ must be backslashed at line 22, near "$planet.uc"␤»
jnthn gah!
rakudo: sub hello(Str $planet) { say "Hello, $planet.ucfirst()!" }; hello('world');
p6eval rakudo b2af27: OUTPUT«Hello, World!␤» 17:26
sorear really likes the command line niecza_eval --language=NULL -c CORE.setting for some reason
17:27 cognomore left
moritz_ jnthn: it's the example from the POD in lib/SVG/Plot.pm 17:27
[Coke] wozzit do?
(sorear)
jnthn moritz_: found it 17:28
sorear compiles my setting, as a single module, without linking the standard setting (CORE)
jnthn moritz_: gah :-( 17:29
moritz_: I get...a load of svg output.
mscha rakudo: say gather for 0..10 { take (1..6).pick(2, :replace) }
jnthn moritz_: And no segfault.
p6eval rakudo b2af27: OUTPUT«3361115242153241545451␤»
jnthn moritz_: That's with nothing pre-compiled.
mscha rakudo: say gather for 0..10 { take (1..6).pick(2, :replace) }.join(', ');
p6eval rakudo b2af27: OUTPUT«===SORRY!===␤Confused at line 22, near "say gather"␤»
mscha rakudo: my @a = gather for 0..10 { take (1..6).pick(2, :replace) }; say @a.join(', '); 17:30
p6eval rakudo b2af27: OUTPUT«6, 4, 5, 4, 5, 3, 2, 2, 1, 4, 6, 2, 5, 4, 1, 3, 2, 1, 4, 5, 5, 4␤»
moritz_ jnthn: try running it a few times 17:31
mscha rakudo: my @a = gather for 0..10 { take (1..6).pick(2, :replace) }; say @a.classify(*).join(', ');
p6eval rakudo b2af27: OUTPUT«1 3 4 1 1 4 1 5 4 4 2 6 5 5 1 3 3 5 2 2 3 2 6␤»
jnthn moritz_: worked a second time too :-(
17:31 tylercurtis joined
jnthn moritz_: Let me try under the debugger 17:31
moritz_ :(
mscha rakudo: my @a = gather for 0..10 { take (1..6).pick(2, :replace) }; say @a.classify(*).perl;
p6eval rakudo b2af27: OUTPUT«("1" => [3, 6, 3, 4, 4, 3, 5, 6, 4, 5, 1, 3, 3, 5, 3, 5, 3, 4, 3, 4, 1, 6])␤»
mscha rakudo: my @a = gather for 0..10 { take (1..6).pick(2, :replace) }; say @a.classify(*.Int).perl; 17:32
p6eval rakudo b2af27: OUTPUT«("3" => [3, 3, 3, 3], "4" => [4, 4, 4, 4], "5" => [5, 5, 5], "6" => [6, 6], "1" => [1, 1, 1, 1], "2" => [2, 2, 2, 2, 2])␤»
jnthn moritz_: Aha!
moritz_: The debug memory libs found heap corruption or some such.
17:33 cognomore joined
moritz_ "cool" 17:34
mscha rakudo: my @a = gather for 0..10 { take [+] (1..6).pick(2, :replace) }; say @a.classify(*.Int).perl; 17:36
p6eval rakudo b2af27: OUTPUT«("5" => [5, 5], "6" => [6, 6, 6, 6], "7" => [7], "8" => [8], "10" => [10, 10], "9" => [9])␤»
dalek ecza: 92aafc8 | sorear++ | (4 files):
Unify handling of setting and programs in CompilerDriver
ecza: c37e91c | sorear++ | (4 files):
Extend niecza_eval with the ability to compile libraries and settings, and use
sorear does this make niecza the first implementation to fully implement --language? 17:37
moritz_ probably :-)
17:37 traqy left
jnthn Oh wow. interp->mem_pools points into invalid memory 17:37
17:37 japhb joined
jnthn er...and it appears to be trying to allocate something 136,204,424 bytes big?! 17:38
mscha rakudo: my @a = gather for 0..10 { take [+] (1..6).pick(2, :replace) }; say @a.classify(*.Int).map({ .key => scalar(.value) }).perl;
p6eval rakudo b2af27: OUTPUT«Method 'key' not found for invocant of class ''␤ in main program body at line 22:/tmp/s84Ky3jHt7␤»
lue goodnow citizens o/
TimToady if [op] is reduce, what does rakudo use for the internal definition of [\op]?
moritz_ TimToady: reducewith(:triangle) 17:39
rakudo: my @a = gather for 0..10 { take [+] (1..6).pick(2, :replace) }; say @a.classify(*.Int).map({; .key => +.value}).perl
p6eval rakudo b2af27: OUTPUT«("7" => 2, "8" => 2, "9" => 1, "10" => 2, "5" => 3, "6" => 1)␤»
moritz_ mscha: it interpreted the block as a hash constructor 17:40
mscha Ah, thanks.
TimToady I think it's a mistake to do that with an option
means reduce doesn't have a consistent return type 17:41
mscha rakudo: (gather for 0..100 { take [+] (1..6).pick(2, :replace) }).classify(*.Int).map({ .key => scalar(.value) }).perl.say;
17:41 Trashlord left
tadzik rakudo: say 'ok' if "0" 17:41
p6eval rakudo b2af27: ( no output )
tadzik :(
17:44 timbunce left 17:45 justatheory left
jnthn moritz_: Ugh. This is going to take some finding. :-/ 17:47
moritz_: I get it blowing up in different ways on different runs too.
tadzik rakudo: say "a string"[3] # why so fatal?
p6eval rakudo b2af27: ( no output )
tadzik crashes the REPL here. Normal?
PerlJam wonders if .3 could be made to work like .[3] ala TT 17:51
TimToady all things are possible, but not all things are expedient 17:52
tadzik rakudo: my $what = "foobar"; $what .= substr(2).chop; say $what 17:53
p6eval rakudo b2af27: OUTPUT«===SORRY!===␤Confused at line 22, near "$what .= s"␤»
tadzik why?
jnthn -> dinner
tadzik std: my $what = "foobar"; $what .= substr(2).chop; 17:56
p6eval std 31794: OUTPUT«===SORRY!===␤Method call found where infix expected (omit whitespace?) at /tmp/8G9dd4TH80 line 1:␤------> my $what = "foobar"; $what .= substr(2)⏏.chop;␤ expecting any of:␤ argument list␤ infix or meta-infix␤ method arguments␤
..prefix or term␤…
tadzik hmm
[Coke] tadzik: . isn't string concat: that's ~
tadzik [Coke]: I didn't mean concat, I mean method call
$what .= substr(2) is ok, but adding .chop spoils everything
TimToady odd 17:58
[Coke] rakudo: my $what = "foobar"; ($what .= substr(2)).chop; say $what #? 17:59
p6eval rakudo b2af27: OUTPUT«obar␤»
TimToady std: my $what = "foobar"; ($what .= substr(2)).chop;
p6eval std 31794: OUTPUT«ok 00:01 118m␤»
18:00 macroron joined, macroron left
[Coke] yah, I suspect that you can't chop the substr when it's just hanging there. 18:00
18:00 oha joined
[Coke] rakudo: "foobar".chop() 18:01
p6eval rakudo b2af27: ( no output )
[Coke] rakudo: "foobar".chop().say
p6eval rakudo b2af27: OUTPUT«fooba␤»
TimToady std: my $what = "foobar"; $what.=substr(2).chop; 18:02
p6eval std 31794: OUTPUT«ok 00:01 118m␤»
TimToady that also works
tadzik without whitespace?
[Coke] rakudo: my $what = "foobar"; $what.=substr(2).chop;
tadzik Weird, why so?
TimToady parsed as a postfix rather than fake infix
p6eval rakudo b2af27: ( no output )
[Coke] rakudo: my $what = "foobar"; $what.=substr(2).chop.say 18:03
p6eval rakudo b2af27: OUTPUT«oba␤»
18:03 mberends joined
tadzik still weird imho 18:04
oha sorry i've found something strange on rakudo while using grammar
18:04 mberends left, mberends joined
oha it seems an heisenbug, sometimes they match correctly, but the Match object .caps aren't always "complete" 18:05
PerlJam oha: why are you sorry about it? Rejoice! you're helping improve things (either improve your understanding or improve rakudo)
oha PerlJam, sorry for the intrusion, i saw you were already speaking.
PerlJam ah. No worries there either. We're a friendly bunch who don't mind interruption :) 18:06
tadzik rakudo: my $what = "foobar"; $what.=substr(1).chop.say
p6eval rakudo b2af27: OUTPUT«ooba␤»
tadzik oh
PerlJam oha: anyhow ... got the code somewhere where we can see it? 18:07
18:07 gbacon left
oha lemme find a place 18:07
[Coke] nopaste.snit.ch works, or a gist @ github. 18:08
(though the nopaste site is not hooked up to this channel.)
18:08 justatheory joined
[Coke] phenny: .u sad 18:08
:P
oha PerlJam, pastebin.com/b9huV5a2
PerlJam, if i execute it more and more, i often get an incomplete outer "sum" 18:10
PerlJam, i could be wrong, but the previous release of rakudo was ok.
18:10 timbunce joined
[Coke] oha: that code generates output with no error. Whats' the problem again? 18:11
ah, it /occasionally/ fails? 18:12
oha i'm a bit laggy it seems. yep occasionally "dump" incompleted data 18:13
and it do twice, so it seems a problem on the Match and not the dump
This is Rakudo Perl 6, version 2010.06-214-g1ef278d 18:14
PerlJam I just reproduced it I think.
[Coke] $ ./parrot_install/bin/parrot --hash-seed=234234234234 perl6.pbc foo.p6 18:15
18:15 eternaleye left
[Coke] that generates different output than : 18:15
$ ./parrot_install/bin/parrot --hash-seed=23423423 perl6.pbc foo.p6
I think this is an issue with LTM or alternation ending up in a hash.
tadzik my ($str) = $what ~~ s/^\e\[//; # what is wrong in here? 18:16
18:17 eternaleye joined
oha [Coke], no idea, btw the missing data is in not where alternations take place, but where [ .. ]* do 18:18
to me, the incomplete .caps array is always in the first level Match 18:19
tadzik ah, nevermind
PerlJam oha: you mean like this: gist.github.com/487820 ? 18:20
moritz_ we have a spectest for incomplete caps in perl 6
that rakudo occasionally fails
oha PerlJam, exactly the same 18:21
[Coke] oha: try this grammar: nopaste.snit.ch/22247
oha moritz_, so it seems a known bug here.
[Coke], that seems always working here 18:22
moritz_ oha: RT #75484
[Coke] er, you can probably remove the || in the pow rule.
oha [Coke], yep i've noticed it. but it should be skipped being the first empty autoremoved for what i know 18:23
moritz_, yep it seems exactly the same issue. 18:28
18:29 auggy joined, auggy is now known as mmmpork
tadzik how are named regexes now used, outside grammars? <&=foo>, or something? 18:30
moritz_ tadzik: <foo=&foo>
PerlJam tadzik: <foo=&foo>
moritz_ or just <&foo> if you don't capture
tadzik awful. Thanks
18:33 nimiezko joined
moritz_ not happy about it 18:33
18:37 _macdaddy joined
oha i've checked rakudo src, it seems that there are named capture missing in .hash() 18:41
tylercurtis std: rx :1th / foo / 18:42
p6eval std 31794: OUTPUT«ok 00:01 115m␤»
moritz_ I've verified that the Regex::Match object is built incorrectly; it's not Match.caps that's wrong
tylercurtis std: rx :9st / foo /
p6eval std 31794: OUTPUT«ok 00:01 115m␤» 18:43
18:43 ashleydev left
moritz_ std: rx :12foo / bar / 18:43
p6eval std 31794: OUTPUT«===SORRY!===␤Unrecognized adverb :foo(12) at /tmp/GRA07KJohl line 1:␤------> rx⏏ :12foo / bar /␤Check failed␤FAILED 00:01 115m␤»
18:45 takadonet left
moritz_ rakudo: say('a' ne 'b') 18:47
p6eval rakudo b2af27: OUTPUT«1␤»
moritz_ nqp: say('a' ne 'b')
p6eval nqp: OUTPUT«1␤»
18:51 justatheory left
tylercurtis rakudo: class Foo { has $.n; }; our multi sub prefix:<-> (Foo $a) { Foo.new(:n(-$a.n)); }; my Foo $f .= new(:n(6)); (-$f).n.say; 18:51
p6eval rakudo b2af27: OUTPUT«-6␤»
tadzik A beer a day keeps the grammar problems away 18:52
magic 18:53
18:57 pwd joined 18:59 pwd left 19:05 cono left 19:06 cono joined 19:09 justatheory joined 19:12 ktne left 19:14 ashleydev joined 19:18 ktne joined 19:20 pochi joined 19:21 justatheory left 19:25 justatheory joined 19:27 grussell left, grussell joined 19:30 M_o_C joined, arthur-_ joined, jql left 19:31 arthur-_ left
pugssvn r31795 | moritz++ | [t/spec] s/todo/skip/ some tests that depend on specific libicu versions 19:31
19:32 timbunce left
dalek kudo: 4195e9a | moritz++ | src/Perl6/ (2 files):
disallow parsing of adverb form in quotes that we can't handle yet
 19:34
kudo: f196e88 | moritz++ | src/ (3 files):
Merge branch 'subst_adverbs'
tylercurtis futureboy.us/frinkdocs/#Superman Perl 6 needs docs like this. 19:35
colomon tylercurtis: nice! 19:37
19:38 timbunce joined, envi^home left
PerlJam tylercurtis++ Rare though it is, I'm actually LOL :) 19:39
19:40 alester joined
jnthn back 19:44
[Coke] espalda! 19:45
19:45 wamba joined 19:46 cognomore left
tylercurtis Now, we just need a unit-correct math library for Perl 6 and we can have funny example code like that for it. Which I would work on if I could understand why my operator overloads aren't working. :) and didn't need to be working on GSoC. 19:46
19:46 cognominal left, justatheory left
[Coke] we have rats, what more do you need! 19:46
sjohnson wow, p5 5.10.0 has state vars like in p6 19:47
well what do ya know
moritz_ rakudo: my $x = 'hello, world'; $x ~~ s:g/<[aeiou]>/X/; say $x 19:48
p6eval rakudo f196e8: OUTPUT«hXllX, wXrld␤»
moritz_ jaffa4 will be happy :-)
jnthn hehe
moritz_++
moritz_ rakudo: my $x = 'hello, world'; $x ~~ s:g(1)/<[aeiou]>/X/; say $x
p6eval rakudo f196e8: OUTPUT«===SORRY!===␤Arguments to adverbs on quotes are not yet implemented at line 22, near "/<[aeiou]>"␤»
jnthn moritz_: Does .subst not yet support them? 19:49
moritz_: Or were you hitting some odd code-gen problem?
moritz_ jnthn: the latter
jnthn: it complained about identifier Capture not being declared
jnthn fleetingly noticed soemthing about an errant Capture making it into the generated code
Aha.
moritz_ and the --target=past output didn't contain any Capture
jnthn Yeah, that's..."fun" to track down
I've hit them before, but rarely. Generally there's some subtle problem with the generated PAST that's hard to notice. 19:50
colomon [Coke]: I think tylercurtis wants units of measure, like in Frink.
jnthn Can happen if a PAST node makes it into the :value of a PAST::Val
iirc
But perhaps other times too
moritz_: If you have a patch or branch that exhibits the problem, I can probably track it down quite quickly. 19:51
moritz_ jnthn: that's what I had. And then I changed the code not generate the outer PAST::Val (and set :name on the inner tree), and got the same error. "huh".
19:51 cognominal joined
jnthn Much quicker tha the segfault. :-/ 19:51
moritz_ jnthn: git checkout e91ac9e836d25e3c19db2bff46fbd10e214aa938
19:56 lulinha joined, lulinha left
jnthn moritz_: If I change things having done this, will I be able to commit them to master somehow? 19:56
moritz_ jnthn: not easily. But you can send me the diff, and I'll fiddle it out 19:57
you'll get some failures if you run spectest, because that commit is a few days behind master
19:59 M_o_C left, lulinha joined, lulinha left 20:01 lulinha joined, lulinha left 20:04 MichaelRWolf joined 20:05 justatheory joined
moritz_ are two Match objects from failed matches against the same string eqv? 20:07
what if they matched against different strings? does it matter?
pmichaud good afternoon, #perl6 20:12
moritz_ hi. How's OSCON?
pmichaud almost over 20:13
pjf++ just did his closing keynote
szabgab I just saw the talk of pjf
live streaming
jnthn o/ pmichaud 20:16
frettled szabgab: entertaining as always? 20:19
20:19 MichaelRWolf left
colomon pmichaud, etc: I thought that I should try mandelbrot again after the Rat speed ups yesterday. 20:19
pmichaud colomon: I'm always in favor of speed ups and bench marks. 20:20
colomon I don't have exact timings, but it looks like things are much slower, and more likely to crash. :(
I may have to borrow your git bisect and time script to track down what has changed 20:21
moritz_ guesses lists
colomon something math-y is more likely, lots of iteration in there. 20:22
(rather than list processing, I mean.)
[Coke] use MONKEY_BRAINS 20:23
jnthn moritz_: Think I grok what's up - hacking.
20:26 timbunce left
pmichaud oscon done 20:26
afk for lunch, then back to hack for the evening
jnthn pmichaud: But don't worry - YAPC::EU soon for more conferencage! :-) 20:27
20:27 justatheory left
moritz_ read that as "conference cage" :-) 20:27
20:27 wamba left
jnthn moritz_: What does :g with a value do, ooc? 20:27
rakudo: my $x = 'xxxxx'; $x = $x.subst('x', 'y', :g); say $x; 20:28
p6eval rakudo f196e8: OUTPUT«yyyyy␤»
jnthn rakudo: my $x = 'xxxxx'; $x = $x.subst('x', 'y', :g(3)); say $x;
p6eval rakudo f196e8: OUTPUT«yyyyy␤»
20:28 bartie joined
moritz_ jnthn: evaluate for truthiness 20:28
jnthn oh.
Are there any more interesting ones? :-) 20:29
moritz_ sure
jnthn rakudo: my $x = 'xxxxx'; $x = $x.subst('x', 'y', :n(3)); say $x;
moritz_ :x(4)
hugme hugs moritz_, good vi(m) user!
p6eval rakudo f196e8: OUTPUT«yxxxx␤»
jnthn rakudo: my $x = 'xxxxx'; $x = $x.subst('x', 'y', :x(3)); say $x;
p6eval rakudo f196e8: OUTPUT«yyyxx␤»
moritz_ :nth
:c, :p
szabgab I need some help with grammars
paste.scsys.co.uk/47047
I am trying to understand how should I do error reporting 20:30
so I started to write a small grammar for a calculator
20:32 bartie left
szabgab this version now does not report the correct error for the "2 + 3 x" case 20:32
are those die calls the way you would add error repoting at all?
szabgab wonders if he should change his nick to screencast to get some attention :) 20:33
moritz_ rakudo: 'a' ~~ / x | <.panic("OH NOEZ")> /
p6eval rakudo f196e8: OUTPUT«Method 'panic' not found for invocant of class 'Cursor'␤ in <anon> at line 22:/tmp/NREnwrOqSI␤ in 'Cool::match' at line 2382:CORE.setting␤ in 'Regex::ACCEPTS' at line 5483:CORE.setting␤ in 'infix:<~~>' at line 401:CORE.setting␤ in main program body at line
..22:/tmp/NREnwrOqSI␤»
moritz_ hum
20:33 grussell left
moritz_ szabgab: I don't think we have anything better at the moment 20:33
szabgab I saw panic but AFAIK it is specifically implemented for STD
tylercurtis rakudo: grammar Foo { rule TOP { <.panic("Foo")> } }; Foo.parse('flksnfalksdlfk') 20:34
p6eval rakudo f196e8: OUTPUT«Method 'panic' not found for invocant of class 'Foo'␤ in 'Foo::TOP' at line 22:/tmp/iEVvJ9FHXB␤ in 'Grammar::parse' at line 5531:CORE.setting␤ in main program body at line 22:/tmp/iEVvJ9FHXB␤»
szabgab I think this is good for now, but it does not catch that specific case
20:34 bphillips left
szabgab so how can I call a die if there are extra characters after the second operand? 20:35
moritz_ szabgab: '' always matches
that's the problem
moritz_ tries something else
[ \S { die "Invalid value after second operand" } ]? 20:36
this works
szabgab actually I now tried to swap the '' || { die } and that way it also works 20:37
diakopter [Coke]: you got "lue" twice in Atlanta annc
szabgab but let me check yours
jnthn > my $x = 'xxxxx'; $x ~~ s:g(1)/x/y/; say $x; 20:40
yyyyy
> my $x = 'xxxxx'; $x ~~ s:g(0)/x/y/; say $x;
yxxxx
moritz_ \o/
jnthn moritz_: Want the diff?
szabgab moritz_: that one does not work well, it does the right thing for the extra character case but also kills the good case
moritz_ jnthn: yes
szabgab I am not yet sure why
[Coke] diakopter: probably a bug in the script, but mea culpa.
moritz_ szabgab: did you include the ? after the ] ?
jnthn moritz_: gist.github.com/488005 20:41
moritz_ rakudo: (-> $x { say $x}).assuming(3).()
jnthn moritz_: It fixes the issue but makes things a bit simpler / more efficient too.
p6eval rakudo f196e8: OUTPUT«3␤»
20:41 Raynes left
szabgab moritz_: not yet :( 20:41
moritz_ szabgab: that's necessary :-) 20:42
colomon mandelbrot 501 keeps running for 2.5 hours and then crashing. :(
szabgab right, with ? it works
jnthn colomon: Epic profiling needed. :S
colomon jnthn: but I wanna write my grant report!
colomon feels certain this one is probably his fault too. 20:43
jnthn colomon: Well, I can hardly argue with that. ;-)
colomon: er, that was too you writing the grant report! :-)
*to
colomon: If you want me to look over it, I'm going to be here-ish for the next several hours. 20:44
szabgab moritz_: thanks
colomon jnthn++
20:44 djr joined 20:45 djr left, skangas left 20:46 astrojp joined 20:48 perlygatekeeper left
moritz_ jnthn: I have your patch fixed up locally, but I don't know if I'll get around to testing and pushing it today... certainly tomorrow 20:49
[Coke] tools/contrib* doesn't handle the date if the release already happened this month. 20:50
IWBNI that worked.
20:50 jferrero joined
jnthn moritz_: No hurry. :-) 20:50
jnthn afk for a break 20:52
21:01 BooK joined, MichaelRWolf joined, pyrimidine left
[Coke] .u clear 21:05
phenny U+239A CLEAR SCREEN SYMBOL (⎚)
[Coke] .u ornate
phenny U+FD3E ORNATE LEFT PARENTHESIS (﴾)
21:05 alester left 21:07 Daenyth left 21:12 kst joined 21:15 alester joined 21:19 dolmen joined, MichaelRWolf left
tadzik g'night 21:22
21:22 tadzik left
[Coke] ~~ 21:22
21:23 gbarr joined 21:24 gbarr left
dolmen Kodi? 21:25
21:27 dotan joined
dolmen diakopter, are you Kodi on svn.pugscode.org? 21:30
dotan does perl6 (July 2010 Rakudo tarball) understand __DATA__ sections? 21:31
PerlJam dotan: doubtful 21:32
they'd be spelt more like =begin DATA though, not __DATA__ as in Perl 5
[particle] i think diakopter is diakopter 21:33
dotan ok, thanks. 21:34
PerlJam dotan: you can say =begin DATA ... =end DATA, but I don't think the linkage is there to provide the contents in $=DATA 21:35
21:36 nimiezko left 21:37 sundar joined 21:38 jql joined
sjohnson hows #perl6 today 21:38
dolmen [particle], do you known who is Kodi? 21:42
[particle] kodi is kodi arfer
rt.perl.org/rt3//Public/Bug/Display...l?id=70105 21:43
sjohnson rakudo: my @a = <one two three>; local $" = ', '; say "@a" 21:46
p6eval rakudo f196e8: OUTPUT«===SORRY!===␤Non-declarative sigil is missing its name at line 22, near "$\" = ', ';"␤»
sjohnson thats what i deserve for not reading the apocolpyses yet.
sjohnson slaps own wrist
PerlJam that error is certainly LTA though 21:47
std: my @a = <one two three>; local $" = ', '; say "@a"
p6eval std 31795: OUTPUT«===SORRY!===␤Unsupported use of $" variable; in Perl 6 please use .join() method at /tmp/Lt2CeKj5VT line 1:␤------> my @a = <one two three>; local $"⏏ = ', '; say "@a"␤Parse failed␤FAILED 00:01 117m␤»
sjohnson what does LTA mean?
PerlJam Less Than Awesome 21:48
sjohnson i broketed it.
PerlJam I bet STD could be made to declare about "local" if it looks like a declarator too
std: local $a = 5; 21:50
p6eval std 31795: OUTPUT«===SORRY!===␤Variable $a is not predeclared at /tmp/Z6nN9slZnr line 1:␤------> local $a⏏ = 5;␤Undeclared routine:␤ 'local' used at line 1␤Check failed␤FAILED 00:01 117m␤»
sjohnson <--- has no idea 21:51
21:51 whiteknight joined
dolmen ** 21:55
dotan So say I want to read the file my program is in, discard all lines until something with __DATA__ and keep all the following lines. What is a better way than using a while loop and a flag?
avar I love the detailed error messages in STD, are those in rakudo too? 21:56
dolmen avar, no
avar dolmen: Isn't it the $?DATA filehandle or something?
dolmen avar: $=DATA See perlcabal.org/syn/S28.html 22:00
PerlJam dotan: What are your criteria for "better"? :) 22:01
dotan: you could do something like this: repeat { $line = $fh.get } until $line eq '__DATA__'; my @data = $fh.lines; 22:02
dotan Perljam: thanks! I put a comment before __DATA__ because perl6 assumed it was a function call. So I need $line =~ /__DATA__/ - does that still work? 22:04
PerlJam $line ~~ /^__DATA__$/;
or even $line ~~ /^^ __DATA__ $$/; 22:05
but you need to use ~~ rather than =~ :)
22:07 molecules left
PerlJam maybe /^^ '#__DATA__' $$/ for you 22:07
dotan That's the whole line, isn't it? might as well just use string equality. 22:08
22:08 alester left
PerlJam yep 22:09
whatever works for you
dotan That link about twigils looks promising. $=__DATA__ perhaps. 22:10
22:10 gbacon joined
lue ohai o/ 22:10
PerlJam It's $=DATA and @=DATA whenever it gets implemented 22:11
there is no more __DATA__
oha i must have done something wrong, but i keep getting PackFile_unpack: This Parrot cannot read bytecode files with version 8.0. after i've recompiled parrot and rakudo. any hint? 22:15
PerlJam oha: I dunno, but I had the same problem. I nuked my previous parrot installations and reinstalled and it went away. 22:16
oha i'll try 22:18
22:26 amkrankruleuen joined
dotan Is there @ARGV? command line arguments in some other shape? 22:27
PerlJam @*ARGS 22:29
dotan thanks
22:32 sundar left
sjohnson rakudo: say "hi there"; 22:40
p6eval rakudo f196e8: OUTPUT«hi there␤»
22:44 colomon left
oha PerlJam, i think i've found the bug 22:45
i would like to have you to check it first 22:46
22:46 stepnem left 22:47 stepnem joined 22:50 dotan left 22:52 tylercurtis left 23:00 meppl left 23:18 colomon joined 23:33 mscha left 23:40 snarkyboojum joined 23:44 supernovus joined
supernovus g'day #perl6 23:44
jnthn o/ supernovus
[particle] jnthn: who's running rakudo.org? i wanna get google analytics working there so we can get metrics on rakudo star traffic 23:46
alester? pmichaud? 23:47
supernovus So what are the chances of getting DateTime::Utils included with R* ?
[particle] does pls know about it, and does it pass its tests? 23:48
if so, then there's not a bad chance
supernovus [particle]: I'm not sure if pls/proto has it listed yet, who would I ask about adding it? And yeah, the tests all pass.
colomon supernovus: masak 23:50
jnthn [particle]: I believe alester
colomon though, hmm... a lot of us may actually have permission to do it.
jnthn [particle]: And yes, good idea. 23:51
[particle] supernovus: fork masak's proto repo at github, modify the pls/poc-projects.list, and send a pull request 23:52
you should modify proto's projects.list, too
colomon what [particle] said, that's the nice way of doing it. :)
supernovus [particle]: Cool, I'll do so.
[particle] proto needs some docs to that effect in the readme
supernovus: care to write them in a separate commit? 23:53
supernovus [particle]: Sure. 23:55
[particle] supernovus++ 23:56
23:57 Psyche^ joined
cxreg [particle]: regarding passing tests, is it troublesome if you need to jump through hoops to make them pass? 23:59
like, say, MiniDBI, requires you to set up database with a given port, user, password, etc
23:59 amkrankruleuen 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!