Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2013-08-29

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:02 benabik joined #perl6
00:07 dayangkun joined #perl6
00:09 jnap joined #perl6
00:12 Psyche^ joined #perl6
00:21 jnap joined #perl6
00:22 dalek perl6-roast-data: 0453525 | coke++ | / (5 files):
00:22 dalek perl6-roast-data: today (automated commit)
00:22 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/04535259fa
00:38 locsmif coke++ indeed
00:38 locsmif ;P
00:43 dalek roast: 3e1cfca | (David Warring)++ | S32-io/io-spec- (2 files):
00:43 dalek roast: fixed bogus test failures. messed up TAP
00:43 dalek roast: review: https://github.com/perl6/roast/commit/3e1cfcac29
00:44 snoopy_ joined #perl6
00:45 dwarring joined #perl6
00:51 FROGGS joined #perl6
01:01 diakopter sorear: ping
01:04 diakopter host08:rakudo mwilson$ time ./perl6 -e 'my $z = "a" x 1_000_000; print(eval("so \"$z\" ~~ /$z/"))'
01:04 diakopter ===SORRY!===
01:04 diakopter java.lang.ClassFormatError: Unknown constant tag 97 in class file 08760A548F985687272891CCAE0291F993893F21
01:04 diakopter real0m22.200s
01:05 diakopter .ask sorear find this spot in the backlog.. jvm bytecode parsing error from rakudo
01:05 yoleaux diakopter: I'll pass your message to sorear.
01:10 diakopter .seen sorear
01:10 yoleaux I saw sorear 19 Aug 2013 04:06Z in #perl6: <sorear> sweet
01:13 diakopter here are the known constant tags: http://docs.oracle.com/javase/spec​s/jvms/se7/html/jvms-4.html#jvms-4.4-140
01:14 diakopter oh, hahaha, 97 is ascii for a
01:17 diakopter host08:rakudo mwilson$ time ./perl6 -e 'my $z = "\x1" x 100_000; print(eval("so \"$z\" ~~ /'\''$z'\''/"))'
01:17 diakopter ===SORRY!===
01:17 diakopter java.lang.ClassFormatError: Illegal UTF8 string in constant pool in class file C2191414E5EAB76038258C2568A4F799EB32FC6F
01:22 dalek roast: f39b0e8 | (David Warring)++ | S32-io/io-spec-win.t:
01:22 dalek roast: fixed typo
01:22 dalek roast: review: https://github.com/perl6/roast/commit/f39b0e8789
01:36 diakopter .tell masak unexpected in rakudo:  say so "a" x 1_0 ~~ /a ** 1_0/  says  False
01:36 yoleaux diakopter: I'll pass your message to masak.
01:36 jaldhar joined #perl6
01:38 FROGGS joined #perl6
01:41 skids joined #perl6
01:52 mtk joined #perl6
01:57 dalek roast: cdb6698 | coke++ | S0 (4 files):
01:57 dalek roast: pugs fudge
01:57 dalek roast: review: https://github.com/perl6/roast/commit/cdb6698b5e
01:57 thou joined #perl6
02:01 cognominal joined #perl6
02:23 dalek perl6-roast-data: a558235 | coke++ | bin/ (2 files):
02:23 dalek perl6-roast-data: Up the timer from 15s to 1m
02:23 dalek perl6-roast-data:
02:23 dalek perl6-roast-data: Should avoid some timeout errors on niecza.
02:23 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/a558235de8
02:24 colomon Hmmm, ABC seems to work on jakudo
02:24 colomon oodles of Passing an object candidate to Mu.bless is deprecated messages
02:25 [Coke] colomon: looks like some of the niecza failures in roast I'm seeing are just timeouts
02:25 colomon ?
02:26 [Coke] will check again after the next run.
02:26 [Coke] all the roast runs use a time/memory limited perl6 wrapper script.
02:27 [Coke] so, https://github.com/coke/perl6-roast-dat​a/blob/master/niecza_summary.out#L5932 is just a timeout error.
02:27 [Coke] Things should look a little nicer after this.
02:30 colomon [Coke]++
02:51 Chat8992 joined #perl6
02:51 Chat8992 Hi
03:04 skids joined #perl6
03:04 FROGGS joined #perl6
03:13 mauke joined #perl6
03:15 preflex_ joined #perl6
03:30 Tene joined #perl6
03:31 djanatyn joined #perl6
03:34 PZt joined #perl6
03:52 xinming joined #perl6
04:26 risou_awy joined #perl6
04:31 cognominal joined #perl6
04:48 hummeleBop joined #perl6
04:56 SamuraiJack__ joined #perl6
05:04 FROGGS joined #perl6
05:13 dalek std: 27326bc | larry++ | STD.pm6:
05:13 dalek std: allow _ in regex ** range integers
05:13 dalek std: review: https://github.com/perl6/std/commit/27326bc097
05:14 ggoebel2 joined #perl6
05:27 mtk joined #perl6
05:37 cognominal joined #perl6
05:41 rindolf joined #perl6
06:06 dmol joined #perl6
06:27 diakopter hi
06:30 FROGGS joined #perl6
06:46 moritz good morning diakopter, *
06:49 FROGGS o/
06:57 domidumont joined #perl6
06:57 berekuk joined #perl6
07:01 qiyong joined #perl6
07:01 qiyong does p6 still have three kinds of variables?
07:06 GlitchMr qiyong: No. Neither Perl 5 or Perl 6 had three kinds of variables.
07:06 GlitchMr But if you mean, scalar, array, and hash, Perl 6 still has these.
07:07 JimmyZ qiyong: 你好
07:07 GlitchMr (but Perl 5 also had typeglobs (*), and subroutines (&))
07:08 GlitchMr (and Perl 6 also has generic package name (::), and subroutines (&))
07:09 GlitchMr Also, Perl 6 has twigils.
07:11 zakharyas joined #perl6
07:14 domidumont joined #perl6
07:16 qiyong GlitchMr: isn't that variable type?
07:16 GlitchMr I'm not sure. What is your definition of variable type?
07:16 qiyong scalar, arry/list, hash
07:16 GlitchMr The documentation says it is.
07:17 nwc10 Perl 5 has 5 things you can put sigils in front of, and about 3 more things that don't have sigils
07:17 nwc10 scalar, array, hash, code, typeglob, format, filehandle, dirhandle
07:17 qiyong it is what? what's the proper term for scalar, arry, ...
07:17 GlitchMr http://feather.perl6.nl/syn/S02.html#Sigils
07:17 GlitchMr array
07:18 qiyong how to test if an item exist? like set operator in, using hash?
07:19 domidumont joined #perl6
07:20 FROGGS r: my %hash = foo => 42; say %hash<bar>:exists; say %hash<foo>:exists
07:20 camelia rakudo 9faa30: OUTPUT«False␤True␤»
07:21 FROGGS r: my %hash = foo => 42; say %hash.exists('bar'); say %hash.exists('foo')
07:21 camelia rakudo 9faa30: OUTPUT«False␤True␤»
07:21 dmol joined #perl6
07:26 qiyong FROGGS: what is :exists?
07:26 qiyong object method?
07:27 JimmyZ 一个动词
07:29 pernatiy joined #perl6
07:34 kaleem joined #perl6
07:35 FROGGS qiyong: sort of, it is called an adverb
07:36 qiyong FROGGS: new in perl6?
07:37 qiyong i want to implement this as simple as possible, in p5 or p6.  if ($user in @users) { }
07:37 FROGGS qiyong: adverbs? at least the hash adverbs are newly added to rakudo within the last months
07:38 JimmyZ std: if $user == any(@users) { }
07:38 camelia std 27326bc: OUTPUT«[31m===[0mSORRY![31m===[0m�Variable $user is not predeclared at /tmp/rjoW0he5bt line 1:�------> [32mif [33m�[31m$user == any(@users) { }[0m�Variable @users is not predeclared at /tmp/rjoW0he5bt line 1:�------> [32mif $user == any([33m�[31m@users) { }[0m�C…
07:38 FROGGS well, @users is not a hash
07:38 JimmyZ you want any
07:39 pdcawley joined #perl6
07:40 FROGGS r: my $user = 4; my @users = ^6; say $user == any @users
07:40 camelia rakudo 9faa30: OUTPUT«any(False, False, False, False, True, False)␤»
07:40 FROGGS so it will be true because one of it is true
07:41 FROGGS r: my $user = 4; my @users = ^6; say so $user == any @users
07:41 camelia rakudo 9faa30: OUTPUT«True␤»
07:43 dagurval r: say so 'a'...'z'
07:43 dagurval so makes it timeout ._.
07:43 camelia rakudo 9faa30: OUTPUT«(timeout)»
07:52 FROGGS ?
07:52 FROGGS r: say 'a'...'z'
07:52 camelia rakudo 9faa30: OUTPUT«a b c d e f g h i j k l m n o p q r s t u v w x y z␤»
07:53 FROGGS r: say ?('a'...'z')
07:53 camelia rakudo 9faa30: OUTPUT«True␤»
07:53 FROGGS r: say so ('a'...'z')
07:53 camelia rakudo 9faa30: OUTPUT«True␤»
07:53 FROGGS r: say so 'a'...'z'
07:53 camelia rakudo 9faa30: OUTPUT«(timeout)»
07:53 FROGGS maybe the ... has lower precedence than so?
07:53 sqirrel joined #perl6
07:54 FROGGS r: say True...'z'
07:54 camelia rakudo 9faa30: OUTPUT«(timeout)»
07:54 lizmat good *, #perl6!
07:55 FROGGS hi lizmat
07:55 lizmat FROGGS: %h.exists is supposed to be an internal method
07:55 lizmat also, the :exists adverb can be applied to slices, whereas %h.exists only takes 1 parameter
07:55 FROGGS lizmat: ahh, okay
07:56 lizmat r: my %h=a=>1,b=>2; say %h<a b c d>:exists
07:56 camelia rakudo 9faa30: OUTPUT«True True False False␤»
07:56 FROGGS dunno if I will remember it though :o)
07:56 qiyong JimmyZ: does p5 have any?
07:56 lizmat r: my %h=a=>1,b=>2; say %h<a b c d>:k   # if you just want to know which keys exist
07:56 camelia rakudo 9faa30: OUTPUT«a b␤»
07:57 FROGGS qiyong: no, since there are no junctions
07:57 lizmat r: my %h=a=>1,b=>2; say %h<a b c d>:v   # if you just want to know the values of existing keys
07:57 camelia rakudo 9faa30: OUTPUT«1 2  ␤»
07:57 lizmat hmmm… that feels wrong
07:57 FROGGS qiyong: but there are other ways to achieve that, see List::Utils or so
07:58 FROGGS r: my %h=a=>1,b=>2; say %h<a b c d>:v.perl
07:58 camelia rakudo 9faa30: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/_qYFvzGoiMâ�¤Two terms in a rowâ�¤at /tmp/_qYFvzGoiM:1â�¤------> [32mmy %h=a=>1,b=>2; say %h<a b c d>:v[33mâ��[31m.perl[0mâ�¤    expecting any of:â�¤        pair valueâ�¤        infix stopperâ�¤        infix or meta-inf…
07:58 FROGGS r: my %h=a=>1,b=>2; say (%h<a b c d>:v).perl
07:58 camelia rakudo 9faa30: OUTPUT«(1, 2, (), ())␤»
07:58 lizmat indeed, rakudobug
08:01 JimmyZ qiyong: no
08:04 qiyong what the hell is adverb?
08:04 qiyong new invention?
08:04 qiyong any is keyword or function?
08:04 FROGGS dagurval: it is about precedence, and I think it is not a bug at all
08:05 lizmat r: my %h=a=>1,b=>2; say %h<a b c d>:v.elems
08:05 camelia rakudo 9faa30: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/FrZy9WBD0iâ�¤Two terms in a rowâ�¤at /tmp/FrZy9WBD0i:1â�¤------> [32mmy %h=a=>1,b=>2; say %h<a b c d>:v[33mâ��[31m.elems[0mâ�¤    expecting any of:â�¤        pair valueâ�¤        infix stopperâ�¤        infix or meta-in…
08:05 lizmat r: my %h=a=>1,b=>2; say (%h<a b c d>:v).elems
08:05 camelia rakudo 9faa30: OUTPUT«2␤»
08:05 FROGGS r: say &any.WHAT # any is a function
08:05 camelia rakudo 9faa30: OUTPUT«(Sub)␤»
08:06 JimmyZ r: say any.WHAT
08:06 camelia rakudo 9faa30: OUTPUT«(Junction)␤»
08:07 FROGGS r: say any.WHAT.gist.subst(/\w/, 'f') # *g*
08:07 camelia rakudo 9faa30: OUTPUT«(function)␤»
08:08 JimmyZ r: constant any = &any; say any.WHAT
08:08 camelia rakudo 9faa30: OUTPUT«(Sub)␤»
08:08 lizmat r: my %h=a=>1,b=>2; say (%h<a b c d>:v).elems; say (%h<a b c d>:v).perl # TimToady: something feels incongruent here
08:08 camelia rakudo 9faa30: OUTPUT«2␤(1, 2, (), ())␤»
08:08 lizmat feels to me, the .perl should say "(1,2)', or am I seeing this wrong ?
08:34 moritz you mean .elems shouldn't flatten?
08:35 daxim joined #perl6
08:35 berekuk joined #perl6
08:36 lizmat well, I'm not sure anymore…  clearly showing 4 elements in .perl, and .elems saying 2, feels incongruent to me
08:37 dalek roast: 1a63bc5 | moritz++ | / (3 files):
08:37 dalek roast: some rakudo.jvm unfudges
08:37 dalek roast: review: https://github.com/perl6/roast/commit/1a63bc54c9
08:38 denis_boyun joined #perl6
08:42 jnthn .elems implies flattening list context
08:43 jnthn .perl probably should not do so, but represent what it has, since if you did other operations on it (like .tree) you'd recover the structure.
08:46 Vlavv joined #perl6
08:46 * lizmat tries to adjust to incongruity :-)
08:50 masak good antenoon, #perl6
08:50 yoleaux 01:36Z <diakopter> masak: unexpected in rakudo:  say so "a" x 1_0 ~~ /a ** 1_0/  says  False
08:50 masak r: say so "a" x 1_0 ~~ /a ** 1_0/
08:50 camelia rakudo 9faa30: OUTPUT«False␤»
08:50 masak r: say so "a" ~~ /a ** 1_0/
08:50 camelia rakudo 9faa30: OUTPUT«False␤»
08:51 moritz it's parsed as /[a ** 1] _0/
08:51 masak r: say so "a_0" ~~ /a ** 1_0/
08:51 camelia rakudo 9faa30: OUTPUT«True␤»
08:51 masak moritz: yes.
08:51 jnthn TimToady patched STD to get that right in the backlog :)
08:51 moritz masak: and TimToady++ already patched it in STD
08:51 masak I think a fairly strong case could be made... OK, that settles it.
08:51 jnthn So just needs somebody to nab the patch :)
08:51 * masak submits rakudobug
08:52 atroxaper joined #perl6
08:52 atroxaper Hello, #perl6
08:52 FROGGS hi atroxaper
08:52 masak hi atroxaper
08:52 arnsholt That looks like something that should be added to roast as well
08:52 masak today I got an email encouraging me to blog the t3 reviews.
08:53 masak I'm going to set time aside for it this afternoon.
08:54 * lizmat finally gets a rakudo built after stubbing Set, KeySet, Bag, KeyBag in src/core/Any
08:55 moritz so it's a missing forward declaration?
08:55 lizmat well, that was only one of the problems in the patch, apparently
08:55 atroxaper Does anybody know how can i call system utilit and print his aoutput to console in real time. I try to say qqx|$command|; # for axample i try to call git command. say qqx|"git clone url"| print git's asking of my password and doesn't print anything after that.
08:56 atroxaper Sorry for offtop
08:56 moritz atroxaper: the problem is likely that git asks for the password on the terminal directly, and not via STDIN/STDOUT
08:56 moritz atroxaper: can't you just call shell("git clone $url") ?
08:57 atroxaper moritz: ok. but is there some way to see output?
08:57 moritz well, the user sees the output that way
08:57 masak r: say so "a" ~~ / a ** 1e0 / # what about this one?
08:57 camelia rakudo 9faa30: OUTPUT«False␤»
08:57 masak r: say so "ae0" ~~ / a ** 1e0 / # what about this one?
08:57 donaldh joined #perl6
08:57 camelia rakudo 9faa30: OUTPUT«True␤»
08:57 bruges joined #perl6
08:58 moritz masak: I'd think it would make sense to require a word boundary after the number that follows a **
08:58 dakkar joined #perl6
08:58 moritz number of reptitions are inherently integers, so I don't see the point of allowing floating-point numbers here
08:59 atroxaper moritz: you are best! shell($command) works! Many thanks!
08:59 masak moritz: I think I agree.
08:59 moritz \o/
08:59 moritz atroxaper: and fwiw Perl 6 questions are never off-topic here :-)
09:00 masak and I've always felt that a warning in the case of A ** BCD parsing as [A ** B]CD -- would be very appropriate.
09:00 bruges_ joined #perl6
09:02 moritz +1
09:03 donaldh rakudo chdir emulation: https://github.com/rakudo/rakudo/pull/202
09:03 bruges_ joined #perl6
09:04 donaldh this helps unblock panda
09:04 FROGGS donaldh: I'm doing the lasts tests on windows/jvm
09:04 donaldh FROGGS: oh cool
09:04 FROGGS donaldh: do you had the chance to check for the failing chdir test?
09:04 donaldh FROGGS: on parrot?
09:05 donaldh or JVM ?
09:05 FROGGS donaldh: https://github.com/rakudo/rakudo​/pull/202#issuecomment-23376687
09:06 donaldh FROGGS: I cannot test on XP
09:06 donaldh s/XP/Windows/
09:06 dalek roast: 16fc9c2 | moritz++ | S (14 files):
09:06 dalek roast: rakudo unfudges
09:06 dalek roast: review: https://github.com/perl6/roast/commit/16fc9c287e
09:06 FROGGS k, I will test your latest commit in a minute
09:07 FROGGS or maybe after the jvm tests
09:08 jnthn wow, lots of unfudge :)
09:09 moritz lizmat++ implemented much of the changes that enabled those
09:09 jnthn lizmat++
09:09 donaldh oh, the chdir-emu PR also contains missing IO ops that allow us to unfudge slurp tests and pass most of the spurt tests.
09:09 jnthn donaldh: On JVM, yes? :)
09:09 jnthn donaldh: Binary IO?
09:09 donaldh jnthn: yes.
09:09 jnthn \o/
09:09 jnthn donaldh++
09:10 JimmyZ \o/ CQRS in perl6!
09:11 donaldh jnthn: the JVM binary IO ops are already committed in nqp
09:12 jnthn donaldh: nice
09:12 * jnthn was a bit distracted for the last few days and only sorta followed what's been going on :)
09:12 donaldh work does that
09:13 masak ain't that the truth.
09:16 lizmat jnthn: can a native ever be aliased to $_ ?
09:17 lizmat I'm asking because of "my int $a will begin { … }"
09:17 lizmat technically, adding a phaser for a native, would be no pb
09:17 lizmat but aliasing $_ might ?
09:18 * lizmat is now down to uncommenting single lines to see what breaks the build
09:21 jnthn lizmat: It's the same issue was rw native things...
09:21 jnthn s/was/as/
09:21 lizmat well, you mean it's NYI?  or that it will be impossible ?
09:21 jnthn lizmat: NYI
09:21 lizmat ok, I can live with that
09:22 lizmat :-)
09:22 jnthn lizmat: Though note that you can't apply mixins to a native "container" as there is no container...
09:22 lizmat yep, got that
09:22 jnthn I'm not sure what .WHAT should say on an rw native thing. We could just pretend it's the underlying native type.
09:23 masak lizmat: in a way, it's nice to see you bumping up against untried/unimplemented use cases. means you're breaking new ground, chopping away djungle for future explorers.
09:24 berekuk joined #perl6
09:24 jnthn Remember to roar while doing so...
09:25 mathw joined #perl6
09:30 sqirrel joined #perl6
09:32 JimmyZ jnthn++, CQRS is useful
09:33 jnthn Wasn't my invention... :)
09:33 denis_boyun joined #perl6
09:36 dalek rakudo/nom: e846e9b | (Donald Hunter)++ | src/core/IO.pm:
09:36 dalek rakudo/nom: Start using absolute paths for IO ops
09:36 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/e846e9bc56
09:36 dalek rakudo/nom: 58964e0 | (Donald Hunter)++ | src/core/IO.pm:
09:36 dalek rakudo/nom: Make chdir slightly less naive. Write tests before elaborating.
09:36 FROGGS dalek: o/
09:37 dalek joined #perl6
09:39 lizmat one of my latest inventions:
09:39 lizmat method delete($a) is hidden_from_backtrace {
09:39 lizmat X::Method::NotFound.new( method => 'delete', typename => self.^name ).throw;
09:39 lizmat }
09:39 lizmat :-)
09:39 jnthn um...why not just let the normal method caller do that?
09:39 lizmat a method telling the world it doesn't exist
09:40 lizmat because it *does* exist because of inheritence, and it shouldn't
09:40 jnthn arrrrghghghghghgh
09:40 jnthn No you can't do that!!
09:40 jnthn That's ALWAYS wrong design.
09:40 moritz lizmat: this will be the most confusing errors ever
09:40 moritz *one of the
09:41 lizmat why?  it will just state it doesn't exist?
09:41 moritz but it exists
09:41 moritz and that's the problem
09:41 jnthn But it does exist
09:41 lizmat well, yes,
09:41 moritz you can check .^can that it exists
09:41 lizmat but Key.delete and Bag.delete are specced to not exist
09:41 moritz lizmat: it's much better to introduce an X::Immutable.new(method => 'delete').throw or so
09:42 lizmat tha's fine by me too  :-)
09:42 moritz well, seems the specs are inconsistent, happens
09:42 lizmat I was just going by the spec
09:42 jnthn Yes, "we don't support this operation" is very different from "method not found"
09:42 lizmat or rather, I was fixing a spectest
09:42 lizmat ok, will fix  :-)
09:42 jnthn But even this seems Liskov-dubious...
09:42 jnthn What delete method is inherited?
09:42 lizmat Any.delere
09:42 lizmat *delete
09:42 moritz putting .delete into Any is dubious, I'd say
09:42 jnthn Why does that exist?
09:43 jnthn r: Any.delete
09:43 camelia rakudo 9faa30: OUTPUT«Not enough positional parameters passed; got 1 but expected 2␤  in method delete at src/gen/CORE.setting:1443␤  in block  at /tmp/PlumjvRbqu:1␤␤»
09:43 jnthn r: Any.delete('x')
09:43 camelia rakudo 9faa30:  ( no output )
09:43 jnthn r: Any.new.delete('x')
09:43 moritz r: say Any.delete('x')
09:43 camelia rakudo 9faa30:  ( no output )
09:43 camelia rakudo 9faa30: OUTPUT«Nil␤»
09:43 lizmat r: my %h; my $b := %h<a><b>; say %h; $b=42; say %h
09:43 camelia rakudo 9faa30: OUTPUT«().hash␤("a" => {"b" => 42}).hash␤»
09:44 lizmat well, actually, that's Any.exists
09:44 lizmat I also needed Any.delete for that...
09:44 lizmat brb
09:44 * jnthn doesn't see where...
09:49 lizmat ok, will try to remove Any.delete after I fixed this Set/Bag business
09:49 lizmat or tell you guys why I couldn't
09:49 itz joined #perl6
09:49 jnthn exists and delete in Any do feel a little odd to me...
09:50 lizmat exists *is* needed for %h<a><b>
09:50 lizmat and binding to that
09:50 jnthn How so?
09:50 jnthn (where do we call .exists, that is...)
09:50 lizmat because bind_key does an exists
09:50 moritz r: my $x; $x<a><b> = 3; say $x
09:50 camelia rakudo 9faa30: OUTPUT«("a" => {"b" => 3}).hash␤»
09:50 jnthn Where do we call bind_key?
09:51 lizmat r: my $x; $x<a><b> := 3; say $x
09:51 camelia rakudo 9faa30: OUTPUT«("a" => {"b" => 3}).hash␤»
09:51 moritz jnthn: ah, you mean we could viviy the hash before calling bind_key
09:51 jnthn Right, that calls bind_key...
09:51 jnthn moritz: Not meaning anything yet, just trying to understand... :)
09:51 lizmat patches welcome  :-)
09:51 jnthn I'd thought exists and delete had been put in Any for :exists/:delete related reasons...
09:52 lizmat it was inspired by spectest todo/skips
09:52 lizmat nope, for that they are not needed
09:52 lizmat well, maybe
09:52 lizmat hmmm
09:52 jnthn Right, I'd expected we could get that out of :U candidates in the postcircumfixes...
09:53 lizmat maybe that can wait until I made subs of them ?
09:53 lizmat r: my %h; say %h<a><b>:delete # this is maybe why I needed Any.delete
09:53 camelia rakudo 9faa30: OUTPUT«Nil␤»
09:54 jnthn lizmat: Yeah, perhaps so. I just worry when our core classes are breaking Liskov so blatantly...
09:55 lizmat the alternative would have been to add a *lot* of candidates for the :U case for all of the adverbs
09:55 lizmat or to fix each candidate up
09:55 lizmat so that it would accept U: as well
09:56 lizmat this was a much simpler solution
09:56 jnthn Yeah, that's the way I'd have leaned towards...
09:57 lizmat that would be an extra .defined check on *every* [] and {} access
09:57 moritz maybe make that an nqp::defined check or so, that's fairly cheap
09:57 lizmat perhaps when they're subs, we can make the extra candidates
09:58 jnthn :U is .DEFINITE, not .defined
09:58 jnthn As is :D
09:58 jnthn That boils down to nqp::isconcrete, which is cheap.
09:58 jnthn (flag check on Parrot and Moar, instanceof in JVM)
09:58 lizmat if that's the way you think I should go, I will
09:59 * lizmat is finally able to uncomment code and still build, yeah!
09:59 jnthn lizmat: It feels righter, but as you say it doesn't deal with the other case where you needed .exists...
09:59 lizmat my mind is now filled with Set and Bag, I will revisit this later, ok?
10:00 jnthn sure, no hurry :)
10:00 jnthn And I'm fine if this waits until the method => sub move.
10:00 lizmat what should be better: nqp::isconcrete or .DEFINITE ?
10:00 lizmat performance wise
10:01 lizmat or which would you prefer :-)
10:01 jnthn nqp::isconcrete, though $x.DEFINITE is just nqp::p6bool(nqp::isconcrete($x)) really
10:02 lizmat but I would have to wrap nqp::isconcrete to be able to use it as a conditional, no?
10:02 rindolf joined #perl6
10:03 jnthn no
10:03 jnthn it returns a native int.
10:03 jnthn 0 or 1
10:03 jnthn You can use that directly in a conditional.
10:03 jnthn if nqp::isconcrete(x) { ... } # totally fine
10:04 lizmat cool
10:05 timotimo heyo :)
10:06 masak heyo, timotimo
10:07 donaldh nqp: nqp::copy('a','b')
10:07 camelia nqp: OUTPUT«Class 'File' not found␤current instr.: '' pc 40 ((file unknown):39) (/tmp/chJDn33FTJ:10)␤»
10:07 timotimo ... wat? :)
10:07 timotimo oh, of course
10:08 jnthn no, still wat :P
10:08 donaldh nqp-jvm: nqp::copy('a','b')
10:08 camelia nqp-jvm: OUTPUT«No registered operation handler for 'copy'␤  in compile_op␤  in as_jast␤  in as_jast␤  in <anon>␤  in compile_all_the_stmts␤  in as_jast␤  in as_jast␤  in <anon>␤  in compile_all_the_stmts␤  in as_jast␤  in as_jast␤  in <anon>␤  in compile_all_the_stmts␤  in <anon>␤  in …
10:08 jnthn I think nqp-jvm here is hopelessly out of date, fwiw.
10:08 donaldh ah
10:08 nwc10 is there an easy way to ask nqp-jvm its version?
10:09 timotimo find a recent commit to nqp and see if it is in? ;)
10:09 jnthn nwc10: I think it's actually old enough that it's from the nqp-jvm-prep repo!
10:09 donaldh rakudo copy silently fails for invalid params on JVM.
10:09 nwc10 nostagia!
10:10 nwc10 is there any way to ask it to upgrade itself?
10:10 donaldh I think it is because nqp::copy returns -1 on JVM but maybe throws an exception on parrot ?
10:15 jnthn donaldh: I suspect we want it to be consistent there.
10:17 salv0 joined #perl6
10:20 SamuraiJack joined #perl6
10:20 firstfire joined #perl6
10:23 * donaldh rebuilds nqp on parrot
10:45 donaldh r: copy('a', 'b')
10:45 camelia rakudo e0b2af: OUTPUT«copy is disallowed in restricted setting␤  in sub restricted at src/RESTRICTED.setting:2␤  in sub copy at src/RESTRICTED.setting:13␤  in block  at /tmp/svvN6_WzMO:1␤␤»
10:46 * donaldh golfs alone
10:49 denis_boyun joined #perl6
11:04 clairence joined #perl6
11:07 denysonique joined #perl6
11:07 berekuk joined #perl6
11:07 denysonique Does perl 6 have something like 3->times { }
11:07 denysonique ?
11:07 denysonique e.g. ruby: 3.times { print 'foo ' }
11:08 timotimo no, but you can build it yourself
11:08 lizmat r: say "for" for ^3
11:08 camelia rakudo e0b2af: OUTPUT«for␤for␤for␤»
11:08 lizmat ?
11:08 timotimo r: sub times($num, *&block) { block for ^$num }; 5.&times { say "test" }
11:08 camelia rakudo e0b2af: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/lepeCFnIlK�Unexpected block in infix position (two terms in a row, or previous statement missing semicolon?)�at /tmp/lepeCFnIlK:1�------> [32m *&block) { block for ^$num }; 5.&times [33m�[31m{ say "test" …
11:09 moritz r: sub times($num, *&block) { block for ^$num }; 5.&times: { say "test" }
11:09 timotimo r: sub times($num, *&block) { block for ^$num }; 5.&times: { say "test" }
11:09 camelia rakudo e0b2af: OUTPUT«test␤test␤test␤test␤test␤»
11:09 camelia rakudo e0b2af: OUTPUT«test␤test␤test␤test␤test␤»
11:09 timotimo :)
11:09 moritz :-)
11:09 lizmat .&times is not really .times, but yeah
11:09 timotimo r: sub times($num, *&block) { block for ^$num }; 10.&times: { say "test"; last if (True, False, False, False).pick }
11:09 camelia rakudo e0b2af: OUTPUT«test␤test␤test␤»
11:10 timotimo even works with control exceptions
11:10 timotimo lizmat: right, but now that i know it's possible to do it like this, you can easily MONKEY_TYPING-enhance Int (or Cool if you must)
11:10 timotimo (except method caches are not properly invalidated when using MONKEY_TYPING ...)
11:10 lizmat probably, Int, but yeah
11:12 timotimo denysonique: does that help you?
11:12 denysonique timotimo: yes thanks
11:12 denysonique timotimo: I was wondering if something like this could become a feature of perl
11:13 denysonique without defining it
11:13 moritz why?
11:13 denysonique e.g. feature of stdlib
11:13 moritz we already have a very easy way to iterate $n times
11:13 moritz for ^$n { ... }
11:13 moritz compared to  $n.each: { ... }
11:13 moritz I don't see the big advantage
11:14 moritz it's also questionable wether numbers should implement control flow
11:15 masak denysonique: if you like it so much, just define a subroutine.
11:15 GlitchMr for ^3 { }
11:15 moritz s/each/times/
11:15 masak r: sub upto($N) { 0 ..^ $N }; for upto(7) { say $_ }
11:15 camelia rakudo e0b2af: OUTPUT«0␤1␤2␤3␤4␤5␤6␤»
11:15 JimmyZ well, it's easy to be a feture
11:15 JimmyZ write it ,submit to panda
11:15 timotimo i don't see a problem with putting it into a module.
11:16 timotimo i don't think it should be in "core"
11:16 masak we already have it in core: it's called prefix:<^>
11:16 JimmyZ then, use feature 'times';
11:16 GlitchMr Perl 6 also has perfectly fine 3.times { print 'foo ' }
11:16 timotimo whereby i mean "available without a 'use'"
11:16 GlitchMr It's called either print 'foo ' x 3, or for ^3 { print 'foo ' }
11:16 brrt joined #perl6
11:16 JimmyZ r: print 'foo ' x 3
11:17 camelia rakudo e0b2af: OUTPUT«foo foo foo »
11:17 GlitchMr for ^3 is the "repeat" you want.
11:17 JimmyZ r: print 3 x 'foo '
11:17 camelia rakudo e0b2af: OUTPUT«Cannot convert string to number: base-10 number must begin with valid digits or '.' in '⏏foo ' (indicated by ⏏)␤  in method Int at src/gen/CORE.setting:11524␤  in method Int at src/gen/CORE.setting:5108␤  in sub infix:<x> at src/gen/CORE.setting:1378␤  in sub infix…
11:17 JimmyZ r: print 'foo ' for ^3
11:17 camelia rakudo e0b2af: OUTPUT«foo foo foo »
11:18 GlitchMr Actually, "base-10 number must begin with valid digits or '.'"
11:18 GlitchMr That's interesting.
11:18 GlitchMr r: print 'foo ' x '3'
11:18 camelia rakudo e0b2af: OUTPUT«foo foo foo »
11:18 GlitchMr r: print 'foo ' x '3.5'
11:18 camelia rakudo e0b2af: OUTPUT«foo foo foo »
11:18 GlitchMr So, you can repeat with dots?
11:19 JimmyZ r: print 'foo ' x '3i'
11:19 camelia rakudo e0b2af: OUTPUT«Can not convert 0+3i to Int: imaginary part not zero␤  in method Int at src/gen/CORE.setting:11524␤  in method Int at src/gen/CORE.setting:9573␤  in method Int at src/gen/CORE.setting:5108␤  in sub infix:<x> at src/gen/CORE.setting:1378␤  in sub infix:<x> at src/ge…
11:19 JimmyZ r: print 'foo ' x '3+2i'
11:19 camelia rakudo e0b2af: OUTPUT«Can not convert 3+2i to Int: imaginary part not zero␤  in method Int at src/gen/CORE.setting:11524␤  in method Int at src/gen/CORE.setting:9573␤  in method Int at src/gen/CORE.setting:5108␤  in sub infix:<x> at src/gen/CORE.setting:1378␤  in sub infix:<x> at src/ge…
11:19 JimmyZ r: print 'foo ' x '3+0i'
11:19 camelia rakudo e0b2af: OUTPUT«foo foo foo »
11:22 JimmyZ r: { print 'foo '} x '3+0i'
11:22 camelia rakudo e0b2af:  ( no output )
11:23 timotimo is CQRS a good pattern for implementing event stores? :3
11:24 timotimo CQRS with an event store*
11:24 JimmyZ I think so
11:25 * timotimo failed trolling apparently
11:26 JimmyZ I can got a Perl 6 one easily :P
11:27 pernatiy joined #perl6
12:02 kbaker joined #perl6
12:08 denis_boyun joined #perl6
12:12 arnsholt *sigh* Programming in Python is just annoying enough to make me grumpy
12:15 * lizmat is grumpy from uncommenting lines all the time, make, fail, repeat
12:16 tadzik programming is one thing, but that toolchain is just bad
12:16 tadzik sorry excuses for NYTProf and Devel::Cover
12:16 lizmat on that thought: can we pipe --profile output to NYTProf already ?
12:17 jnthn The stuff it outputs can be viewed in kcachegrind, I believe.
12:17 arnsholt Perhaps the most annoying right now is just the fact that def is an assignment, not a declaration. That way all the little utility functions have to live at the top of the file
12:17 arnsholt Rather than at the bottom, where they belong
12:18 moritz speaking of NYTProf, the flamegraph feature really rocks
12:20 tadzik didn't see it yet
12:20 tadzik I don't remeber the last time I worked with Perl 5 :|
12:20 tadzik well, to the extent of having to do profiling
12:21 moritz tadzik: http://blog.timbunce.org/2013/04/​08/nytprof-v5-flaming-precision/ has an example
12:21 moritz (and in the generated HTML the bars have a hover tooltip with the complete routine name + time spent)
12:22 tadzik wow, interesting to see that in practice
12:23 jnthn ooh, pretty!
12:24 moritz tadzik: I already had a good use for them; we had a case where a slowdown between two profiles occurred, and it was visually obvious where it came from
12:24 tadzik nice
12:24 jnap joined #perl6
12:25 berekuk joined #perl6
12:27 grondilu joined #perl6
12:29 moritz (it was also obvious from the list of exclusive call times, where a routine changed from ~0s to 15.5s)
12:29 moritz ok, it's such a "nice" story, I simply have to tell it :-)
12:30 moritz OTRS has a feature where, for reasons of extensibility, classes are dynamically loaded, and added as superclasses to the current class
12:30 tadzik woosh
12:30 tadzik ;)
12:30 jnthn wait what?
12:30 jnthn :)
12:31 moritz wait for the really weird part :-)
12:31 * lizmat has finally built something that can be spectested
12:31 lizmat I know, it's weird  :-)
12:31 moritz the really weird part is that's done INSIDE THE CONSTRUCTOR
12:31 moritz and it does basically   require $modulepath; push @ISA, $class;
12:32 moritz and that for the template object, which is instantiated about 1 to 5 times per request
12:32 lizmat cycling&
12:32 moritz so, there are about 10 such dynamically loaded classes
12:33 moritz and 3 such dynamic loadings per request
12:33 moritz so @ISA grows by 30 elements per request
12:33 moritz and of course we don't start a new process per request
12:33 moritz so after two hours, @ISA is large that each requests spends more than 1s just pushing to @ISA
12:34 jnthn o. m. g.
12:34 moritz http://bugs.otrs.org/show_bug.cgi?id=9686
12:34 arnsholt Whoops! =D
12:34 jnthn I was about to say, isn't there a class vs instance confusion there... :)
12:34 moritz jnthn: just a slight one :-)
12:34 moritz and it took me about 16 hours to locate the problem
12:35 moritz because the problem was just "OTRS gets slower, until we restart apache"
12:35 jnthn argh
12:35 tadzik one once said, "other will fix memory leaks, I'll just restart apache"
12:35 jnthn .oO( Oh That's Really "Special" )
12:35 moritz and first I checked things like cache response times, DB response times, system load etc.
12:37 nwc10 why does bugs.otrs.org run on bugzilla, and not on OTRS itself?
12:38 * nwc10 finds this strange.
12:38 nwc10 http://www.bestpractical.com/rt/issues.html -- Unsurprisingly, we use RT itself to track bugs and feature requests for RT.
12:38 FROGGS hmmm, I fear we have heisenbugs now in rakudo@parrot
12:38 jnthn nwc10: It got too slow after a while :P
12:38 tadzik why does rakudo.org use php? :P
12:38 moritz nwc10: I wondered too
12:38 FROGGS when I have my Windows 7 vm running an spectesting rakudo on linux, some tests of t/spec/S02-magicals/ abort
12:38 FROGGS like args.t and env.t
12:39 FROGGS (when using TEST_JOBS=4)
12:47 FROGGS r: slurp "does-not-exist"; say "alive"; say $!
12:47 camelia rakudo e0b2af: OUTPUT«Unable to open filehandle from path '/home/p6eval/does-not-exist': No such file or directory(2)␤  in method open at src/gen/CORE.setting:12519␤  in method open at src/gen/CORE.setting:12508␤  in method slurp at src/gen/CORE.setting:12628␤  in sub slurp at src/gen/C…
12:47 FROGGS star: slurp "does-not-exist"; say "alive"; say $!
12:47 camelia star 2013.08: OUTPUT«Unable to open filehandle from path 'does-not-exist': No such file or directory(2)␤  in method open at src/gen/CORE.setting:12479␤  in method open at src/gen/CORE.setting:12469␤  in method slurp at src/gen/CORE.setting:12586␤  in sub slurp at src/gen/CORE.setting:12…
12:55 dalek roast: 9f12744 | moritz++ | S02-types/undefined-types.t:
12:55 dalek roast: two more rakudo unfudges
12:55 dalek roast: review: https://github.com/perl6/roast/commit/9f12744da7
12:55 moritz hopefully LHF: get tools/autounfudge.pl working with rakudo-jvm
12:55 jnap joined #perl6
13:01 ajr joined #perl6
13:11 PacoAir joined #perl6
13:23 kaare_ joined #perl6
13:26 pmurias joined #perl6
13:27 pmurias fun nqp-parrot bug of the day nqp::escape doesn't work with literal arguments ;)
13:27 pmurias nqp: say(nqp::escape("foo"))
13:27 camelia nqp: OUTPUT«error:imcc:The opcode 'escape_s_sc' (escape<2>) was not found. Check the type and number of the arguments␤  in file '(file unknown)' line 155524891␤␤»
13:28 grondilu nqp: my $x := "foo"; say(nqp::escape($x))
13:28 camelia nqp: OUTPUT«foo␤»
13:28 jnthn Oops :)
13:29 jnthn It's signature probably needs tweaking in QSATOperations.
13:29 jnthn An S where there's an s
13:34 skids joined #perl6
13:37 btyler joined #perl6
13:38 pmurias jnthn: I'm fixing that atm
13:38 jnthn pmurias++
13:38 jnthn pmurias: Guess you're improving our test coverage again? :D
13:46 FROGGS t/spec/S32-io/IO-Socket-INET.t hangs on my linux box when running with TEST_JOBS=4 (spectests), has anybody the same issue?
13:47 moritz FROGGS: doesn't hang here
13:47 FROGGS hmmm
13:47 FROGGS moritz: what kernel do you use?
13:47 FROGGS I suspect that could make a difference
13:48 moritz FROGGS: uname -a says   Linux mlenz-workstation 3.5.0-36-generic #57-Ubuntu SMP Wed Jun 19 15:10:49 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux
13:48 FROGGS k, I have a 3.8, I'll switch to 3.5 and see what happens
13:49 dalek joined #perl6
13:50 dalek joined #perl6
13:51 dalek roast: f7dd7e8 | moritz++ | integration/ (2 files):
13:51 dalek roast: rakudo.jvm unfudges
13:51 dalek roast: review: https://github.com/perl6/roast/commit/f7dd7e8f01
13:51 bloonix joined #perl6
13:54 * flussence untravels
13:54 flussence I didn't get anything I planned to done on the way, but I did fix surrogate pairs in JSON::Tiny instead
13:54 moritz \o/
13:54 moritz the one thing missing for feature completeness
13:54 flussence and I didn't have niecza to test on so it's probably broken right now :)
13:55 moritz well, there's another thing, but I won't implement it
13:55 moritz iirc json has an arbitrary limit on nesting depths
14:00 bluescreen10 joined #perl6
14:03 bluescreen10 joined #perl6
14:05 PerlJam moritz: I'm pretty sure that JSON-the-standard does not have any arbitrary limitation on nesting depths.  But certianly, a particular implementaiton of JSON may.
14:19 FROGGS joined #perl6
14:25 moritz PerlJam: you are right. http://www.ietf.org/rfc/rfc4627.txt?number=4627 says An implementation may set limits on the maximum depth of nesting.
14:27 hummeleBop joined #perl6
14:31 nebuchadnezzar joined #perl6
14:33 Maddingu1 joined #perl6
14:35 flussence whoops, no wonder the JSON tests were failing on my desktop... had a branch checked out that's almost 2 months old
14:36 dalek nqp: 93aea23 | (Pawel Murias)++ | src/vm/parrot/QAST/Operations.nqp:
14:36 dalek nqp: Fix nqp::escape("literal").
14:36 dalek nqp: review: https://github.com/perl6/nqp/commit/93aea230d7
14:36 dalek nqp: db474ef | (Pawel Murias)++ | t/nqp/59-nqpop.t:
14:36 dalek nqp: Add test for nqp::escape.
14:36 dalek nqp: review: https://github.com/perl6/nqp/commit/db474ef142
14:50 FROGGS joined #perl6
15:00 ajr joined #perl6
15:01 pmurias how can I get the unicode value of a character in nqp?
15:02 jnthn nqp::ord
15:05 benabik joined #perl6
15:07 FROGGS joined #perl6
15:11 berekuk joined #perl6
15:16 colomon hmmm…. when I throw a larger collection of ABC tunes at Rakudo, it fails under both Parrot and JVM (maybe bug of mine?) but is 5x faster on JVM up to the fail.
15:18 colomon hmmm… whereas on Niecza, it runs a bit then dies with a completely different error.  :\
15:21 jnthn That might be your fault too :P
15:21 colomon oh, absolutely
15:21 jnthn :)
15:21 colomon In the case of "fault" here, I meant ABC bugs, not compiler bugs
15:21 EvanTeitelman joined #perl6
15:22 colomon I think it's been a while since I threw this larger file at the ABC code.
15:22 colomon and I've made quite a few changes to the ABC stuff in the meantime.
15:23 colomon though of course, if it were a compiler issue, that might be my fault to.  In all of the cases.  ;)
15:23 REPLeffect joined #perl6
15:24 colomon looks like it's the dreaded tuplet parser in the grammar.
15:24 diakopter jnthn: did you see the jvm classfile problem in the backlog?
15:24 colomon ABC grammar
15:24 jnthn diakopter: Yes, wasn't sure where that was coming from...
15:24 colomon (in the niecza case)
15:25 * masak .oO( why would you parse dreaded tuplets? ) :P
15:26 jnthn 'cus tuplets think dreads are trendy
15:26 colomon :p
15:28 arnsholt jnthn: What's AUTOGEN supposed to do?
15:28 arnsholt Looks like something's missing when PARAMETERIZE_TYPE gets called?
15:29 jnthn arnsholt: Context?
15:30 arnsholt Trying to fix Zaovlaj
15:31 arnsholt I twiddled the code a bit, and now the error message is "Cannot call 'AUTOGEN'; none of these signatures match:\n:(CStr : Str:D $encoding, Mu *%_)\nin method AUTOGEN at src/gen/CORE.setting:510\n..."
15:38 pmurias nqp: say("\a")
15:38 camelia nqp: OUTPUT«Unrecognized backslash sequence: '\a' at line 2, near "\\a\")"␤current instr.: 'panic' pc 14693 (src/stage2/gen/NQPHLL.pir:5223) (src/stage2/gen/NQPHLL.nqp:279)␤»
15:38 pmurias nqp: say(nqp::escape("\x1b"))
15:38 camelia nqp: OUTPUT«error:imcc:The opcode 'escape_s_sc' (escape<2>) was not found. Check the type and number of the arguments␤  in file '(file unknown)' line 42␤␤»
15:39 pmurias nqp: my $str := "\x1b";say(nqp::escape($str));
15:39 camelia nqp: OUTPUT«\e␤»
15:39 pmurias nqp: my $str := "\x07";say(nqp::escape($str));
15:39 camelia nqp: OUTPUT«\a␤»
15:39 pmurias jnthn: nqp doesn't accept \a in literals yet nqp::escape emits it
15:40 masak pmurias: that could still be perfectly legal. sounds like a case of Postel's to me.
15:40 masak hm, unless the output of nqp::escape is explicitly meant to function as literals.
15:41 colomon jnthn: just found a longer ABC file that still works.  jakudo is 8x faster than parrakudo on it.
15:44 colomon (total jakudo running time 88 seconds, and that's slowed down by a zillion "Passing an object candidate to Mu.bless is deprecated" messages.)
15:44 colomon oh!
15:44 pmurias masak: I seems more a case of reverse Postel we emit something we don't accept
15:45 colomon that was with garbage collection counted.  clock time was 53 seconds
15:45 masak pmurias: yes. dangit! this is the second time this year I make this mistake.
15:46 masak evidently I have my inner Postel reversed somehow.
15:47 jnthn arnsholt: That's...very weird. I guess it's the proto auto-generation
15:48 jnthn .oO( Going Postel )
15:48 jnthn arnsholt: I've never see it go wrong that way, however...
15:48 TimToady .oO( Postel Coming )
15:51 jnthn colomon: 88 seconds is still a while :)
15:51 colomon jnthn: I'm not saying you can stop making improvements
15:51 colomon ;)
15:51 jnthn :)
15:51 xenoterracide joined #perl6
15:52 colomon jnthn: note that was 88 processor-seconds, wall-clock was 53 seconds
15:52 colomon I forgot the distinction was important under JVM
15:52 jnthn ah :)
15:52 jnthn yes
15:52 arnsholt jnthn: Apply https://gist.github.com/arnsholt/6379906 to an up-to-date Zavolaj and Rakudo and you should get the error I'm seeing
15:54 colomon jnthn: anyway, saw a chromatic tweet being a dick about the JVM startup time, and wanted to emphasize again that jakudo crushes parrakudo on real world tasks that are big enough to get by that startup penalty.
15:56 jnthn arnsholt: Trying to box to a CStr[$encoding] is extremely dubious, fwiw.
15:57 jnthn arnsholt: Because that gives you a CurriedRoleHOW-based type, not a class
15:57 jnthn And you can't instantiate a role, and the box op is too low level to pun for you.
15:58 abnorman joined #perl6
15:58 jnthn colomon: Right, and that's not specific to our use of JVM, it's just the nature of the beast.
15:59 colomon jnthn: understood
15:59 jnthn Not that we can't make startup time better, but it never was going to be a Rakudo on JVM strength.
16:00 colomon that's MoarVM's bailiwick, eh?  ;)
16:00 jnthn We should do rather better there, yes. :)
16:01 jnthn *rather
16:01 jnthn MoarVM is smart enough to allocate deserialized objects directly in the old generation since they'll probably live for the whole program, for example...
16:03 * TimToady wonders if anyone has tried bringing unexec functionality to JVM...
16:04 colomon unexec?
16:04 TimToady or undump
16:04 colomon ah
16:04 TimToady startup a process, load everything in you want, do a core dump, transform that to an executable
16:06 dmol joined #perl6
16:08 TimToady looks like nobody's done it yet
16:09 TimToady too many "frameworks" with untrackable external state, seems
16:10 arnsholt jnthn: Right, right. That's likely the previously accidentally working bit then. Thanks!
16:14 * colomon just realized his C++ bug is because OS X's GCC is no longer GNU
16:15 geekosaur mmm, -fheinous-gnu-extensions
16:15 colomon geekosaur: actually, the problem is because I was using #ifdef __GCC__ to make sure we weren't on Windows
16:17 colomon actually, #ifdef __GNU__
16:17 colomon looks like I used #ifdef __GNUC__ elsewhere in the code?
16:20 pmurias jnthn: re nqp::escape should it mangle unicode characters?
16:20 diakopter jnthn: "coming from"?
16:20 colomon __GNUC__ it is
16:20 diakopter the repro case was there
16:21 pmurias nqp: say(nqp::escape("¢"))
16:21 camelia nqp: OUTPUT«error:imcc:The opcode 'escape_s_sc' (escape<2>) was not found. Check the type and number of the arguments␤  in file '(file unknown)' line 39␤␤»
16:21 jnthn diakopter: Yes, I more meant if it's a malformed constant pool entry or just malformed bytecode in general...
16:22 jnthn pmurias: I don't think the JVM one does, at least...
16:22 pmurias it doesn't
16:22 pmurias the parrot one mangles
16:22 pmurias nqp: my $cent:="¢";say(nqp::escape($cent))
16:22 camelia nqp: OUTPUT«\x{a2}␤»
16:24 jnthn pmurias: I don't immediately have a good argument either way. I think it's mostly used for debugging output today anyway...
16:25 diakopter jnthn: I don't know the difference of what you're asking/wondering... it's probably a bug in that assembler library
16:25 jnthn diakopter: aye, I was just speculating on the nature of it
16:27 pmurias jnthn: it's used for escaping string while targeting parrot
16:27 jnthn pmurias: Ah...
16:27 pmurias and I'm trying to use it for escaping string while targeting js
16:27 jnthn pmurias: Yeah, I remember now. I think that it needs to stay the way it is at lest on Parrot...
16:28 jnthn *least
16:29 Ayiko joined #perl6
16:29 pmurias doe we have a infix:<x> equivalent on parrot?
16:30 jnthn Whatever nqp::x is mapped to, I guess
16:31 pmurias meant on nqp ;)
16:35 berekuk joined #perl6
16:35 moritz nqp: say(nqp::x('abc', 5))
16:35 camelia nqp: OUTPUT«abcabcabcabcabc␤»
16:38 kbaker joined #perl6
16:38 pmurias Parrot_str_escape docs are incorrect and crazy
16:39 diakopter s/_str_escape//
16:39 diakopter s/_str_escape docs//
16:39 jnthn diakopter: Your grammar are disappointing :P
16:40 diakopter s/grammar//
16:42 domidumont joined #perl6
16:46 [Coke] niecza failures down from 1999 yesterday to 255 today
16:47 colomon \o/
16:49 SamuraiJack_ joined #perl6
16:49 pmurias who is working on niecza nowdays?
16:52 * [Coke] realizes he can run visualvm against a spec test run these days.
16:53 [Coke] pmurias: colomon
16:53 colomon in theory
16:53 colomon very loose theory this summer
16:53 colomon but I'm planning on taking a look at changing bless this weekend
16:54 [Coke] I didn't mean to imply a level of commitment. ;)
16:55 nebuchadnezzar joined #perl6
16:55 colomon I should be committed, is that what you're saying?  ;)
16:56 jnthn No, just commit ;)
16:58 PerlJam btw, speaking of neicza ... has anyone seen sorear lately? I haven't noticed him in a long while (since right after YAPC::NA even)
16:59 diakopter .seen sorear
16:59 jnthn Last time I saw him here, he mentioned something about lots of $dayjob
16:59 diakopter preflex: seen sorear
16:59 preflex sorear was last seen on #perl6 10 days, 12 hours, 52 minutes and 28 seconds ago, saying: sweet
16:59 diakopter abnorman: yer fault ;)
17:00 diakopter commutativitiness &
17:01 colomon r: say (4 * 60 + 48 ) / 36
17:01 camelia rakudo e0b2af: OUTPUT«8␤»
17:02 colomon oooo, exactly 8x faster.  :)
17:02 diakopter 8x as fast... 7x or 700% faster :PPP
17:03 colomon fair enough
17:04 diakopter .oO( enuf fayryes to go around )
17:06 panchiniak joined #perl6
17:07 [Coke] f
17:07 virtualsue joined #perl6
17:08 FROGGS joined #perl6
17:12 Ayiko joined #perl6
17:17 berekuk joined #perl6
17:18 yoleaux joined #perl6
17:20 gdey joined #perl6
17:21 xenoterracide joined #perl6
17:27 stevan__ joined #perl6
17:34 berekuk joined #perl6
17:35 dalek rakudo-star-daily: bc65cfd | coke++ | log/version.log:
17:35 dalek rakudo-star-daily: today (automated commit)
17:35 dalek rakudo-star-daily: review: https://github.com/coke/rakudo​-star-daily/commit/bc65cfd605
17:35 dalek rakudo-star-daily: 6871921 | coke++ | log/version.log:
17:35 dalek rakudo-star-daily: today (automated commit)
17:35 dalek rakudo-star-daily: review: https://github.com/coke/rakudo​-star-daily/commit/68719215fc
17:35 dalek rakudo-star-daily: 5debe70 | coke++ | log/version.log:
17:35 dalek rakudo-star-daily: today (automated commit)
17:35 dalek rakudo-star-daily: review: https://github.com/coke/rakudo​-star-daily/commit/5debe70935
17:36 [Coke] ... man, that just keeps getting more screwed up.
17:37 moritz yes, that looks very screwed up
17:38 [Coke] started happening post release. still need to get back to it. :|
17:38 panchiniak left #perl6
17:38 abnorman joined #perl6
17:39 [Coke] -> offline
17:54 jaldhar joined #perl6
17:59 stevan_ joined #perl6
18:04 FROGGS the chdir emulation doesnt play well with require :o(
18:04 FROGGS https://gist.github.com/FR​OGGS/ebdc0c7809d1b2902533
18:10 berekuk joined #perl6
18:11 yoleaux joined #perl6
18:12 moritz it's an all-or-nothing approach, and we're somewhere at 80%...
18:14 * FROGGS is testing a fix
18:14 SamuraiJack__ joined #perl6
18:20 FROGGS r: chdir "t"; say $*CWD
18:20 camelia rakudo e0b2af: OUTPUT«IO::Path</home/p6eval/t>␤»
18:20 FROGGS r: chdir "t"; say $*CWD; say nqp::gethllsym('perl6', 'PROCESS')
18:20 camelia rakudo e0b2af: OUTPUT«IO::Path</home/p6eval/t>␤(PROCESS)␤»
18:20 FROGGS r: chdir "t"; say $*CWD; say nqp::gethllsym('perl6', 'PROCESS')<$CWD>
18:21 camelia rakudo e0b2af: OUTPUT«IO::Path</home/p6eval/t>␤(Any)␤»
18:21 FROGGS r: chdir "t"; say $*CWD; say nqp::gethllsym('perl6', 'PROCESS')<$*CWD>
18:21 camelia rakudo e0b2af: OUTPUT«IO::Path</home/p6eval/t>␤(Any)␤»
18:21 FROGGS hmm
18:22 FROGGS r: chdir "t"; say $*CWD; say nqp::gethllsym('perl6', 'PROCESS').WHO
18:22 camelia rakudo e0b2af: OUTPUT«IO::Path</home/p6eval/t>␤("\$FATAL" => Bool::False, "\&cwd" => sub cwd() { ... }, "\&chdir" => sub chdir() { ... }, "\$IN" => IO::Handle.new(path => "-", ins => 0, chomp => Bool::True), "\$OUT" => IO::Handle.new(path => "-", ins => 0, chomp => Bool::True), "\$ERR" …
18:22 FROGGS r: chdir "t"; say $*CWD; say nqp::gethllsym('perl6', 'PROCESS').WHO<$CWD>
18:22 camelia rakudo e0b2af: OUTPUT«IO::Path</home/p6eval/​t>␤IO::Path</home/p6eval/t>␤»
18:25 kivutar joined #perl6
18:26 donaldh joined #perl6
18:27 frdmn joined #perl6
18:28 masak ok, I was delayed a bit in my $dayjob plans today... but now I'm off duty.
18:28 jnthn FROGGS: Please add a test as well as a fix :)
18:28 * masak digs into the last of the t3 reviews/blog post
18:28 FROGGS jnthn: I will
18:29 jaldhar joined #perl6
18:29 timotimo i have reached tge great outdoors
18:30 FROGGS ?
18:30 jnthn Tree Grammar Engine?
18:30 timotimo the
18:30 masak sounds like a scary place.
18:31 pecastro joined #perl6
18:31 timotimo trees everywhere and a nice river with huge white stones in it
18:32 jnthn Isn't it almost dark? :)
18:32 timotimo for a nerd like me that is already outdoorsy enough
18:32 timotimo we are in a house :D
18:32 diakopter only big houses have rivers in them
18:33 masak not to mention trees everywhere.
18:33 * colomon always wanted a river in his house.  Do Not Tell his 4yo.
18:34 moritz [Coke]: fwiw I just did a rakudo-star-daily run, and it went fine
18:35 moritz ... except for the "Passing an object candidate to Mu.bless is deprecated" warnings
18:36 moritz oh, and XML::Writer has a few "use of uninitialized value of type Nil in numeric context"
18:36 ssutch joined #perl6
18:37 dhaivat joined #perl6
18:37 dhaivat Are there any startups using Perl 6? Any corporate support?
18:40 diakopter dhaivat: not really.. it's not really useful or dependable or stable enough yet
18:41 teslos joined #perl6
18:41 * geekosaur can think of a number of things that companies actually base their stuff around that that applies to...
18:41 jnthn :)
18:42 moritz dhaivat: there was one guy from a startup around here last year who used Perl 6 for $work
18:43 moritz haven't heard of him for a while
18:43 masak I use Perl 6 for $work.
18:43 jnthn Are we still a startup? :P
18:43 moritz dhaivat: and colomon uses Perl 6 for some work stuff (involving CAD files and 3D models, iirc)
18:43 diakopter masak, the Startup
18:43 diakopter .. more like Upstart
18:43 * moritz used Perl 6 to get $work
18:43 colomon dhaivat: my use of p6 for that is still pretty lightweight
18:44 colomon dhaivat: I'm using p6 as a convenient occasional tool, not relying heavily on it -- at the heart, my company is still based on C++ and p5.
18:44 dhaivat Ah, so, not yet stable then? What are the primary roadblocks for Perl 6? Why is taking so long to get a stable implementation?  I don't mean this is in a derogatory way, I'm genuinely curious.
18:45 diakopter dhaivat: the past few years, simply a very small amount of availabile development resources, comparatively
18:45 diakopter *available
18:45 moritz dhaivat: the primary roundblocks for production use are 1) lack of good performance 2) lack of solidity in stuff like IO and 3) few modules/libraries available
18:45 jnthn Ambitious project, constrained resources...and it's hard to get fast. :)
18:45 PerlJam dhaivat: Why are you asking around about startups using Perl 5/6 ?
18:45 PerlJam dhaivat: Are you thinking about using perl for your startup?
18:45 moritz dhaivat: and it takes that long because it's a fairly complicated language with lots of research aspects
18:46 dhaivat PerlJam: yes, I am actually
18:46 PerlJam dhaivat: If you're adventurous, use perl 6, otherwise, use perl 5  :)
18:46 dhaivat PerlJam: I used to be a Perl guy a while back, with all the perldocs swirling around my head
18:47 dhaivat PerlJam: But I left because I finally got tired of checking every damn argument in procedures in Perl 5
18:47 colomon dhaivat: for me, it's all about moritz's point 1.  It's only with the recent JVM port that p6 is fast enough to keep up at all with my CAD work -- it's still slow, but it's doable
18:47 dhaivat colomon: JVM port? I haven't noticed this.
18:47 masak I'd advise using predominantly Perl 5, but keeping eyes and doors open for Perl 6 opportunities.
18:48 colomon dhaivat: yes, there are still a few things that need to be finished, but Rakudo has a version which works on JVM and seems to be at least an order of magnitude faster than the Parrot version.
18:48 dhaivat I guess so. I was hoping Perl 6 might be coming around especially since Perl 5 is quite an annoyance at times
18:48 diakopter Perl 6 has its annoyances too
18:48 dhaivat colomon: I see
18:49 colomon dhaivat: it's new, it didn't work at all back in early June.  :)
18:49 dhaivat colomon: wow, that's some quick development
18:49 colomon dhaivat: well, most of the pieces were already in place at that point.
18:49 donaldh joined #perl6
18:52 colomon jnthn++ moritz++ sorear++ etc++
18:52 diakopter donaldh++ too
18:53 PerlJam dhaivat: There are modules on CPAN that make up for many of Perl 5's warts
18:53 * moritz didn't really do much for the JVM port
18:54 PerlJam dhaivat: you can get a type system, subroutine signatures, parameter checking, a nice object oriented system, etc.
18:54 colomon FROGGS++ arnsholt++ pmurias++ jnthn++
18:54 PerlJam dhaivat: Ask more about it on #perl if you're interested
18:54 jferrero joined #perl6
18:54 * FROGGS wakes up... what what?
18:55 dhaivat PerlJam: cool, will do that after lunch
18:55 colomon FROGGS: sorry to bother you, just trying to credit people involved with the JVM port.  I'm sure I've missed some...
18:55 * colomon is mostly going by NQP commit graphs, admittedly.
18:56 FROGGS so I have did something for JVM port? my brain isnt that good these days :o)
18:56 PerlJam colomon++ (for giving karma to all those people  ;)
18:57 colomon FROGGS: you definitely have a worthwhile number of recent commits.  If they are not NQP commits, well, FROGGS++ anyway.  ;)
18:58 FROGGS *g*
18:58 FROGGS thank you sir
19:00 * colomon is feeling the rigidness of the C++ type system very much at the moment.  O for the ability to use p6 instead...
19:03 dalek Perlito: 26aa3e9 | (Flavio S. Glock)++ | / (3 files):
19:03 dalek Perlito: Perlito5 - js - implement some variants of array slices
19:03 dalek Perlito: review: https://github.com/fglock/​Perlito/commit/26aa3e9170
19:15 ssutch joined #perl6
19:19 * lizmat is back and checking backlog
19:20 abnorman joined #perl6
19:25 PerlJam Um ... why did http://perl6advent.wordpress.co​m/2011/12/20/paired-up-hashes/ show up in my newsfeed?
19:25 colomon lizmat++ # just because
19:25 lizmat *blush*
19:26 lizmat I'm in two minds of committing my Set/Bag changes
19:26 lizmat there's quite a lot of set/bag tests failing, and some TODO's passing
19:27 lizmat and I wonder if having more eyes on it would be better than going at it alone
19:27 flussence PerlJam: ...hash collision? :)
19:27 dalek Perlito: d52afd7 | (Flavio S. Glock)++ | / (4 files):
19:27 dalek Perlito: Perlito5 - js - array slice tests
19:27 dalek Perlito: review: https://github.com/fglock/​Perlito/commit/d52afd7fe2
19:27 lizmat otoh, I don't think it is worth a branch
19:27 colomon lizmat: just so happens I have eyes.
19:27 dalek rakudo/nom: 7e80cd5 | (Elizabeth Mattijsen)++ | / (4 files):
19:27 dalek rakudo/nom: Move precedence setting of operators into its own dedicated file
19:27 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/7e80cd5fe6
19:27 dalek rakudo/nom: ea06360 | (Elizabeth Mattijsen)++ | src/core/ (6 files):
19:27 dalek rakudo/nom: First version of Set/Bag changes from niecza, ported to rakudo
19:27 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/ea06360429
19:28 lizmat jnthn: please have a look at src/core/precedence.pm
19:28 * colomon has run out of folk festivals, and while still catching up on $work, has some time to look at p6 stuff again.
19:28 lizmat the commented out trait_mod's make the build fail in odd ways
19:29 masak ladies and gentlebots, t3 reviews are *complete*.
19:29 lizmat basically, trait mods with \x22xx and trait mods with ()
19:29 masak moritz++ # just the right amount of nagging ;)
19:29 PerlJam masak: congrats :)
19:29 masak now for the blog post.
19:29 * lizmat will now focus on fixing the tests
19:30 dalek rakudo/nom: 8897b01 | (Tobias Leich)++ | src/vm/parrot/ModuleLoaderVMConfig.nqp:
19:30 dalek rakudo/nom: take cwd into account when loading modules
19:30 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8897b01994
19:30 FROGGS test comes in a minute (and jvm patch after that)
19:30 masak lizmat: I think merging into master sounds like the right choice here, fwiw.
19:30 masak lizmat: worst case, we revert before release ;)
19:30 lizmat only set/bag tests are affected
19:31 lizmat it's not like itr's all over the place
19:31 lizmat and quite a few TODO's are now passing  :-)
19:33 masak win some, lose some.
19:33 masak no, I guess my point is, "keep master passing at all time" could well be seen as a guideline more than a rule.
19:34 lizmat :-)
19:34 lizmat we have failures now  :-)
19:34 masak and one could consider committing things which fail the tests as taking a loan, of sorts.
19:34 lizmat fixing them up now
19:34 PerlJam masak: technical debt?  :)
19:34 masak as long as you commit to paying back the loan, everything's fine.
19:34 masak PerlJam: yes but no :)
19:34 masak PerlJam: technical debt and failing tests feel orthogonal in my mind.
19:34 masak but it's the same principle, yes.
19:35 diakopter <- is a failing test
19:35 * masak submits diakopter as a rakudobug
19:36 colomon can he be reproduced?
19:36 diakopter only by budding
19:36 PerlJam I think chromatic even published something along the lines of re-explaining technical debt as a loan that you should repay
19:36 * masak .oO( resulting in a Bud Light )
19:37 masak PerlJam: it's a common metaphor.
19:37 PerlJam indeed
19:37 masak PerlJam: and an interesting one, because it's not 100% about repaying, it's about keeping the debt under control.
19:39 diakopter if your interest rate on your technical debut is usurious...
19:39 arnsholt jnthn: What's the nqp:: equivalent of pir::repr_box_str__PsP?
19:40 dalek rakudo/nom: e1f2558 | (Elizabeth Mattijsen)++ | src/core/Bag.pm:
19:40 dalek rakudo/nom: Oops, new for Bag is not like new for Set
19:40 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/e1f2558821
19:40 jnthn nqp::box_s
19:40 diakopter er,.
19:40 diakopter technical debut, indeed.
19:41 arnsholt Oh, derp. Thanks!
19:42 masak diakopter: technical debut is caused by technical debutantes.
19:44 lizmat .oO( technical dillettantes cause techincal dill )
19:44 * colomon cannot keep up with lizmat++
19:44 * ingy waves o/
19:44 colomon ingy!  \o/
19:45 ingy hey there
19:45 * diakopter wavers
19:45 ingy :)
19:46 pmurias masak: re commiting things that fail, a lot of that is caused by tests running slowly
19:46 jnthn lizmat: does only sub infix:<<"\x228D">> work?
19:47 lizmat not sure yet, it compiles is all I know so far
19:47 lizmat let me see
19:47 masak ingy: hejhej! :D
19:47 Newman joined #perl6
19:48 masak pmurias: how do you mean? it was caused by lizmat deciding to, AFAICS.
19:48 jnthn lizmat: oh, the ones that all fail in the prec changes are &infix:<<"\x228D">> right?
19:49 lizmat also the ones with parens in the op name
19:49 lizmat fg
19:49 lizmat oops, ww
19:49 jnthn r: sub infix:<<"\x228D">>($a, $b) { say "$a $b" }; &infix:<<"\x228D">>(1, 2)
19:49 camelia rakudo e0b2af: OUTPUT«1 2␤»
19:49 jnthn r: sub infix:<<"\x228D">>($a, $b) { say "$a $b" }; say UNIT::.keys
19:49 camelia rakudo e0b2af: OUTPUT«$=pod !UNIT_MARKER GLOBALish EXPORT $?PACKAGE ::?PACKAGE $_ $/ $! %?LANG &infix:<⊍>␤»
19:51 FROGGS jnthn: why the extra quotes?
19:51 FROGGS : sub infix:«x228D»($a, $b) { say "$a $b" }; say UNIT::.keys
19:51 FROGGS r: sub infix:«\x228D»($a, $b) { say "$a $b" }; say UNIT::.keys
19:51 camelia rakudo e0b2af: OUTPUT«$=pod !UNIT_MARKER GLOBALish EXPORT $?PACKAGE ::?PACKAGE $_ $/ $! %?LANG &infix:<⊍>␤»
19:51 jnthn FROGGS: Was just copying what lizmat did in the commit :)
19:51 FROGGS and it is even shorter if you strip the "r:" :o)
19:52 lizmat FWIW, the only version that does not blow up in the SETTING, is the one with <<"\xhhhh">>
19:55 diakopter ..
20:00 BooK joined #perl6
20:00 colomon interesting that the set constructor doesn't work in just that one case.
20:01 jaldhar joined #perl6
20:01 dalek Perlito: 7b7f08f | (Flavio S. Glock)++ | / (5 files):
20:01 dalek Perlito: Perlito5 - js - hash slice tests
20:01 dalek Perlito: review: https://github.com/fglock/​Perlito/commit/7b7f08fdd9
20:01 colomon oh, wait
20:01 colomon roast is wrong, not rakudo
20:02 colomon I think?
20:02 lizmat checking…
20:02 colomon rn: my $b = set { foo => 10, bar => 17, baz => 42 }.hash; say $b.elems; say $b.perl
20:03 camelia rakudo e0b2af: OUTPUT«3␤set("foo", "bar", "baz")␤»
20:03 camelia ..niecza v24-90-gd827fa8: OUTPUT«3␤Set.new("foo\t10", "bar\t17", "baz\t42")␤»
20:09 dalek nqp/file-op-retcodes: 57da120 | (Donald Hunter)++ | src/vm/jvm/ (2 files):
20:09 dalek nqp/file-op-retcodes: Add :tc to signatures for file ops.
20:09 dalek nqp/file-op-retcodes: review: https://github.com/perl6/nqp/commit/57da120197
20:09 dalek nqp/file-op-retcodes: 536c063 | (Donald Hunter)++ | src/vm/jvm/stage0/ (10 files):
20:09 dalek nqp/file-op-retcodes: Update bootstrap for file ops with :tc signatures.
20:09 dalek nqp/file-op-retcodes: review: https://github.com/perl6/nqp/commit/536c063b4c
20:09 dalek nqp/file-op-retcodes: 47a64bf | (Donald Hunter)++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
20:09 dalek nqp/file-op-retcodes: Throw exceptions from file ops.
20:09 dalek nqp/file-op-retcodes: review: https://github.com/perl6/nqp/commit/47a64bf61d
20:09 dalek nqp/file-op-retcodes: 7df1c7a | (Donald Hunter)++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/ (2 files):
20:09 dalek nqp/file-op-retcodes: Better exception message generation.
20:09 dalek nqp/file-op-retcodes: review: https://github.com/perl6/nqp/commit/7df1c7aba0
20:12 dalek roast: 8e4f6ce | (Elizabeth Mattijsen)++ | S02-types/ (4 files):
20:12 dalek roast: First set of unfudges/new fudges for Set/Bag related changes
20:12 dalek roast: review: https://github.com/perl6/roast/commit/8e4f6ce8b0
20:12 colomon I completely fail to understand why set { foo => 10, bar => 17, baz => 42 }.hash and set { foo => 10, bar => 17, baz => 42 } should be differnet
20:12 colomon rn: say { foo => 10, bar => 17, baz => 42 }.WHAT
20:12 camelia rakudo e0b2af, niecza v24-90-gd827fa8: OUTPUT«(Hash)␤»
20:12 colomon rn: say { foo => 10, bar => 17, baz => 42 }.hash.WHAT
20:13 camelia rakudo e0b2af, niecza v24-90-gd827fa8: OUTPUT«(Hash)␤»
20:13 dalek nqp: 57da120 | (Donald Hunter)++ | src/vm/jvm/ (2 files):
20:13 dalek nqp: Add :tc to signatures for file ops.
20:13 dalek nqp: review: https://github.com/perl6/nqp/commit/57da120197
20:13 dalek nqp: 536c063 | (Donald Hunter)++ | src/vm/jvm/stage0/ (10 files):
20:13 dalek nqp: Update bootstrap for file ops with :tc signatures.
20:13 dalek nqp: review: https://github.com/perl6/nqp/commit/536c063b4c
20:13 dalek nqp: 47a64bf | (Donald Hunter)++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
20:13 dalek nqp: Throw exceptions from file ops.
20:13 dalek nqp: review: https://github.com/perl6/nqp/commit/47a64bf61d
20:13 dalek nqp: 7df1c7a | (Donald Hunter)++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/ (2 files):
20:13 dalek nqp: Better exception message generation.
20:13 dalek nqp: review: https://github.com/perl6/nqp/commit/7df1c7aba0
20:13 dalek nqp: da7b7b1 | (Donald Hunter)++ | src/vm/jvm/ (13 files):
20:13 dalek nqp: Merge pull request #119 from perl6/file-op-retcodes
20:13 dalek nqp:
20:13 dalek nqp: Update the file ops to throw exceptions on failure.
20:13 dalek nqp: review: https://github.com/perl6/nqp/commit/da7b7b1f67
20:13 arnsholt Oooh. Segfault
20:13 colomon I'm 95% certain that the set { foo => 10, bar => 17, baz => 42 }.hash tests are a leftover from the old spec.
20:14 lizmat colomon: I'm just looking at tests, to make them clean in the spectest, not at why they're failing
20:14 colomon lizmat: I'm looking at the big picture.
20:14 arnsholt Segfault in the binder, even. Odd.
20:14 lizmat and I'm not (yet)
20:14 diakopter arnsholt: jvm?
20:15 pmurias jnthn: do you remeber how does (foo)+ now to capture (foo) as an array?
20:15 lizmat r: say Set(Any.new)   # what should this do if it doesn't fail ?
20:15 camelia rakudo e0b2af: OUTPUT«Cannot coerce object of type Capture to Set. To create a one-element set, pass it to the 'set' function␤  in sub to-set at src/gen/CORE.setting:14701␤  in sub to-set at src/gen/CORE.setting:14694␤  in method postcircumfix:<( )> at src/gen/CORE.setting:14693␤  in bl…
20:15 pmurias s/now/know/
20:15 arnsholt diakopter: Parrot, so a bit less surprising =)
20:16 arnsholt Pretty sure it's my fault too
20:16 lizmat ah. camelia not up to date yet
20:16 lizmat $ perl6 -e 'say Set(Any.new)'
20:16 lizmat set(Any<-9213374975699074855>)
20:16 jferrero joined #perl6
20:16 lizmat r: say Any.new
20:16 camelia rakudo e0b2af: OUTPUT«Any.new()␤»
20:18 colomon lizmat: I don't know how well it works yet -- we need Sets which can handle any object type -- but it should create a Set that has a single Any in it.  (Assuming Any makes sense as an object.)
20:19 BooK joined #perl6
20:19 lizmat r: say Set(Any.new).keys
20:19 camelia rakudo e1f255: OUTPUT«Any<1696284401>␤»
20:19 lizmat r: say Set(Any.new).keys[0].exists
20:19 camelia rakudo e1f255: OUTPUT«Not enough positional parameters passed; got 1 but expected 2␤  in method exists at src/gen/CORE.setting:1447␤  in block  at /tmp/eMWxtwdhPx:1␤␤»
20:19 lizmat r: say Set(Any.new).keys[0].exists('foo')
20:19 camelia rakudo e1f255: OUTPUT«False␤»
20:20 colomon it probably adds Any.new.Str into the Set right now
20:20 lizmat seems to be a real object in there, so that's ok I guess
20:20 colomon nr: say Any.new.Str
20:20 camelia niecza v24-90-gd827fa8: OUTPUT«Any()<instance>␤»
20:20 camelia ..rakudo e1f255: OUTPUT«Any<-1364598985>␤»
20:21 masak blog post written.
20:21 dalek roast/origin/master: dafc74f | (Donald Hunter)++ | S32-io/slurp.t:
20:21 dalek roast/origin/master: Unfudge slurp read tests on JVM.
20:21 dalek roast/origin/master: review: https://github.com/perl6/roast/commit/dafc74ffb4
20:21 masak but apparently I have to rebuild Rakudo on this computer before I can run my blog software.
20:21 * masak crosses fingers and hopes for no regressions!
20:21 * lizmat crosses her fingers too .L;KAOAFKY.SDHGX/LAISk
20:21 lizmat hmm, better uncross them before I go on
20:22 TimToady for a moment there I thought my eyes were crossed
20:22 dalek roast: ced8cd2 | (Elizabeth Mattijsen)++ | S32-exceptions/misc.t:
20:22 dalek roast: Remove bogus Set(Any.new) test
20:22 dalek roast: review: https://github.com/perl6/roast/commit/ced8cd2538
20:23 dalek v5: d1a0a0b | (Tobias Leich)++ | STATUS.md:
20:23 dalek v5: status update, after rakudos chdir changes
20:23 dalek v5: review: https://github.com/rakudo-p5/v5/commit/d1a0a0bad6
20:23 lizmat only two more test files with failures to go
20:23 zwut00 joined #perl6
20:23 colomon TimToady: can you think of any reason set { foo => 10, bar => 17, baz => 42 }.hash would be different from set { foo => 10, bar => 17, baz => 42 } ?
20:23 TimToady you mean like, {} doesn't interpolate in list context?
20:24 masak r: constant True = 42; say +True
20:24 camelia rakudo e1f255: OUTPUT«42␤»
20:24 masak r: constant True = "yeah, well, you know, that's just, like, your opinion, man"; say True
20:24 camelia rakudo e1f255: OUTPUT«yeah, well, you know, that's just, like, your opinion, man␤»
20:24 masak \o/
20:24 colomon TimToady: should {}.hash interpolate in list context?
20:24 * masak adds that to https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=111734
20:25 TimToady colomon: I believe so
20:26 pmurias jnthn: I'm figurint out myself how it's done
20:26 colomon okay, so that is the rakudo bug
20:26 colomon TimToady++
20:26 TimToady just as .[] or .{} would deref
20:26 jnthn pmurias: Sorry, missed question while hutning GC bug...
20:27 jnthn pmurias: I know where it's worked out, moment
20:27 TimToady r: say [ [<a b c>] ].elems
20:27 lizmat jnthn: it would appear that the "\xhhh" operators indeed have not arrived
20:27 camelia rakudo e1f255: OUTPUT«1␤»
20:27 lizmat $ perl6 a.pm6
20:27 lizmat ===SORRY!=== Error while compiling a.pm6
20:27 lizmat Two terms in a row
20:27 lizmat at a.pm6:2
20:27 lizmat ------> my $b=bag( <a b c> ); say $b; say $b ⏏∪ $b
20:27 TimToady r: say [ [<a b c>][] ].elems
20:27 camelia rakudo e1f255: OUTPUT«3␤»
20:27 colomon rn: .say for { foo => 10, bar => 17, baz => 42 }.hash
20:27 TimToady r: say [ [<a b c>].array ].elems
20:27 camelia rakudo e1f255: OUTPUT«("foo" => 10, "bar" => 17, "baz" => 42).hash␤»
20:27 camelia ..niecza v24-90-gd827fa8: OUTPUT«"foo" => 10␤"bar" => 17␤"baz" => 42␤»
20:27 camelia rakudo e1f255: OUTPUT«No such method 'array' for invocant of type 'Array'␤  in block  at /tmp/V9zw8FSPGy:1␤␤»
20:27 colomon rn: .say for { foo => 10, bar => 17, baz => 42 }
20:27 * lizmat is going to rewrite those tests to use texan
20:27 camelia rakudo e1f255: OUTPUT«("foo" => 10, "bar" => 17, "baz" => 42).hash␤»
20:27 camelia ..niecza v24-90-gd827fa8: OUTPUT«{"bar" => 17, "baz" => 42, "foo" => 10}␤»
20:27 jnthn pmurias: sub capnames
20:28 jnthn pmurias: In src/QRegex/P6Regex/Actions.nqp
20:28 lizmat .u ∪
20:28 TimToady r: say [ { :a, :b } ].elems
20:28 camelia rakudo e1f255: OUTPUT«1␤»
20:28 TimToady r: say [ { :a, :b }{} ].elems
20:28 camelia rakudo e1f255: OUTPUT«2␤»
20:28 TimToady r: say [ { :a, :b }.hash ].elems
20:28 camelia rakudo e1f255: OUTPUT«1␤»
20:28 lizmat .oO( where is yoleaux when you need her )
20:29 TimToady yes, I'd call that a bug, I think
20:29 masak dammit, https://rt.perl.org/rt3/Tic​ket/Display.html?id=117841 is still open. I'll have to downgrade my Rakudo to get the blog post out.
20:30 TimToady r: say [ %({ :a, :b }) ].elems
20:30 camelia rakudo e1f255: OUTPUT«1␤»
20:30 TimToady that's the same bug
20:31 TimToady % in list context should flatten
20:31 * masak guesses the bug didn't exist a year ago, and downgrades to Rakudo 2012.09
20:31 TimToady r: say [ (1,2) ].elems
20:31 camelia rakudo e1f255: OUTPUT«2␤»
20:31 TimToady r: say [ (1,(2,3)) ].elems
20:31 camelia rakudo e1f255: OUTPUT«3␤»
20:31 TimToady okay, just the %/.hash
20:31 dalek perl6-roast-data: cfc5b0d | coke++ | / (5 files):
20:31 dalek perl6-roast-data: today (automated commit)
20:31 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/cfc5b0dd3b
20:32 FROGGS masak: it works for me now...
20:32 FROGGS masak: https://gist.github.com/FR​OGGS/e26c9b14854e3cc94936
20:32 jferrero joined #perl6
20:33 masak FROGGS: huh.
20:34 dalek roast: 3fa3416 | (Solomon Foster)++ | S02-types/set.t:
20:34 dalek roast: Add comments and new fudging string.
20:34 dalek roast:
20:34 dalek roast: With luck this will eliminate future confusion over this subject.
20:34 dalek roast: review: https://github.com/perl6/roast/commit/3fa3416318
20:35 [Coke] LHF: look for "todo PASSED" in https://github.com/coke/perl6-roast-da​ta/blob/master/rakudo.jvm_summary.out and untodo them.
20:35 donaldh joined #perl6
20:36 [Coke] OHEY, rakudo.parrot has no failures today.
20:36 masak \o/
20:36 lizmat colomon: this probably also applies to some S02-types/bag.t tests ?
20:36 colomon maybe?  let me look
20:37 masak FROGGS: `icuinfo` says I have Compiled-Version: 4.8.1.1 installed.
20:37 Rotwang1 joined #perl6
20:39 dalek nqp: 197640d | (Arne Skjærholt)++ | src/vm/parrot/6model/reprs/CStr.c:
20:39 dalek nqp: Mark CStr as being able to box strings and implement get_str.
20:39 dalek nqp: review: https://github.com/perl6/nqp/commit/197640def9
20:40 masak argh, 2012.09 also has the bug :(
20:40 * masak downgrades to 2011.09
20:41 BooK joined #perl6
20:41 dalek perl6-roast-data: bc4a6a5 | coke++ | cull:
20:41 dalek perl6-roast-data: use ISO dates going forward
20:41 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/bc4a6a5aef
20:41 dalek roast: 2907bc6 | (Solomon Foster)++ | S02-types/bag.t:
20:41 dalek roast: Add comments and new fudging string.
20:41 dalek roast:
20:41 dalek roast: With luck this will eliminate future confusion over this subject.
20:41 dalek roast: review: https://github.com/perl6/roast/commit/2907bc6c83
20:41 * lizmat is glad it's not something she did recently
20:41 lizmat colomon++
20:42 lizmat colomon: I see you wrote tests for both ∪ and (|)
20:42 lizmat did you do that for all the new ops ?
20:42 colomon lizmat: I think so?  I certainly intended to
20:42 colomon lizmat: they're both part of the spec
20:42 dalek zavolaj: 6f912ac | (Arne Skjærholt)++ | / (2 files):
20:42 dalek zavolaj: Unbust explicitly managed strings.
20:42 dalek zavolaj: review: https://github.com/jnthn/zavolaj/commit/6f912aca5f
20:42 lizmat okidoki, then my work is almost done :-)
20:43 arnsholt moritz: I think Zavolaj should be fixed now. Could you give your DB things a whirl and see if everything works properly?
20:43 stevan_ joined #perl6
20:44 arnsholt jnthn++ # Root cause discovery
20:44 moritz arnsholt: will do tomorrow
20:44 arnsholt Thanks!
20:45 FROGGS masak: same
20:46 dalek perl6-roast-data: 2ba3ef4 | coke++ | perl6_pass_rates.csv:
20:46 dalek perl6-roast-data: fix stats from previous day that went in twice.
20:46 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/2ba3ef4645
20:46 dalek perl6-roast-data: 0985dab | coke++ | perl6_pass_rates.csv:
20:46 dalek perl6-roast-data: Fixup non-ISO dates that slipped in
20:46 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/0985dab6fc
20:47 FROGGS masak: $ perl6 -e 'say qx[perl6 -e "say 1 x 4095,q|—|" | cat]; say "alive"'
20:47 FROGGS Unaligned end in UTF-8 string
20:47 FROGGS my chunksize is just different now
20:48 masak FROGGS: oh!
20:48 masak FROGGS: I feel so powerless in face of that bug.
20:48 masak FROGGS: it really messes up my ability to use Perl 6 for something real and useful.
20:49 FROGGS :/
20:49 masak though maybe I can now explore the JVM backend for this.
20:49 FROGGS >.<
20:49 FROGGS rakudo-jvm$ ./perl6 -e 'say qx[./perl6 -e "say 1 x 4095,q|—|" | cat]; say "alive"'
20:49 FROGGS Unhandled file open mode 'rp'
20:49 masak dang.
20:49 diakopter ETOOPIPEY
20:49 jnthn ah, we still didn't get the pipe open in...
20:49 FROGGS if you'd time you could hunt that bug down
20:50 masak bug, as in, missing feature? :)
20:50 FROGGS it is up to you :o)
20:50 jnthn NYI
20:50 FROGGS you have a choice at least
20:50 masak because I've been looking at the Unicode bug in the Parrot source code, and it's... well-hidden.
20:50 masak they have loads of layers of abstraction, making everything look real neat and bug-free.
20:50 masak except it isn't.
20:51 diakopter masak: that's the several layers of obfuscators
20:52 diakopter Bogged Down by Smoke and Mirrors
20:52 masak It's Not A Refactor If You Break Someone's Production Code(TM).
20:52 diakopter i don't recognize that backronym
20:53 masak oh my, I think I just made a chromatic mumble "if only they saw the irony of it all" somewhere in the world :P
20:53 * masak waves to chromatic
20:53 * diakopter irons out your irony
20:54 masak oh phew, 2011.09 doesn't have the regression.
20:54 diakopter but.. it has a few other regressions
20:54 lizmat .oO( chromatic, is that about shades of grey? )
20:55 diakopter 50 of those takes 6 bits
20:56 masak heh, apparently 2011.09 is too old to run my blog, though...
20:56 * masak upgrades slightly
20:56 masak this will take all evening.
20:56 perigrin lizmat: only when he was sick with mono.
20:56 diakopter *goran*
20:57 masak kind of an off-color joke, don't you think?
20:57 diakopter masak: it's a sliding scale
20:57 lizmat and that on a day like this, 50 years hence ?
20:58 * arnsholt <3 git
20:58 masak diakopter: so I guess there's no way to be discrete about it, then. :)
20:58 lizmat .oO( some people don't have a dream anymore )
20:59 diakopter masak: but it's pretty accident-prone
20:59 diakopter <- # ftw; two music jokes in a row
20:59 arnsholt lizmat: I think #perl6 is more concerned with colourless green ideas
21:00 lizmat I guess we're all super green on that
21:00 xenoterracide joined #perl6
21:00 * cognominal goes to sleep furiously
21:01 masak furieusement!
21:01 lizmat bonne nuit, cognominal!
21:02 cognominal thx, good night to you,  all * cast.
21:05 bruges joined #perl6
21:07 diakopter masak: put away your rainbow flag..
21:10 masak yay! finally it runs!
21:10 masak I settled on 2012.05, by the way.
21:11 masak note to self, in case irclog's search function ever becomes reliable again.
21:13 masak lol, I blug! \o/ http://strangelyconsistent.​org/blog/t3-wire-crossings
21:13 * donaldh golfs all the rakudo JVM regressions
21:13 masak can I do t4 within... a week? yes, I think I can, actually.
21:15 dalek roast: f70457d | (Tobias Leich)++ | S11-modules/require.t:
21:15 dalek roast: added chdir+require test
21:15 dalek roast: review: https://github.com/perl6/roast/commit/f70457dc48
21:15 * donaldh looks like is_run is bust:
21:15 donaldh sh: perl6: command not found
21:16 dalek rakudo/nom: c45ddb9 | (Tobias Leich)++ | src/vm/jvm/ModuleLoaderVMConfig.nqp:
21:16 dalek rakudo/nom: take cwd into acount when looking up modules
21:16 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/c45ddb96e3
21:18 dalek roast: 6682a3b | (Elizabeth Mattijsen)++ | S03-operators/ (2 files):
21:18 dalek roast: Fudge all the failing Set/Bag tests
21:18 dalek roast: review: https://github.com/perl6/roast/commit/6682a3b2ec
21:18 lizmat spectest should be clean again now
21:18 FROGGS lizmat++
21:18 donaldh haha, busted by the nqp JVM $*EXECUTABLE_NAME changing from "java" to "perl6"
21:18 lizmat however, it appears that most of the new operators simply do not work
21:19 lizmat :-(
21:19 donaldh r: say($*EXECUTABLE_NAME)
21:19 camelia rakudo e1f255: OUTPUT«./nom-inst/bin/perl6␤»
21:19 donaldh ah.
21:20 donaldh no path on JVM, just the "perl6" bare word. Would require . to be in PATH
21:20 lizmat donaldh: does PR 119 mean we now have a nqp:tc ?
21:21 celesta joined #perl6
21:21 celesta ahoi #perl6
21:22 lizmat celesta o/
21:22 donaldh lizmat: no, that means I added the :tc ThreadContext parameter to all the nqp file ops for JVM.
21:22 lizmat ah, ok
21:23 donaldh so that I could use die_s(…, tc);
21:24 lizmat okidoki, it's just that I recently implemented .tc (titlecase) rather naively
21:24 diakopter r: $OUTER::OUTER::OUTER::OUTER::OUTER::OUTER::True
21:24 camelia rakudo e1f255: OUTPUT«Null PMC access in get_attr_str()␤  in sub  at src/gen/CORE.setting:8786␤  in method at_key at src/gen/CORE.setting:8839␤  in method postcircumfix:<{ }> at src/gen/CORE.setting:1861␤  in method postcircumfix:<{ }> at src/gen/CORE.setting:1849␤  in block  at /tmp/OB…
21:24 jferrero joined #perl6
21:24 diakopter whee
21:24 diakopter masak: NPMCA
21:24 lizmat and put it on my list for after the RaNIW
21:24 lue masak++ # continuing the review
21:24 Khisanth joined #perl6
21:25 celesta r: my %h = ( 1 => <foo bar> ); my @baz = %h<1>; say @baz[0]
21:25 camelia rakudo e1f255: OUTPUT«foo bar␤»
21:25 celesta i expected output "foo"
21:26 panchiniak joined #perl6
21:26 lizmat r: my %h = ( 1 => <foo bar> ); say %h
21:26 camelia rakudo e1f255: OUTPUT«("1" => $("foo", "bar")).hash␤»
21:26 celesta r: my %h = ( 1 => <foo bar> ); my @baz = %h<1>; say @baz.WHAT
21:26 camelia rakudo e1f255: OUTPUT«(Array)␤»
21:27 * masak submits rakudobug
21:27 masak diakopter++
21:27 lizmat r: my %h = ( 1 => <foo bar> ); say %h; my @baz=%h<1>; say @baz
21:27 camelia rakudo e1f255: OUTPUT«("1" => $("foo", "bar")).hash␤foo bar␤»
21:27 lizmat r: my %h = ( 1 => <foo bar> ); say %h; my @baz=%h<1>; say @baz.elems
21:27 camelia rakudo e1f255: OUTPUT«("1" => $("foo", "bar")).hash␤1␤»
21:27 masak r: $OUTER::OUTER::OUTER::OUTER::OUTER::True
21:27 camelia rakudo e1f255:  ( no output )
21:27 celesta me confused now
21:27 lizmat r: my %h = ( 1 => <foo bar baz > ); say %h; my @baz=%h<1>; say @baz.elems
21:27 masak r: say $OUTER::OUTER::OUTER::OUTER::OUTER::True
21:27 camelia rakudo e1f255: OUTPUT«("1" => $("foo", "bar", "baz")).hash␤1␤»
21:27 camelia rakudo e1f255: OUTPUT«(Any)␤»
21:30 lizmat r: my %h = ( 1 => <foo bar baz > ); say %h; my @baz=%h<1>.list; say @baz.elems
21:30 camelia rakudo e1f255: OUTPUT«("1" => $("foo", "bar", "baz")).hash␤3␤»
21:30 lizmat r: my %h = ( 1 => <foo bar baz > ); say %h; my @baz=%h<1>.WHAT; say @baz.elems
21:30 camelia rakudo e1f255: OUTPUT«("1" => $("foo", "bar", "baz")).hash␤1␤»
21:30 lizmat r: my %h = ( 1 => <foo bar baz > ); say %h; say %h<1>.WHAT
21:30 camelia rakudo e1f255: OUTPUT«("1" => $("foo", "bar", "baz")).hash␤(Parcel)␤»
21:31 lizmat r: my %h = ( 1 => [<foo bar baz >] ); say %h; say %h<1>.WHAT
21:31 camelia rakudo e1f255: OUTPUT«("1" => ["foo", "bar", "baz"]).hash␤(Array)␤»
21:31 lizmat r: my %h = ( 1 => [<foo bar baz >] ); say %h; my @baz= %h<1>; say @baz.perl
21:31 camelia rakudo e1f255: OUTPUT«("1" => ["foo", "bar", "baz"]).hash␤Array.new(["foo", "bar", "baz"])␤»
21:32 lizmat celesta: I *think* it's according to spec, as a Parcel doesn't flatten in that context, an Array does
21:32 celesta lizmat: thank you
21:33 * lizmat finds that distinction still rather hard to grok
21:33 * diakopter still glazes over when the word parcel appears
21:33 ssutch joined #perl6
21:33 * masak too
21:34 masak it's a bit like Python's tuple type, I guess.
21:34 lue masak: Reading through your reviews, I wonder if I shouldn't start future-proofing my Int->Bool coercions and just write +(True/True) :)
21:34 masak lue: way to miss the core complaint :P
21:34 lue er, Bool->Int
21:35 lue r: say (0/0).WHAT
21:35 camelia rakudo e1f255: OUTPUT«Divide by zero␤  in method sink at src/gen/CORE.setting:11557␤  in method BUILDALL at src/gen/CORE.setting:893␤  in method bless at src/gen/CORE.setting:834␤  in method new at src/gen/CORE.setting:819␤  in method new at src/gen/CORE.setting:817␤  in sub DIVIDE_NUMB…
21:41 denisboyun joined #perl6
21:42 yoleaux joined #perl6
21:51 dalek roast/jvm-prog-name: 3128fcd | (Donald Hunter)++ | packages/Test/Util.pm:
21:51 dalek roast/jvm-prog-name: Another JVM process name workaround.
21:51 dalek roast/jvm-prog-name: review: https://github.com/perl6/roast/commit/3128fcdd32
21:52 dalek roast: 3128fcd | (Donald Hunter)++ | packages/Test/Util.pm:
21:52 dalek roast: Another JVM process name workaround.
21:52 dalek roast: review: https://github.com/perl6/roast/commit/3128fcdd32
21:52 dalek roast: c5c0a68 | (Donald Hunter)++ | packages/Test/Util.pm:
21:52 dalek roast: Merge pull request #33 from perl6/jvm-prog-name
21:52 dalek roast:
21:52 dalek roast: Another JVM process name workaround. Should resolve a lot of JVM regressions.
21:52 dalek roast: review: https://github.com/perl6/roast/commit/c5c0a6879f
21:52 ajr joined #perl6
21:59 celesta thanks and good night #perl6
22:12 stevan_ joined #perl6
22:18 FROGGS .tell donaldh does ./perl6 work on windows at all? https://github.com/perl6/roast/commit/c5c0a6879f
22:18 yoleaux FROGGS: I'll pass your message to donaldh.
22:28 skids joined #perl6
22:41 pernatiy joined #perl6
22:44 alexghacker joined #perl6
22:45 snoop`nqr joined #perl6
22:48 snoopy` joined #perl6
22:54 dalek rakudo/nom: 7b26cbf | (Elizabeth Mattijsen)++ | src/core/ (2 files):
22:54 dalek rakudo/nom: Refinement on use of flattening
22:54 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/7b26cbfa96
22:56 lizmat and on that thought, I wish you all a good night!
22:56 lizmat sleep&
22:58 jnthn 'night, lizmat
23:00 PZt joined #perl6
23:06 jnthn sleep also &
23:08 FROGGS joined #perl6
23:11 gdey joined #perl6
23:19 masak 'night, #perl6
23:40 abnorman joined #perl6

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo