»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:09 uasi left, jamtech_ joined, jamtech left
dalek ast: 53b43a6 | Coke++ | S05-metasyntax/repeat.t:
rakudo/nom fudging
 00:12
kudo/nom: 0d73bac | Coke++ | t/spectest.data:
run fudged test
 00:13
00:14 replore_ joined, Chillance left 00:30 uasi joined 00:33 soh_cah_toa joined, y3llow left, y3llow joined
[Coke] rakudo: sub foo { say "bar" }; my $x = 'foo'; ::$x() 00:39
p6eval rakudo b10ef5: OUTPUT«===SORRY!===␤Cannot look up empty name␤»
diakopter does it work without the :: 00:40
oh, of course not
sorear should be ::($x) anyway
[Coke] sorear: perhaps you can close out RT#64846 00:41
00:46 jamtech_ left
sorear [Coke]: no. 00:47
00:47 jamtech_ joined
sorear hmm, rakudo needs to become self-aware 00:48
00:49 buubot_backup left
[Coke] hey, rakudodevs. can we reject #68636 ? 00:53
rakudo: %*h.push: $*IN.slurp.join.comb.map(-> $k {$k=>1}); say %*h.perl 00:54
p6eval rakudo b10ef5: OUTPUT«Method 'push' not found for invocant of class 'Failure'␤ in <anon> at /tmp/HOfa8x5OBj:1␤ in <anon> at /tmp/HOfa8x5OBj:1␤␤»
sorear niecza: class B0rk { say $.a } 00:55
p6eval niecza v9-32-g380d891: OUTPUT«===SORRY!===␤␤Variable $.a used where no 'self' is available at /tmp/QbuhzpkBAZ line 1:␤------> class B0rk { say ⏏$.a }␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 685 (CORE di…
colomon woah, slew of bug e-mails from [Coke]
:) 00:56
00:57 uasi left
[Coke] perl6: role woowoo {}; multi trait_mod:<is>(Routine $c, woowoo) { $c.wrap: sub { say "Woowoo" } }; sub foo is woowoo { say "foo" }; foo(); 00:59
p6eval niecza v9-32-g380d891: OUTPUT«===SORRY!===␤␤Cannot extend category:trait_mod with subs at /tmp/BMLhA2C9Nk line 1:␤------> role woowoo {}; multi trait_mod:<is>⏏(Routine $c, woowoo) { $c.wrap: sub { sa␤␤Unhandled exception: Unable to resolve method ast…
..pugs: OUTPUT«*** ␤ Unexpected ":<is"␤ expecting "::", "handles", "is", bare trait, subroutine parameters, trait or block␤ at /tmp/nyWkzg6sgM line 1, column 32␤»
..rakudo 0d73ba: OUTPUT«===SORRY!===␤Method 'wrap' not found for invocant of class 'Sub'␤»
soh_cah_toa that reminds me, in other trait_mod definitions (in src/core/traits.pm), i see signatures w/ what look like smiley faces attached to the type. for instance, 'multi trait_mod:<of>(Routine:D $target, Mu:U $type)'. what are the :D and :U for? i don't think i've ever seen those in the spec before 01:04
[Coke] t/spec/S02-types/enum.t - any reason this file isn't even in spectest.data ? 01:06
01:06 uasi joined 01:09 buubot_backup joined 01:14 thou left
dalek ast: 74ac196 | Coke++ | S02-types/enum.t:
rakudo nom fudging
 01:18
[Coke] guesses no.
dalek kudo/nom: cbdd9b0 | Coke++ | t/spectest.data:
run fudged test.
[Coke] rakudo: ({nextsame})() 01:19
p6eval rakudo 0d73ba: OUTPUT«No dispatcher in scope␤ in block <anon> at /tmp/eaLBhdn45F:1␤ in <anon> at /tmp/eaLBhdn45F:1␤ in <anon> at /tmp/eaLBhdn45F:1␤␤»
dalek ecs: fe723c3 | larry++ | S12-objects.pod:
rm "Conjectural" from :D and :U
 01:21
TimToady soh_cah_toa: see S12:1590 01:22
01:22 woosley joined 01:23 wolfman2000 joined
soh_cah_toa ah, i knew it! i guessed d = defined and u = undefined 01:23
so :T as a type object means an object that can't be instantiated, correct? 01:24
01:26 vlixes joined 01:29 arnsholt joined
TimToady well, it means an object *isn't* instantiated at that spot... 01:35
but you can certainly instantiate one out of a type object using the .new method :D
soh_cah_toa oh, so 'my Int $a' doesn't actually instantiate it until i say '$a = 42' 01:37
TimToady doesn't instantiate it then either, since the 42 is already instantiated 01:39
nom & 01:40
soh_cah_toa how is 42 instantiated? it's a constant
donri but still an object right? 01:41
soh_cah_toa oh right, you can do stuff like 42.abs() 01:42
everything is an object now, that's right 01:43
so then still, what's the difference between a type object and say a user-defined object? if a type object isn't instantiated when it's declared, when is it instantiated? 01:45
donri does that work, it's commonly a syntax error in many languages 01:46
rakudo: 42.abs
p6eval rakudo 0d73ba: ( no output )
donri rakudo: say 42.abs
p6eval rakudo 0d73ba: OUTPUT«42␤»
donri rakudo: say 42.
p6eval rakudo 0d73ba: OUTPUT«===SORRY!===␤Confused at line 1, near "say 42."␤»
donri rakudo: say .5
p6eval rakudo 0d73ba: OUTPUT«0.5␤»
01:50 buubot_backup left
[Coke] rakudo: say one.5 01:53
p6eval rakudo 0d73ba: OUTPUT«===SORRY!===␤Confused at line 1, near "say one.5"␤»
donri this aint ruby ;) 01:59
02:11 buubot_backup joined 02:18 mykhal left 02:20 donri left
TimToady perl6: say slurp; 02:22
p6eval niecza v9-32-g380d891: OUTPUT«Unhandled exception: No value for parameter $path in CORE slurp␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (CORE slurp @ 0) ␤ at /tmp/m4v6NNxa25 line 1 (MAIN mainline @ 1) ␤ at /home/p6eval/niecza/lib/CORE.setting line 2060 (CORE C968_ANON @ 2) ␤ a…
..rakudo cbdd9b: OUTPUT«No applicable candidates found to dispatch to for 'slurp'. Available candidates are:␤:(Any $filename)␤␤ in sub slurp at src/gen/CORE.setting:4892␤ in <anon> at /tmp/aDXoSGkc7x:1␤ in <anon> at /tmp/aDXoSGkc7x:1␤␤»
..pugs: OUTPUT«*** ␤ Unexpected ";"␤ at /tmp/geUmRzq0SQ line 1, column 10␤»
diakopter niecza: say slurp 'Makefile' 02:23
p6eval niecza v9-32-g380d891: OUTPUT«Unhandled exception: CORE slurp may not be used in safe mode␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (CORE slurp @ 0) ␤ at /tmp/Ew4ZEILOpb line 1 (MAIN mainline @ 1) ␤ at /home/p6eval/niecza/lib/CORE.setting line 2060 (CORE C968_ANON @ 2) ␤ at /…
TimToady slurp should default to ARGFILES
dinner really &
02:25 am0c joined 02:35 jamtech_ left 02:36 jamtech_ joined
TimToady perl6: say slurp($*ARGFILES) 02:44
p6eval niecza v9-32-g380d891: OUTPUT«Unhandled exception: CORE slurp may not be used in safe mode␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (CORE slurp @ 0) ␤ at /tmp/bYV_c9HJnQ line 1 (MAIN mainline @ 1) ␤ at /home/p6eval/niecza/lib/CORE.setting line 2060 (CORE C968_ANON @ 2) ␤ at /…
..rakudo cbdd9b: OUTPUT«This type cannot unbox to a native string␤ in method open at src/gen/CORE.setting:4761␤ in sub open at src/gen/CORE.setting:4874␤ in sub slurp at src/gen/CORE.setting:4894␤ in <anon> at /tmp/KMqpS6_q5U:1␤ in <anon> at /tmp/KMqpS6_q5U:1␤␤»
..pugs: OUTPUT«*** Unsafe function 'slurp' called under safe mode␤ at /tmp/e04AMhVxhP line 1, column 5 - line 2, column 1␤»
diakopter I think slurp and other file reading things should be allowed in safe mode; it's my p6eval box, after all 02:45
02:46 packetknife joined 02:50 __sri left 02:51 Patterner left, jfried joined, Gothmog_ joined, Psyche^ joined, Psyche^ is now known as Patterner
sorear diakopter: noted. 02:51
02:53 _sri joined
TimToady rosettacode.org/wiki/Letter_frequency#Perl_6 02:53
02:59 jnthn joined 03:00 vlixes left 03:04 cotto left 03:05 cotto joined 03:13 uasi left 03:18 jamtech__ joined 03:19 jamtech_ left, envi_|2 joined 03:28 packetknife left 03:31 bluescreen10 left
[Coke] rakudo: say index("uuúuúuùù", "úuù") 03:31
p6eval rakudo cbdd9b: OUTPUT«4␤»
[Coke] perl6: my $*x = 42; say $*x.WHAT 03:33
p6eval pugs: OUTPUT«Scalar␤»
..rakudo cbdd9b, niecza v9-32-g380d891: OUTPUT«Int()␤»
[Coke] rakudo: $*x = 42; say $*x.WHAT 03:34
p6eval rakudo cbdd9b: OUTPUT«Cannot assign to a non-container␤ in <anon> at /tmp/meGTbjLflb:1␤ in <anon> at /tmp/meGTbjLflb:1␤␤»
[Coke] rakudo: my Str %s = {a => 'b'} 03:35
p6eval rakudo cbdd9b: ( no output )
[Coke] rakudo: my %s of Str; %s<a> = 'b';
p6eval rakudo cbdd9b: ( no output )
[Coke] rakudo: my Str %s = a => 'b' 03:36
p6eval rakudo cbdd9b: ( no output )
03:38 uasi joined
[Coke] rakudo: enum A::B <a b c>; say b 03:44
p6eval rakudo cbdd9b: OUTPUT«A::B::b␤»
[Coke] rakudo: $a = (my $a) 03:49
p6eval rakudo cbdd9b: ( no output )
[Coke] rakudo: my $a = $_ given $a
p6eval rakudo cbdd9b: ( no output )
[Coke] rakudo: my @a = (1,2) Z (3,4); say @a.perl 03:50
p6eval rakudo cbdd9b: OUTPUT«Array.new(1, 3, 2, 4)␤»
[Coke] wonders why "testsneeded" comes last. 03:53
rakudo: &infix:(+)(3,4).say 03:54
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Malformed parameter at line 1, near "+)(3,4).sa"␤»
[Coke] rakudo: &infix:("+")(3,4).say
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Symbol '&infix:("+")' not predeclared in <anonymous> (/tmp/Sz2DJWxqL6:1)␤»
[Coke] rakudo: my $foo="+";&infix:($foo)(3,4).say
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Symbol '&infix:($foo)' not predeclared in <anonymous> (/tmp/qTMVLz747s:1)␤»
[Coke] rakudo: class A { method f(*%_) { |%_ } }' 03:55
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Confused at line 1, near "class A { "␤»
[Coke] rakudo: for 1..5 -> $a, $b { say $a, $b } 03:56
p6eval rakudo cbdd9b: OUTPUT«12␤34␤Not enough positional parameters passed; got 1 but expected 2␤ in block <anon> at /tmp/DxYdxC7jfN:1␤ in method reify at src/gen/CORE.setting:3690␤ in method reify at src/gen/CORE.setting:3595␤ in method reify at src/gen/CORE.setting:3595␤ in method gimme…
[Coke] rakudo: for 1..5 -> $a, $b { say $a, $b? }
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 1␤»
[Coke] rakudo: for 1..5 -> $a, $b? { say $a, $b }
p6eval rakudo cbdd9b: OUTPUT«12␤34␤5Mu()␤»
[Coke] rakudo: say %("foo" ~~ /foo/).exists("foo") 03:57
p6eval rakudo cbdd9b: OUTPUT«Bool::False␤»
03:59 satyavvd joined
[Coke] rakudo: ok %( 'foo' ~~ /<alpha> oo/ ){ 'alpha' }:exists; say "alive" 03:59
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Confused at line 1, near "ok %( 'foo"␤»
[Coke] rakudo: Inf.int.say 04:00
p6eval rakudo cbdd9b: OUTPUT«Method 'int' not found for invocant of class 'Num'␤ in <anon> at /tmp/qPsFlg_0FW:1␤ in <anon> at /tmp/qPsFlg_0FW:1␤␤»
[Coke] rakudo: Inf.Int.say
p6eval rakudo cbdd9b: OUTPUT«-9223372036854775808␤»
04:01 birdwindupbird joined
[Coke] sorear: ping. 04:08
phenny: ask sorear if we can close 75030 - it was originally rejected, and rakudo has autoprint in REPL now. 04:09
phenny [Coke]: I'll pass that on when sorear is around.
[Coke] rakudo: sub foo() { return 1,2,:c<3> }; say foo().perl 04:10
p6eval rakudo cbdd9b: OUTPUT«(1, 2, "c" => Mu)␤»
[Coke] rakudo: say '16'{'%x'}
p6eval rakudo cbdd9b: OUTPUT«Method 'at_key' not found for invocant of class 'Str'␤ in method postcircumfix:<{ }> at src/gen/CORE.setting:1002␤ in <anon> at /tmp/MHZjKseyZi:1␤ in <anon> at /tmp/MHZjKseyZi:1␤␤»
sorear [Coke]: you may close 75030.
phenny sorear: 04:09Z <[Coke]> ask sorear if we can close 75030 - it was originally rejected, and rakudo has autoprint in REPL now.
sorear [Coke]: I don't know if you can - I know I can't. Who can give out close bits? 04:11
moritz sorear: I think pmichaud can
good morning
[Coke] I can. 04:12
I am an rt p6 admin from the dark times.
but to give them out, I have to git-bisect parrot until I find the instructions. ;)
enum A <a b c>; enum B <a b c>; say &a.perl 04:13
rakudo: enum A <a b c>; enum B <a b c>; say &a.perl
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Symbol '&a' not predeclared in <anonymous> (/tmp/4cTc1McdVJ:1)␤»
[Coke] rakudo: enum A <a b c>; enum B <a b c d>; say &d.perl
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Symbol '&d' not predeclared in <anonymous> (/tmp/vOGTaBd5w0:1)␤»
[Coke] rakudo: my ::foo $x; say $x 04:19
p6eval rakudo cbdd9b: OUTPUT«Cannot call a method on type variable foo␤ in find_method at src/gen/Metamodel.pm:261␤ in sub say at src/gen/CORE.setting:4722␤ in sub say at src/gen/CORE.setting:4729␤ in <anon> at /tmp/ormgC9k2j3:1␤ in <anon> at /tmp/ormgC9k2j3:1␤␤»
[Coke] rakudo: my ::foo $x;
p6eval rakudo cbdd9b: ( no output )
[Coke] rakudo: my ::foo $x; say "alive"
p6eval rakudo cbdd9b: OUTPUT«alive␤»
04:22 jamtech joined, jamtech__ left
dalek ecza/serialize: 06f4a53 | sorear++ | / (6 files):
Code execution in new model, checkpoint 2
 04:22
[Coke] rakudo: rakudo: say (4...^5).perl 04:23
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Confused at line 1, near "rakudo: sa"␤»
[Coke] rakudo: say (4 ... 0, 1, 2, 3, 4).perl
p6eval rakudo cbdd9b: OUTPUT«(4, 3, 2, 1, 0, 1, 2, 3, 4).list␤»
[Coke] rakudo: rakudo: say (-5...^5).perl 04:24
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Confused at line 1, near "rakudo: sa"␤»
[Coke] rakudo: rakudo: say (4...^5).perl
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Confused at line 1, near "rakudo: sa"␤»
[Coke] rakudo: rakudo: say (4 ... ^5).perl
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Confused at line 1, near "rakudo: sa"␤»
[Coke] there's a list branch but I can't see it on github? 04:26
sorear it was deleted a while ago 04:27
git pull doesn't delete locally-saved branches
it only pulls new ones
[Coke] can I see when it was deleted?
sorear maybe by running reflog on the server 04:28
[Coke] eh. not worth it. Danke. 04:30
rakudo: die 04:33
p6eval rakudo cbdd9b: OUTPUT«␤ in <anon> at /tmp/FgvcEoz0pL:1␤ in <anon> at /tmp/FgvcEoz0pL:1␤␤»
[Coke] rakudo: ... 04:34
p6eval rakudo cbdd9b: ( no output )
[Coke] rakudo: fail
p6eval rakudo cbdd9b: ( no output )
[Coke] rakudo: exit
p6eval rakudo cbdd9b: ( no output )
[Coke] rakudo: .die
p6eval rakudo cbdd9b: OUTPUT«Method 'die' not found for invocant of class 'Any'␤ in <anon> at /tmp/0jzYTcsKvO:1␤ in <anon> at /tmp/0jzYTcsKvO:1␤␤»
[Coke] rakudo: class AtestOfSay { ... }; my AtestOfSay $t; say $t; 04:35
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤The following packages were stubbed but not defined:␤ AtestOfSay␤␤»
[Coke] rakudo: Seq.new.^methods>>.say 04:36
p6eval rakudo cbdd9b: OUTPUT«join␤unshift␤at_pos␤exists␤rindex␤say␤dispatch:<.*>␤acos␤dispatch:<::>␤infinite␤grep␤gist␤max␤atanh␤sqrt␤fmt␤cos␤sinh␤match␤BUILDALL␤cis␤infinite␤any␤pairs␤asech␤Array␤tan␤DUMP-ID␤trim-trailing␤classify␤iterator␤Num␤postcircumfix:<[ ]>␤cosech␤Int␤values␤capitalize␤…
[Coke] die Exception.new("Test"); 04:37
04:37 jaldhar left
[Coke] rakudo: die Exception.new("Test"); 04:37
p6eval rakudo cbdd9b: OUTPUT«Default constructor only takes named arguments␤ in method new at src/gen/CORE.setting:502␤ in <anon> at /tmp/Vb7LIiy8VA:1␤ in <anon> at /tmp/Vb7LIiy8VA:1␤␤»
04:38 jaldhar joined
[Coke] rakudo: sub foo() { succeed }; say foo 04:38
p6eval rakudo cbdd9b: OUTPUT«Warning␤0␤»
[Coke] rakudo: say <1 2 3>.>>.perl 04:40
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Confused at line 1, near "say <1 2 3"␤»
04:42 packetknife joined
[Coke] rakudo: say "abcd".index("xyz") == 0 04:43
p6eval rakudo cbdd9b: OUTPUT«Use of uninitialized value in numeric context␤Bool::True␤»
[Coke] rakudo: say Mu.perl 04:46
p6eval rakudo cbdd9b: OUTPUT«Mu␤»
[Coke] rakudo: class A {}; A.new.perl
p6eval rakudo cbdd9b: ( no output )
[Coke] rakudo: class A {}; A.new.perl.say
p6eval rakudo cbdd9b: OUTPUT«A<95569160>␤» 04:47
[Coke] rakudo: say (1..10).sort(&rand)
p6eval rakudo cbdd9b: OUTPUT«Too many positional parameters passed; got 1 but expected 0␤ in sub rand at src/gen/CORE.setting:2609␤ in method reify at src/gen/CORE.setting:3690␤ in method reify at src/gen/CORE.setting:3595␤ in method gimme at src/gen/CORE.setting:3932␤ in method eager at …
[Coke] rakudo: say 357+498+7 04:50
p6eval rakudo cbdd9b: OUTPUT«862␤»
04:52 packetknife left 04:53 soh_cah_toa left 05:03 mberends left
moritz nom: enum SomeEnum < a b c>; say SomeEnum::.keys 05:04
p6eval nom cbdd9b: OUTPUT«0␤»
moritz nom: enum SomeEnum < a b c>; say SomeEnum::.enums.keys
p6eval nom cbdd9b: OUTPUT«a b c␤»
moritz (that was RT #70894) 05:05
05:05 packetknife joined 05:07 jamtech left
moritz [Coke]++ # going wild in RT queue 05:09
dalek ast: a4a120d | moritz++ | S0 (3 files):
random rakudo unfudges
 05:19
05:26 koban` joined, koban` left 05:39 daniel-s joined 05:41 cookys joined 06:02 wtw joined, lestrrat left 06:03 lestrrat joined 06:08 orafu left 06:10 orafu joined 06:11 JimmyZ joined 06:13 wolfman2000 left
JimmyZ rakudo: trait_mod:<is-a>(Routine:D $r, :$rw!) { $r.set_rw(); } # bug ? 06:14
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 1␤»
06:14 uasi_ joined 06:15 uasi left, uasi joined 06:16 kaleem joined
sorear JimmyZ: you need "sub " at the beginning of the line. 06:18
JimmyZ LTA error?
sorear JimmyZ: rakudo is parsing (...) as an argument and getting confused, because argument lists can't have types
It's a technically correct error, and I'm not sure how it can be improved. 06:19
06:19 uasi_ left
sorear except maybe by unifying parameter and argument syntax, but for a language like Perl 6 that would waterbed a geyser... it could probably be made to work in C# 06:20
06:20 Sarten-X left
JimmyZ rakudo: class A { method trait_mod:<is-a>( $r, :$rw!) { $r.set_rw(); } }; # how can I use is-a here? 06:21
p6eval rakudo cbdd9b: ( no output )
sorear A.trait_mod:<is-a>(@args)
implicit trait syntax only works for subs, not methods
general rule - if you're changing the syntax, it needs to be through something with the my-scope 06:22
jnthn It's just "is", not "is-a" 06:23
JimmyZ rakudo: class A { sub trait_mod:<is-a>( $r, :$rw!) { $r.set_rw(); } }; # and this one?
p6eval rakudo cbdd9b: ( no output )
jnthn Unless you're trying to introduce a new one...
mornin', btw
JimmyZ yes, I'm trying to introduce new one
sorear JimmyZ: Why do you keep saying class A?
06:23 zby_home_ joined
JimmyZ good morning, jnthn 06:23
sorear mornin', jnthn 06:24
cotto hi jnthn
JimmyZ sorear: I'm trying to find bug
jnthn JimmyZ: Then it needs to be declared multi trait_mod:<blah>(...) { ... }, but Rakudo won't augment the parser for new ones yet.
sorear JimmyZ: I'm fairly sure that introducing new trait keywords is not supported by any compiler yet
(this will change next month) 06:25
06:25 packetknife left
JimmyZ rakudo: class A { sub trait_mod:<is>( $r, :$rw!) { $r.set_rw(); } }; # how can I use is here? 06:25
p6eval rakudo cbdd9b: ( no output )
JimmyZ I didn't is not supported ,hehe
I didn't know it is not supported ,hehe
jnthn YOu can declare custom trait mods for sure.
For now you should explicitly mark them multi. 06:26
JimmyZ rakudo: class A { mutil trait_mod:<is>(Class:D $r, :$rw!) { $r.set_rw(); } }; # how can I use is here?
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 1␤»
jnthn JimmyZ: See my grammar debugger module for an example of a module that adds, and exports, extra traits.
multi, not mutil :)
YOu use traits by putting them on declarative thins. 06:27
06:27 Sarten-X joined
JimmyZ rakudo: class A { multi trait_mod:<is>(Class:D $r, :$rw!) { $r.set_rw(); } }; 06:28
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Invalid typename in parameter declaration at line 1, near " $r, :$rw!"␤»
jnthn rakudo: multi trait_mod:<is>(Routine:D $r, :$blah!) { $r does role { method blah { True } } }; sub foo() is blah { }; say &foo.blah
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Lexical '$r' not found␤»
jnthn uh...huh. 06:29
rakudo: multi trait_mod:<is>(Routine:D $r, :$blah!) { $r does role { method blah { True } } }
p6eval rakudo cbdd9b: ( no output )
jnthn meh, too little coffee to debug stuff... :)
JimmyZ rakudo: class A { multi trait_mod:<is>(Routine:D $r, :$rw!) { $r.set_rw(); } }; 06:30
p6eval rakudo cbdd9b: ( no output )
JimmyZ jnthn: How can I use it when it's defined in a class
jnthn JimmyZ: By applying it to things in the class. 06:32
method foo() is rw { ... } # should hit your trait mode
*mod
It may not, however, because I didn't do the tie-breaking on scope depth yet. 06:33
JimmyZ rakudo: class A { multi trait_mod:<is>( $r, :$rwr!) { $r.set_rw(); }; sub b is rwr { }; }; 06:34
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Lexical '$r' not found␤»
jnthn JimmyZ: That should work. I'm a bit confused about why it doesn't. 06:36
rakudo: sub foo($a) { say $a }; BEGIN foo(42) 06:37
p6eval rakudo cbdd9b: OUTPUT«42␤»
jnthn rakudo: multi sub foo($a) { say $a }; BEGIN foo(42) 06:38
p6eval rakudo cbdd9b: OUTPUT«42␤»
JimmyZ rakudo: class A { multi method trait_mod:<is>( :$rwr!) { }; Class b is rwr { }; };
p6eval rakudo cbdd9b: OUTPUT«Could not find sub &rwr␤ in <anon> at /tmp/fu78dmR3EG:1␤ in <anon> at /tmp/fu78dmR3EG:1␤ in <anon> at /tmp/fu78dmR3EG:1␤␤»
jnthn oh...it's that bug :( 06:39
moritz uhm, is it write to declare trait_mod as a methodß 06:40
s/ß/?/
jnthn moritz: no
moritz: I just found why the multi sub case isn't working though. 06:41
JimmyZ uses rakudo with a bit violence
how to user multi method trait_mod:<is> ?
use 06:42
jnthn JimmyZ: You're not meant to write it as a method.
JimmyZ: Traits are dispatched to as subs.
JimmyZ rakudo: class A { multi method trait_mod:<is>( :$rwr!) { }; }
p6eval rakudo cbdd9b: ( no output )
JimmyZ rakudo allows me to write it as a method ;) 06:43
jnthn So what?
It's just a name.
If you do class Foo { method m() { ... }; m() }
Then it'll let you write m as a method too.
But the dispatch will fail. 06:44
It's not that you can't write a method called trait_mod:<is>. You can - we do it in Actions.pm. It's that trait dispatch is a sub dispatch, not a method one.
moritz if it were a method, in which class would it look for the trait method? 06:46
sorear also, JimmyZ, trait subs should be declared at top level, not in a brace group 06:47
unless you want to limit the use of the trait to the group
06:47 frhodes joined 06:50 uasi left
TimToady perl6: [{a => 1}, {a => 2}]»<a>.say 06:52
p6eval pugs: OUTPUT«decodeUTF8': bad data: '\187'␤*** ␤ Unexpected "\187<"␤ expecting term postfix or operator␤ at /tmp/y0k53zSN_O line 1, column 21␤»
..rakudo cbdd9b: OUTPUT«Method 'at_key' not found for invocant of class 'Int'␤ in method postcircumfix:<{ }> at src/gen/CORE.setting:1002␤ in method reify at src/gen/CORE.setting:3690␤ in method reify at src/gen/CORE.setting:3595␤ in method reify at src/gen/CORE.setting:3595␤ in meth…
..niecza v9-32-g380d891: OUTPUT«Unhandled exception: Cannot use hash access on an object of type Int␤ at line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/lib/CORE.setting line 101 (CORE Any.at_key @ 8) ␤ at line 0 (ExitRunloop @ 0) ␤ at /tmp/8xEerIsAOB line 0 (MAIN C3_ANON @ 0) ␤ …
TimToady b: [{a => 1}, {a => 2}]»<a>.say
p6eval b 1b7dd1: OUTPUT«12␤»
TimToady only old rakudo manages to do it right
or is there some reason why hyper should be iterating the key and value of the pair separately? 06:53
jnthn TimToady: hypers descend into data structures. 06:54
TimToady: Here it desends into the inner hashes.
TimToady: I thought that was what it was meant to do, fwiw.
TimToady it makes little sense to treat the keys the same as the values, since they're likely of different types 06:55
JimmyZ if I want to define method trait_mod:<is>, how can I use it?
TimToady so I think it should not descend into the pairs of a hash
jnthn TimToady: I don't think it does
JimmyZ dosesn't how Actions.pm uses it
jnthn TimToady: I think it's calling .<a> on 1 and 2 06:56
The values of the hash.
oops, meeting!! &
06:58 uasi joined
sorear JimmyZ: if you want to define a method named trait_mod:<is>, you need to use it as an ordinary method. 06:58
rakudo: class A { method trait_mod:<is>() { say "hi" } }; A.new.trait_mod:<is>();
p6eval rakudo cbdd9b: OUTPUT«hi␤»
JimmyZ oh
thanks sorear 07:00
07:03 JimmyZ left 07:04 JimmyZ joined
TimToady jnthn: nevermind, I was thinking about it wrong; what I really want is multidimensional subscripting someday... 07:05
07:06 mj41 joined
dalek ecza/serialize: 37d7895 | sorear++ | lib/ (2 files):
Finish initial cases of running code
 07:12
sorear stefan@stefans:~/rel/niecza$ mono-sgen run/Niecza.exe -L NULL -e 'Q:CgOp { (prog (say (s "Hello world")) {0}) }'
Hello world
sorear sleep
07:13 zby_home_ left 07:17 mberends joined
mberends 'night sorear 07:18
07:24 JimmyZ left 07:26 mishin_ joined
woosley Hi, #perl6. How is undef defined in Perl6? 07:33
Does Mu stand for undefined object in perl6?
moritz woosley: for every class there is a type object. The type object is undefined 07:34
so Int, Str and Mu are perfectly fine undefined objects
Mu is the "most undefined" object in the sense that it defines the least behavior 07:36
woosley moritz: Can I get a Mu object, by "$x.WHAT.say" ?
moritz rakudo: my $x = Mu; $x.WHAT.say 07:38
p6eval rakudo cbdd9b: OUTPUT«Mu()␤»
woosley rakudo: my $x = FO; $x.WHAT.say 07:39
p6eval rakudo cbdd9b: OUTPUT«Could not find sub &FO␤ in <anon> at /tmp/VvCCif3zXi:1␤ in <anon> at /tmp/VvCCif3zXi:1␤␤»
woosley moritz: what is the relationship between Cool, Any, Mu, it there any document about these object type? 07:41
moritz S02 I think 07:42
rakudo: say Cool.^mro 07:43
p6eval rakudo cbdd9b: OUTPUT«Cool() Any() Mu()␤»
08:02 y3llow left, y3llow joined 08:17 cougnut joined 08:26 bakedb joined 08:27 y3llow left, y3llow_ joined, y3llow_ is now known as y3llow 08:30 dakkar joined, am0c left
jnthn back 08:30
08:41 am0c joined 08:42 SHODAN joined
moritz welcome back. 08:46
tadzik backtoo 08:49
mls_ morning! 08:51
jnthn: CATCH/CONTROL still have to have their own $/ $! $_, right? 08:53
moritz I'm pretty sure that they have their own $_ at least 08:54
because that's aliased to $!
mls_ yes. It's more about the other two, cause jnthn++ patched the way they work yesterday 08:55
moritz but we expect do { die 'foo'; CATCH { default { 1 } } }; say $! to access the $!
(because it's the same as try { die 'foo' }; say $! # which we certainly want to continue working)
mls_ What about CONTROL? I don't think we want to leak $! $/ ? 08:57
moritz has no idea about CONTROL, because I don't know any use cases for it
mls_ warn() et al
moritz that's not exactly a "use case", just a keyword :-) 08:58
mls_ I'm talking about the "resume exception" case. I don't think we should change $/ $! in that case.
S04 talks about warnings throwing a control exception 08:59
afk for half an hour... dayjob... 09:00
09:04 mberends left
jnthn mls_: Well, the other thing is that we're discussing whether $! should be looked up lexically or dynamically. Which also has some bearing on this. 09:08
I don't have a good feeling for which way that one will go.
It does fix the long-standing .subst closure issue.
09:27 mishin_ left 09:28 daxim joined 09:39 replore_ left 09:51 bakedb left 10:08 bakedb joined
mls_ jnthn: I'd prefer lexical 10:13
could grep et al do something like local($CALLER::</>) ? 10:14
moritz that was my proposal as well :-)
mls_ (local being the perl5 local, I don't know if it's the same in perl6)
moritz called temp() in p6
erm, probably just temp, it's not a function 10:15
10:16 cyfi joined
mls_ regarding setting $! in CATCH: S04's catch lambda contains an explicit set_outer_caller's_bang(@handled) call 10:21
moritz ok, so that's not an argument for missing $! in CATCH blocks 10:22
mls_ perl6: say do {die; CATCH {"hello"}}; 10:25
p6eval rakudo cbdd9b: OUTPUT«Method 'gist' not found for invocant of class 'Exception'␤ in sub say at src/gen/CORE.setting:4729␤ in <anon> at /tmp/znwlmr7cjz:1␤ in <anon> at /tmp/znwlmr7cjz:1␤␤»
..niecza v9-32-g380d891: OUTPUT«Unhandled exception: Died␤ at /home/p6eval/niecza/lib/CORE.setting line 742 (CORE die @ 2) ␤ at /tmp/RL2zpdsmL6 line 1 (MAIN C1_ANON @ 1) ␤ at /tmp/RL2zpdsmL6 line 1 (MAIN mainline @ 2) ␤ at /home/p6eval/niecza/lib/CORE.setting line 2060 (CORE C968_ANON…
..pugs: OUTPUT«*** Died␤ at /tmp/hSTOj_AHu1 line 1, column 9-12␤»
cougnut hello moritz, could you assit me with ilbot? 10:26
10:26 am0c left
moritz cougnut: I can try, please /msg me 10:26
mls_ afk -> lunch 10:29
10:37 lateau_ joined 10:44 mishin joined 11:00 lateau_ left 11:02 uasi left
ingy phenny: tell pmurias alias TOP to a rule called something else and start on that rule? Realize that I am writing language grammars where I almost always want to parse an entire text. requiring ^$ is not only annoying, it is going to cause problems when people forget it. 11:03
phenny ingy: I'll pass that on when pmurias is around.
bbkr enum.t passing \o/, RT queue will decrease to 850 :) 11:08
11:23 uasi joined 11:26 nsh joined, sbp joined 11:34 envi_|2 left 11:35 satyavvd left 11:36 Psyche^ joined 11:39 benabik joined, Patterner left, Psyche^ is now known as Patterner 11:48 envi_laptop joined 11:58 mtk left 12:03 mtk joined, woosley left
[Coke] yawns from $DAYJOB 12:11
12:12 envi_|2 joined
ingy moritz: gist.github.com/1228947 as promised 12:13
moritz: basically want to think up operators for these use cases. also more use cases if you think of any. 12:14
moritz thinks about it 12:15
12:15 envi_laptop left
[Coke] wonders why the grants queue is always filled with spam whenever he happens to look at it. 12:16
12:19 bluescreen10 joined 12:20 cyfi left
[Coke] rakudo: sub incr1 (*@v is rw) { @v[0]++; @v[1]++; } 12:21
p6eval rakudo cbdd9b: ( no output )
12:23 cougnut left
moritz rakudo: sub incr1 (*@v is rw) { @v[0]++; @v[1]++; }; incr1 my $x; say $x 12:23
p6eval rakudo cbdd9b: OUTPUT«Any()␤»
[Coke] rakudo: $?:: # known? 12:24
p6eval rakudo cbdd9b: OUTPUT«===SORRY!===␤Non-declarative sigil is missing its name at line 1, near "$?:: # kno"␤»
moritz std: $?::
p6eval std bb4f150: OUTPUT«===SORRY!===␤Bogus statement at /tmp/UmSFYMvwJz line 1:␤------> <BOL>⏏$?::␤Parse failed␤FAILED 00:01 120m␤»
moritz ingy: somehow I dislike ?**?, but I'm not sure what a better operator for it would be
usually the ? implies that it's ok to have fewer atoms in your string, here it means it's ok have more 12:25
[Coke] rakudo: class A { method foo {} }; say A.can("foo").WHAT
p6eval rakudo cbdd9b: OUTPUT«Parcel()␤»
felher [Coke]++ #35 Tickets. Wow!
[Coke] 35 tickets... worked on? 12:26
tadzik remaining? :P
felher Just checked his mails and thought: 35 Mails tagged with Perl6BugTracker... what the.. has to be spam. 12:28
12:29 woosley joined
jnthn moritz: Not sure if "is rw" is really valid in slurpy...does the spec call it either way? 12:29
[Coke] who manages rakudo.org, accounts and sthings?
jnthn [Coke]: pmichaud, I believe. 12:30
moritz jnthn: I think the spec is silent on that
aye
jnthn moritz: Feared so ;)
I'll just hope TimToady++ swings by with a ruling ;)
[Coke] ... in your favor!
jnthn [Coke]: could go either way ;) 12:31
12:32 satyavvd joined, rhr joined 12:33 wolverian joined
moritz fwiw I just deployed some small but untested changes to the IRC logs -- please notify me if you see something odd (erm, odder than usual :-) 12:33
12:36 cyfi joined
felher just looks at S04-statements/for.t line 47-55 and wonders: where should the '4' in 'is $str, "(1 5)(2 4)(3 6)",' come from? 12:37
12:38 cougnut joined
moritz that test looks wrong. 12:38
felher *nods* 12:39
moritz felher: do you want to fix it, or should I?
felher guess it should be 1..3 and 4..6, and (1 4), (2 5) and (3 6)
moritz: i'll fix it
Woodi how i can check general type of param ? eg: sub( $k, $v) { if $v.is_numeriv {} else ... ? 12:49
12:49 cougnut left
jnthn if $v ~~ Numeric { ... } 12:49
tadzik .WHAT 12:50
jnthn Don't really need to resort to .WHAT
tadzik I think I do sotmehing like this in File::Find, hmm
jnthn Plus it won't give you the general type
Woodi yes, what gives to many posibilities
tadzik oh well, if it's about Numeric
jnthn All numeric types do Numeric. Anything string-like does Stringy. etc.
Woodi: What does?
moritz .WHAT should not be used for type checking 12:51
Woodi WHAT gives: Int, Rat, ... - so many for 'if'
tadzik github.com/tadzik/perl6-File-Tools...ind.pm#L16
moritz Woodi: use if $thing ~~ Numeric { ... } instead
jnthn Woodi: Yes, that's why I suggested you use Numeric. 12:52
moritz Woodi: that will catch them all
tadzik it's not really WHAT here
jnthn Doing .WHAT === ... is a very bad idea.
Woodi k, thank you
jnthn As you break polymorphism.
moritz tadzik: that's really just type checking with smart matching
tadzik yef
moritz (given/when does smart matching)
tadzik yep
and no ifs :)
12:54 cougnut joined
dalek ast: 3c9d437 | (Felix Herrmann)++ | S04-statements/for.t:
[S04-statements/for.t] fix test: 'for zip(@a; @b) -> $x, $y'
 12:56
12:58 benabik left, uasi left 13:01 benabik joined 13:04 Holy_Cow joined 13:05 Holy_Cow left 13:11 jimmy1980 joined
lucs lslcom 13:14
ww, sorry 13:15
13:16 uasi joined, skangas__ left, skangas joined 13:32 bluescreen10 left 13:46 robins joined 13:47 bluescreen10 joined, robins is now known as robinsmidsrod 13:48 Juerd joined
Juerd Say, is anyone still using feather3 for anything? 13:48
Judging by the charts its filesystem has been full for quite a while; it's managed to exhaust its inodes, even! 13:49
moritz I think try.rakudo.org runs on feather3
erm, tried to run :-)
Juerd: have you cleaned it up again? doesn't look full to me now 13:50
Juerd No; I'm just looking at munin charts
jnthn try.rakudo.org seems to basically work.
Juerd It's using 99% of available inodes and it's been topped out a few times
moritz ah, I was just looking at df -h 13:51
Juerd I'll try to find where all those (temporary?) files are
Try df -i
moritz /dev/vdb 655360 647369 7991 99% /
wee
Juerd You'd think that 655360 ought to be enough... :D
13:52 skangas left, skangas joined
moritz no lsof installed :( 13:52
13:54 dakkar left
Juerd It is now :) 13:54
moritz thanks 13:55
doesn't show me anything interesting though
Woodi rakudo: my $v = "Hallo World!"; $v ~~ s/(\s)/\\$1/; say $v;
13:55 dakkar joined
p6eval rakudo cbdd9b: OUTPUT«Hallo\World!␤» 13:55
Woodi how to get escaped whites ? 13:56
rakudo: my $v = "Hallo World!"; $v ~~ s/(\s)/\\$0/; say $v;
p6eval rakudo cbdd9b: OUTPUT«Hallo\World!␤»
Woodi perl6: my $v = "Hallo World!"; $v ~~ s/(\s)/\\$0/; say $v; 13:57
p6eval niecza v9-32-g380d891: OUTPUT«Hallo\ World!␤»
..rakudo cbdd9b: OUTPUT«Hallo\World!␤»
..pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.16/blib6/pugs/perl5/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** '<HANDLE>' trapped b…
Woodi wooo, niecza works
perl6: my $v = "Hallo World!"; $v ~~ s/(\s)/\\$1/; say $v;
p6eval niecza v9-32-g380d891: OUTPUT«Use of uninitialized value in string context␤ at /home/p6eval/niecza/lib/CORE.setting line 589 (CORE warn @ 2) ␤ at /home/p6eval/niecza/lib/CORE.setting line 21 (CORE Mu.Str @ 12) ␤ at line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/lib/CORE.setting …
..rakudo cbdd9b: OUTPUT«Hallo\World!␤»
..pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.16/blib6/pugs/perl5/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** '<HANDLE>' trapped b…
Juerd feather3:/home/tene/.ccache# find | wc -l 13:59
354474
That is a bit much.
Can anyone write a nice perl or find oneliner to delete old files, say, everything older than a month? 14:00
14:00 soh_cah_toa joined
Juerd I'm rm -rf'ing ~tene/.ccache 14:00
Tene: ping
benabik find <DIR> -mtime +30 -delete # I suggest running it without -delete at least once to see what it'll delete 14:02
Juerd: ^^ 14:03
Juerd Thanks :)
benabik Juerd: Note: untested…
(Which is why I recommend not using -delete at first. ;-) 14:04
Juerd That's okay; we'll find out if it works in a month or so
The rm is still running :)
benabik Living on the edge. :-D
Juerd Inode usage has dropped to 95% 14:05
This isn't going very fast.
benabik Deleting 354k files takes a while...
Juerd Apparently :)
14:07 uasi left, uasi joined
mls_ jnthn: patch that makes Rakudo_cont_store throw an exception if the cont == PMCNULL: gist.github.com/1229149 14:07
(guards against perl6_container_store misuse) 14:08
jnthn mls_: oh, good one. 14:09
mls_: Will apply later today
Juerd rm: cannot remove `.ccache/3': Directory not empty
But... but... but... that's why I did -r
moritz Juerd: maybe permission defined for one file? 14:10
Juerd moritz: I'm root
moritz Juerd: or a new file has been created while the others have been deleted?
Juerd I'm guessing that something is compiling while I'm deleting stuff. The compile will probably fail.
benabik Juerd: Probably not? ccache is pretty conservative. 14:11
moritz htop doesn't show any compilation running
.oO( not anymore )
Juerd moritz: That's exactly what I thought too :)
mls_ jnthn: store_lex_skip_current doesn not exist, right?
Juerd 88% and still dropping 14:12
\o/
mls_ (btw, find_lex_skip_current doesn't seem to get used anymore)
Juerd By the way, it has had this many files for at least half a year.
It's a miracle that it kept working.
PerlJam Juerd: is this just feather3 you're talking about? 14:13
Juerd Yes.
[Coke] Juerd++
jnthn mls_: Maybe not. 14:14
mls_: Do you need it for something?
If so, can always add it... :)
mls_: Note that we assign $! in Rakudo now, not bind.
mls_: So if it's for that, you should use find and the assign.
(totally taking a wild guess at what you might be working on :)) 14:15
mls_ not sure. seems I have to set $! of the outer ctx in the exception handler somehow
jnthn Yeah, you'd do find, and then assign to it.
store_lex_skip_current would bind, not assign.
(if it existed)
mls_ true, we want assign nowadays ;) 14:16
jnthn Well, it was only last night that I went through and tried to get us to be consistent on that. :)
14:17 woosley left
mls_ (that's why I tried to use perl6_container_store) 14:17
(which segfaulted because the find_lex returned PMCNULL)
(cause I also have a ".lex "$!") 14:18
(and it doesn't matter that the .lex comes after the find_lex)
jnthn .lex is declarative, not imperative. So yes, order won't help you. :) 14:19
mls_ Yeah, I saw that when I checked the imcc sources... 14:20
14:21 donri joined
mls_ So I use find_lex_skip_current to get the $! container of the outer? 14:23
14:23 jimmy1980 left
jnthn Should work, yes. 14:24
mls_ ok, thanks!
14:25 wtw left 14:28 thou joined, kaleem left 14:34 wamba joined 14:37 localhost left
moritz now down to 285K used inodes on feather3 14:39
that's still more than on the root partition of my laptop
but not excessive 14:40
Juerd A few code repositories fills it up easily.
That was my first guess. Fortunately I was wrong.
moritz du -sch /home/*/.cpan/build
127M 14:41
Juerd That's not quite as many files as full git repositories with history.
14:41 kaare_ joined
moritz right 14:41
Juerd I wonder which filesystems they use at github 14:42
moritz didn't some people invent file systems for nntp servers that were optimized for many small files? 14:43
mux I think you're thinking of ifs which has long since been removed from base; and it wasn't really optimized for small files but for applications who don't actually need a namespace 14:45
that is, applications using ifs needed to be explicitely modified to open files by their inode number instead 14:46
14:47 soh_cah_toa left
mux oh damn, wrong channel, what I just said was actually FreeBSD-specific 14:47
I fail at IRC.
14:48 abercrombie joined
TimToady well, it's freenode at least :) 14:50
sorear good * #perl6
hi mux!
14:52 espadrine joined
TimToady mls_: I'm trying hard to get rid of $! inside of CATCH, since $! should only mean "the exception that was just handled", and the ones in flight inside CATCH are not handled 14:54
mux sorear: hi there :)
14:54 lateau_ joined
TimToady and we already have $_ bound to the current in-flight exception (in a loop) 14:54
I think use of temp to modify a caller's variable is very smelly. 14:56
moritz agreed
TimToady and we've already mentioned that control functions need to be transparent to $/ and $! as dynvars 15:00
15:04 im2ee joined
im2ee Hello :) 15:04
mikemol Hi 15:06
15:08 birdwindupbird left
moritz #phasers in 2h20min, right? 15:10
sorear more like 26h20min, no? 15:11
moritz oh right, Wed, not Tue :-) 15:12
glad I asked
mls_ Hi TimToady! 15:22
TimToady: what are your thoughts on the return value of blocks that caught an exceptions? 15:24
dalek ecs: 58a9ca5 | larry++ | S05-regex.pod:
get hybrid ** out of my head and into spec
mls_ perl6: say do {die; CATCH {"hello"}};
p6eval rakudo cbdd9b: OUTPUT«Method 'gist' not found for invocant of class 'Exception'␤ in sub say at src/gen/CORE.setting:4729␤ in <anon> at /tmp/dLQUlyFNma:1␤ in <anon> at /tmp/dLQUlyFNma:1␤␤»
..pugs: OUTPUT«*** Died␤ at /tmp/Qn9K0lODHD line 1, column 9-12␤»
..niecza v9-32-g380d891: OUTPUT«Unhandled exception: Died␤ at /home/p6eval/niecza/lib/CORE.setting line 742 (CORE die @ 2) ␤ at /tmp/1ms6CMP83Y line 1 (MAIN C1_ANON @ 1) ␤ at /tmp/1ms6CMP83Y line 1 (MAIN mainline @ 2) ␤ at /home/p6eval/niecza/lib/CORE.setting line 2060 (CORE C968_ANON…
mls_ should that say hello?
moritz since the exception is not handled, it should rethrow as niecza does 15:25
TimToady I don't think so, offhand
moritz more interesting:
perl6: say do {die; CATCH { default { "hello" }}};
p6eval niecza v9-32-g380d891: OUTPUT«Nil␤»
..pugs: OUTPUT«*** Died␤ at /tmp/itEC3UwkJm line 1, column 9-12␤»
..rakudo cbdd9b: OUTPUT«Method 'gist' not found for invocant of class 'Exception'␤ in sub say at src/gen/CORE.setting:4729␤ in <anon> at /tmp/E18saprxiD:1␤ in <anon> at /tmp/E18saprxiD:1␤␤»
mls_ moritz: assume that is handled. I'm wondering if there's a nice way for the handler to set the block return value 15:26
thus an easy way to return an "error" result
TimToady CATCH is never implicitly going to set the return value, though we might allow explicit return 15:27
15:27 jaldhar left
mls_ well, return leaves the Routing. We might need OUTER::leave or something 15:28
TimToady that is, not implicitly set to the return value of the handler. we should implicitly set it to something undefined, as in p5, I think
dalek ecs: e639b7d | moritz++ | S05-regex.pod:
[S05] fix typo
moritz TimToady: did you see the suggestion to use ~** for the <a> ** <delimiter> form, and +** for the <a> ** 1..* form? 15:29
moritz liked that, and it's orthogonal to the last S05 patch
TimToady it seems a bit crufty 15:30
15:31 envi_|2 left
moritz less crufty than using the same operator for two different functions 15:31
15:31 envi_laptop joined
TimToady there's an underlying sense in which it is the same operator with a different test 15:31
moritz still the problem remains that a mis-typed range on the is usually valid syntax for a separator, resulting in LTA misbehavior 15:32
TimToady trew 15:34
making it a separate operator still doesn't help with the fact that we have two independent criteria for the same repetition 15:37
I've also been playing with ways of combining the two right args
<a> ** [0..* ','] or something 15:38
but don't really like that one enough
moritz well, <a> +** 0..* ~** ',' would help, but it's not pretty at all 15:39
mux arg, my eyes
TimToady as for the typo problem, restricting the syntax of the sep form would help
for instance, if we required brackets: 15:40
<a> ** [',']
or something that a person would never use near 0..*
'course, you still have the problem that a malformed 0..* can be taken for normal match material
moritz requiring brackets would work, yes 15:41
TimToady std: / <a> ** 0.. foo /
p6eval std bb4f150: OUTPUT«===SORRY!===␤Malformed range at /tmp/rZ7j34PsYE line 1:␤------> / <a> ** 0..⏏ foo /␤ expecting quantifier␤Parse failed␤FAILED 00:01 121m␤»
TimToady std actually catches some of 'em
I wish the other regex folx hadn't stolen ++ for possesive + 15:43
15:43 jaldhar joined
TimToady ah well, can't be undone 15:44
still need to come up with a way to combine them though 15:49
<a> ** {0..*}[','] or something like that
flussence the first idea that comes to mind for me is [',' xx 0..*]. Is xx already taken in regex? 15:51
TimToady well, they'd be taken literally by default 15:52
but we can claim keywords if we need to 15:53
however, your form looks like it's replicating the comma
flussence I was thinking along the lines of having it look like "$regex ** $list-of-separators"... 15:54
TimToady well, but 0 commas still means 1 of the lhs
flussence oh, good point. 15:55
TimToady and I think that varying separators is rare enough that we can dehuffmanize it
and just make people write it out long 15:56
15:57 dukeleto left 15:58 dukeleto joined
TimToady there are also approaches where we make the separator a modifier on any quantifier 15:58
then you could write <a>* via ',' or some such
and for the current <a> ** ',' we'd either force people to write <a>+ via ',' or assume the + 16:00
PerlJam It took me a while to get used to (and even come to like) A ** B, and now you're gonna change it?!? 16:02
:-)
TimToady I'd be inclined to require the + in order to reduce accidental recognition of whatever keyword we use
PerlJam <a>+:sep(',') ? 16:04
TimToady it's not a string, it's a regex
PerlJam sure ... <a>+:sep(<foo>)
TimToady that form is a normal P6 expression inside ()
I wouldn't steal a normal adverbial for that 16:05
PerlJam okay, let's just say the "via" in "<a>+ via ','" doesn't quite work for me :)
just that it's wordy, not the word itself
though, looking at the scroll back, some form that combines range and separator would be nice. 16:06
16:07 MayDaniel joined
TimToady prejudiced against romance languages, eh? :P 16:07
wordy is good when the things around it are symbolic
16:08 mkramer1 joined
PerlJam Except that I have this nice rule in my head that says wordy things are literal in a regex and I kind of like it that way 16:09
TimToady and we have, in fact, already reserved all the punctuation
we need something that doesn't look like either side 16:10
16:10 satyavvd left
TimToady <a>+ % ',' # modulo! 16:11
PerlJam notes that #perl6 once again has me staring at my keyboard 16:12
slavik1 TimToady: I think I realized the trouble you have with designing a language ... 16:13
TimToady std: / <a> % <b>
p6eval std bb4f150: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter % (must be quoted to match literally) at /tmp/DT0IXDt_L3 line 1:␤------> / <a> %⏏ <b>␤Unable to parse regex; couldn't find final '/' at /tmp/DT0IXDt_L3 line 1:␤------> / <a> %[33…
TimToady std: / <a> % <b> /
p6eval std bb4f150: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter % (must be quoted to match literally) at /tmp/idkFhHLC3z line 1:␤------> / <a> %⏏ <b> /␤Unable to parse regex; couldn't find final '/' at /tmp/idkFhHLC3z line 1:␤------> / <a> %[…
TimToady I think I like it
slavik1 we don't have enough symbols ...
PerlJam I like it, but I worry a bit about that +
slavik1 on the keyboard
TimToady which way do you worry about it? 16:14
requiring it, or not requiring it?
PerlJam Basically just differentiating between aaa,aaaa,aaa and a,a,a,a,a
TimToady there's always [<a>*] % ',' 16:15
moritz % is cute 16:16
TimToady and has two parts, like key/value
16:17 uasi left
TimToady and means "constrained by" in colloquial usage of "modulo" 16:17
in fact, I expect people would pronounce it that way in their heads
arnsholt I agree with moritz. Definitely cute. The modulo mnemonic complements it nicely 16:18
PerlJam and so we'd have / [<a> % ','] ** 3..5 / for combinations
(for instance) 16:19
moritz TimToady: so we keep ** $range?
TimToady that would be the idea
moritz +1
TimToady PerlJam: no, inside out from that
so % could modify any quantifier 16:20
including **
PerlJam Aye, I realized the difficulty as soon as I'd hit enter :)
TimToady <a> ** 3..5 % ','
PerlJam so yes ... % is really nice TimToady++
TimToady the only remaining question is whether to require a quantifier 16:21
PerlJam I vote no
TimToady well, and where to allow whitespace, and what whitespace means
<a> +% ',' ain't so bad 16:22
though looks like a numeric prefix
<a>+ % ',' is probably better from that perspective 16:23
16:23 JimmyZ joined
TimToady with ws dependency before and after %, we need to worry just a bit about accidental hashes, as in <a>%foo 16:24
16:25 bluescreen10 left
PerlJam TimToady: Where would the ? mod go in either rendition? / <a> +%? ',' / or / <a>+? % ',' / 16:25
16:25 tokuhiro_ joined, JimmyZ_ joined
PerlJam (or disallowed?) 16:26
16:26 mj41 left 16:28 JimmyZ left, JimmyZ_ is now known as JimmyZ
TimToady the latter 16:28
er, the second 16:29
+? %
because of the possibility of a hash, I think we have to require the real quantifier
it's not like people will write the form all that often that we need to huffmanize it down to nothing 16:30
and <a>+ indicates we expect some number of them even before we read the %
mls_ jnthn: your steal_outer hack in perl6lexpad.pmc doesn't seem to work if the outer block doesn't know the lexical 16:31
TimToady and we can give a decent error message: "% doesn't modify a quantifier"
jnthn mls_: It's not really a hack...
mls_: Also, it's only meant to be for $_ these days.
(Since $! and $/ are per routine now.) 16:32
mls_ Ok, so you need to remove thr $! $/ part
jnthn Makes sense...got a test case?
16:32 cosimo joined
jnthn It'd have been nice if the spectests had reminded me of this ;) 16:32
PerlJam TimToady: I don't know ... I think most of my uses of ** are really of the % variety rather than ranges. 16:33
oh, never mind. My brain is a tad slow right now. 16:34
TimToady: in any case, it doesn't chunk right in my head. / <a>+ % ',' / still "looks funny" 16:35
TimToady: but that's why you're the languages designer ... :-) 16:36
16:36 Chillance joined, bluescreen10 joined 16:42 tokuhiro_ left
im2ee Can i iterate through 001, 002, 003 and so on using only maths? You know.. something like 001+1 = 002 16:44
:)
say 001+1 returns 2. 16:45
I want 002. :)
Tene Juerd: I haven't logged into feather3 in years, iirc, and I've never configured ccache
wolverian rakudo: "002"+1
p6eval rakudo cbdd9b: ( no output )
mls_ jnthn: I can't find a simple testcase. I need an uninitialized .lex $/ ... 16:46
JimmyZ rakudo: say "001" ... "005"
p6eval rakudo cbdd9b: OUTPUT«001 002 003 004 005␤»
Juerd Tene: Interesting :)
wolverian I always forget saying it. thanks, JimmyZ.
bbkr rakudo: say "01" .. "there is a cake in last test chamber" 16:47
p6eval rakudo cbdd9b: OUTPUT«"01".."there is a cake in last test chamber"␤»
JimmyZ rakudo: say "001.jpg" ... "005.jpg"
p6eval rakudo cbdd9b: OUTPUT«001.jpg 002.jpg 003.jpg 004.jpg 005.jpg␤»
Tene Juerd: at least, I don't remember doing so. It's a bit long ago for me to be certain. You could check my .bash-history. 16:49
im2ee Thanks, i will try it. :)
PerlJam im2ee: you can also just use ++
16:49 JimmyZ left
[Coke] if you want numbers that are formatted as strings instead of strings that are incremented like numbers, you could do that too. 16:50
PerlJam rakudo: my $str = "001"; for 0..5 { say $str++ }
p6eval rakudo cbdd9b: OUTPUT«001␤002␤003␤004␤005␤006␤»
im2ee PerlJam, ++ on this channel (for karma) or in perl (for my program)? :) 16:51
16:53 Chillance left 16:54 espadrine left 16:55 MayDaniel left, espadrine joined
jnthn mls_: Ah, that's not likely to happen... 16:58
mls_: But anyway, removing it...it's just cruft now.
16:59 saaki left
dalek kudo/nom: 7552eb9 | jnthn++ | src/pmc/perl6lexpad.pmc:
Remove now unrequired calls to steal_outer (mls++).
 17:00
17:03 Chillance joined 17:10 saaki joined 17:11 mberends joined, mj41 joined 17:16 zby_home_ joined 17:20 sayu joined, thou left
dalek ecs: a455b11 | larry++ | S05-regex.pod:
old <a>**',' is now written <a>+%','

The separator syntax is now required to be a modifier on a quantifier, so the old <a> ** ',' is now written <a>+ % ','. This allows for other quantifiers like * or ** on the left as the basis of the constraint, while disambiguating use of ** to only work as a quantifier without overloading separator matching. The C<%> because it can be pronounced
  'modulo' and because it looks like the relationship of two things. And
also because it's unlikely to be confused with other regex forms.		 17:23
17:23 espadrine left
mls_ jnthn: current state of exception handler change: gist.github.com/1229726 17:24
flussence % looks vaguely pythonish :)
im2ee karma larry
aloha larry has karma of 134.
mls_ jnthn: gets rid of the extra block, assigns to OUTER::$! 17:25
I'll do some testing tomorrow
(some more testing, first tests look good)
jnthn TimToady: Should CATCH blocks declare a $? ? 17:26
er
a $/
?
mls_ (I hope it's ok to modify $handler<past_block>...)
jnthn $handler<past_block>.unshift($handler_start);
Probably should be
$handler<past_block>[1].unshift($handler_start);
So you change the imperative bit, not the declaritive bit 17:27
TimToady jnthn: probably, since when clauses may be doing pattern matching
jnthn OK, then mls++ :)
+ PAST::Var.new( :scope('lexical_6model'), :name('$_'), :isdecl(1) ),
PAST::Op.new( :pasttype('bind_6model'),
PAST::Var.new( :scope('lexical_6model'), :name('$_') ),
mls_: ^^ here, why the separate :isdecl? Why not just put :isdecl on the one inside the bind? 17:28
dalek kudo/nom: 7fdc6b4 | jnthn++ | src/binder/container.c:
Apply patch from mls++ to harden container storage.
TimToady should probably declare a $! too, just to keep it empty, and force people to say OUTER::<$!> if they mean it
jnthn OK
TimToady but the SIMPLECATCH doesn't set $!
mls_ isdecl: I just copied that from the old code 17:29
TimToady it's not like CATCH has to be all that efficient, if expections are really exceptional
jnthn mls_: In the case of what TimToady just said, the binding to $! should probably be removed.
s/case/light/
mls_ ok, removing...
jnthn mls_: Ah, OK...feel free to clear up too :)
mls_: Feels like the right direction. 17:30
TimToady CONTROL, on the other hand, should probably be transparent
17:30 Sarten-X left
TimToady and as lightweight as possible 17:30
since control exceptions are not unusual 17:31
17:31 espadrine joined
jnthn *nod8 17:31
mls_ so CONTROL should not mess with the outers $!? 17:32
$! ?
TimToady outer shouldn't have a $! either
mls_ Ok, the Routine's $! 17:33
17:33 thou joined
TimToady statement_control:<while> should not be declared such that it defines a $! or $/ 17:33
it should be transparent to the dynvar
if we're going to set the user's $! or $/ dynamically 17:34
which it looks like we need to do
17:34 benabik left
TimToady (note that dynvars are still lexically scoped; it's just the lookup that is dynamic) 17:34
dynvars are temp done right 17:35
felher TimToady++ for '**' in regex. /me searched for that 1 1/2 months ago. :)
TimToady we could just say that any routine that defines a CONTROL block doesn't get $/ or $! by default 17:36
17:36 Sarten-X joined 17:37 benabik joined 17:42 dakkar left
mls_ new version of patch: gist.github.com/1229760 17:43
17:43 espadrine left 17:45 Chillance left
mls_ afk -> home... 17:45
jnthn mls_: Thanks 17:46
17:50 mishin left 17:52 bakedb left
tadzik perl6lexpad.c: In function ‘Parrot_Perl6LexPad_get_number_keyed_str’: 17:52
perl6lexpad.c:211:5: warning: passing argument 2 of ‘register_number_for_get’ discards ‘const’ qualifier from pointer target type [enabled by default]
perl6lexpad.c:69:15: note: expected ‘struct Hash *’ but argument is of type ‘const struct Hash *’
I got plenty of those with the newest patches
flussence hm 17:55
those showed up between 2011-09-15 and 2011-09-16 according to my cron output 17:56
17:56 Chillance joined 18:00 envi_laptop left 18:06 daniel-s left 18:07 araujo left
Woodi perl6 async { say "my tid is ", +self; }; 18:07
perl6: async { say "my tid is ", +self; }; 18:08
p6eval rakudo 7fdc6b: OUTPUT«===SORRY!===␤Symbol 'self' not predeclared in <anonymous> (/tmp/pca6xM4Cyz:1)␤»
..niecza v9-32-g380d891: OUTPUT«===SORRY!===␤␤'self' used where no object is available at /tmp/Xv1BMaOA_H line 1:␤------> async { say "my tid is ", +self⏏; };␤␤Action method statement_prefix:async not yet implemented at /tmp/Xv1BMaOA_H line 1:␤------> …
..pugs: ( no output )
18:08 sayu left
benabik tadzik: Do you compile parrot with g++ or gcc? 18:08
Woodi perl6: async { say "my tid is ", ~ '1213'; };
p6eval rakudo 7fdc6b: OUTPUT«Could not find sub &async␤ in <anon> at /tmp/bOS42Lcful:1␤ in <anon> at /tmp/bOS42Lcful:1␤␤»
..niecza v9-32-g380d891: OUTPUT«===SORRY!===␤␤Action method statement_prefix:async not yet implemented at /tmp/NLgVB6SjkH line 1:␤------> async { say "my tid is ", ~ '1213'; }⏏;␤␤Unhandled exception: Unable to resolve method statement_level in class St…
..pugs: OUTPUT«my tid is 1213␤»
tadzik benabik: gcc 18:09
dalek kudo/nom: 79344e0 | jnthn++ | src/pmc/perl6lexpad.pmc:
Try to eliminate a bunch of C compiler warnings tadzik++ is seeing.
 18:10
jnthn tadzik: Lemme know if ^^ helps.
Woodi pugs: my $thr = async { say "my tid is ", ~ '1213'; }; $thr.started;
p6eval pugs: OUTPUT«my tid is 1213␤*** No such method in class Thread: "&started"␤ at /tmp/e_rxPalyOf line 1, column 50-62␤»
tadzik jnthn: will see
18:11 benabik left
jnthn some of the consting was bogus, we were gonna modify the hash... :) 18:11
tadzik oh you :) 18:13
PerlJam TimToady: I'm still trying to puzzle out the edges of the new % .... A non-sigspace repitition separated by comma would look like this? <a>**2..*%',' Doesn't "*%" run afoul of LTM? Seems like a space between quantifier and % should be mandatory and not contribute to sigspace 18:16
tadzik jnthn: yeah, that removed the warning from perl6lexpad 18:20
18:22 localhost joined
jnthn tadzik: nice :) 18:24
18:26 frhodes left 18:27 espadrine joined 18:33 espadrine left, espadrine joined
TimToady PerlJam: there is no *% token available at that point, so I don't know what would try to match it 18:39
the next thing must be * or a number
and that's a Whatever star, not a quantifier star 18:40
so it happilly parses **2..* as a quantifier, then %',' as a quant modifier
well, it's not a quant modifier in the sense that ? or ! is, since you can have <a>+? % ',' 18:42
I guess a quant mod in the sense of being a quant modulus :) 18:43
PerlJam :-)
TimToady++ good call on using % :) 18:47
18:54 daxim left 18:57 odoacre left 19:01 lateau_ left 19:03 MayDaniel joined 19:05 worr joined 19:09 odoacre joined, [Coke] left 19:11 [Coke] joined 19:12 espadrine left
dalek kudo/nom: 091c262 | jnthn++ | src/ (2 files):
Fix issue with dynamic compilation of multis belonging to a derived dispatcher. Fixes issues with declaring custom traits and using them in the same compilation unit.
 19:15
19:15 frhodes joined 19:21 birdwindupbird joined, birdwindupbird left
dalek d: c22b2ee | larry++ | STD.pm6:
first step to +% transition, deprecate ** sep

and also take a whack at parsing the +% form		 19:25
19:29 bbkr1 joined 19:34 espadrine joined 19:41 MayDaniel left 19:44 jaldhar left 19:45 benabik joined 19:50 jaldhar joined 19:56 thou left 20:08 SHODAN left 20:11 araujo joined
cognominal when I do a break from C to Perl, that does not last. But when I do a last in C, everything breaks. :( 20:25
20:30 soh_cah_toa joined
mberends cognominal: that is almost true in a philosophical sense as well # double entendre curieux :) 20:32
jnthn :) 20:34
Woodi there is trink in perl5 to generate accessors with closure: for my $field (qw(name first second)) { my $slot = __PACKAGE__."::$field"; ... how to do something like this in v6 ? 20:42
*trick
PerlJam Woodi: you're asking how to dynamically create methods? 20:43
Woodi yes i think
with nearly same code, different name 20:44
PerlJam Woodi: you use the mop to add methods to a class/object
Woodi it is documentet somewhere ? 20:46
bbkr1 Woodi: you may check github.com/bbkr/jsonrpc/blob/maste.../Client.pm - I install there RPC methods in runtime by composing roles into class 20:47
Woodi k 20:48
PerlJam Woodi: I don't know if it's documented ... try github.com/jnthn/6model maybe
jnthn The introspecty bits of the MOP are in S12, but the declarative bits arne't. 20:49
*aren't
PerlJam perhaps there needs to be a S12.5 - the MOP :) 20:50
jnthn rakudo: class Foo { }; for <bar baz waz> -> $name { Foo.^add_method($name, method () { "method for $name" }); }; Foo.baz 20:51
p6eval rakudo 091c26: ( no output )
jnthn rakudo: class Foo { }; for <bar baz waz> -> $name { Foo.^add_method($name, method () { "method for $name" }); }; say Foo.baz
p6eval rakudo 091c26: OUTPUT«method for baz␤»
jnthn Woodi: ^^
Though, that adds them at runtime 20:52
Woodi is reading && parsing
jnthn rakudo: class Foo { BEGIN { for <bar baz waz> -> $name { Foo.^add_method($name, method () { "method for $name" }); } }; say Foo.baz
p6eval rakudo 091c26: OUTPUT«===SORRY!===␤Method 'dispatch:<.^>' not found for invocant of class 'Foo'␤»
jnthn oh, yes :)
rakudo: class Foo { BEGIN { for <bar baz waz> -> $name { Foo.HOW.add_method(Foo, $name, method () { "method for $name" }); } }; say Foo.baz
p6eval rakudo 091c26: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 1␤»
jnthn tries to spot his syntax error... 20:53
rakudo: class Foo { BEGIN { for <bar baz waz> -> $name { Foo.HOW.add_method(Foo, $name, method () { "method for $name" }); } } }; say Foo.baz
p6eval rakudo 091c26: OUTPUT«method for baz␤»
jnthn That adds them at BEGIN time. 20:54
Which basically gives the same effect as if the compiler added them.
(So, they're available to participate in role compositon conflict resolution, etc.)
PerlJam jnthn: when does "dispatch:<.^>" get installed?
jnthn PerlJam: It's inherited from Mu.
PerlJam: But we don't set a default parent until composition time. 20:55
PerlJam gotcha
Woodi thanx jnthn++ bbkr1++ PerlJam++ :)
jnthn PerlJam: We could maybe cheat on it a bit. But there's a chance we should be doing .^ by a syntactic re-write anyway. :) 20:56
PerlJam Woodi: why are you creating methods dynamically btw?
bbkr1 bets on "database ORM" answer :) 20:57
Woodi PerlJam: to save place in file in some typing. just do not want to have same methods with different names
PerlJam Woodi: so ... you're just making aliases? 20:58
Woodi bbkr1: s/ORM// :)
PerlJam: i didn't write it yet, so don't know exactly. "alias" sugests identical behaviour 21:00
21:00 kaare_ left
PerlJam Woodi: not necessarily. 21:00
Woodi do alias means something specyfic in perl6 ?
bbkr1 Woodi: If you're making aliases just install the same method (it's an object) under many names. my $m = method () {}; Foo.^add_method('name1', $m); Foo.^add_method('name2', $m); 21:01
PerlJam Woodi: you could parameterize a method appropriately then create aliases to curried versions of that method
Woodi my maind isn't curied yet... what is difference to jnthn++ add_method ? 21:02
arnsholt Or you could write multis, or use keyword arguments/default values in general
Reassuringly, there's more than one way to do it =)
21:03 zby_home_ left
Woodi i want something like in v5 trick :) 21:03
PerlJam It's all TimToady's fault
mberends (there's only one way to blame it) 21:04
PerlJam Woodi: Perl 6 has things like the Perl 5 trick, but better/safer. But the exact nature of those things really depends on what you're trying to achieve. 21:05
bbkr1 off topic - do we have any P6 GTK libs under development?
PerlJam bbkr1: are you volunteering? :) 21:06
mberends bbkr1: yes, over here, with Niecza!
im2ee Good night!
21:06 im2ee left
PerlJam mberends: Shh! I wanted him to replicate it with Rakudo 21:06
:)
mberends oops 21:07
jnthn Probably need to fix nativecall first...
bbkr1 PerlJam: no, i plan to spent time on improving Mongo driver after NOM release.
mberends: do you need any GTK dev packages installed? ssh X forwarding is already enabled so you can test them on MacOS 21:08
21:08 japhb left
PerlJam what's a nom-based R* release waiting on exactly? pmichaud regex fixes? something else? 21:09
21:09 japhb joined
jnthn PerlJam: Largely regexes and grammars. 21:11
There's other bits in nommap, but we're gradually picking away at them.
PerlJam Is there any LHF wrt grammars and regex on nom? 21:13
bbkr1 I miss pack/unpack and Str.encode/decode most. That's huge regression for all streaming and serialization usage.
21:15 wamba left 21:19 masak joined
masak hi sweet #perl6 21:19
just wanted to stop by and wave before I go to sleep :) 21:20
mberends masak: \o/
masak ingy++ and TimToady++ for the % changes.
I approve of all of it... although I'm not sure I think the negative-numbers stuff was worth the extra complexity. 21:21
jnthn o/ masak
21:21 bluescreen10 left
bbkr1 i flagged 9 RT tickets assigned to me as testneeded, feel free to steal if you want to close them faster, as I won't be able to take care of them as planned during weekend :( 21:27
21:28 y3llow_ joined, bbkr1 left 21:29 pothos_ joined 21:30 y3llow left 21:31 pothos left, y3llow_ is now known as y3llow, pothos_ left, bbkr1 joined 21:32 pothos_ joined, pothos_ is now known as pothos
mberends bbkr1: after some testing, I'll be able answer what Gtk libs are needed, not right now. I look forward to testing it soon. Meantime you could try this mostly working Gtk xclock for Niecza: gist.github.com/1224163 21:33
bbkr1 testing... 21:41
21:45 dukeleto left 21:46 dukeleto joined, dukeleto left, dukeleto joined
bbkr1 tons of dependencies for gtk-sharp, installing 21:46
21:46 dukeleto left 21:47 dukeleto joined, dukeleto left, dukeleto joined, dukeleto left
tadzik good evening 21:47
masak tadzik! \o/
21:48 dukeleto joined, dukeleto left
jnthn o/ tadzik 21:48
tadzik lol, masak!
masak lol!
tadzik and jnthn! lol!
masak lolnthn!
tadzik how are you guys doing?
21:48 dukeleto joined, dukeleto left
masak <-- giddy, happy, tired 21:48
21:49 dukeleto joined, dukeleto left
jnthn kinda tired too 21:49
21:49 dukeleto joined, dukeleto left 21:50 dukeleto joined, dukeleto left
tadzik well-fed, notmuch tired, quite happy 21:50
21:50 dukeleto joined, dukeleto left 21:51 dukeleto joined, dukeleto left, dukeleto joined, dukeleto left
masak dukeleto: dude, you need an ark of something. 21:51
or* 21:52
21:52 dukeleto joined, dukeleto left 21:53 dukeleto joined, dukeleto left, dukeleto joined, dukeleto left 21:54 dukeleto joined, dukeleto left 21:55 dukeleto joined, dukeleto left, dukeleto joined 21:56 dukeleto left, dukeleto joined, dukeleto left 21:57 dukeleto joined, dukeleto left, dukeleto joined, dukeleto left 21:58 dukeleto joined, dukeleto left 22:00 dukeleto joined, dukeleto left, mberends left, dukeleto joined, dukeleto left 22:01 dukeleto joined, dukeleto left, tokuhiro_ joined 22:02 dukeleto joined
masak 'night, channel 22:03
22:03 masak left, mj41 left 22:04 Juerd left 22:10 mkramer1 left, tokuhiro_ left, tokuhiro_ joined 22:11 Ledrug[RCo] joined 22:13 Ledrug[RCo] is now known as Ledrug 22:16 tokuhir__ joined, tokuhiro_ left 22:18 Sarten-X left 22:24 donri left 22:26 benabik left 22:28 Sarten-X joined 22:36 satyavvd joined
Woodi rakudo: class Foo::Bar { BEGIN { for <bar baz waz> -> $name { Foo::Bar.HOW.add_method( Foo::Bar, $name, method () { "method for $name" }); } } }; say Foo::Bar.baz 22:40
p6eval rakudo 091c26: OUTPUT«method for baz␤»
Woodi this don't work on debian rakudo, build with parrot 3.6.0 0 22:41
means Star 2011.07 ? 22:42
tadzik rakudo: is the development version, not yet released 22:43
p6eval rakudo 091c26: OUTPUT«Could not find sub &version␤ in <anon> at /tmp/oBHMhTwiip:1␤ in <anon> at /tmp/oBHMhTwiip:1␤␤»
tadzik yeah, right
what you have in debian is most probably b:
b: class Foo::Bar { BEGIN { for <bar baz waz> -> $name { Foo::Bar.HOW.add_method( Foo::Bar, $name, method () { "method for $name" }); } } }; say Foo::Bar.baz
p6eval b 1b7dd1: OUTPUT«Method 'baz' not found for invocant of class ''␤ in main program body at line 22:/tmp/k4hI48T8GD␤»
Woodi other error
Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 10 22:44
rakudo: class Foo::Bar { BEGIN { for <bar baz waz> -> $name { Foo::Bar.HOW.add_method( Foo::Bar, $name, method { "method for $name" }); } } }; say Foo::Bar.baz
p6eval rakudo 091c26: OUTPUT«method for baz␤»
jnthn Only the current (unreleased) development branch has good BEGIN support.
tadzik that's just a parse error, I'd bet it's not the same code
jnthn That's one of the things we've drastically improved in the nom refactors.
Woodi i spend hour looking on that line
22:45 REPLeffect joined
Woodi it isn't the same :) 22:45
rakudo: class Foo::Bar { BEGIN { for <bar baz waz> -> $name { Foo::Bar.HOW.add_method( Foo::Bar, $name, method() { "method for $name" }); } } }; say Foo::Bar.baz 22:57
p6eval rakudo 091c26: OUTPUT«===SORRY!===␤Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 1␤»
Woodi give me perl5 perl back... 22:58
stupid space
22:59 d4l3k_ joined, Maddingu1 joined 23:00 JodaZ_ joined 23:01 kthakore_ joined, jdv79_ joined, szabgab_ joined, tewk_ joined 23:02 daemon_ joined, apejens_ joined, TimToady_ joined, ruz joined 23:03 Tene_ joined, Tene_ left, Tene_ joined, wooden_ joined, wooden_ left, wooden_ joined, yath_ joined 23:07 wooden left, TimToady left, Tene left, dalek left, tewk left, daemon left, jdv79 left, apejens_ is now known as apejens, daemon_ is now known as daemon, d4l3k_ is now known as dalek, TimToady_ is now known as TimToady, ChanServ sets mode: +v dalek 23:08 bbkr joined, phenny joined 23:11 pmichaud joined, hugme joined, jrockway_ joined, zelazny.freenode.net sets mode: +v hugme 23:13 pernatiy joined, benabik joined 23:14 molaf_ left
jnthn sleep & 23:16
Woodi rakudo: my $a = 2; given $a { when 2, 3 { say "2 or 4" } } 23:19
p6eval rakudo 091c26: ( no output )
Woodi rakudo: my $a = 2; given $a { when 2 { say "2 or 4" } }
p6eval rakudo 091c26: OUTPUT«2 or 4␤»
TimToady Woodi: earlier you were asking about checking the type of a parameter; the easiest way to do that is to just put the type in the signature along with the parameter 23:21
Woodi so just use ? :) 23:23
sjohnson given technology!
TimToady sub foo ( $k, Numeric $v ) {...} 23:24
and if you want to conditionalize on it, that's when you want multi subs 23:25
Woodi now understud
23:26 molaf_ joined
Woodi rakudo: my $a = 2; given $a { when 2|3 { say "2 or 3" } } 23:28
p6eval rakudo 091c26: OUTPUT«2 or 3␤»
23:30 Alias joined 23:32 espadrine left 23:35 dorlamm joined 23:36 pmichaud left, hugme left, jrockway_ left 23:39 pmichaud joined, hugme joined, jrockway_ joined, zelazny.freenode.net sets mode: +v hugme
TimToady phenny: tell masak the negative quantification has been there a long time; it just got relocated to stay near the ** text 23:43
phenny TimToady: I'll pass that on when masak is around.
Woodi rakudo: sub a() { 1 }; my $r = a() or die; 23:48
p6eval rakudo 091c26: ( no output )
Woodi rakudo: sub a() { 0 }; my $r = a() or die;
p6eval rakudo 091c26: OUTPUT«␤ in <anon> at /tmp/u1Zv8l6sPh:1␤ in <anon> at /tmp/u1Zv8l6sPh:1␤␤»
23:49 dorlamm left
TimToady std: method() {...} 23:49
p6eval std c22b2ee: OUTPUT«===SORRY!===␤Word 'method' interpreted as 'method()' function call; please use whitespace around the parens at /tmp/uHKDkEkCO9 line 1:␤------> method⏏() {...}␤Unexpected block in infix position (two terms in a row) at /tmp/uHKDkEk…
TimToady Woodi: note that rakudo gives a substandard error there
Woodi rakudo: sub a() { 0 }; my $r = a() or say( 'die' );
p6eval rakudo 091c26: OUTPUT«die␤»
Woodi yes, i wondered what exactly was that. 23:50
rakudo: sub a() { 1 }; my $r = a() or say( 'die' );
p6eval rakudo 091c26: ( no output )
Woodi just translating from v5; 23:51
TimToady niecza: method() {...}
p6eval niecza v9-32-g380d891: OUTPUT«===SORRY!===␤␤Word 'method' interpreted as 'method()' function call; please use whitespace around the parens at /tmp/MSktHY09ty line 1:␤------> method⏏() {...}␤Unexpected block in infix position (two terms in a row) at /…
TimToady note that niecza has a more standard parser, so maybe you want to syntax check with niecza, then get it to run under rakudo 23:52
Woodi i will
btw. can Perl6 be more user friendly ? :) 23:53
TimToady well, if you can figure out a way to do that, I'm all ears 23:54
std: say $?
Woodi 'my' not needed, self!method is harder then .
p6eval std c22b2ee: OUTPUT«===SORRY!===␤Unsupported use of $? variable as child error; in Perl 6 please use $! at /tmp/x1zw1pWca0 line 1:␤------> say $?⏏<EOL>␤Parse failed␤FAILED 00:01 119m␤»
23:54 Ledrug left
TimToady $!method is not much harder than $.method 23:55
Woodi thats for readability
23:55 ledrug joined
Woodi yes, but it is much different then c/c++/perl5 23:55
TimToady unfortunately we can't just have !method :) 23:56
sometimes different is better, eventually :)
Woodi i talk only about call !method, not declaration
yes
TimToady you can call it as $!method(args)
supposed to be able to, anyway 23:57
23:57 uasi joined
Woodi but why ! ? :) 23:57
TimToady ran out of characters :) 23:58
Woodi . for calling ?
TimToady but also it's a mnemonic for negativity, as in "Don't touch!"
Woodi aha
hard to switch mind to perl6
so potentially less users... 23:59
TimToady potentially different users
we expect to leave some Perl 5 programmers behind, but gain others

Powered by Cro and the Raku® Programming Language.

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

Thank you!