Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2012-02-11

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:24 flussence nom: subset Even of Int where * %% 2; multi MAIN(Even $) { ... } # doesn't work locally...
00:24 p6eval nom a39d86: OUTPUT«use of uninitialized value of type Even in string context␤Usage:␤  --setting=SAFE <Int> ␤»
00:24 flussence # well, it does, but it complains
00:33 skids .oO(through some strange quantum entanglement, jnthn++ keeps fixing things about two weeks after I try to use them.)
00:33 [Coke] skids: KEEP TRYING THINGS!
00:34 skids :-)
00:41 colomon o/
00:43 [Coke] ho, colomon.
00:43 colomon [Coke], my friend.
00:43 colomon I added 20 tests to set.t while sitting at the tire store today.
00:43 colomon haven't pushed yet because I figure I'll add some more tonight.
00:49 sorear colomon: more pushes more karma
00:49 colomon sorear!
00:51 colomon after your ops patch, niecza spectests clean except for that one Unicode test that's been failing
00:53 colomon oooo, they're making pan fried dumplings on Triple D!
00:53 [Coke] hee!
00:55 sorear colomon: yay
00:56 colomon I've been making my own for a decade and a half now, but it's nice to see how a pro does it.
01:14 yertalert joined #perl6
01:18 slavik joined #perl6
01:18 TimToady sigh, I'm trying to do some "production code", and nom keeps giving me strange errors without any line numbers
01:18 TimToady Cannot assign to a non-container
01:19 skids means it was't a scalar, and didn't have ST
01:19 skids STORE
01:20 TimToady my program runs several minutes, and has many assignments, and likely executes many more assignments buried down in CORE; it's just be nice to know which of those assignments is failing...
01:22 sorear no backtrace !?
01:23 TimToady nope
01:24 TimToady I'd try niecza on it, but I need Date
01:25 colomon if only someone had Date running on niecza... (kicks [Coke] under the table)
01:25 skids Hmm.  Seeing that I'm an expert on causing that particular error message, it has always has a line number for me.
01:25 ggoebel joined #perl6
01:33 colomon niecza: my $s = 10; my $s; say $s
01:33 p6eval niecza v14-34-g1b3d1e8: OUTPUT«Potential difficulties:â�¤  Useless redeclaration of variable $s (see line 1) at /tmp/cLqwUUbtrE line 1:â�¤------> [32mmy $s = 10; my $s[33mâ��[31m; say $s[0mâ�¤â�¤10â�¤Â»
01:34 am0c joined #perl6
01:42 mikemol joined #perl6
01:50 TimToady niecza: multi infix:<+>(Str $x, Str $y) { "$x + $y" }; say 1 + 1
01:50 p6eval niecza v14-34-g1b3d1e8: OUTPUT«Unhandled exception: No candidates for dispatch to &infix:<+>; candidates are:␤    Str, Str␤  at /tmp/9YFkKKhkws line 1 (mainline @ 2) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3751 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3752 (mod…
01:51 TimToady sorear: ^^ same problem nom had; needs the CORE's proto to govern
01:51 colomon TimToady: I don't think infix:<+> is a multi in niecza.
01:52 geekosaur shouldn't it complain when it sees "multi infix:<+>" then?
01:52 TimToady no, it just shadows the outer one, and defines a new implicit proto
02:04 dalek niecza: 1a0d0a8 | (Solomon Foster)++ | lib/CORE.setting:
02:04 dalek niecza: Update Set.new, fix Set.pick, Set.roll, Key.pick, KeySet.roll.
02:04 dalek niecza: review: https://github.com/sorear/niecza/commit/1a0d0a8e86
02:05 dalek roast: a35fbd5 | (Solomon Foster)++ | S02-types/set.t:
02:05 dalek roast: 51 additional tests inspired by bag.t.
02:05 dalek roast: review: https://github.com/perl6/roast/commit/a35fbd5454
02:14 geekosaur hm, right, keep forgetting the setting is an outer scope to the "outermost" scope
02:16 bluescreen10 joined #perl6
02:30 _jaldhar_ joined #perl6
02:47 yertalert joined #perl6
03:12 wolfman2000 joined #perl6
03:16 dalek niecza: 8b82f01 | (Solomon Foster)++ | lib/CORE.setting:
03:16 dalek niecza: Update KeySet.new, fix KeySet.Str, KeySet.gist, and KeySet.perl.
03:16 dalek niecza: review: https://github.com/sorear/niecza/commit/8b82f01f90
03:17 dalek roast: 9da8a25 | (Solomon Foster)++ | S02-types/keyset.t:
03:17 dalek roast: 59 additional tests ported over from set.t.
03:17 dalek roast: review: https://github.com/perl6/roast/commit/9da8a25e98
03:22 cognominal_ nom:  say [1].map: { "{$_.WHAT}" }
03:22 p6eval nom a39d86: OUTPUT«use of uninitialized value of type Int in string context␤␤»
03:22 colomon niecza: say 1
03:22 p6eval niecza v14-36-g1a0d0a8: OUTPUT«1␤»
03:23 cognominal_ nom:  say [1].map: { $_.WHAT  }
03:23 p6eval nom a39d86: OUTPUT«use of uninitialized value of type Int in string context␤␤»
03:24 * colomon needs to learn how to get p6eval to rebuild
03:28 TimToady evalbot rebuild niecza
03:28 p6eval OK (started asynchronously)
03:28 colomon TimToady++
03:29 colomon so, I just tried using your operators with a KeySet, and it doesn't quite work
03:29 dalek Rebuild of niecza complete.
03:29 colomon sweet!
03:29 colomon niecza: say "b" ∈ KeySet.new(<a b c>)
03:29 p6eval niecza v14-37-g8b82f01: OUTPUT«Unhandled exception: Ambiguous dispatch for &to-set; matched candidates are:␤    Any␤    Any␤  at /home/p6eval/niecza/lib/CORE.setting line 2008 (Set.postcircumfix:<( )> @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 2036 (infix:<∈> @ 3) ␤  at /tmp/Y…
03:30 colomon If I comment out multi to-set ($elem --> Set) it works
03:32 colomon apparently niecza can't decide between to-set(%elems) and to-set($elem), given a KeySet
03:39 orafu joined #perl6
03:57 dalek niecza: ac280b0 | (Solomon Foster)++ | lib/CORE.setting:
03:57 dalek niecza: Add explicit to-set multis for KeySet, Bag, and KeyBag, allowing you to use the special Set operators with those types.
03:57 dalek niecza: review: https://github.com/sorear/niecza/commit/ac280b0051
03:58 colomon evalbot rebuild niecza
03:58 p6eval OK (started asynchronously)
03:59 dalek Rebuild of niecza complete.
03:59 colomon niecza: say "b" ∈ KeySet.new(<a b c>)
03:59 p6eval niecza v14-38-gac280b0: OUTPUT«True␤»
03:59 colomon \o/
04:00 TimToady yayness
04:00 colomon I reckon if no one beats me to it, I'll start writing set operator tests over the weekend.  :)
04:00 colomon bedtime now.
04:01 TimToady night
04:23 Su-Shee_ joined #perl6
04:38 PacoAir joined #perl6
04:53 lestrrat joined #perl6
05:00 retupmoca joined #perl6
05:44 moritz o/
05:51 fasad joined #perl6
06:03 sorear /o
06:05 TimToady ø
06:15 dalek roast: 3eb0231 | moritz++ | S12-class/attributes.t:
06:15 dalek roast: rakudo unfudge
06:15 dalek roast: review: https://github.com/perl6/roast/commit/3eb02314bc
06:18 aindilis joined #perl6
06:35 Trashlord joined #perl6
06:46 lestrrat joined #perl6
06:49 kaare_ joined #perl6
06:54 takesako joined #perl6
07:05 Bucciarati joined #perl6
07:14 dalek roast: 6763664 | moritz++ | S32-exceptions/misc.t:
07:14 dalek roast: fix test for X::Attribute::Undeclared
07:14 dalek roast: review: https://github.com/perl6/roast/commit/6763664024
07:19 lestrrat joined #perl6
07:25 bkolera joined #perl6
07:31 Trashlord joined #perl6
07:40 moritz nom: .say for item 1, 2, 3
07:40 p6eval nom a39d86: OUTPUT«1 2 3␤»
07:40 dalek roast: 5f45cb4 | moritz++ | S (4 files):
07:40 dalek roast: more rakudo unfudges
07:40 dalek roast: review: https://github.com/perl6/roast/commit/5f45cb4cb7
07:43 araujo joined #perl6
07:43 araujo joined #perl6
07:48 moritz niecza: say item(<a b c>.list).WHAT
07:48 p6eval niecza v14-38-gac280b0: OUTPUT«List()␤»
07:49 noam_ joined #perl6
07:50 moritz nom: sub i($x) { $x }; for i(<a b c>) {.say}
07:50 p6eval nom a39d86: OUTPUT«a b c␤»
07:53 sorear pushing onto TODO stack: handles, lazy constant @foo
07:54 y3llow_ joined #perl6
07:55 dalek rakudo/nom: 8080589 | moritz++ | src/core/Exceptions.pm:
07:55 dalek rakudo/nom: fix exception type thinko
07:55 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8080589238
07:56 birdwindupbird joined #perl6
07:57 REPLeffect joined #perl6
08:08 dalek rakudo/nom: 20cae50 | moritz++ | src/core/Any.pm:
08:08 dalek rakudo/nom: make &item a tiny bit more correct
08:08 dalek rakudo/nom:
08:08 dalek rakudo/nom: the slurpy case still changes the type to Array, which is wrong.
08:08 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/20cae50d48
08:09 dalek roast: 5c0802d | moritz++ | S (2 files):
08:09 dalek roast: more exception tests
08:09 dalek roast: review: https://github.com/perl6/roast/commit/5c0802db67
08:09 dalek roast: c59b775 | moritz++ | S03-operators/context.t:
08:09 dalek roast: rakudo unfudge/refudge of context.t
08:09 dalek roast: review: https://github.com/perl6/roast/commit/c59b77554e
08:12 dalek roast: bdc62b4 | moritz++ | S29-any/isa.t:
08:12 dalek roast: remove dubious tests
08:12 dalek roast: review: https://github.com/perl6/roast/commit/bdc62b4951
08:17 _jfried joined #perl6
08:18 moritz phenny: ask TimToady if hash zen slices return the whole hash or just the values. S02:3954 says "entire hash", S09:636 seems to imply that only the values are returned (though that S09 thing is a bit weird too, it {}-indexes an array)
08:18 phenny moritz: I'll pass that on when TimToady is around.
08:19 xinming_ joined #perl6
08:19 dalek roast: 4015e8c | moritz++ | S02-literals/hash-interpolation.t:
08:19 dalek roast: get rid of hash ordering assumption in interpolation test
08:19 dalek roast: review: https://github.com/perl6/roast/commit/4015e8ccfb
08:21 gfldex joined #perl6
08:21 overrosy joined #perl6
08:29 overrosy_ joined #perl6
08:29 y3llow_ joined #perl6
08:30 zamolxes joined #perl6
08:31 betterworld joined #perl6
08:42 dalek roast: cdfdee6 | moritz++ | S32-list/ (2 files):
08:42 dalek roast: remove dubious tests (.pick and .roll on junctions)
08:42 dalek roast: review: https://github.com/perl6/roast/commit/cdfdee6db0
08:42 dalek roast: 34d6705 | moritz++ | integration/99problems-21-to-30.t:
08:42 dalek roast: layout; remove superstitious use of parens
08:42 dalek roast: review: https://github.com/perl6/roast/commit/34d670592f
08:42 GlitchMr joined #perl6
08:42 jrockway_ joined #perl6
08:45 jlaire joined #perl6
08:48 birdwindupbird joined #perl6
08:49 REPLeffect joined #perl6
08:57 sftp_ joined #perl6
09:16 Tedd1 joined #perl6
09:22 Trashlord joined #perl6
09:29 ab5tract joined #perl6
09:39 travis-ci joined #perl6
09:39 travis-ci [travis-ci] parrot/parrot#57 (master - 4c49a78 : Jonathan "Duke" Leto): The build passed.
09:39 travis-ci [travis-ci] Change view : https://github.com/parrot/parrot/compare/1b91854...4c49a78
09:39 travis-ci [travis-ci] Build details : http://travis-ci.org/parrot/parrot/builds/661068
09:39 travis-ci left #perl6
09:40 moritz dukeleto: do we really need parrot CI reports in #perl6?
09:42 dukeleto i can remove it
09:42 dukeleto moritz: would you prefer only failures?
09:43 dukeleto moritz: or nothing?
09:43 moritz dukeleto: I'd prefer no parrot reports in #perl6
09:44 moritz dukeleto: when you set up rakudo CI, I'm happy to have the reports in here
09:44 dukeleto moritz: ok
09:44 moritz dukeleto++
09:44 dukeleto moritz: i just removed #perl6 from the .travis.yml
09:47 dalek rakudo/nom: 1733c1f | moritz++ | src/ (2 files):
09:47 dalek rakudo/nom: typed exceptions for errors from has_self branch
09:47 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/1733c1f1e3
09:47 dalek roast: 478ed38 | moritz++ | S32-exceptions/misc.t:
09:47 dalek roast: more tests for availabilty-of-self related errors
09:47 dalek roast: review: https://github.com/perl6/roast/commit/478ed38dff
09:58 noam joined #perl6
10:02 serg211 joined #perl6
10:04 travis-ci joined #perl6
10:04 travis-ci [travis-ci] parrot/parrot#58 (master - ae4bf3d : Jonathan "Duke" Leto): The build passed.
10:04 travis-ci [travis-ci] Change view : https://github.com/parrot/parrot/compare/4c49a78...ae4bf3d
10:04 travis-ci [travis-ci] Build details : http://travis-ci.org/parrot/parrot/builds/661181
10:04 travis-ci left #perl6
10:06 mj41_nb joined #perl6
10:07 sorear dukeleto: um?
10:11 dalek roast: 3551795 | moritz++ | S14-roles/basic.t:
10:11 dalek roast: remove dubious test
10:11 dalek roast: review: https://github.com/perl6/roast/commit/355179544f
10:23 xinming joined #perl6
10:40 snearch joined #perl6
10:44 jnthn moritz: ping
10:45 jnthn phenny: tell moritz it seems that things like https://gist.github.com/1798689 get no backtrace or indication of line number. If subs are used they do, but if there's no routines in there then nothing.
10:45 phenny jnthn: I'll pass that on when moritz is around.
10:47 sorear o/ jnthn
10:47 jnthn o/ sorear
10:47 sorear mmm, probably what TimToady++ was seeing
10:48 jnthn sorear: Yes, saw it while scanning backlog and wanted to hunt it down.
10:52 benabik joined #perl6
10:53 moritz jnthn: yes, I noticed; that was the result from experimenting with different backtrace modes
10:53 phenny moritz: 10:45Z <jnthn> tell moritz it seems that things like https://gist.github.com/1798689 get no backtrace or indication of line number. If subs are used they do, but if there's no routines in there then nothing.
10:53 moritz jnthn: I guess I'll revert that part
10:55 jnthn moritz: Yeah. I think we need something slightly smarter (more)
10:56 moritz one problem is that even if I dump $!.backtrace.full, there are cases where the calling location of the error doesn't appear anywhere
10:56 jnthn When inside a nested block within a sub, it'd be nice to show the current line number you're at *inside* the sub (even if within, say, an if statement's block) along wiht the sub's name. But eliminating non-routines isn't enough either - we need to consider closures etc.
10:57 jnthn One way to detect that is by looking one frame down, and seeing if the code object there is the outer of the current code object.
10:58 jnthn moritz: (location doesn't appear anywhere) I'd be interested to see/look at such cases.
10:58 jnthn In this one, though, --ll-exception contains all the info, so it seems it is there.
11:01 MayDaniel joined #perl6
11:07 birdwindupbird joined #perl6
11:14 moritz jnthn: if I come across such a case again, I'll make a gist
11:14 moritz jnthn: note that it's not so easy with inner blocks
11:14 moritz for example
11:14 moritz sub foo { for 1..3 { die $_ } }
11:15 moritz there the block has a call chain of depth ~10 before it hits the outer sub
11:15 tadzik good morning
11:15 tadzik (I hope you have one :))
11:15 moritz we should really walk the outer chain right away to find the next enclosing sub
11:15 moritz good morning tadzik
11:19 tokuhirom joined #perl6
11:24 moritz nom: say Block.^methods
11:24 p6eval nom 1733c1: OUTPUT«Method 'gist' not found for invocant of class 'Sub'␤␤»
11:25 moritz nom: my @b := Block.^methods
11:25 p6eval nom 1733c1:  ( no output )
11:25 moritz nom: my @b := Block.^methods; for ^(+@b) { try say @b[$_] }
11:25 p6eval nom 1733c1: OUTPUT«arity␤count␤signature␤of␤returns␤ACCEPTS␤Str␤»
11:26 moritz nom: my @b := Block.^methods; for ^(+@b) {  say @b[$_] }
11:26 p6eval nom 1733c1: OUTPUT«Method 'gist' not found for invocant of class 'Sub'␤␤»
11:26 moritz nom: my @b := Block.^methods; for ^(+@b) {  say ~@b[$_] }
11:26 p6eval nom 1733c1: OUTPUT«(signal SEGV)»
11:26 moritz wow.
11:26 moritz jnthn: can I walk the outer chain in Perl 6? if so, how?
11:30 jnthn moritz: That SEGV is ticketed already...
11:30 jnthn moritz: You...probably can though it's a bit magical
11:31 jnthn moritz: Maybe I should give Code a get_outer method or something.
11:31 jnthn Or just .outer
11:32 moritz .outer would be great to have
11:34 Chillance joined #perl6
11:34 moritz jnthn: https://gist.github.com/1798902
11:35 jnthn moritz: https://gist.github.com/1798905
11:36 jnthn in block <anon> at foo.pl:4
11:36 jnthn ...hmm. That's off-by-one.
11:39 jnthn moritz: oh, no, it's more involved than that.
11:41 moritz jnthn: which part is more involved?
11:44 jnthn moritz: The thingy you gisted me with the wrong line number reporting
11:44 jnthn moritz: The 4 is actually the location inside the CATCH block
11:44 moritz yes, I though so
11:45 jnthn The annotations show up in the right places in the compiled output.
11:45 moritz the original example had a statement between the die and the CATCH
11:46 moritz jnthn: the p6ification of the backtrace has 'next'
11:46 moritz two actually
11:46 moritz next if pir::isnull($bt[$_]<sub>);
11:46 moritz next unless $line && $file;
11:48 moritz maybe some of the annotations don't survive, and then that code filters out the whole frame
11:49 jnthn Ah, I think I can guess what's going on.
11:51 jnthn We don't unwind frames when entering an exception handler, but we do end up back in the block which has the CATCH. It executes a couple of instructions back in that block in order to enter the CATCH block, but in doing so shifts its PC, which means that when we come to build the backtrace for that frame, the location has shifted.
11:54 moritz afk
11:56 Su-Shee juddididai. good morning everyone.
11:57 whiteknight joined #perl6
11:59 jnthn morning, Su-Shee
12:02 isBEKaml joined #perl6
12:03 isBEKaml lag--
12:03 isBEKaml hola, #perl6!
12:07 isBEKaml since when was ^methods behaviour changed? complete with signatures, that's nice.
12:07 jnthn isBEKaml: Forget exactly, but fairly recently
12:08 jnthn Like, within the last month or so
12:09 isBEKaml jnthn: is there any way I can get the full listing in a private msg window to p6eval? What do I have to do if I just wanted method names?
12:10 isBEKaml I tried :local, that gave method signatures too, albeit local ones..
12:12 isBEKaml jnthn: Ah, gist. Nvm. :)
12:20 isBEKaml joined #perl6
12:25 jnthn isBEKaml: What exactly are you doing?
12:25 jnthn nom: say Int.^methods>>.name
12:25 p6eval nom 1733c1: OUTPUT«Int Num Rat abs Bridge chr succ pred sqrt base floor round ceiling WHICH perl Bool Str Rat abs sign conj sqrt sin asin cos acos tan atan atan2 sec asec cosec acosec cotan acotan sinh asinh cosh acosh tanh atanh sech asech cosech acosech cotanh acotanh floor ceiling ro…
12:26 isBEKaml nom: Int.^methods.gist.say
12:26 p6eval nom 1733c1: OUTPUT«Int Num Rat abs Bridge chr succ pred sqrt base floor round ceiling WHICH perl Bool Str Rat abs sign conj sqrt sin asin cos acos tan atan atan2 sec asec cosec acosec cotan acotan sinh asinh cosh acosh tanh atanh sech asech cosech acosech cotanh acotanh floor ceiling ro…
12:26 isBEKaml jnthn: same thing
12:26 jnthn OK, that just gives you method names...
12:26 jnthn I think I missed your question somehow :)
12:27 isBEKaml jnthn: yes, what I wanted was a listing of all methods and signatures in a _private_ p6eval window? I guess p6eval has no way of second guessing if that's a private instance or something. Stupid question, anyway. :)
12:28 jnthn isBEKaml: /msg p6eval nom: ...
12:29 isBEKaml jnthn: yes, and I'm already in a separate window - but that wasn't my question...
12:30 jnthn OK. I clearly didn't drink enough coffee yet today :)
12:31 isBEKaml jnthn: :) # Hacking coffee!
12:31 isBEKaml hmm, <XX> ** <YY> is deprecated in grammar rules?
12:31 jnthn isBEKaml: Yes; the latest Star release still supports it, but we also support the new syntax.
12:32 jnthn It's now done with % on a quantifier
12:32 jnthn <XX>+ % <YY> # means at least one of XX, separated by YY
12:32 isBEKaml jnthn: I was just trying to precompile a module with niecza. It promptly shot it down. :)
12:33 isBEKaml jnthn: Yes, I'll try that
12:36 isBEKaml niecza: grammar Foo { regex blah { <yate> % <bates> }; };
12:36 p6eval niecza v14-38-gac280b0: OUTPUT«[31m===[0mSORRY![31m===[0m��Unrecognized regex metacharacter % (must be quoted to match literally) at /tmp/50nNnRLwhw line 1:�------> [32mgrammar Foo { regex blah { <yate> %[33m�[31m <bates> }; };[0m��Unable to parse regex; couldn't find right brac…
12:36 isBEKaml niecza: grammar Foo { regex blah { <yate>+ % <bates> }; };
12:36 p6eval niecza v14-38-gac280b0:  ( no output )
12:36 isBEKaml niecza: grammar Foo { regex blah { <yate> ** <bates> }; };
12:36 p6eval niecza v14-38-gac280b0: OUTPUT«Potential difficulties:â�¤  Unsupported use of atom ** <bates> as separator; nowadays please use atom+ % <bates> at /tmp/SBcbW2rzPa line 1:â�¤------> [32mmar Foo { regex blah { <yate> ** <bates>[33mâ��[31m }; };[0mâ�¤â�¤Â»
12:37 isBEKaml hmm, niecza doesn't support %, but <X>+ % <Y> ?
12:38 jnthn isBEKaml: Niecza is right here, I think - % isn't used on its own, but as a modified on another quantifier.
12:40 isBEKaml jnthn: Okay, I see
12:41 jnthn Pretty sure that's how I implemented it in Rakudo too...
12:41 jnthn nom: grammar Foo { regex blah { <yate> % <bates> }; }
12:41 p6eval nom 1733c1: OUTPUT«===SORRY!===␤Malformed regex at line 1, near "blah { <ya"␤»
12:42 jnthn nom: grammar Foo { regex blah { <yate>+ % <bates> }; }
12:42 p6eval nom 1733c1:  ( no output )
12:42 jnthn *nod*
12:43 isBEKaml phenny: ask sorear if he can have a look at this?  https://gist.github.com/1799248. I got that when I tried to precompile a module with niecza -C.
12:44 phenny isBEKaml: I'll pass that on when sorear is around.
12:45 isBEKaml jnthn++
12:52 isBEKaml phenny: ignore my last message. I didn't read the error message at all. :/
12:52 isBEKaml phenny: tell sorear to ignore my last message. I didn't read the error message at all. :/
12:52 pochi joined #perl6
12:52 phenny isBEKaml: I'll pass that on when sorear is around.
12:54 isBEKaml colomon: if you're around, can you tell what @path in niecza is? and, how do I set it? I guess you got this thing set with panda-niecza.
12:55 isBEKaml niecza: @*PATH.perl.say;
12:55 p6eval niecza v14-38-gac280b0: OUTPUT«Any␤»
12:56 isBEKaml niecza: use Foo; say 42;
12:56 p6eval niecza v14-38-gac280b0: OUTPUT«Unhandled exception: Unable to locate module Foo in @path␤  at /home/p6eval/niecza/boot/lib/CORE.setting line 1344 (die @ 3) ␤  at /home/p6eval/niecza/src/NieczaPathSearch.pm6 line 23 (NieczaPathSearch.load_module @ 12) ␤  at /home/p6eval/niecza/src/Niecza…
12:56 isBEKaml colomon: ^^that
12:57 isBEKaml oh, that's just NieczaPathSearch's variable array? I could just use @*INC....
13:02 sisar joined #perl6
13:05 * jnthn afk for a bit, errands
13:13 colomon isBEKaml: yes, use @*INC
13:14 colomon panda-niecza uses additional -I command line arguments in the niecza invocation to set @*INC
13:14 colomon be warned, last time I checked @*INC seemed to be write-only
13:15 isBEKaml colomon: write-only?
13:15 colomon It doesn't reflect the command line arguments, at the very least.
13:15 isBEKaml niecza: my @*INC=%*ENV{"HOME"} ~ "/.perl6/lib"; @*INC.perl.say;
13:15 p6eval niecza v14-38-gac280b0: OUTPUT«["/home/p6eval/.perl6/lib"].list␤»
13:16 colomon Wynne:niecza colomon$ mono run/Niecza.exe -Ifoobar -e"say @*INC"
13:16 colomon that prints a blank line
13:16 colomon but it still searches directory foobar
13:17 moritz niecza: @*INC.push: 'foo'
13:17 p6eval niecza v14-38-gac280b0:  ( no output )
13:19 isBEKaml colomon: oh, but it still reads the include path?
13:21 isBEKaml colomon: I see what you mean :)
13:27 ggoebel left #perl6
13:38 xinming joined #perl6
13:45 moritz jnthn: I've experimented a bit with walking the outer chain...
13:45 snearch joined #perl6
13:47 moritz jnthn: what I've found is that the same sub produces different code objects (as determined by .WHICH), depending on whether you get there via call chain or outer chain
13:47 moritz probably becaues the one from the call chain holds information about the lexpad?
13:48 jnthn moritz: Much more likely because of closure cloning.
13:49 moritz how do I compare them in a way that treats them as equal? nqp::where of their $!do objects?
13:49 jnthn moritz: nqp::where of $!do.get_lexinfo() would do it
13:51 moritz jnthn: thanks. That code is slowly getting nasty :-)
14:03 moritz \o/
14:03 jnthn Nastiness FTW! \o/
14:03 jnthn hey, wait...
14:03 jnthn :)
14:03 moritz I've correctly found the outer Routine in the call chain
14:04 jnthn :)
14:04 moritz and I've added a Code.static_id method that makes the code look a little less nasty
14:04 jnthn ok
14:05 moritz ie I don't have to litter the backtrace printer with nqp::getattr statements
14:05 jnthn It may be that .WHICH should use that.
14:05 moritz well, the question if $closure === $cloned_closure should be True or not
14:06 jnthn Yeah, not sure on that. Don't recall a spec decision on it either.
14:07 moritz so far the policy for === has been rather strict
14:10 moritz I mean if you do thing like:
14:10 moritz sub gen($x) { sub () { $x } }; my $a = gen('a'); my $b = gen('b'), say $a === $b
14:10 jnthn Right.
14:10 jnthn Yeah, .WHICH probably shouldn't use .static_id :)
14:11 moritz then that would give True, even though both subs are pure, and give different results when executed
14:11 jnthn *nod*
14:11 moritz and neither should eqv
14:21 ccc joined #perl6
14:45 TimToady moritz: vvv first approximation without caffiene: [] should have same effect as leading @
14:45 phenny TimToady: 08:18Z <moritz> ask TimToady if hash zen slices return the whole hash or just the values. S02:3954 says "entire hash", S09:636 seems to imply that only the values are returned (though that S09 thing is a bit weird too, it {}-indexes an array)
14:46 moritz TimToady: I'm asking about { } zen slice, not []
14:49 colomon TimToady: while snowblowing the driveway this morning, I had grave misgivings about using to-set to convert Bags, KeySets, and KeyBags to Sets for the purposes of the Set operators.  Seems like most (all?) of the code would be identical if the arguments were any of those types, and it would be considerably more efficient.  I suspect we need a Setty role, or something like that, to make the organization simple and elegant...
14:52 TimToady colomon++ is addicted to power tools :)
14:53 colomon TimToady: in particular, the code is implemented using just Set.exists and Set.keys, and those are available on all those types.
14:53 colomon having a snowblower is indeed pretty awesome
14:53 TimToady well, roles are kinda like snowblowers...
14:53 colomon as is having perl 6
14:54 colomon little guy is requesting  more Dr Seuss tongue-twisters, but I'll be right by my laptop the whole time...
14:55 jnthn .oO( Finally, I know how to explain roles to people! )
14:55 PacoAir joined #perl6
14:58 TimToady moritz: I'm not sure that using Zen slice for :v is going to stay in S09; it doesn't seem like a good idea this morning--but then not much does this early... :)
14:59 moritz TimToady: let's try to make it simple... what should   my %h = a => 1; say "%h{}";  print?
14:59 moritz a\t1\n or 1\n ?
14:59 TimToady same as {~%h}
15:00 moritz ok, thanks
15:00 TimToady that is, Zen slice really has no effect when interpolating, but outside of a string, has same effect as % rather than $
15:00 TimToady so you can use it for a postfix "sigil"
15:01 ggoebel joined #perl6
15:05 colomon niecza: my $a = set <a b c d>; say $a<c>:exists
15:05 p6eval niecza v14-38-gac280b0: OUTPUT«Unhandled exception: Parameter '' in 'Set.perl' requires a defined argument␤  at /home/p6eval/niecza/lib/CORE.setting line 0 (Set.perl @ 1) ␤  at /home/p6eval/niecza/lib/CORE.setting line 310 (Any.exists_key @ 4) ␤  at <unknown> line 0 (ExitRunloop @ 0) ␤ …
15:05 dalek rakudo/nom: 4b92201 | moritz++ | src/core/Code.pm:
15:05 dalek rakudo/nom: Code.oute; patch from jnthn++
15:05 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/4b922015f5
15:05 dalek rakudo/nom: 84e6cbc | moritz++ | src/core/Code.pm:
15:05 dalek rakudo/nom: Code.static_id which ignores closure cloning
15:05 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/84e6cbc1e6
15:05 dalek rakudo/nom: 13529c7 | moritz++ | src/core/Any.pm:
15:05 dalek rakudo/nom: fix Hash zen slice to return the whole Hash, not just the values
15:05 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/13529c7a88
15:05 dalek rakudo/nom: 1200dd8 | moritz++ | src/core/Exception.pm:
15:05 dalek rakudo/nom: switch back to fuller stacktraces until smarter traces are implemented
15:05 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/1200dd84fd
15:06 ksi joined #perl6
15:06 colomon niecza: my $a = set <a b c d>; say $a.hash()<c>:exists
15:06 p6eval niecza v14-38-gac280b0: OUTPUT«True␤»
15:12 colomon niecza> my $a = set <a b c d>; say $a<c>:exists
15:12 colomon True
15:12 colomon that's after a pretty trivial patch.
15:14 colomon suggests to me that maybe instead of making a new role, the Set operators should work on Associative instead of Set?
15:19 shinobicl___ joined #perl6
15:24 dalek roast: 4e2fc98 | moritz++ | S02-types/hash.t:
15:24 dalek roast: hash slices return the whole hash, not just values
15:24 dalek roast: review: https://github.com/perl6/roast/commit/4e2fc98b4f
15:29 moritz https://gist.github.com/1800832 # current (uncommitted) state of nicer backtraces in rakudo
15:29 moritz (output at the bottom of the gist)
15:31 moritz it searches for the first caller that is "interesting", remembers its line number, then searches outwards for a routine (and if none exists, a block) that also appears in the call chain, and takes its name
15:31 moritz and starting from there it searches for the next caller, rinse and repeat
15:32 Simas_J joined #perl6
15:39 tadzik moritz: very nice!
15:39 HashWeb joined #perl6
15:40 tadzik moritz++
15:42 moritz perl6: my %h = a => 1; print %h
15:42 p6eval pugs b927740, rakudo 1200dd, niecza v14-38-gac280b0: OUTPUT«a   1»
15:43 Psyche^ joined #perl6
15:45 dalek roast: ecd13f4 | moritz++ | S02-literals/ (2 files):
15:45 dalek roast: fix and unfudge zen hash slice tests
15:45 dalek roast: review: https://github.com/perl6/roast/commit/ecd13f455d
15:45 moritz it seems that github still suffers from that DDOS; every second request I make gets "connection refused"
15:45 colomon TimToady: any reason why Set et al shouldn't support :exists, and KeySet/KeyBag also :delete  ?
15:45 * colomon has just implemented this...
15:48 colomon niecza: my %h = a=>1, b=>2; say %h.perl; %h<b>:delete; say %h.perl
15:48 p6eval niecza v14-38-gac280b0: OUTPUT«{"a" => 1, "b" => 2}.hash␤{"a" => 1}.hash␤»
15:49 colomon niecza: my %h = a=>1, b=>2; say %h.perl; %h.delete_key("b"); say %h.perl
15:49 p6eval niecza v14-38-gac280b0: OUTPUT«{"a" => 1, "b" => 2}.hash␤{"a" => 1, "b" => 2}.hash␤»
15:50 shinobicl___ left #perl6
15:50 colomon niecza: my %h = a=>1, b=>2; say %h.perl; say %h<b>:delete; say %h.perl
15:50 p6eval niecza v14-38-gac280b0: OUTPUT«{"a" => 1, "b" => 2}.hash␤2␤{"a" => 1}.hash␤»
15:59 dalek specs: d28c8ef | moritz++ | S32-setting-library/Exception.pod:
15:59 dalek specs: [S32::Exception] X::AdHoc naming conventions
15:59 dalek specs: review: https://github.com/perl6/specs/commit/d28c8ef33b
16:04 moritz ok, there's one problem with the new backtrace code
16:04 moritz it doesn't show any location for exceptions thrown directly from the mainline
16:21 * moritz founds http://journal.dedasys.com/2010/03/30/where-tcl-and-tk-went-wrong/ quite interesting
16:28 colomon joined #perl6
16:30 dalek niecza: 5857ccb | (Solomon Foster)++ | lib/CORE.setting:
16:30 dalek niecza: Fix apparent (probably meaningless) bug in Hash.delete_key.  Add exists_key and delete_key as appropriate in Set, KeySet, Bag, and KeyBag.
16:30 dalek niecza: review: https://github.com/sorear/niecza/commit/5857ccba8f
16:34 dalek roast: df27e0d | (Solomon Foster)++ | S02-types/ (4 files):
16:34 dalek roast: Add tests for Set et al :exists and :delete.
16:34 dalek roast: review: https://github.com/perl6/roast/commit/df27e0df31
16:44 TimToady colomon: that's fine; sets and bags are allomorphically hashes when used that way
16:44 colomon good, now I don't have to undo that patch.  ;)
16:45 colomon TimToady: any notion of a good name for a role common to Set, KeySet, Bag, and KeyBag?
16:47 moritz Keyed?
16:47 * moritz not good at naming
16:48 felher tadzik++ # 2251cbb0b89256d884c9f00c4c6946b0a3bc4a30
16:48 * felher adds #= to a few little scripts.
16:51 colomon moritz, TimToady: Setty?
16:52 colomon Settish?
16:53 colomon this is the role that has .exists and .keys, and the goal is to mean it's okay to use it as "set" input for the Set operators
16:53 Su-Shee Beagle as in Setter ;)
16:54 jnthn moritz: Nice work on the backtrace stuff :)
16:58 TimToady well, I kinda like Baggy
16:59 colomon The thing of it is, there's also the notion that we might create a role to handle much of the Set/KeySet functionality, and another to handle the Bag/KeyBag functionality; and I thought we'd use Baggy for the latter.
17:00 colomon naming things is almost the hardest part of computer programming...
17:01 Su-Shee Birkenbag is the ultimate bag. ;)
17:03 yertalert joined #perl6
17:03 moritz naming and corner cases :-)
17:07 noam joined #perl6
17:07 moritz and error handling :-)
17:30 TimToady and everything else :)
17:33 unobe joined #perl6
17:40 araujo joined #perl6
17:50 timotimo my fibonacci implementation with memoisation is able to calculate up to 974 numbers
17:50 timotimo with 975 it will not terminate
17:50 timotimo i wonder why
18:03 dukeleto o/
18:04 arnsholt :q
18:04 arnsholt Er, opps
18:04 dukeleto Baggy++
18:04 arnsholt Wrong window ^_^
18:09 birdwindupbird joined #perl6
18:09 am0c joined #perl6
18:37 buubot_backup joined #perl6
18:38 mj41_nb joined #perl6
18:50 am0c joined #perl6
18:51 cognominal_ nom: say Parcel == Array
18:51 p6eval nom 1200dd: OUTPUT«use of uninitialized value of type Parcel in numeric context␤use of uninitialized value of type Array in numeric context␤True␤»
18:52 cognominal_ I don't think == should accept to compare undefined values.
18:52 cognominal_ nom: say Parcel === Array
18:52 p6eval nom 1200dd: OUTPUT«False␤»
18:52 cognominal_ it should propose to use === and fail.
18:54 moritz well, == is for numbers, so it makes numbers out of the objects
18:54 moritz if you don't want that, don't use ==
18:54 moritz and it even warns you
18:55 cognominal_ the warning is without filename and  line number. So it is useless..
18:56 moritz well, that can be changed
18:57 geekosaur insufficient clarity in undefined value warnings is my biggest bugaboo with p5; let's not keep it in p6...
18:58 cognominal_ geeksosaur++  # I feel less lonely
18:59 cognominal_ and we get new ones in Perl6, like in stringifying undefined values wich triggered no warnings in Perl 5.
18:59 cognominal_ *which
19:00 geekosaur hm, I recall them producing warnings, but have always been the least useful ones...
19:02 cognominal_ moritz, I complain but I should applaude because there is been a great effort toward better error messages lately.
19:03 cognominal_ s/is/has/
19:04 buubot_backup joined #perl6
19:06 moritz eval: use warnings; undef ~ ""
19:06 buubot_backup moritz: ERROR: Can't modify constant item in undef operator at (eval 20) line 2, at EOF
19:06 moritz eval: use warnings; undef . ""
19:06 buubot_backup moritz: Use of uninitialized value in concatenation (.) or string at (eval 20) line 1.
19:06 birdwindupbird joined #perl6
19:07 moritz geekosaur: which part of the undef warnings do you find insufficient clear?
19:10 moritz nom: sub f() { 3; CATCH { } }; say f
19:10 p6eval nom 1200dd: OUTPUT«Nil␤»
19:10 moritz OH
19:10 geekosaur sorry, I was talking about (mostly older) perl 5 there, since cognominal_ seemed to be saying p5's undef warnings for undef-to-string weren't so useful when there were a bunch of them in the same line.  come to think of it, those messages suggest it's still true in that case, if you have 8 stringifications in the same line, which one threw the warning?
19:11 moritz geekosaur: which perl 5 are you using?
19:11 moritz geekosaur: perl 5.10 and higher tries to give you the variable names (though it doesn't always succeed)
19:11 geekosaur I always end up breaking the line into individual statements, one operation per, so I get a line number specific to the one that is failing...
19:11 geekosaur yeh, I'm mostly thinking 5.8 but I have seen it in 5.12 sometimes as well
19:11 geekosaur maybe I'm just good at running into the "doesn't always succeed" cases...
19:12 moritz perl -wE 'my $x;  my $y = 3; say "$x;$y"'
19:12 moritz Use of uninitialized value $x in concatenation (.) or string at -e line 1.
19:15 geekosaur (I keep threatening to get "human fuzzer" shirts made up; I seem to be inordinately good at running headlong into corner cases)
19:15 jnthn nom: my $x; say $x + 42
19:15 p6eval nom 1200dd: OUTPUT«use of uninitialized variable $x of type Any in numeric context␤42␤»
19:15 jnthn Rakudo gets the variable name too sometimes :)
19:16 moritz geekosaur: then you should join the ranks of the perl 6 testers :-)
19:19 geekosaur I was :p now I'm trying to reconstruct what I had had on the external drive which is now ... somewhere
19:19 geekosaur (macbook air, tinny ssd already mostly full, sigh)
19:20 geekosaur s/tinny/tiny/
19:21 geekosaur ...actually I did manage to find most of the other netbook yesterday, maybe I should set stuff up on that
19:21 geekosaur (then again I have this crappy notwork to deal with too, sigh.  and can't go anywhere with better net today.  rrrgh)
19:30 moritz ARGH
19:30 moritz there's this problem that I'm working on for the last 3.5 hours (not full time though, on and off)
19:30 moritz and now I've found an off-by-one error
19:31 moritz I should use either  $idx < $.elems OR $idx <= $.end  when iterating over the array indexes
19:31 moritz $idx < $.end won't give me the last one
19:32 moritz and thus not the callframe of the mainline
19:32 moritz and thus an empty backtrace for a simple  perl6 -e 'die "foo"'
19:32 jnthn argh!
19:33 jnthn The simplest things are always the hardest to spot :/
19:35 moritz first I've suspected that an exception in the backtrace printer is eating the exception
19:35 moritz erm, eating the backtrace
19:35 moritz then I've added a CATCH block
19:35 renormalist joined #perl6
19:35 moritz with some debugging
19:35 moritz which ate the return value, and returned Nil instead
19:35 moritz then I fixed that, and added lots of debugging output, and safety band-aid
19:38 MayDaniel joined #perl6
19:38 moritz hm, but it doesn't seem to fix the whole thing :/
19:38 wolfman2000 joined #perl6
19:40 tarch joined #perl6
19:43 moritz prototyping the whole thing was fun, because it was easy to do outside the setting
19:43 moritz but I can't do that for the case of just a single callframe
19:46 whiteknight joined #perl6
19:52 moritz I'm just listening to Pink Floyd's "Brick in the Wall", and heard a part of the lyrics as "we don't need no education, we don't need no source control" :-)
19:53 cognominal_ :)
19:54 jnthn .oO( No Darcs sarcasm in the hack room... )
19:55 * geekosaur *eyeroll*
19:57 retupmoca joined #perl6
19:57 dukeleto .oO( No non-atomic version control systems in the bike shed ... )
19:58 benabik joined #perl6
20:02 sorear good * #perl6
20:02 phenny sorear: 12:44Z <isBEKaml> ask sorear if he can have a look at this?  https://gist.github.com/1799248. I got that when I tried to precompile a module with niecza -C.
20:02 phenny sorear: 12:52Z <isBEKaml> tell sorear to ignore my last message. I didn't read the error message at all. :/
20:15 moritz AARGH AGAIN
20:17 wentam joined #perl6
20:17 jnthn moritz: What this time?
20:18 moritz jnthn: -e doesn't add file name annotations
20:18 moritz called from Sub '_block1002' pc 68 ((file unknown):123221183) (:1)
20:18 moritz called from Sub '_block1000' pc 14 ((file unknown):123221125) (:1)
20:18 moritz and so we filter them out while creating the Perl 6 backtraces in the first place
20:18 jnthn moritz: ugh
20:18 moritz which is why I've been debugging the wrong place all the time
20:18 jnthn :(
20:19 jnthn Guess we could make it something like "<-e ...>:1" or some other way to indicate where the code came from.
20:20 sorear moritz: Where is 123221183 coming from?
20:20 sorear niecza: say 123221183.unbase(16)
20:20 p6eval niecza v14-39-g5857ccb: OUTPUT«Unhandled exception: Unable to resolve method unbase in class Int␤  at /tmp/P5KhRqoQJ7 line 1 (mainline @ 2) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3773 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 3774 (module-CORE @ 65) ␤  at /home…
20:20 sorear niecza: say 123221183.base(16)
20:20 p6eval niecza v14-39-g5857ccb: OUTPUT«75834BF␤»
20:21 jnthn It's what Parrot thinks the current PIR line is...
20:21 jnthn ...though I think it's a bit off ;)
20:21 whiteknight joined #perl6
20:24 moritz where do the file annotations come from anyway?
20:25 flussence outer space?
20:25 moritz through pir::find_caller_lex__ps('$?FILES');
20:30 jnthn moritz: Set in comp_unit action method, iirc
20:32 moritz jnthn: well, set in NQP first
20:32 moritz jnthn: and I'll see if I can fix it there first
20:32 moritz oh well, the file annotations are set in comp_unit, based on $?FILES
20:32 moritz and $?FILES is set in NQP
20:33 moritz hm, eval should set $?FILES too
20:33 moritz nom: my $?FILES = 3
20:33 p6eval nom 1200dd:  ( no output )
20:33 moritz nom: sub f ( ++state $ }; say f, f, f;
20:33 p6eval nom 1200dd: OUTPUT«===SORRY!===␤Malformed parameter at line 1, near "++state $ "␤»
20:34 moritz nom: sub f { ++state $ }; say f, f, f;
20:34 p6eval nom 1200dd: OUTPUT«123␤»
20:34 moritz \o/
20:34 jnthn Note that
20:34 jnthn nom: sub f { state $++ }; say f, f, f
20:34 p6eval nom 1200dd: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix + instead at line 1, near " }; say f,"␤»
20:34 jnthn :(
20:34 jnthn I tried to hunt that one down yesterday and failed.
20:35 moritz nom: sub f { (state $)++ }; say f, f, f
20:35 p6eval nom 1200dd: OUTPUT«===SORRY!===␤Contextual $*GOAL not found␤»
20:35 jnthn eek
20:35 jnthn moritz: btw
20:35 moritz std: sub f { (state $)++ }; say f, f, f
20:35 jnthn nom: sub f { ++state $ }; say f xx 3;
20:35 p6eval std 52f3895: OUTPUT«ok 00:01 113m␤»
20:35 p6eval nom 1200dd: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&xx' called (line 1)␤»
20:35 jnthn er :)
20:35 jnthn nom: sub f { ++state $ }; say f() xx 3;
20:35 moritz f() xx 3
20:35 p6eval nom 1200dd: OUTPUT«1 2 3␤»
20:35 jnthn I did the thunking of xx last night :)
20:35 moritz yes, I've noticed
20:40 flussence nom: sub f { (state $ )++ }; say f, f, f
20:40 p6eval nom 1200dd: OUTPUT«012␤»
20:42 moritz meh, still "file unknown"
20:43 snearch joined #perl6
20:44 Psyche^ joined #perl6
20:45 moritz ah, because I did it in the wrong routine. Well.
20:47 moritz http://it.isagit.com/rakudo/rakudo
20:48 jnthn Not Found?
20:48 jnthn oh, worked on second try
20:49 flussence javascript-only site?
20:49 moritz flussence: nope
20:49 jnthn Awesome, we're nearly at 10000 :)
20:49 moritz flussence: I have noscript here, and nothing allowed
20:50 moritz but it's... weird.
20:50 flussence oh, the errors just happened to coincide with me switching it on/off several times :)
20:50 Psyche^ joined #perl6
20:56 bbkr joined #perl6
20:57 dalek nqp: fd864b6 | moritz++ | src/HLL/Compiler.pm:
20:57 dalek nqp: add "-e" file annotation where appropriate
20:57 dalek nqp: review: https://github.com/perl6/nqp/commit/fd864b606f
20:59 bbkr left #perl6
20:59 bbkr joined #perl6
21:01 bbkr left #perl6
21:02 bbkr joined #perl6
21:02 bbkr left #perl6
21:17 moritz I don't think I'll manage to finish that backtrace stuff tonight
21:17 dalek rakudo/bt: 3f3fd18 | moritz++ | docs/ChangeLog:
21:17 dalek rakudo/bt: update ChangeLog
21:17 dalek rakudo/bt: review: https://github.com/rakudo/rakudo/commit/3f3fd18a7a
21:17 dalek rakudo/bt: 15ed52e | moritz++ | src/core/Backtrace.pm:
21:17 dalek rakudo/bt: improve backtraces a lot
21:17 dalek rakudo/bt: review: https://github.com/rakudo/rakudo/commit/15ed52e4ab
21:17 dalek rakudo/bt: cb3b61d | moritz++ | src/core/Backtrace.pm:
21:17 dalek rakudo/bt: bt off-by-one fixes and more robustness
21:17 dalek rakudo/bt: review: https://github.com/rakudo/rakudo/commit/cb3b61da53
21:17 dalek rakudo/bt: 55444bc | moritz++ | src/core/control.pm:
21:17 dalek rakudo/bt: give eval()s a separate file annotation
21:17 dalek rakudo/bt: review: https://github.com/rakudo/rakudo/commit/55444bcbdb
21:17 dalek rakudo/bt: cf4be76 | moritz++ | tools/build/NQP_REVISION:
21:17 dalek rakudo/bt: bump to newest NQP to get -e file annotation fix
21:17 dalek rakudo/bt: review: https://github.com/rakudo/rakudo/commit/cf4be76948
21:18 moritz so I'm pushing it as a branch, so that it's not lost in case of hard disc failure or so
21:18 moritz and maybe an interested soul even finds out why ./perl6 -e 'die "foo"' doesn't produce sensible output
21:33 jnthn moritz: Here it produces...no output.
21:33 moritz not even the 'foo'?
21:34 moritz maybe I've broken that in cb3b61da536060e27d8401526baf2004020cc68c; will debug tomorrow
21:34 jnthn Not even the foo
21:34 jnthn OK
21:34 jnthn I'll leave it for you, lemme know if you want me to look at anything
21:34 jnthn My Perl 6 tuits didn't really work out today. Hopefully tomorrow is better.
21:42 tadzik ~~
21:50 HashWeb left #perl6
22:16 sisar joined #perl6
22:17 sisar Has Larry Wall given up on Perl 5 ? #sorry for the harsh term :)
22:18 tadzik I don't think Larry works on Perl 5 anymore
22:20 sorear Perl 5 exists quite independantly of Larry
22:23 sisar how can i get coloured text in my output ?
22:24 tadzik use Term::ANSIColor
22:25 sorear welcome to #perl6, sisar
22:25 sisar oops, i should have mention, i'm fasad :)
22:25 tadzik ha
22:25 sisar now with a permanent nick hopefully
22:26 sisar tadzik: you might have as well guessed by my questions ;-)
22:26 tadzik mebbe
22:28 colomon Larry did work on the most recent revision (like, is it even out yet?) of Programming Perl (5).
22:38 sisar how can i run panda from behind a proxy? sh ./bootstrap.sh is stuck at 'Connecting to feather.perl6.nl|193.200.132.135|:3000...  '
22:38 tadzik hmm, I have no idea, I'm not good at proxies
22:39 whiteknight joined #perl6
22:40 sisar :(
22:40 colomon it just uses curl, right?  figure out how curl deals with proxies....
22:42 sisar colomon: curl, the programming language ?
22:43 tadzik I think it uses wget
22:45 sisar tadzik: my terminal output for bootstrap.sh: https://gist.github.com/1804831
22:45 tadzik heh
22:45 tadzik understandable :)
22:45 tadzik You can try downloading the file manually
22:46 sisar all because of the proxy? or some bug ?
22:47 tadzik I think it assures the file is in place
22:47 tadzik *assumes
22:47 tadzik and proceeds, without checking if it really is
22:48 sisar tadzik: which file should i download? i've already cloned the panda repository... #git works flawlessly behind a proxy :)
22:48 tadzik sisar: http://feather.perl6.nl:3000/projects.json
22:50 sisar well wether 'the network is unreachable' or not, it should better tell me why it needed to use the internet in the first place
22:51 sisar may I open a issue regarding proxies on github for panda?
22:51 tadzik of course, go ahead
23:07 sisar tadzik: so put my proxy info in the ~/.wgetrc file, but i still get same errors, without the 'failed to connect part'. The error begins at line #7
23:07 sisar well atleast some improvement :)
23:12 sisar phenny: tell tadzik assuming bootstrap.sh uses wget,  I put my proxy info in the ~/.wgetrc file, but i still get the same errors, this time without the 'failed to connect' part i.e. the error begins at line #7 of https://gist.github.com/1804831 this time
23:12 phenny sisar: I'll pass that on when tadzik is around.
23:12 tadzik I am around :)
23:12 phenny tadzik: 23:12Z <sisar> tell tadzik assuming bootstrap.sh uses wget,  I put my proxy info in the ~/.wgetrc file, but i still get the same errors, this time without the 'failed to connect' part i.e. the error begins at line #7 of https://gist.github.com/1804831 this time
23:13 tadzik I see
23:14 masak good morning, #perl6.
23:15 tadzik good... wait, what
23:15 masak :)
23:15 masak yeah. not entirely correct, I guess.
23:15 tadzik sisar: ok, I see where it is
23:16 tadzik it's in https://github.com/tadzik/panda/blob/master/lib/Panda/Ecosystem.pm#L32
23:16 masak it's night, but if I say "good night", it sounds like I'm saying good bye. I'm not -- I just arrived.
23:16 Teratogen g'day
23:16 Timbus good evening
23:16 tadzik no error checking in update, from-json or others I suppose
23:16 tadzik sisar: do you want to open that issue, or shall I do it?
23:17 tadzik I'm probably not going to fix it today, ETOOTIRED
23:17 sisar tadzik: you should do it
23:17 tadzik okay
23:18 tadzik oh, I have an issue opened 2 months ago which I didn't notice :/
23:18 sisar tadzik: so is the issue with wget or with panda itself ?
23:18 tadzik with panda
23:18 sisar tadzik: issue no. ?
23:18 tadzik still, downloading the file manually should help
23:18 tadzik https://github.com/tadzik/panda/issues/11
23:18 tadzik I'm quite sure it's not panda's fault, but still
23:20 tadzik your case: https://github.com/tadzik/panda/issues/12
23:21 sisar tadzik: well you are tired ;-)
23:21 tadzik heh, a bit hung-over too
23:21 tadzik but, you know, the end of exams :)
23:22 sisar i've got exams this coming week :(
23:22 tadzik good luck with them. Many, hard?
23:23 tadzik I had the last one on friday morning, owned it totally ;)
23:24 localhost joined #perl6
23:24 sisar many. should not be hard if I get away from my laptop and pickup the damn textbook
23:24 sisar *just pickup
23:27 sisar um, where should i put projects.json
23:28 tadzik ~/.panda/ I think
23:28 tadzik yep
23:33 masak tadzik: c'gratulations on end of exams!
23:33 tadzik yay thanks :)
23:33 tadzik I also got to enjoy a suprise vacation, in shape of skiing trip
23:34 sorear masak \o/
23:35 masak sorear! \o/
23:37 sisar tadzik: yay! panda installed ! but see line #2 and #43 of https://gist.github.com/1805099
23:37 tadzik #43 is expected
23:37 tadzik the tests checks, if git fails
23:38 tadzik I think #2 is a rakudobug, ticketed
23:38 araujo moritz, so you guys got pugs development bootstrapped??
23:39 sisar tadzik: #2 looks out of place with no info, without a line no. at which the error/warning was seen
23:40 tadzik indeed, but I seem to recall tracking it down and ticketing
23:40 sisar tadzik++
23:41 ccc left #perl6
23:43 tadzik I can't seem to find it though
23:47 ksi joined #perl6
23:49 tadzik huh, maybe I didn't
23:50 tadzik may have something to do with this: https://gist.github.com/1623088
23:51 tadzik may have not :) Feel free to open another issue for this, my consciousness is slowly drifting away
23:52 sisar tadzik: sorry, don't have the technical know-how for that
23:53 sisar before you sleep, can II have commit access to panda, would like to add the bit about proxies to the README
23:53 sisar *can I
23:53 tadzik sure thing
23:53 sisar github username: Siddhant
23:55 * sisar taking advantage of tadzik's drowsiness to extort commit access ;p
23:55 tadzik it never goes wrong :)
23:55 tadzik and if it would, I would just revert your commits and banish you forever and stuff
23:56 tadzik ;)
23:56 tadzik and re technical know-how, it's just a matter of visiting https://github.com/tadzik/panda/issues/new and writing some random words in text fields
23:57 tokuhirom joined #perl6
23:58 sisar tadzik: oh. That i can do. i thought you wanted me fill in a rakudobug
23:58 sisar *me to
23:58 sisar tadzik++ #for commit access
23:59 tadzik filing a rakudobug is a matter of seding one email with random words to a specific e-mail address ;)

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

Perl 6 | Reference Documentation | Rakudo