pugscode.org/ planetsix.perl.org/ | nopaste: sial.org/pbot/perl6 | evalbot: perl6: say 3; (or rakudo:, pugs:, elf:, etc) | irclog: irc.pugscode.org/ | ~315 days 'til Xmas
Set by mncharity on 8 February 2009.
00:02 eternaleye left
shinobi-cl rakudo: class A { method BUILD { say "constructor"; } }; my $a = new A; 00:03
p6eval rakudo a0a390: OUTPUT«./parrot: error while loading shared libraries: libparrot.so.0.9.0: cannot open shared object file: No such file or directory␤»
shinobi-cl pugs: class A { method BUILD { say "constructor"; } }; my $a = new A;
p6eval pugs: OUTPUT«constructor␤»
00:05 eternaleye joined 00:06 gravity left 00:09 Limbic_Region joined 00:10 DemoFreak left 00:11 alexn_org joined, justatheory joined 00:15 c9s_mv_ is now known as c9s_mv
pugs_svn r25335 | frew++ | [t] tests for polar 00:19
00:22 hanekomu left
pugs_svn r25336 | frew++ | [t] removed test debug code 00:24
00:25 maerzhase left
frew pugs: 0 but Bool::True 00:25
p6eval pugs: OUTPUT«*** Cannot cast from VBool True to Pugs.AST.Types.VCode (VCode)␤ at Prelude.pm line 541, column 5-16␤»
frew rakudo: 0 but Bool::True
rakudo: 0 but Bool::True 00:26
that can't be good
oh dear, that causes some kind of infinite loop
sorry
p6eval rakudo a0a390: No output (you need to produce output to STDOUT)
jnthn frew: Don't worry, there's limits in place for things like that. ;-) 00:30
frew: File a ticket. I seem to remember there's somthing funny about that...it should work.
frew I hate to ask this, but where is the bug tracker 00:32
SamB perl6: 0 but Bool::True 00:33
p6eval rakudo a0a390: No output (you need to produce output to STDOUT)
..elf 25336: OUTPUT«Undefined subroutine &GLOBAL::infix_but called at (eval 119) line 3.␤ at ./elf_h line 4346␤»
..pugs: OUTPUT«*** Cannot cast from VBool True to Pugs.AST.Types.VCode (VCode)␤ at Prelude.pm line 541, column 5-16␤»
frew pugs: [1,2,3].shape 00:35
p6eval pugs: OUTPUT«*** No such method in class Array: "&shape"␤ at /tmp/g9yEDfyvQD line 1, column 1 - line 2, column 1␤»
jnthn frew: mailto:rakudobug@perl.org 00:36
frew ah, thanks
jnthn That creates a ticket.
alexn_org hi, I'm a newbie :) I tried compiling parrot and rakudo (the latest from svn/git) and while parrot compiled ok with all tests passed, I got an error from rakudo: sial.org/pbot/35058 00:41
jnthn Ooooh... 00:42
who broke the dynops compiler... :-|
alexn_org: What platform?
alexn_org debian, 2.6.26-1-686, gcc 4.3.2 00:43
00:43 nihiliad joined
jnthn Hmm. That's nothing unusual. 00:44
SamB where is the index, anyway?
jnthn SamB: index?
SamB for the synopses
[particle]1 i don't think an index file has been generated in html, or if it has, it's not online 00:45
you can use spack here for searching
spack: PROGRAM_NAME
jnthn alexn_org: Just asked on #parrot...seems you're not the first person to notice this problem.
[particle]1 spack PROGRAM_NAME 00:46
hrmm....
00:46 [particle]1 is now known as [particle]
jnthn alexn_org: I don't see a workaround right now...it seems it's a breakage in Parrot's dynop generator...or we're doing something it doesn't support... 00:47
alexn_org jnthn: don't worry, I'll be as eager to try it out when it gets fixed :) 00:48
jnthn alexn_org: OK, great!
Sorry it's broken...am trying to work out why...
frew this makes me want to not have rakudo and parrot rebuild every night... 00:49
SamB frew: which ?
frew the issue that is making alexn_org's rakudo not build 00:50
jnthn frew: This is why in Rakudo we're going to hopefully target Parrot revisions that we know are fine, and so on...
frew well, it should be more stable when it hits 1.0 right?
jnthn (And eventually Parrot releases, but that's premature for now...)
bacek jnthn: github.com/bacek/rakudo/commit/08ff...fce0f32fbf initial implementation of Flipflop. Can you take a look?
frew because that's soon right?
jnthn Yes and no.
bacek away &
jnthn Parrot 1.0 doesn't mean "Parrot stops getting developed now" 00:51
frew yeah, but the interface should be stable right?
jnthn Aye.
That's the aim.
Thing is, Rakudo may need more "cutting edge" stuff.
Especially considering things like PCT and PGE ship with Parrot.
frew I don't know what those are 00:52
SamB maybe they shouldn't ?
jnthn Parts of the compiler toolchain.
SamB doesn't know what they are either
jnthn PGE is the parser engine.
SamB and why does that ship with Parrot?
jnthn PCT is the compiler toolchain - the thing defining the PAST nodes, translating them to bytecode etc.
00:53 justatheory left
jnthn SamB: I guess because since many Parrot-targetting compilers are built on top of them, it makes sense. 00:53
SamB and why does that ship with Parrot?
hmm.
jnthn Not that by toolchain I don't mean something that you use to generate a compiler, then ship the compiler without it. 00:54
You can see it more of as a library of common stuff.
SamB still, if one thing is "stable" and the others aren't, and they aren't actually intimate ...
jnthn alexn_org: Ah, got an answer. Apparently it's a line ending issue.
SamB: Depends on your definition of stable. 00:55
frew does anyone know why running 'perl Configre.pl' would fail in a script even though git pull succeeded?
alexn_org jnthn: where can I find more info on PCT and PGE?
jnthn If it's "we won't take things away that work or change things that used to do X to do Y toher than bug fixes", then yes, that's often going to be true. 00:56
00:56 kanru left
jnthn alexn_org: If you run something like dos2unix on src/ops/perl6.ops (I think that's the file), it may help. 00:56
SamB jnthn: well, I mean, if it's reasonable not to use the HEAD
but to use a release 00:57
jnthn SamB: But it doesn't mean there won't be new things.
And Rakudo will often be one of the first compilers to want to use such New Stuff.
SamB new things like what ?
jnthn alexn_org: In Parrot see perhaps docs/book/ or also the PDD (docs/pdd/) on AST, or compilers/pct/ 00:58
frew if anyone was curious it was because of the hardcoded paths in Configure.pl
jnthn SamB: Mostly new things in PGE. 00:59
SamB: We're working towards using STD.pm.
SamB jnthn: not new thins in Parrot itself?
jnthn Generally, not.
SamB well, that sounds like splitting would be good
jnthn Though occasionally, I've found need to twiddle things in Parrot itself...
SamB: Perhaps. It's not really may call...it's not something I've thought about a lot either to be honest. 01:00
But as we work towards a release mentality rather than an svn HEAD one, yes, it may well be worth considering.
We are going to start doing Rakudo releases shortly. 01:02
Those I expect *will* target a specific Parrot release.
SamB at the very least they should target a specific "r"
jnthn Last I heard, the plan was that Rakudo monthly releases would follow shortly after Parrot ones for now. 01:03
So if that is what happens, we can target a specific Parrot release.
alexn_org jnthn: thanks, the line ending was the issue
jnthn Most probably, anyway.
alexn_org: OK, cool. We'll fix that...
rurban++ from #parrot for knowing what to do :-) 01:04
01:05 M_o_C left
frew should S06-advanced_subroutine_features/wrapping be failing? 01:10
SamB so, are you certain that haskerl was a joke ? 01:14
jnthn frew: Did you run that manually? 01:17
But I am pretty sure, Rakudo does not do .wrap yet...
frew jnthn: yeah
it's not in the spectest apparently 01:18
jnthn frew: Right. We don't run all test scripts.
Just the ones we know we can pass entirely or partially.
frew ok
jnthn We keep a list of those in t/spectest.data
frew right
jnthn SamB: huh?
frew but if it's in the spectest it will eventually be supported right? 01:19
jnthn Right.
frew becasue .wrap is very cool
jnthn Indeed.
SamB @go haskerl
lambdabot No Result Found.
jnthn If you are wanting it soonish, a ticket is a good way.
SamB damn buggy search plugin, I should have the author shot ...
frew jnthn: nah. I just got curious
I have been trying to do things that are in t/TASKS and one of the tasks mentioned it, so I looked at it and got excited 01:20
but then when I tried it on rakudo it wouldn't work
jnthn While we certainly don't want tickets for every missing bit of Perl 6, generally ones for small-ish features that people want - especially those with code examples to make work - are welcomed.
And those who are implementing Rakudo often find implementing cool exciting stuff fun too. ;-) 01:21
frew at some point I'd like to contribut to more than the test suite
and when I feel I have done a good amount of the test suite I'll try helping to add to rakudo myself
jnthn The test suite is a very valuable place to contribute, but of course contributions elsewhere are welcome too. :-)
I think things will get easier in Rakudo when we have a Perl 6 prelude. 01:22
Both pmichaud and I are *very* much wanting this.
frew don't we have an incomplete one?
jnthn Not that we're actually using in Rakudo.
frew oh
I didn't know that
jnthn But we really need t.
*it
Because there's a bunch of stuff that we need to do, that gets really hairy to do in PIR now. 01:23
frew I believe it
jnthn (Generic types...)
frew plus the more that's done in perl6 the easier it is for implementors at some point
right?
jnthn Right.
But right now, we've hit the point where it's a real need. 01:24
frew who is working on that?
jnthn I seriously cannot be bothered, to try and write parametric types in PIR!
frew hahaha
jnthn pmichaud will lead it, but I expect I'll contribute heavily.
frew so it will be part of rakudo? or will it go with pugs? because that would make sense to go with the test suite as it's implementation independant 01:25
jnthn I have a grant depending on getting typed arrays and hashes working, and doing that sanely depends on us having a Perl 6 prelude.
frew I thought they both already worked...
jnthn I think perhaps the best thing I can say on that, is that while there's interest in a cross-implementation prelude, there's difficulties in realizing it for now. 01:26
No. You can do type constriants on scalars.
But my Int @x; is not supported.
Yet.
frew oooooh
gotcha
jnthn frew: It's a parametric type. 01:28
frew: We do support those now though. :-)
frew right
01:28 alester joined
jnthn ...and my oh my was it a brain ache to do so (and continues to be once you consider classes within roles...) 01:29
01:29 eternaleye_ joined
frew haha 01:30
alester hey, jnthn , at some pooint we're going to be converting rakudo.org to drupal
Are you blogging in HTML?
I guess i could look myself.
SamB jnthn: maybe you should reconsider this whole OO/imperative deal 01:31
alester oh good, html 01:32
jnthn alester: I was, yes.
alester that's good
'cause Drupal only does HTML blogging.
jnthn alester: It works on use.perl.org and rakudo.org, so for me it was The Win.
alester I do all the perlbuzz stuff in markdown
jnthn SamB: Hmm? 01:33
01:33 chid joined
SamB jnthn: it sounds like it makes your head hurt ;-P 01:33
jnthn SamB: Perl 6 is a multi-paradigm language. I'm very much in favor of multi-paradigm stuff. 01:34
SamB: I know what OO is, and what imperative is...I just don't see what you're getting at.
SamB just teasing
jnthn SamB: I do value the functional programming paradigm. I was once fairly fluent in ML. I know that purity is good for multi-threading stuff. And so forth... 01:36
SamB I repeat: teasing
jnthn I know.
:-)
SamB now I wonder what all this imperative code in this Haskell compiler does ...
jnthn I remember being at a Haskell tutorial by spj and his hilarious analysis of that situation... :-) 01:38
spj's presentation rocked.
SamB only got to see the videos 01:39
jnthn I felt very lucky to be in it.
SamB but yes, he's cool
he's extremely animated
jnthn I studied for my degree in a building that was under 100m from the building where his office was. :-)
Or maybe still is...I don't know if he's still at the same place... 01:40
But in the end, the first time I saw him present was in Oregon!
SamB afaik he's still at MSR
jnthn Aye, then same place. 01:42
I'm a good thousand kilometers from there now, though.
01:43 eternaleye left
jnthn -> sleep 01:55
02:00 kate21de left 02:08 wknight8111 left 02:24 eternaleye_ left 02:32 eternaleye joined
meppl good night 02:39
02:41 meppl left 02:42 alexn_org left 02:57 bigal joined
alester is writing perl:;critic rules for Rakudo 03:10
literal uh, how? 03:13
using STD.pm instead of PPI or something?
03:18 justatheory joined 03:33 justatheory left
TimToady rakudo is the name of an implementation, not the name of a language 03:44
SamB TimToady: I suppose you think the same thing about GHC? 03:46
TimToady C is for Compiler 03:48
alester SamB: Why argue about it?
TimToady: Yes, I know. I'm cleaning up the Perl 5 code that is in the rakudo repo. 03:49
SamB because most implementations end up forming a dialect, that's why!
TimToady my comment was answering literal 03:50
alester TimToady: We've started putting actual usefulness on rakudo.org
03:58 shinobi-cl left
literal ah, I thought he meant something else, then 04:06
04:12 Limbic_Region left 04:24 eternaleye left 04:32 eternaleye joined 04:49 chid left
bacek TimToady: what should "(1..5).map: { $_ if 1 ff 1 }" produce? (1..5) or (1,3,5)? 05:00
05:14 jferrero left 05:20 c9s_mv left 05:25 eternaleye_ joined 05:26 eternaleye left 05:27 nihiliad left
bacek perl6: sub foo { 1;; }; say foo 05:33
p6eval rakudo a0a390: OUTPUT«Null PMC access in isa()␤current instr.: 'parrot;List;!flatten' pc 5976 (src/classes/List.pir:283)␤»
..elf 25336, pugs: OUTPUT«1␤»
bacek heh.
05:40 araujo left
pugs_svn r25337 | putter++ | [elfparse] Renamed elfgreen to elfrx for clarity (r25333 commit message was wrong). Also elfgreenstd to elfdev. Updated README. 05:50
05:52 nihiliad joined 05:53 fjord41 left
pugs_svn r25338 | putter++ | [elfparse] std.pm: Hacking out a path down from comp_unit towards EXPR. Aiming to parse '3'. 05:55
TimToady bacek: 1..5 06:02
bacek TimToady: thanks. It means that I have to check "flop" right after "flip"? 06:03
TimToady correct, it's exactly the same as .. in p5 (except p5 would implicitly compare 1 with $.)
bacek and "1 .. *" actually same as "1 .. 0"? 06:05
TimToady no, 1..Inf 06:06
bacek hmm...
oh, sorry. 06:07
1 ff 0 vs 1 ff *
"1 ff 0" vs "1 ff *"
TimToady yeah * is taken as false
on the theory that "whatever" is the opposite of the known endpoint 06:08
bacek thanks again. And (hopefully) last question: what should produce "(1..5).map: { $_ if 1 ^ff 1 }"? 06:09
06:09 eternaleye_ left
TimToady () I suspect 06:10
06:10 Tene joined
bacek TimToady: my implementation suspect same. But S03/flip-flop.t suspect something else... 06:10
TimToady depends on how it's implemented, I suppose 06:11
I can see how they'd think the other one
depends on whether you think the rhs is evaluated for the ^ case 06:12
but I think of it as $a .. $b always picking the same range for each variant 06:13
and ^ then throws something out
ff rather
seems more like .. that way, in an abstract sense 06:14
bacek TimToady: thanks for clarification
06:15 shinobi-cl joined, eternaleye joined
bacek hmm. So, ^ff technically becomes just ff after first flip? 06:17
or I misunderstand last sentence? 06:18
TimToady ff ^ff ff^ and ^ff^ all start and stop on the same elements, but the front ^ forces false if front of run, while back ^ forces false at end 06:19
bacek for (1..5) -> { $_ 1 ff 1 } will have 5 flipflops. Each of them true. 1 ^ff 1 will have 5 falses? 06:20
TimToady it still finds the same 5 ranges, but the first element of each is false 06:21
bacek ok, thanks.
SamB perl6: (1..5).map: { $_ if 1 ^ff 1 }
p6eval rakudo a0a390: OUTPUT«Could not find non-existent sub ff␤current instr.: '_block25' pc 156 (EVAL_20:74)␤»
..elf 25338: OUTPUT«Can't call method "map" without a package or object reference at (eval 120) line 3.␤ at ./elf_h line 4346␤»
..pugs: OUTPUT«*** No such subroutine: "&ff"␤ at /tmp/unfFboqXAA line 1, column 21-29␤»
bacek SamB: it is not implemented yet :)
TimToady maybe ff should smartmatch against $_ 06:22
06:22 Tene_ left
SamB p6eval: (1..5) 06:22
bacek SamB: infix:ff
SamB bacek: I figured ...
perl6: (1..5) 06:23
p6eval elf 25338: RESULT«"1E0"␤»
..rakudo a0a390: RESULT«1..5»
..pugs: RESULT«(1, 2, 3, 4, 5)»
SamB perl6: (1..) 06:24
p6eval elf 25338: OUTPUT«/home/evalenv/pugs/misc/STD_red/match.rb:117:in `block in to_dump0': undefined method `to_dump0' for nil:NilClass (NoMethodError)␤ from /home/evalenv/pugs/misc/STD_red/match.rb:117:in `map'␤ from /home/evalenv/pugs/misc/STD_red/match.rb:117:in `to_dump0'␤ from
../home/evalenv/pugs/...
..pugs: OUTPUT«*** ␤ Unexpected end of input␤ at /tmp/L7RBTutg66 line 1, column 6␤»
..rakudo a0a390: OUTPUT«Statement not terminated properly at line 1, near "..)"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
SamB perl6: (1..*)
p6eval rakudo a0a390: OUTPUT«Use of uninitialized value␤»
..elf 25338: OUTPUT«Undefined subroutine &GLOBAL::whatever called at (eval 121) line 3.␤ at ./elf_h line 4346␤»
..pugs: RESULT«pugs: out of memory (requested 1048576 bytes)␤»
SamB perl6: (1 .. *)
p6eval rakudo a0a390: OUTPUT«Use of uninitialized value␤» 06:25
..elf 25338: OUTPUT«Undefined subroutine &GLOBAL::whatever called at (eval 121) line 3.␤ at ./elf_h line 4346␤»
..pugs: RESULT«pugs: out of memory (requested 1048576 bytes)␤»
SamB perl6: (1 .. */2)
p6eval rakudo a0a390: OUTPUT«Use of uninitialized value␤»
..elf 25338: OUTPUT«Undefined subroutine &GLOBAL::whatever called at (eval 121) line 3.␤ at ./elf_h line 4346␤»
..pugs: RESULT«pugs: out of memory (requested 1048576 bytes)␤»
bacek std: (1 .. */2) 06:26
p6eval std 25338: OUTPUT«ok 00:02 33m␤»
bacek How that?
SamB what was that ?
SamB was just making that up
SamB didn't think it would work
> [0..]
lambdabot [0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,...
06:26 eternaleye left 06:27 eternaleye joined
bacek I didn't expect "*/2" to be parsed... 06:28
SamB perl6: <<0..*>> 06:29
p6eval elf 25338: RESULT«["0"]␤»
..pugs, rakudo a0a390: RESULT«"0..*"»
06:30 kanru joined
SamB perl6: 0..* 06:30
p6eval rakudo a0a390: OUTPUT«Use of uninitialized value␤»
..elf 25338: OUTPUT«Undefined subroutine &GLOBAL::whatever called at (eval 121) line 3.␤ at ./elf_h line 4346␤»
..pugs: RESULT«pugs: out of memory (requested 1048576 bytes)␤»
pugs_svn r25339 | lwall++ | [STD] provide guidance on (1..) 06:34
SamB std: (1..)
p6eval std 25338: OUTPUT«############# PARSE FAILED #############␤Unable to parse parenthesized expression; couldn't find final ')' at /tmp/2NVonsCrgd line 1:␤------> (1..)␤ expecting any of:␤ prefix or noun␤ whitespace␤FAILED 00:02 33m␤»
SamB perl6: (1..Inf).perl 06:37
06:37 alester left
p6eval elf 25339: RESULT«"\"1E0\""␤» 06:37
..rakudo a0a390: RESULT«"1..Inf"»
..pugs: OUTPUT«pugs: out of memory (requested 1048576 bytes)␤»
TimToady of course */2 parses, it just doesn't mean anything that a range can use 06:38
1 .. */2 means 1.. { $^x / 2 } 06:39
which probably doesn't find a matching signature
SamB perl6: $^x
p6eval rakudo a0a390: OUTPUT«too few arguments passed (0) - 1 params expected␤current instr.: '_block14' pc 40 (EVAL_10:28)␤»
..pugs: OUTPUT«*** Undeclared variable: ("$^x",MkPad (padToList [("$_",PELexical {pe_type = (mkType "Scalar"), pe_proto = <Scalar:0xb6615e91>, pe_flags = MkEntryFlags {ef_isContext = True}, pe_store = <ref:0xb662816d>}),("@_",PELexical {pe_type = (mkType "Array"), pe_proto = <Array:0xb661c725>,...
..elf 25339: OUTPUT«Global symbol "$x" requires explicit package name at (eval 118) line 3.␤ at ./elf_h line 4346␤»
SamB perl6: * 06:40
p6eval elf 25339: OUTPUT«Undefined subroutine &GLOBAL::whatever called at (eval 119) line 3.␤ at ./elf_h line 4346␤»
..rakudo a0a390: RESULT«*»
..pugs: RESULT«Inf»
TimToady and now you see why pugs blows up on 1..* 06:41
SamB er, not really ;-P
TimToady it just always translate * to Inf
*translates
SamB it's written in Haskell ;-P
TimToady that's no excuse :)
std: (1..) 06:42
SamB no, I mean it should handle lazy stuff even with stuff like that
p6eval std 25339: OUTPUT«############# PARSE FAILED #############␤Please use ..* for indefinite range at /tmp/XWp4ZBHl51 line 1:␤------> (1..)␤FAILED 00:02 33m␤»
TimToady there you go
SamB std (1..*)
std: (1..*)
p6eval std 25339: OUTPUT«ok 00:02 33m␤»
pugs_svn r25340 | bacek++ | [spec] Fix flipflops tests. TimToady++ for help
06:43 eternaleye left
pugs_svn r25341 | lwall++ | [flip-flop.t] fix parsefail 06:44
bacek oops. 06:46
It successfully parsed by rakudo... 06:47
TimToady rakudo isn't the normative parser :) 06:48
bacek TimToady: :)
bacek still don't understand fudged test in flipflop.t... 06:49
SamB so how come STD doesn't return a syntax tree ?
bacek SamB: it is very big.
TimToady that fudges all of the tests
it does return a syntax tree
we choose not to send it to irc 06:50
SamB oh
bacek TimToady: "that fudges all of the tests"? 06:51
TimToady but you can cd src/perl6; make; and then run viv to see the tree
bacek bacek@icering:~/src/pugs/src/perl6$ ./try5 TOP -e '(1..*)' | wc -l
1397
TimToady a lot of those lines are the CORE symbol table these days 06:52
which we wouldn't have to attach to the top node, but we do
bacek TimToady: question again about flipflop :) Is in current S03/flipflop.t supposed to be '&always_true' and '&always_false'? 07:17
07:19 shinobi-cl left
TimToady no, the way it's written is correct 07:21
in fact, &always_true without () would be wrong
though &always_true() would work
07:23 digitalicarus_aw left
bacek but those functions invoked once only... 07:24
(in way how they written now)
07:25 kate21de joined
TimToady they're invoked repeatedly 07:27
I don't know why the tests use those functions though, unless they were worried about p5 semantics re line numbers
maybe the tests were copied out of the p5 test suite? 07:28
bacek no idea.
but why they invoked repeatedly?
07:29 SamB left
TimToady multiple tests, and the tester function has a loop in it 07:30
07:31 Eevee left
bacek hm... 07:31
Looks like mine implementation of ff is incorrect. I capture "flip" and "flop" operands... 07:33
07:36 Nom- left 07:37 Nom- joined
pugs_svn r25342 | bacek++ | [spec] Add more tests for flipflop. TimToady++ once again 07:44
bacek TimToady: can you check last change to flipflop.t. Is it correct test? 07:45
TimToady looks pretty good, though my brain is too frazzled to run it in detail 07:46
looks like a good way to test that the condition is tested even if false is returned 07:47
bacek no worries :)
pugs_svn r25343 | lwall++ | [STD] better diagnostics on if()
bacek And I failing those tests... :(
TimToady std: given(42) {...} 08:05
p6eval std 25343: OUTPUT«############# PARSE FAILED #############␤given() interpreted as function call at line 1 ; please use whitespace after keyword␤Unexpected block in infix position (two terms in a row) at /tmp/tr56HQeeRB line 1:␤------> given(42) {...}␤ expecting any of:␤ infix or
..m...
pugs_svn r25344 | bacek++ | [spec] Refactor flipflop.t test for readability 08:12
08:34 iblechbot joined 08:43 cognominal joined, kanru left, cognominal_ joined 08:45 mberends joined 08:49 cognominal left 09:00 c9s_mv joined
pugs_svn r25345 | lwall++ | random test cleanup 09:12
mberends TimToady: I cannot figure out in what time zone your brain is ticking 09:15
09:23 DemoFreak joined 09:34 zamolxes joined 09:37 barney joined 09:47 ejs joined 09:49 chid joined 09:50 M_o_C joined 09:55 hanekomu joined 09:59 meppl joined 10:09 c9s_mv left
bacek perl6: my @a=(1..3); say ~(1..@a) 10:17
p6eval pugs, rakudo a0a390: OUTPUT«1 2 3␤»
..elf 25345: OUTPUT«Out of memory!␤»
bacek perl6: my @a=(1..3); say ~(1..^@a) 10:18
p6eval pugs, rakudo a0a390: OUTPUT«1 2␤»
..elf 25345: OUTPUT«Out of memory!␤»
bacek perl6: my @a=(1); say ~(@a..3) 10:20
p6eval pugs: OUTPUT«1 2 3␤»
..rakudo a0a390: OUTPUT«Multiple Dispatch: No suitable candidate found for 'cmp', with signature 'PP->I'␤current instr.: 'parrot;Range;!to_test' pc 8402 (src/classes/Range.pir:391)␤»
..elf 25345: OUTPUT«No viable candidate for call to multimethod prefix__126() at (eval 122) line 4␤ at ./elf_h line 4346␤»
10:30 simplex joined 10:33 kane__ joined 10:36 maerzhase joined
bacek perl6: say 1 does Positional 10:38
p6eval rakudo a0a390: OUTPUT«1␤»
..pugs: OUTPUT«*** Unknown syntactic construct: Syn "does" [Val (VInt 1),App (Var "&Positional") Nothing []]␤ at /tmp/cHUaXhqCdO line 1, column 5 - line 2, column 1␤»
..elf 25345: OUTPUT«Undefined subroutine &GLOBAL::infix_does called at (eval 120) line 3.␤ at ./elf_h line 4346␤»
bacek perl6: say "a" does Positional
p6eval elf 25345: OUTPUT«Undefined subroutine &GLOBAL::infix_does called at (eval 122) line 3.␤ at ./elf_h line 4346␤»
..pugs: OUTPUT«*** Unknown syntactic construct: Syn "does" [Ann (Cxt (CxtItem (mkType "Str"))) (Val (VStr "a")),App (Var "&Positional") Nothing []]␤ at /tmp/Aj3EXzyvKz line 1, column 5 - line 2, column 1␤»
..rakudo a0a390: OUTPUT«a␤»
bacek perl6: say ("a" does Positional)
p6eval elf 25345: OUTPUT«Undefined subroutine &GLOBAL::infix_does called at (eval 122) line 3.␤ at ./elf_h line 4346␤»
..rakudo a0a390: OUTPUT«a␤»
..pugs: OUTPUT«*** Unknown syntactic construct: Syn "does" [Ann (Cxt (CxtItem (mkType "Str"))) (Val (VStr "a")),App (Var "&Positional") Nothing []]␤ at /tmp/1Q6YaTKsWa line 1, column 5 - line 2, column 1␤»
bacek hmm....
perl6: say "a".does(Positional) 10:39
p6eval pugs: OUTPUT«*** No such subroutine: "&Positional"␤ at /tmp/EmKz3JNDGF line 1, column 14-24␤»
..rakudo a0a390: OUTPUT«0␤»
..elf 25345: OUTPUT«␤»
bacek perl6: say 42.does(Positional)
p6eval pugs: OUTPUT«*** No such subroutine: "&Positional"␤ at /tmp/ULrRDhKx3f line 1, column 13-23␤»
..rakudo a0a390: OUTPUT«0␤»
..elf 25345: OUTPUT«␤»
bacek perl6: say 42.does(Int) 10:40
p6eval pugs, rakudo a0a390: OUTPUT«1␤»
..elf 25345: OUTPUT«␤»
bacek perl6: say "a".does(Int)
p6eval rakudo a0a390: OUTPUT«0␤»
..elf 25345, pugs: OUTPUT«␤»
bacek perl6: say "1".does(Int)
p6eval rakudo a0a390: OUTPUT«0␤»
..elf 25345, pugs: OUTPUT«␤»
10:42 simplex left 11:05 sri_kraih joined 11:10 araujo joined 11:35 kane__ left 11:44 kate21de left 11:48 masak joined
mberends hi masak, I'm interested in doing some work on proto today 11:54
masak mberends: cool! you have my blessing. 11:55
I'd prefer to stand by and provide opinions, if that's ok.
I'll try not to get in the way of your creativity, though.
mberends this rather goes against your earlier expression of desire to do it yourself, or was that meant (at least partly) in jest? 11:56
(why am I asking? I have his blessing!) 11:57
masak what I meant was that I looked forward to doing it myself, and wasn't primarily fishing for collaborators, like with November and Druid. 11:58
but I won't say no to help either. :)
mberends heh
masak if nothing else, you have the README to guide you as to my plans.
mberends ok, I'll conform to that. please pull my recent contribution for a start. 11:59
12:00 maerzhase left
mberends I have created a new 'proto' user account on my system to test starting from scratch. 12:00
masak cool. 12:03
masak pulls
mberends aka 'proto' clones 12:05
masak jnthn: didn't answer your question 'proto hacking?' yesterday, sorry. this is it. github.com/masak/proto 12:06
12:07 alanhaggai joined, alanhaggai left 12:08 alanhaggai joined 12:09 maerzhase joined 12:13 kate21de joined
masak mberends: no, env is not redundant and export doesn't work in this case. 12:18
mberends ok, your version still stands
masak removing your comment. 12:19
re conditionally doing something with the log; no, I only plan to refer to it if the make fails.
just as in the proto shell script.
I like the unlink() version, uncommenting it.
12:20 cognominal_ left
masak also removing the commented-out 'say' above the 'warn'. 12:20
12:20 cognominal_ joined
masak in summary, thanks. it's great to be reviewed. 12:20
mberends 'source proto' here b0rked on the parrot ftp download. why not http instead?
masak sure.
also, I'm not sure about 'source proto' any more. 12:21
better to let the user manually set environment variables, despite the slight annoyance of doing so.
we can give them the option of eval-ing a short file with the correct values in them, perhaps. 12:22
mberends good idea
masak I think about such things on my bike. :)
mberends Or in the shower :)
12:23 alanhaggai left
masak that, too. or in the pool. 12:23
12:23 alanhaggai joined
masak in the future, when improving crap I wrote, please just remove it rather than comment it out. if I want it back, I'll hit the logs instead. 12:24
mberends ok. a leave-behind script could also offer to alias perl6
jnthn masak: proto looks helpful :-)
masak mberends: aye, good.
jnthn: it's just a prototype. :)
but it's meant to soften the coming deluge of interdependencies on github. 12:25
jnthn *nod*
mberends git push please, then I'll pull and s/ftp/http/
masak one minute. 12:27
still editing.
mberends 59, 58, 57, 56 ...
masak pushed.
mberends pulled. 12:28
12:30 alanhaggai left, alanhaggai__ joined 12:31 kane__ joined
mberends the parrot ftp download actually succeeded, but the script thought it failed in the if (($? != 0)) immediately after. Perhaps wget is funny with exit codes. 12:31
masak either that, or my condition is funny somehow.
could you run wget manually and check the exit code after? 12:32
i.e. wget ...; echo $?
mberends will try
with another url the exit status is 0, which is what we want 12:33
.oO shell comparison operators are the wrong way round. We may need ne instead of != 12:34
masak with another url? but not with the current one?
mberends: aye, but (()) means the evaluation is arithmetic. 12:35
mberends I was just saving time for the test, didn't need 5 megabytes.
masak ah.
mberends so (( != )) === [ ne ] ? I want to check and learn that. It seems so inconsistent. 12:36
masak I'll see if I can find chapter and verse in the bash manpage. 12:37
aye, go 'man bash' and grep (/) for ^ARITH
12:38 ihrd joined
mberends oh, and an offline thought was, cast the proto script language into Bourne subset, that way AIX users with Korn shell etc are not left out. 12:38
mberends greps
ihrd hi there
masak ihrd: HI!
mberends: I have an unfinished commit to bring proto down to sh level.
it's hard.
mberends worth it imo. 12:39
what's the blocker?
masak bah, the usual. time. 12:40
mberends I have today.
masak I'll test-run it and see if the patch works. if so, I'll commit it. 12:41
mberends cool
masak oh! seems I use a bash-only string prefix test on line 53. 12:42
hm...
mberends that is !( 12:43
(what a face)
masak yup, now it works. (except for that tricky case.) 12:49
committing.
I also discovered a new case which we need to test for -- adding that. 12:51
mberends what's nice about the bash manual is that everything is in there. what's not so nice about the bash manual is that it is so long. go figure.
masak it's the case where the user has Parrot installed via svn, and uses Rakudo in the languages/rakudo dir, not in languages/perl6
mberends: bash is really well documented. and very widespread. on some level though, it's insane. 12:52
12:52 aindilis` left 12:53 aindilis` joined
ihrd rakudo: class A { has @.foo; method bar { @.foo[0] ="Hey"; say @.foo[*-1] } }; A.new.bar 12:54
p6eval rakudo a0a390: OUTPUT«Hey␤»
ihrd looks like 61766 resolved 12:55
12:55 drbean left
masak yup. and so it says, too. 12:56
jnthn rakudo: my @a = <a b c>; for @a -> { .say }
p6eval rakudo a0a390: OUTPUT«Use of uninitialized value␤␤Use of uninitialized value␤␤Use of uninitialized value␤␤»
jnthn rakudo: my @a = <a b c>; for @a { .say }
p6eval rakudo a0a390: OUTPUT«a␤b␤c␤»
13:02 Whiteknight joined
mberends masak: proto successfully downloaded parrot, now building. spurious message: -su: test: too many arguments, but seems unimportant 13:08
masak does it have a line number? 13:09
I got that error on line 51, so I kept that test as it was.
mberends no line number (there is probably a trace option for something like that). np 13:10
Building Perl 6... 13:11
done. env includes PARROT_DIR=/home/proto/proto/parrot-0.9.0 13:13
masak cool. 13:14
mberends ./proto install perl6-examples 13:17
Downloading perl6-examples...
Building perl6-examples...
done.
masak nice, isn't it? :)
mberends oh yes! 13:18
now we just have to port the decent parts of CPAN ;)
masak ruhh, don't talk like that. :P 13:19
we're _not_ a CPAN substitute.
just a tiny, insignificant, hyper-lightweight package manager for the dark dawn between CPAN and CPAN6.
mberends Running Right Now is what matters. TODO? 13:20
masak sorry, what's the question?
the TODO right now is the things in the README below the "not yet impl" notice. 13:21
if you want something fun to do today, we could branch November, and remove HTML::Template from the branch. 13:22
then we create a new github repo html-template, and create a dependency from November to this new repo.
mberends and do it without hardcoding HTML::Template's address 13:23
masak right. 13:24
then (1) make sure proto recofnizes that the November branch needs the external project, and (2) merge to master and make sure November won't install without first installing the external project.
13:24 alanhaggai__ is now known as alanhaggai
masak that sounds like about a day's work. 13:24
and the fun part will be handling the deps.proto file format.
mberends and the recursion is the fun
masak mberends: indeed. :)
it'll have to be download, check, download, check ..., build, build, build 13:25
mberends did you draft a deps.proto file layout?
masak sure in the README. 13:26
s/ /, /
it's dead-simple by design.
mberends those are the best
masak I hope so.
we don't need much right now. 13:27
an obvious thing that I consciously ignore is versioning.
proto is meant to make life easier for the end user, but not _that_ much easier :P
when Rakudo has versioning, maybe we'll build it into proto as well.
en.wikipedia.org/wiki/Directed_acyclic_graph 13:28
mberends ah yes, used that once in project planning, critical path as well 13:31
masak to me, the module deps form a DAG.
(which raises the concern that it might not be; we have to detect cycles) 13:32
mberends initially we can detect and avoid manually, such a feature can be retrofitted like deadlock detection in a database server. 13:33
masak mberends: I have limited time today, but I'll do this: I'll create the html-template project on github, and a November branch that doesn't have HTML::Template.
mberends: I think automatic detection is not that hard. just follow all deps and make sure you don't get yourself.
but if you don't want to write the code, I'll gladly do it. :) 13:34
mberends okay, you start it and I'll take over when you must go. I have tomorrow and Tuesday available as well.
masak oh, I thought we might parallelize in the way outlined above.
I'll do the November part and you do the proto part. 13:35
if you want.
it's not so much that I must go, it's that I have at least three other writing tasks today. I'll happily remain online while doing those, but I won't be able to actively participate in programming. 13:36
ihrd I can do so for URI.pm 13:37
masak ihrd: good idea. please also do it in a branch.
ihrd masak: okay
masak ihrd: can I try and move everything into p6w/lib in master before we begin?
I've been wanting to do that for a long time.
ihrd mb just /lib? 13:38
do we realy need p6w dir?
masak ihrd: and remove p6w entirely? yes, I've been thinking about that too.
ihrd: agreed. we don't.
making November more accessible by removing useless directories :) 13:39
I'll just put it all in the root dir.
ihrd masak: mb #november-wiki is better choice for this disscussion
masak sure.
13:39 drbean joined 13:44 icwiener joined 13:45 barney left 13:46 aindilis` left
mberends masak: each line in deps.proto should have two words, not one: the github account name, then the project name. In fact a third field would help distinguish github from other stores likve svn sites. 13:46
masak ok.
I propose this order: site accname project 13:47
mberends ok. three fields for lots more minefields
masak and only 'registered' sites are allowed, and only github is registered so far.
in the long run (if there's such a thing for proto), we should use YAML or JSON or similar for deps.proto 13:48
mberends yes
13:50 barney joined 13:58 mikehh joined 13:59 rhr_ joined 14:04 shinobi-cl joined 14:05 cai joined 14:07 gravity joined 14:12 DJ-DONGDOT joined 14:13 rhr left
masak mberends: on second thought, this account name thing is very github-specific. 14:26
it might feel like a nice thing to parametrize on, but it might also give us more trouble than it's worth. 14:27
given that proto must have a list of all its known installable projects anyway, along with a list of project specifics, I'd advocate just the one-word-per-line approach.
unless you or someone else has a good counter-argument. 14:28
14:29 barney left, chid left 14:30 nihiliad left
mberends the counter is that it's easy to add a little wiggle room now, or more work to change it later. understood that account may be meaningless for some repos, but then it can have a dummy value like '.' 14:34
14:34 nihiliad joined
mberends it does not feel right to make proto github-centric either. 14:35
masak well, it does not really address the fact that this info is stored anyway elsewhere in proto.
mberends: proto already isn't github-centric.
take a look at how things are installed in installer. 14:36
no, the three-word specifier goes against the simplicity of proto. 14:37
it makes the identifier github-viklund-november instead of just november
that's too close to versioning for me.
mberends the installcmd is too hard coded for me
masak hard coding is the objective of proto! :)
it's its only goal in life. 14:38
think of it this way: proto needs a data base of where to find its modules.
that database just happens to be a hash in installer at present.
the idea here is to put as little changeable information out in the projects as possible. 14:39
mberends I plan to fly some of my code out of the eric256 nest one day, and would like to declare that in deps.proto
masak good, good.
14:39 Tene_ joined
masak but is that an argument for a three-word deps.proto format? 14:41
oh, you plan to give the repo the same name? :)
well, proto doesn't support that.
or rather, you'd have to give the project a different name in proto.
mberends the project name can be the third part of a three part name, afaics 14:42
masak mberends: the difference between our opinions right now lies in our different expectations on deps.proto 14:43
mberends there is another database entity that would be nice to add to the design: a single list of all projects offered by an account holder. 14:44
masak it's important to me that deps.proto is independent of the actual locations of projects.
that way, if a project changes location, this fact can be updated centrally in the proto repo.
mberends maybe I should be planning proto+1
masak mberends: re list of all projects by an account holder: sure, why not? but that is IMHO less vital than handling deps. 14:45
I'm about to upload a html-template project to github, by the way.
mberends nice. was it a difficult siamese separation? 14:46
masak well, no. just straight copying. 14:47
mberends tries to square the circle of differing expectations
masak github.com/masak/html-template/
now preparing the November branch...
mberends square-circle suggestion: deps.proto has optional 1-field and 3-field entries, and proto handles 1-word and 3-word arguments (somehow). one day should still suffice. 14:50
mberends doesn't think masak will like the suggestion 14:51
masak no, you don't understand :)
the _point_ of the deps.proto file is that it should be location-independent.
that way, a single central change is enough when location changes.
your suggestion destroys that advantage. 14:52
14:52 alanhaggai left
masak the square-circle patch complicates things, and destroys the advantage. 14:52
14:52 Tene left
mberends I agree that central control is good for launching the core, but disagree about optional modules. 14:53
masak sorry, what was the optional modules part again?
mberends it's like hosts files versus DNS. the system should allow contibutors to extend the addressable field independent of a central authority. 14:54
and proto is a root authority. 14:55
masak well, now it sounds to me that you should be desigining proto+1 :)
I just want to make this one thing work. nothing more.
mberends it works. if it works, change it! 14:56
masak realizes the difficulties in keeping a project down-to-earth
mberends: ok, here you go: github.com/viklund/november/tree/ht...plate-gone
try and check out this branch, and 'make test' in it. 14:57
all tests should pass except the integration tests (because HTML::Template is missing)
mberends ok, will do. today I'll stay within your scope, and get dependencies working within a centrally managed database.
masak mberends++ 14:58
mberends otherwise we'll talk until tomorrow and have no code to show for it.
masak you're a wise man, man.
mberends pulled. but no deps.proto in the root dir. 14:59
masak checking. 15:00
did you check out the branch?
mberends in november, I suppose now that you meant HTML::Template
masak there's a branch in the november project on github, called html-template-gone 15:01
that's where I removed HTML::Template and added deps.proto
mberends ah, the branch ;)
masak aye. 15:02
we try to keep experiments out of master if we can.
15:02 maerzhase left
mberends the github.com/viklund/november/tree/ht...plate-gone page has a public clone url of git://github.com/viklund/november.git 15:05
masak 'git checkout html-template-gone' on the console? does that work? 15:06
15:07 literal left, avar left
mberends error: pathspec 'html-template-gone' did not match any file(s) known to git. 15:07
masak strange. 15:08
'git pull origin html-template-gone' ?
15:08 Whiteknight left
mberends worked. 15:09
masak 'git checkout -b html-template-gone origin/html-template-gone' 15:10
masak is just making stuff up as he goes along 15:11
mberends is a git-n00b desperately in need of git-fu
masak it's a bit zen-like insofar that there's less to understand than it might at first appear. 15:12
some cup-emptying required.
15:12 literal joined, avar joined
mberends emptying another beer atm 15:13
masak that might help.
mberends watching git-speak here is pleasant and fascinating 15:14
masak same way I feel about Haskell. 15:16
mberends btw, it looks like Nordic Perl Workshop are not interested in my Pod talk proposal, but I'm thinking about attending anyway.
masak too bad and excellent, respectively. 15:18
mberends looks forward to the hackathon as well
masak too 15:19
jnthn three 15:21
masak mberends: I'm going to re-focus on my writing tasks for a while now.
15:21 |MoC| joined 15:22 broquaint left, broquaint joined
masak mberends: probably won't have time to fiddle with proto today, but I'll eagerly review and discuss any advances you make. 15:22
mberends masak: ok, after the Ireland-Italy rugby game I'll start hacking 15:23
masak :)
15:24 jan_ left 15:25 maerzhase joined 15:28 alanhaggai__ joined 15:29 alanhaggai__ is now known as alanhaggai 15:34 M_o_C left, |MoC| is now known as M_o_C, DJ-DONGDOT left 15:36 nihiliad left, nihiliad joined 15:46 ihrd left
masak walk & 15:46
16:07 Limbic_Region joined 16:14 maerzhase1 joined, pmurias joined 16:15 maerzhase left 16:18 chid joined 16:34 jan_ joined 16:45 dominiko joined
dominiko What is a good site to learn about perl6? 16:45
masak dominiko: perlgeek.de/blog-en/perl-5-to-6/ 16:46
16:46 kane__ left
dominiko Thanks, I already saw that one, it's pretty good for those who know perl5 16:46
masak ah, you're new to Perl? 16:47
dominiko Does not go into details, but it's simple and good enough to whet my appetite about Perl6
masak this seems good: en.wikibooks.org/wiki/Perl_6_Programming
dominiko I know perl5
masak but it's far from complete.
dominiko Thanks, following your link...
masak dominiko: if you want depth, I'd recommend the specs: perlcabal.org/syn/ 16:48
I don't think something with the same depth as the specs exists.
OTOH, the specs are a bit terse and do not always spell out the consequences of design decisions.
dominiko the wiki link seems quite descent 16:49
masak good. best of luck. 16:50
if you still have questions after reading that, please don't hesitate to ask here.
16:51 eternaleye joined, kate21de1 joined
dominiko The examples use parentheses in the if (...) and unless (...) which is not very perl6ish 16:51
masak indeed.
I removed some of them in a recent read-through of that wiki. 16:52
not sure if those edits were accepted, though.
s/some/many/
dominiko @a = (1, 2, 3); in wiki is also perl5ish 16:56
lambdabot Maybe you meant: activity activity-full admin all-dicts arr ask . ? @ v
masak aye.
some people seem to like the parens even when told they are not necessary. 16:57
dominiko yes, I have to admit, I like them too. I makes it clearer. Also, using vim, it allows to use % to jump to the end of the array. 16:58
But I've never written anything in Perl6 so I don't know enough to have a strong opinion 16:59
masak I'm not sure a strong opinion is required in this case. both ways work, most people can read both ways easily, and they are not very different. 17:00
mberends is 50/50 on that, but the pro-paren half seems, er, uncool 17:01
dominiko The thing I like about Perl is the { } ( ) $ # ! etc. I find that python looks too much like text.
mberends do you have a C or java background? 17:02
dominiko Having { ... } is attracts more to my visual brain
C
C & c++
mberends recommends being open to both styles for a while, following the path of least resistance. 17:04
masak mberends: it would be really nice if there were some way to assert in the Makefile of projects whether PERL6LIB points to the lib/ directory in that project.
I know almost enough about Makefile things to make that happen, but not quite. 17:05
maybe it's not possible.
mberends good suggestion. my make-fu is stronger than my git-fu. 17:06
masak together, we shall rule the known universe.
17:06 mikehh left
masak er, I mean, good. 17:07
mberends well corrected :)
17:08 kate21de left, avar left
dominiko being able to avoid parentheses will help writing poems in Perl6 17:09
17:10 aindilis joined
masak dominiko: I tried to write a Perl 6 test file as poetically as possible some days ago. 17:12
mberends dominko: now that's a benefit some people never even imagine
masak the tests themselves came out like this:
assert :that($board), :yields<b1>, :desc('Go for the missing link');
not as poetic as I'd imagined.
mberends masak++ 17:13
masak maybe I should have called 'desc' 'because' instead.
mberends maybe you will, in the next version 17:14
masak sure. since it's a local branch, I'll change it now.
all in all, it feels a bit Smalltalkish. 17:15
(which I kinda like)
jnthn see :if($board), :is(1), :to-know('It goes for the missing link'); 17:24
;-)
masak heh :)
to each his own, I guess.
jnthn is($board, 1, 'It goes fro the missing link'); # is fine to me :-) 17:25
masak jnthn: you'll note the difference between 'is' and 'yields', I hope.
jnthn I like making tests *work*, not writing tests themselves.
masak there's a transformation function hidden in there.
17:25 eternaleye left
jnthn masak: No, that wasn't obvious to me, though the choice of word made me curious... 17:25
...so not having to write too much is a Good Thng. 17:26
*Thing
masak yes, that was part of the idea. hiding away the complexity.
not sure if I'll keep that syntax, though. it's basically just another rehash of Test::InputOutput (in November), which is a rehash of Test::Base (on CPAN) 17:28
17:28 eternaleye joined
jnthn eats a bar of Swedish Plopp - his favorite named chocolate bar. 17:29
masak :)
17:30 cai left, bacek left
masak www.flickr.com/photos/richardgin/1430394351/ 17:31
jnthn Yes, but mine is "med smak av tutti-frutti" :-) 17:32
17:32 shinobi-cl left
masak 哈哈 17:32
jnthn I'm guessing smak must mean flavor. :-)
17:32 hanekomu left
masak jnthn: aye. 17:32
17:32 hanekomu joined, bacek joined
masak so let me see if I got this straight. it's a chocolate bar... with taste of tutti-frutti, some kind of fruity flavour? 17:33
how do you even get chocolate to taste fruity?
masak adds a trip to grocery store today to find out 17:34
jnthn masak: I'm not sure, but they managed it some how... 17:38
17:38 shinobi-cl joined
jnthn It doesn't taste too bad. 17:38
masak jnthn: is it runny inside?
jnthn Better than the name might lead you to think, in fact.
Yes.
masak that explains it for me.
mberends sniggers that dominiko has not realised that Perl 6 is the path of least resistance 17:42
jnthn masak: My sudden up-surge in Swedish vocabularly is mostly due to one of my clients haven given me a task of adding some features of a system...and all of the comments are written in Swedish! 17:43
s/of/to/
s/haven/having/ ... # English fail.
Normally I read the comments to figure out what the code means...it's odd doing it the other way around. ;-)
masak I can imagine. 17:44
mberends: my writing is going very slowly. how's the rugby match/proto hacking going? 17:45
mberends Ireland 38, Italy 9, and unremarkable walkover. I'm currently pondering the git branch issue with regard to %project_info 17:50
17:51 SamB joined
mberends git checkout seems to be designed to operate within a clone, so it's two separate git invocations to get to the branch. 17:53
17:53 [particle] left, [particle] joined
masak mberends: is there a known case where proto will need to check out a branch? 17:57
17:57 eternaleye left
mberends it depends ;) 17:57
17:57 eternaleye joined
mberends on you ;) 17:58
masak mberends: as far as I know, you can do two git invocations by doing && or ; in the string.
mberends: I'd say leave it for now, since it doesn't seem vital for our existing git projects.
mberends ok, I'll live with that during development and hack it out after proof of concept.
masak mberends: in the future, I'll want to at least support checking out git repos with the commit bit set, to make it easier for people to push. 17:59
not sure about the branching thing.
mberends not sure about the commit thing
masak to me, it'd be sufficient to just enter the directory after installation and switch branches manually.
mberends: oh, it'll just be a service to those who check out their own projects with proto. 18:00
they'd want them to be writeable, so to speak.
the fix might be as easy as trying the rw repo first, and then falling back on the readonly variant.
(but even pondering that shows how the hard-codedness of installcmd might be a problem) 18:01
mberends all the git clone url's are currently public, meaning readonly.
masak right. 18:02
due to 'release early'.
mberends Many of us will remain readonly in many repos. 18:03
masak ah, you're thinking there might be a time penalty during installation?
I doubt that. failure is fairly rapid. 18:04
try and check out someone else's github repo, and you'll see :)
mberends yes, the error is quick. ok, git@github first, git://github second 18:05
masak but, still. hard-coded installcmd. 18:06
the repetition in that hash indicates that we could use your three-word suggestion to ease things up a little. :)
something like hub, account, project 18:07
then there's no problem creating both kinds of downloads.
mberends sorry, afk couple mins 18:10
mberends ponders again, reluctant to enter the hazy realm of 'scope creep' 18:20
18:20 c9s_mv joined, c9s_mv left
masak mberends: indeed. it is more difficult than first appears! 18:21
but I do think that breaking up installcmd into several keys is for the better. 18:22
even if we don't actually use that refactor to fix git@github downloads right now.
it just feels more orthogonal.
masak must go home and make food soon 18:26
18:29 eternaleye left 18:30 eternaleye joined, dominiko left 18:33 eternaleye left
masak mberends: ah! you need the 'check out a branch of a project' feature to check that November installs correctly. sorry, I'm a bit slow, it seems... :/ 18:33
maybe I should just merge that branch back to trunk. 18:34
18:35 Psyche^ joined 18:37 literal left, mberends_ joined 18:38 literal joined, mberends left
masak mberends_: network trouble? 18:38
mberends_ yes, not sure where. will re-connect. 18:39
18:39 mberends_ left, mberends joined
masak did you receive the thing I said starting with 'mberends: ah!' ? 18:40
mberends no, seeing it on perlgeek now 18:41
masak I can merge back to master if that makes things easier for you.
must go in a couple of minutes.
mberends leave master for now, I've worked around the branch temporarily 18:42
masak mberends++
looking forward to seeing what you come up with.
heading home to make dinner.
mberends good plan, we're going to Chinese restaurant later. 18:43
masak eet smakelijk. 18:44
mberends dank je wel :) 18:45
masak :)
masak waves
18:45 masak left 18:46 Patterner left, Psyche^ is now known as Patterner 18:54 bigal left 18:55 Tene joined 19:02 eternaleye joined 19:05 literal left 19:08 Tene_ left 19:09 p6paste left 19:19 literal joined
mberends Chinese restaurant & 19:25
19:33 eternaleye left
pmurias j 19:35
sorry
19:38 sri_kraih left 19:46 aindilis` joined 19:47 alester joined 19:49 pmurias left 19:54 alester left
bacek good morning 19:55
perl6: enum Day <Sunday Saturday>; say ~(Day::Sunday..Day::Saturday) 19:56
p6eval elf 25345: OUTPUT«Parse error in: /tmp/ORZuCtvFMG␤panic at line 1 column 0 (pos 0): Can't understand next input--giving up␤WHERE: enum Day <Sunday Saturday>; sa␤WHERE:/\<-- HERE␤ STD_red/prelude.rb:99:in `panic'␤ STD_red/std.rb:76:in `scan_unitstopper'␤ STD_red/std.rb:224:in `comp_unit'␤
..STD_r...
..rakudo a0a390: OUTPUT«Multiple Dispatch: No suitable candidate found for 'cmp', with signature 'PP->I'␤current instr.: 'parrot;Range;!to_test' pc 8402 (src/classes/Range.pir:391)␤»
..pugs: OUTPUT«*** No such subroutine: "&enum"␤ at /tmp/3LbgPZsVrH line 1, column 1-27␤»
19:56 jrockway left
bacek perl6: enum Day <Sunday Saturday>; say Day::Sunday == Day::Sunday; say Day::Sunday cmp Day::Sunday 19:58
p6eval pugs: OUTPUT«*** No such subroutine: "&enum"␤ at /tmp/QBDZWblzaK line 1, column 1-27␤»
..rakudo a0a390: OUTPUT«1␤0␤»
..elf 25345: OUTPUT«Parse error in: /tmp/gbvgx5smi6␤panic at line 1 column 0 (pos 0): Can't understand next input--giving up␤WHERE: enum Day <Sunday Saturday>; sa␤WHERE:/\<-- HERE␤ STD_red/prelude.rb:99:in `panic'␤ STD_red/std.rb:76:in `scan_unitstopper'␤ STD_red/std.rb:224:in `comp_unit'␤
..STD_r...
bacek summon masak for reporting bug :) 19:59
19:59 aindilis left 20:00 avar joined, jrockway joined 20:01 chid left 20:05 oskie joined, kane__ joined 20:07 jrockway left, jrockway joined 20:08 hanekomu left 20:13 alester joined 20:19 alester left 20:25 maerzhase joined 20:27 maerzhase1 left 20:29 alester joined 20:33 PZt left 20:34 jrockway_ joined 20:35 jrockway_ left 20:41 alester left 20:43 eternaleye joined 20:46 jrockway_ joined 20:47 icwiener left 20:55 jrockway_ left 20:57 jrockway left 21:02 eternaleye left 21:04 eternaleye joined 21:10 alanhaggai_ joined, rindolf joined 21:16 pmurias joined 21:20 shinobi-cl left 21:24 eternaleye left 21:25 Eevee joined 21:26 alanhaggai left, zostay left 21:27 justatheory joined 21:31 zostay joined 21:39 hanekomu joined 21:42 eternaleye joined
pugs_svn r25346 | putter++ | [elfparse] Added <![ ]>. std.pm: more hackish unpodding, and setup for emacs outline mode. With EXPR crowbarred, comp_unit accepts '3;'. 21:48
21:49 jrockway joined 22:05 shinobi-cl joined 22:06 shinobi-cl left, riffraff joined 22:23 rindolf left 22:25 riffraff left 22:30 eternaleye left 22:34 eternaleye joined 22:38 M_o_C left 22:55 mikehh joined 22:56 iblechbot left 22:57 kane__ left 23:21 aindilis` left 23:22 aindilis` joined 23:24 pmurias left 23:29 maerzhase left 23:36 ejs left 23:39 justatheory left 23:46 meppl left 23:54 bacek_ joined 23:55 bacek left 23:56 aindilis` left

Powered by Cro and the Raku® Programming Language.

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

Thank you!