»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by masak on 12 May 2015.
00:07 firstdayonthejob left 00:08 AlexDaniel left, pmurias_ left 00:11 adu left 00:17 Ven_ joined 00:20 AlexDaniel joined 00:25 Ven_ left 00:32 aborazmeh joined, aborazmeh left, aborazmeh joined 00:40 raiph left 00:42 espadrine left 00:58 perlawhirl joined, eliasr joined
dalek ast: 799f8f6 | TimToady++ | S32-exceptions/misc.t:
routine suggestions now leave off &
01:03
01:04 yqt left, yeahnoob joined
dalek p: 3fc7722 | TimToady++ | src/vm/moar/QAST/QASTOperationsMAST.nqp:
tab damage
01:10
p: 4888aa6 | TimToady++ | / (27 files):
Merge branch 'master' of github.com/perl6/nqp
p: c199c14 | TimToady++ | / (48 files):
Merge branch 'master' of github.com/perl6/nqp
p: 64f13fc | TimToady++ | / (8 files):
Merge branch 'master' of github.com/perl6/nqp
p: 237fd82 | TimToady++ | src/ (2 files):
Merge branch 'master' of github.com/perl6/nqp
p: 9cc4494 | TimToady++ | / (15 files):
Merge branch 'master' of github.com/perl6/nqp
p: 972432c | TimToady++ | src/QRegex/P6Regex/Actions.nqp:
Catch missing + or - between cclass elems
TimToady wow 01:11
dalek kudo/nom: c0fa4af | TimToady++ | tools/build/NQP_REVISION:
bump NQP to get missing + or - between cclass elems
01:13 finanalyst joined 01:19 finanalyst left 01:21 Ven_ joined
TimToady those merges are all just from doing periodic ". config.status" 01:25
maybe there's a way to tune that out somehow? 01:26
01:29 Ven_ left 01:31 leont left
dalek c: de9e7f3 | (Stéphane Payrard)++ | doc/Language/faq.pod:
tyop
01:32
psch TimToady: that usually happen when you're not on origin HEAD, then commit and pull after 01:37
TimToady: that's assuming less git-familiarity than you might have, though
TimToady: rebasing usually purges merge commits
the funny bit is that git log only shows one merge commit 01:38
mrf urgh Its far to early.
psch so, yeah, something is weird in there
psch shrugs 01:39
probably not really something i know enough about - neither it being early nor why there's so much merge commit spam 01:40
01:43 Actualeyes joined
mrf Its 1.45 in the UK. Thats early enough I think :D 01:44
01:48 tokuhiro_ joined 01:52 tokuhiro_ left
mrf m: say "hello" ~~ /l ** 0..2/ 01:53
camelia rakudo-moar 273e89: OUTPUT«「」␤»
mrf m: say "hello" ~~ /l ** 1..2/ 01:54
camelia rakudo-moar 273e89: OUTPUT«「ll」␤»
mrf ^ is that a bug? I would expect the 0**2 to match 2 l's as well
TimToady first match takes priority over longest, and it matches '' at the beginning 02:05
mrf TimToady: How come the 1..2 doesn't only match 1 l in the second example? 02:06
timotimo mrf: if you want that, you'd need frugal instead of greedy matching 02:08
yoleaux 2 Nov 2015 16:42Z <nine> timotimo: found another issue with the for range optimization. Now there's only one left (last/next support) till we have a clean spec test again :)
02:09 dayangkun joined
mrf timotimo: Yeah I am just confused why 0..2 appears as non greedy but 1..2 is greedy 02:09
timotimo it's not greedy vs non-greedy
it's what TimToady said. early match trumps all
so when it can match 0 times at the beginning, it's satisfied
m: say "lhello" ~~ / l ** 1..2 / 02:10
camelia rakudo-moar 273e89: OUTPUT«「l」␤»
timotimo er
m: say "lhello" ~~ / l ** 0..2 /
camelia rakudo-moar 273e89: OUTPUT«「l」␤»
timotimo you can see it greedily matches as many ls as it can. just turns out the most ls it could match at the beginning of "hello" is 0
mrf ...Right! so if it matches without having to step further along the string then it stops becuase its all ready matched 02:11
timotimo that's right
mrf Thats likely to have a rather large impact on my patch allowing ^x syntax 02:12
becuase 0..* will always match surely
on the plus side it means that the two failing tests in my patch are just erroneous tests. Which means its good to go 02:18
TimToady it's relatively useless for scanning, sure, but most parsing doesn't scan, but anchors each match at the end of the previous
mrf TimToady: Yeah. I can see its use. Just a lack of understanding on my part 02:19
02:20 sprocket joined 02:25 Peter_R left
mrf TimToady: Don't worry about that earlier gist. I have pushed a PR for NQP to support using 1^ and ^2 for the ** quantifier 02:28
02:28 xpen joined
awwaiid Hello. Are any of the caller semantics for EVAL documented/standardized? I want to use things like use caller(...).line and have it use the 'context' passed in, but so far no dice 02:29
02:29 kaare_ joined
awwaiid as in EVAL('callframe().file ~ ":" ~ callframe().line') vs EVAL('callframe().file ~ ":" ~ callframe().line', context => CALLER::) 02:29
mrf timotimo: You might be interested in that PR tbh 02:33
awwaiid I'm prrreetttyyyy sure the answer is "caller as a param for EVAL kinda works and is a rakudo-only feature" 02:34
02:38 BenGoldberg joined 02:45 atweiden joined
sprocket i’m writing a NativeCall binding to a struct whose structure differs on various platforms (ie. OS X vs Linux) 02:51
how can i write the definition so that the correct struct is used on different architectures?
at the moment i’ve got: if ($*KERNEL eq 'darwin') { … } else { …. }
but when i attempt to run, it complains that it “Could not locate compile-time value for symbol” 02:52
02:55 finanalyst joined 02:59 softmoth left 03:00 softmoth joined
BenGoldberg m: my uint8 $x = 1; $x +<= 7; say $x; 03:03
camelia rakudo-moar 273e89: OUTPUT«-128␤»
BenGoldberg m: my uint8 $x = 1; $x += 255; $x += 0; say $x; 03:04
camelia rakudo-moar 273e89: OUTPUT«0␤»
03:05 uruwi joined
dalek osystem: 23b1db8 | (Fayland Lam)++ | META.list:
Add Cache::LRU to ecosystem

See github.com/fayland/perl6-Cache-LRU
03:11
03:13 raiph joined, kid51 left 03:16 aborazmeh left
psch mrf: fwiw, 1^ looks really weird to me 03:17
m: say 1^
camelia rakudo-moar 273e89: OUTPUT«5===SORRY!5=== Error while compiling /tmp/D2Luq2j7sD␤Missing required term after infix␤at /tmp/D2Luq2j7sD:1␤------> 3say 1^7⏏5<EOL>␤ expecting any of:␤ prefix␤ term␤»
psch mrf: i assume it's short for 1 ^.. *
mrf psch: yes.
psch well, it's not like i can parse every syntax in perl6 without having to think about it... :) 03:18
mrf I am not married to it. Was just easier to implement than catching another special case 03:19
psch i don't really object, it's probably a useful enough shorthand
considering /a{M,}/ is gone 03:20
mrf It certainly seemed useful enough as a 'more than' quantifier
psch so having /a ** M^/ as non-whatever-or-infix:<..>-for is neat
to rephrase, having /a ** M^/ as shorthand for /a ** M..*/ seems useful enough, considering perl5 allowed leaving either end out 03:21
timotimo except M^ would be M+1..*
with parens
i am late for bed 03:22
psch huh
yeah, that
also i apparently misremembered leaving M out in {M,N}
perl5 doesn't allow that
timotimo "perl5 doesn't do it" isn't a good reason to not do something, however, by now i'm afraid it might be a tiny bit more confusing to put ^ in there when it implies a +1 (or -1) 03:23
03:27 adu joined
Hotkeys Also while @foo[^10] easily reads as "the first 10 elements of foo", @foo[10^] doesn't seem as intuitive 03:28
I'm sure one could get used to it though
(thanks zero based arrays) 03:29
mrf Hotkeys: I am not sure that the patch I have added would support that but it does sound like it would be something that could be added.
Hotkeys if @foo[10^..*] is possible then I don't see why the shorthand for it wouldn't work 03:30
then again I'm no parsing expert
mrf I would certainly argue that it should exist if it doesn't already 03:31
raiph m: say 10^..*
camelia rakudo-moar 273e89: OUTPUT«10^..Inf␤»
raiph m: say list 10^..*
camelia rakudo-moar 273e89: OUTPUT«(11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 …»
mrf but in a regex it seems to make send to want /<token> ** 10^/ for more than 10. with + being synonymous with 0^ I suppose 03:33
Hotkeys it seems indexing with infinite ranges on infinite sequences gets unhappy
m: my @foo = 1..*; say @foo[10^..*]; 03:34
m: my @foo = 1..20; say @foo[10^..*];
camelia rakudo-moar 273e89: OUTPUT«(timeout)»
rakudo-moar 273e89: OUTPUT«(12 13 14 15 16 17 18 19 20)␤»
03:35 Ven_ joined
Hotkeys but it would still be reasonable for 10^ to work the same in that case 03:35
mrf m: my @foo = 1..20; say @foo[^10]; say @foo[10]; say @foo]
camelia rakudo-moar 273e89: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ywWJqbmNgW␤Unexpected closing bracket␤at /tmp/ywWJqbmNgW:1␤------> 030; say @foo[^10]; say @foo[10]; say @foo7⏏5]␤»
mrf m: my @foo = 1..20; say @foo[^10]; say @foo[10]; say @foo[10^..*]
camelia rakudo-moar 273e89: OUTPUT«(1 2 3 4 5 6 7 8 9 10)␤11␤(12 13 14 15 16 17 18 19 20)␤»
mrf m: my @foo = 1..20; say @foo[^10]; say @foo[10]; say @foo[10^] 03:37
camelia rakudo-moar 273e89: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ERfC_OKXTl␤Missing required term after infix␤at /tmp/ERfC_OKXTl:1␤------> 3ay @foo[^10]; say @foo[10]; say @foo[10^7⏏5]␤ expecting any of:␤ prefix␤ term␤»
03:37 eliasr left
mrf Hotkeys: it would look neatly symetrical 03:38
03:40 Ven_ left 03:44 sprocket left
PerlJam just spent about a minute staring at some P5 code wondering why it wasn't compiling ... 03:49
Turns out I had used the P6 ternary op rather than the P5 one 03:50
mrf :D
03:50 sprocket joined
Hotkeys lol 03:52
atweiden m: class HOH is Hash {*}; my %h := HOH.new('a', 1); my %toml; %toml<a> = %h; say %toml<a>.WHAT; say %toml.perl; 03:54
camelia rakudo-moar 273e89: OUTPUT«(HOH)␤{:a(${:a(1)})}␤»
03:54 Oatmeal left
atweiden is there any way to customize Hash assignment so that the HOH<a> replaces the original %toml<a>? 03:54
mrf can anyone tell me how camelia outputs ␤ rathern than \n. Would really like if my tests could output a more readable form 04:04
04:07 Oatmeal joined
PerlJam .u ␤ 04:08
yoleaux U+2424 SYMBOL FOR NEWLINE [So] (␤)
PerlJam mrf: by outputting the appropriate unicoe char 04:09
awwaiid mrf: github.com/perl6/evalbot/blob/mast...ot.pl#L289 and L291
PerlJam er, unicode
psch timotimo: i'm not saying that "perl5 doesn't allow it" is a good reason to not allow it. it just makes my attempt at drawing parellels fall apart. so that's more to be understood as "my argument isn't particularly stable under scrutiny", not "we shouldn't do that"
skids atweiden: I'm not sure what exactly you want to happen in the above. 04:12
atweiden skids: essentially i'm trying to label this as type HOH github.com/atweiden/config-toml/bl...ns.pm#L838 04:13
it's pretty hacky
mrf awwaiid: cheers. I was hoping there would be some magic. Regex is simple enough though
PerlJam: Yeah. Was hoping there was a method like .perl that would do it 04:14
awwaiid mrf: you could like: sub justsayin($x) { $x ~~ s/\n/␤/; say $x } . Could monkey-patch that sucker onto String 04:15
BenGoldberg m: operator postfix<^> { $n ^.. * }; m: my @foo = 1..20; say @foo[^10]; say @foo[10]; say @foo[10^]; 04:16
camelia rakudo-moar 273e89: OUTPUT«5===SORRY!5=== Error while compiling /tmp/pwcuCT1a9Z␤Unexpected block in infix position (missing statement control word before the expression?)␤at /tmp/pwcuCT1a9Z:1␤------> 3operator postfix<^>7⏏5 { $n ^.. * }; m: my @foo = 1..20; say @␤ …»
BenGoldberg m: operator postfix<^> { $n ^.. * }; my @foo = 1..20; say @foo[^10]; say @foo[10]; say @foo[10^];
camelia rakudo-moar 273e89: OUTPUT«5===SORRY!5=== Error while compiling /tmp/s5wImMF2M5␤Unexpected block in infix position (missing statement control word before the expression?)␤at /tmp/s5wImMF2M5:1␤------> 3operator postfix<^>7⏏5 { $n ^.. * }; my @foo = 1..20; say @foo␤ …»
mrf awwaiid: Yeah. I have a is_match and not_match test methods so it should be easy enough to just add it there for now. There is so much new p6 syntax I just wanted to check I wasn't missing something before I did
BenGoldberg m: sub postfix<^> { $n ^.. * }; my @foo = 1..20; say @foo[^10]; say @foo[10]; say @foo[10^];
camelia rakudo-moar 273e89: OUTPUT«5===SORRY!5=== Error while compiling /tmp/yAKqKcqW6k␤Missing block␤at /tmp/yAKqKcqW6k:1␤------> 3sub postfix7⏏5<^> { $n ^.. * }; my @foo = 1..20; say @␤ expecting any of:␤ new name to be defined␤»
mrf awwaiid: thanks though
awwaiid mrf: cool, have fun :)
mrf: interestingly I've been thinking that I might like an irssi pluging that un-does ␤, printing output nicely :) 04:17
04:17 Actualeyes left
mrf awwaiid: I will admit I am bored of writing tests for grammars 04:17
awwaiid: really. I quite like the concise-ness. to each their own I suppose 04:18
awwaiid mrf: maybe you should reverse it and write some grammars for tests. Then write a generator that uses the grammar to generate compliant output.
mrf awwaiid: :D 04:19
awwaiid ok. I should clearly go to bed
mrf I should as well. Just waiting for the game to finish then I will go back to bed 04:20
BenGoldberg debug-cat: this is a test/ 04:22
camelia debug-cat: OUTPUT«this is a test/»
BenGoldberg debug-cat: ooh, look, it thinks it's an echo-bot. 04:23
camelia debug-cat: OUTPUT«ooh, look, it thinks it's an echo-bot.»
BenGoldberg evalbotcontrol version 04:27
camelia This is evalbot revision 3235487
adu what's the difference between () and []? 04:33
04:34 Actualeyes joined
adu m: ().Array 04:34
camelia ( no output )
adu m: say ().Array
camelia rakudo-moar 273e89: OUTPUT«[]␤»
adu m: say [].Seq 04:35
camelia rakudo-moar 273e89: OUTPUT«()␤»
04:36 Ven_ joined 04:41 Ven_ left 04:42 BenGoldberg left 04:44 AlexDaniel left
Hotkeys m: say ().WHAT 04:45
camelia rakudo-moar 273e89: OUTPUT«(List)␤»
Hotkeys m: say [].WHAT
camelia rakudo-moar 273e89: OUTPUT«(Array)␤»
Hotkeys adu: ^
adu Hotkeys: thanks 04:46
m: say [1, 2] Z, [3, 4]
camelia rakudo-moar 273e89: OUTPUT«((1 3) (2 4))␤»
04:46 cognominal left
adu m: say [1, 2] >>,<< [3, 4] 04:46
camelia rakudo-moar 273e89: OUTPUT«[(1 3) (2 4)]␤»
adu it looks like one of those preserves type and the other doesn't, are there any other differences between Z and hyper? 04:47
moritz adu: hyper is eager (even to the point where it's allowed to parallelize), Z is lazy 04:49
adu interesting 04:50
Hotkeys also note that Z alone does Z, 04:51
m: say <1 2> Z <3 4> 04:52
camelia rakudo-moar 273e89: OUTPUT«((1 3) (2 4))␤»
adu I didn't know if >><< defaults to ,
basically, I only know about grammars, so I'm trying to learn about the rest of Perl6
Hotkeys I don't believe it does
m: say <1 2> >><< <3 4>
camelia rakudo-moar 273e89: OUTPUT«5===SORRY!5=== Error while compiling /tmp/dEdagzv9iO␤Missing << or >>␤at /tmp/dEdagzv9iO:1␤------> 3say <1 2> >><7⏏5< <3 4>␤ expecting any of:␤ infix␤ infix stopper␤»
adu you know, to smooth out my knowledge 04:53
Hotkeys smooth knowledge is the best knowledge
adu I know a little bit about classes, but only enough to get the parsers to run
TimToady 10^ is not gonna fly either inside or outside of regex 04:54
way too ambiguous
adu TimToady!
TimToady and it's just not all that useful
adu what's next, FFI? 04:55
TimToady hmm, camelia ain't recompilin' 04:58
05:04 tokuhiro_ joined
psch probably broken jvm build 05:05
setinputlinesep{,s} is missing on jvm 05:06
adu how do I make a type alias?
psch someone recently brought up that camelia should build indepent for both backends, 'cause she apparently doesn't
adu like can I do type GString = OpaquePointer? 05:07
psch m: use NativeCall; constant GString = Pointer; my GString $x;
camelia ( no output )
adu psch: ah ok 05:08
psch hm, Pointer and OpaquePointer are probably different things. the concept still applies
adu constant works
05:09 tokuhiro_ left
psch m: use NativeCall; constant GString = Pointer; my GString $x; $x.VAR.of.say 05:09
camelia rakudo-moar 273e89: OUTPUT«(Pointer)␤»
psch adu: note, it still contains a Pointer, not a GString
(where "it" is the container)
dalek ast: dc98136 | TimToady++ | S05-metasyntax/charset.t:
test that cclass elems require + or - between
adu psch: what am I doing wrong: pastie.org/10525745 05:14
I got "(Mu)" but I was expecting "!dlroW ,olleH" 05:17
o wait, I needed "returns Str" never mind 05:18
05:19 sprocket left
adu wow 05:19
that was super easy
adu <3 NativeCall
05:19 lucs left
adu you know what would be so great, like some kind of Advent Calendar, or something 05:20
TimToady we could, like, do it leading up to Christmas and all
05:20 lucs joined
adu I think people would like it 05:20
psch yeah, maybe host it on wordpress
cause that's easy
adu: it exists, perl6advent.wordpress.com/ 05:21
adu I know
I was being facesious, or sardonic
I'm not sure which
psch well, that went over my head :)
that second paragraph from last years 24th is still pretty amazing, TimToady++ 05:22
adu "facetious" sorry my sppeling is bad 05:23
psch well, actually the whole post is great, but i'm a sucker for silly puns
adu psch: have you ever heard of Spoonerisms? 05:25
05:25 khw left
psch adu: yes, i have. i speak them occassionally, sometimes accidentally 05:26
adu c2.com/cgi/wiki?WorstPunEver
but yes, TimToady++ 05:27
just because
psch now i'm getting a tad sad i can't share one of my favorite word plays 05:29
well, and expect people to understand it
"Das Leben gleicht in manchem Sinne einer Brille - man macht viel durch." 05:30
adu I wonder if there's going to be a post about exception handling this year.
psch adu: you can draft one :)
(the great thing about that quote is lost in translation... :P ) 05:31
adu well I've been wondering Python has 4 sections to it's handlers, (1) try (2) catch (3) else (4) finally, and 1, 2 and 4 are obvious in Perl6, but I'm not sure how one would do (3)
actually, maybe 4 isn't obvious 05:33
psch m: sub f { fail }; with f() -> $res { say "did it" } else { say "nyet" }
camelia rakudo-moar 273e89: OUTPUT«nyet␤»
psch m: sub f { 1 }; with f() -> $res { say "did it" } else { say "nyet" } 05:34
camelia rakudo-moar 273e89: OUTPUT«did it␤»
psch that's what i can make up for (3)
(4) is just a LEAVE in the block that has the CATCH, i think
adu so else is run if no exception is thrown, finally is run regardless 05:35
psch almost, with is if for definedness
TimToady there's also KEEP and UNDO
psch oh, right. UNDO is the better finally
well no 05:36
KEEP and UNDO seperate the possible cases for finally
so we have more granularity
m: if 0 { .say }; with 0 { .say }
camelia rakudo-moar 273e89: OUTPUT«0␤»
psch m: if 0 { say "if'd" }; with 0 { say "with'd" } # actually... 05:37
camelia rakudo-moar 273e89: OUTPUT«with'd␤»
psch m: if fail() { say "if'd" }; with fail() { say "with'd" }
camelia ( no output )
TimToady fail is a variant of return
05:38 Ven_ joined
psch oh, so i'm spreading intuited half-knowledge again... 05:38
i wanted to stop that, actually :/
adu psch: lol
TimToady just start spreading the ohter half instead
*other 05:39
05:39 lucs left, lucs joined
adu I was thinking you could like set $cought = True in the CATCH block, which could be used to implement Python's else block 05:40
05:40 raiph left 05:41 vendethiel joined
psch afair there's a bit missing still with exception handling 05:41
m: $!.pending 05:42
camelia ( no output )
adu it sounds like KEEP is similar to (try ... else), and KEEP and UNDO are like finally?
psch i think that doesn't actually hold the pending exceptions, if there were any
TimToady KEEP is like else, LEAVE is like finally
adu ah ok
TimToady iiuc
psch isn't UNDO like else? 05:43
i'm guessing python does it as "try this, if it dies do the else"
05:43 skids left
psch goes back to S04 05:43
hmm 05:44
TimToady maybe else is more like the default {} inside CATCH
psch is a block exit through a CATCH that handles an exception unsuccessful..?
05:45 Ven_ left
psch yeah, i think else is CATCH { default {} } 05:45
(to answer my own question: "probably not" }
adu psch: no, "catch" is run if an exception is thrown AND it matches the symbol given (if any), and "else" is run if an exception is NOT thrown
psch adu: ah, okay, then TimToady was right
adu and actually, UNDO reminds me of Go's defer 05:46
Perl6 is like every language I love rolled into one :) 05:47
psch and the missing bits you can just slang in :)
adu or maybe it's more like LEAVE
psch (depending on how comfortable you are with extending the grammar...) 05:48
adu psch: I've been thinking of writing a new slang
adu <3 grammars
psch adu: did you look at the ones FROGGS++ implemented yet?
adu nope 05:49
psch github.com/FROGGS/p6-Slang-Piersing
adu psch: this is one of the grammars I wrote: github.com/andydude/p6-c-parser/bl...rammar.pm6
psch adu: oh, alright. i guess that means the missing bit is knowing which pegs to hang your tokens on 05:51
adu I think the missing bit for me is learning how all the $* vars work
psch because (currently at least) a slang depends on knowing which tokens already exist, and how to replace the parts you want to change 05:52
llfourn and figuring out the basics of QAST :) 05:53
psch although i can't imagine not having to know what part of the language one wants to replace...
well, for finding pegs we at least have the --target=parse 05:54
llfourn: yeah, that's definitely useful, although i'm not sure it's always neccessary
adu like I would want to make a dramatically different slang, like put "use css; " at the top and the rest of the file would be css 05:55
llfourn psch: thanks I didn't even know aboout --target=parse
psch adu: github.com/tony-o/perl6-slang-sql/ is sort of a step in that direction 05:56
adu: it's of course different in that it exposes a keyword to parse the rest of the statment as a sublanguage 05:57
llfourn adu: I think a slang is a superset of whatever the main language is where as use css would be just replacing the MAIN language
psch doesn't actually know how "use css;" could be useful, but maybe css is just not the best example 05:58
adu psch: that's pretty cool
perhaps
I mean, if you're making a web server, you probably want it at the statement level anyways 05:59
llfourn mmmm for CSS you would just want a grammar I think
psch i'd probably want a simplified subset, because all the newfangled magic css does nowadays seems a little silly to me :) 06:00
it's probably all the writing-a-compiler-in-java i do in my free time ;)
adu CSS is mostly dictionaryies anyways, except the occational gradient and/or url link
psch ...well, *helping* to write a compiler, don't want over-emphasize my contribution :P
+to
adu :)
llfourn I made a slang a while ago but I don't think it works anymore: github.com/LLFourn/perl6-slang-dotty 06:03
adu psch: there was some talk a few months ago of integrating NativeCall, Inline::C, and C::Parser into some kind of hybrid thing 06:04
06:05 vendethiel left, ShimmerFairy left
psch adu: yeah, i remember 06:06
adu: as in, i remember the idea using C::Parser to autogen most of the NativeCall declarations
if not all of it
adu psch: something like that
llfourn: that's pretty involved 06:08
llfourn: where is Dotty?
llfourn adu: Dotty is a method call op?
. .? etc
adu Dotty::Slang::Grammar 06:09
Dotty::Slang::Actions
where are these defined?
llfourn github.com/LLFourn/perl6-slang-dot...g/Dotty.pm 06:10
I build them dynamically
inside EVALs -- quite ugly :\
notice .^add_method etc
adu *mind blown* 06:11
llfourn I think it's less complicated than it looks... I wonder if I could write it better now :) 06:12
06:12 retupmoca joined
llfourn ..maybe using parameterized roles 06:13
06:14 adhoc left
llfourn goes afk for a while 06:15
06:15 tokuhiro_ joined 06:20 tokuhiro_ left 06:21 adu left
mrf TimToady: Will look at removing support for 10^ from my patch when I get in to work. 06:42
06:48 Ven_ joined 06:49 ShimmerFairy joined 06:55 Ven_ left 06:58 [Sno] left 07:03 finanalyst left 07:08 telex left 07:10 telex joined
dalek line-Perl5: ffeb71a | (Stefan Seifert)++ | t/ (2 files):
Fix tests on platforms where objectids can be negative

Fixes GH #46
07:17
07:20 softmoth left
nine llfourn: github.com/niner/Inline-Perl5/issu...-149920242 is a different issue 07:21
07:28 firstdayonthejob joined
llfourn nine: ok shall I make a new issue? :) 07:33
07:34 [Sno] joined
llfourn checks if Inline::Perl5 still doesn't work 07:34
07:34 atweiden left
nine llfourn: yes, if it still doesn't work, please open a new issue. Output of perl6 -Ilib t/use.t may help diagnosing 07:35
07:35 rindolf joined
llfourn nine: sure thing 07:35
07:39 camelia left, camelia joined 07:40 ChanServ sets mode: +v camelia
nine Does perl6-gdb-m work on OS X? 07:47
llfourn nine: like in general or on the test? 07:48
nine in general
llfourn nine: I find it's hit or miss
nine If it works in general, I'd of course be interested in its output on the test :)
llfourn I'll have a go :) 07:49
07:49 xpen left, xpen joined
llfourn nine: nah debug aint working at all 07:51
[Tux] m:$*OUT.nl = "";
m: $*OUT.nl = Str;
camelia rakudo-moar 273e89: OUTPUT«Cannot unbox a type object␤ in block <unit> at /tmp/S17UPVihvy:1␤␤»
[Tux] so half of RT#123978 has been addressed 07:52
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=123978
llfourn nine: would a core dump be useful?
nine llfourn: I doubt as I don't have OS X and Linux's tools probably won't be able to do anything with it. 07:53
07:54 RabidGravy joined
[Tux] test 50000 37.191 37.078 07:55
test-t 50000 39.284 39.171
RabidGravy I think that's getting better 07:56
lizmat good *, #perl6! 07:57
nine Good morning lizmat! Thanks for the weekly :)
lizmat yesterday was better, I think
you're welcome!
07:59 finanalyst joined
[Tux] is the «$*OUT.nl = ""» fix a result of jnthn++'s grapheme work? 07:59
lizmat not grapheme, but IO work on MoarVM 08:00
m: &&::{}[];; # this used to fail, but doesn't anymore 08:01
camelia rakudo-moar 273e89: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ePpNWTE8KT␤Undeclared routine:␤ & used at line 1␤␤»
lizmat ah, it does?
08:01 abraxxa joined
lizmat star-m &&::{}[];; 08:01
star-m: &&::{}[];;
camelia star-m 2015.09: OUTPUT«5===SORRY!5=== Error while compiling /tmp/PAXzNrMLiL␤Undeclared routine:␤ & used at line 1␤␤»
lizmat well, it doesn't fail for me locally anymore 08:02
test #19 in t/spec/integration/weird-errors.t