»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
lue Why is there a perl6-debug-p.exe in my /usr/local? (It's an ELF, just named funny) 00:36
jnthn No idea...sounds like an accident in the Makefile-Parrot.in additions I did today. 00:37
jnthn ah yeah 00:38
darn :)
lue no worries, I just mv'd it :)
dalek kudo/nom: 1bd1ca3 | jnthn++ | tools/build/Makefile-Parrot.in:
Use $(EXE) macro, not hard-coded .exe.
 00:39
jnthn Didn't try it, but pretty sure it's right :)
lue++
lue jnthn: I much prefer having a lead developer who accidentally does windows-isms on occasion, than having no lead developer provide Windows support :) 00:40
jnthn: so you know, I still get a segfault on my pod 5-to-6 script in perl6-debug-m "This is perl6 version built on MoarVM version 2014.01-224-g10580a0". debug-p and debug-j work fine though. 00:44
jnthn lue: OK. If you can get a backtrace out of gdb or something, that'd may help. 00:47
For now, though, I need to go and sleep :)
'night, #perl6
lue jnthn: gist.github.com/lue/68248a2d8370f6506b7c :) 00:48
timotimo lue: backtraces from moarvm are easy, just print dump_backtrace(tc)
lue ♞ jnthn o/
timotimo: in gdb? That just causes moar segfaults :P 00:49
and I can't use a literal "tc"
timotimo ... huh?
well, you've got to be in a frame where "tc" is available 00:50
lue print tc fails in every frame. Do I need to turn on some sort of debug flag in MVM compilation? 00:51
timotimo yeah, print tc isn't right 00:53
--debug=3 is the best you can do in configure.pl
lue can I recompile just Moar, or would --debug=3 require recompiling nqp and rakudo too? 00:54
dalek kudo/pod-code-allow: c940e3f | Mouq++ | src/Perl6/ (2 files):
Make sure =code and =for code take newlines and spaces
 01:00
Mouq That took **way** to long to get right
BUT 01:01
I'd say this is about merge-ready
lue Mouq: what does "Make sure =code and =for code take newlines and spaces" do? 01:03
Mouq The newlines and spaces were being thrown away by the actions 01:04
Because the default mode for the parser is to put consecutive lines together
That's the main thing I've been fighting against in this branch
lue oh, I see. I was thinking blank lines for some reason 01:05
Mouq lue: In hindsight I probably should have done a more descriptive message, but I finally made certain it worked right and went YES! COMMIT, PUSH, DONE, BABY 01:06
lue :) 01:07
dalek Heuristic branch merge: pushed 20 commits to rakudo/pod-code-allow by Mouq
Mouq spectests and gets dinner 01:12
BenGoldberg Could anyone tell me what this error means: 01:22
r: my $x = 0; say sort { $x ^= @_.elems ?? Less !! More }, 1 .. 10
camelia rakudo-jvm 1d2c43: OUTPUT«This representation can not unbox to a native int␤ in method sort at gen/jvm/CORE.setting:7905␤ in sub sort at gen/jvm/CORE.setting:1854␤ in sub sort at gen/jvm/CORE.setting:1851␤ in block at /tmp/tmpfile:1␤␤»
..rakudo-moar 1d2c43: OUTPUT«cannot numify this␤ in any at src/vm/moar/Perl6/Ops.nqp:340␤ in any at src/vm/moar/Perl6/Ops.nqp:327␤ in any p6sort at src/vm/moar/Perl6/Ops.nqp:324␤ in method sort at src/gen/m-CORE.setting:7905␤ in sub sort at src/gen/m-CORE.setting:1854…»
..rakudo-parrot 1d2c43: OUTPUT«This type cannot unbox to a native integer␤ in any at gen/parrot/BOOTSTRAP.nqp:95␤ in method sort at gen/parrot/CORE.setting:7925␤ in sub sort at gen/parrot/CORE.setting:1858␤ in sub sort at gen/parrot/CORE.setting:1855␤ in block at /tmp…»
BenGoldberg n: my $x = 0; say sort { $x ^= @_.elems ?? Less !! More }, 1 .. 10 01:23
camelia niecza v24-109-g48a8de3: OUTPUT«(timeout)»
Mouq BenGoldberg: I think it means Rakudobug :) 01:27
Mouq r: my $x = 0; say sort { $x ^= +@_ ?? Less !! More }, 1 .. 10 01:29
camelia rakudo-jvm 1d2c43: OUTPUT«This representation can not unbox to a native int␤ in method sort at gen/jvm/CORE.setting:7905␤ in sub sort at gen/jvm/CORE.setting:1854␤ in sub sort at gen/jvm/CORE.setting:1851␤ in block at /tmp/tmpfile:1␤␤»
..rakudo-moar 1d2c43: OUTPUT«cannot numify this␤ in any at src/vm/moar/Perl6/Ops.nqp:340␤ in any at src/vm/moar/Perl6/Ops.nqp:327␤ in any p6sort at src/vm/moar/Perl6/Ops.nqp:324␤ in method sort at src/gen/m-CORE.setting:7905␤ in sub sort at src/gen/m-CORE.setting:1854…»
..rakudo-parrot 1d2c43: OUTPUT«This type cannot unbox to a native integer␤ in any at gen/parrot/BOOTSTRAP.nqp:95␤ in method sort at gen/parrot/CORE.setting:7925␤ in sub sort at gen/parrot/CORE.setting:1858␤ in sub sort at gen/parrot/CORE.setting:1855␤ in block at /tmp…»
Mouq r: my $x = 0; say sort { $x ^= $_ ?? Less !! More }, 1 .. 10
camelia rakudo-parrot 1d2c43, rakudo-jvm 1d2c43, rakudo-moar 1d2c43: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
Mouq isn't really sure what sort { $x ^= +@_ ?? Less !! More } is supposed to do 01:30
BenGoldberg rn: my $x = 0; say sort { $x ^= 1 ?? Less !! More }, 1 .. 10
camelia niecza v24-109-g48a8de3: OUTPUT«(timeout)»
..rakudo-parrot 1d2c43, rakudo-jvm 1d2c43, rakudo-moar 1d2c43: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
BenGoldberg I expect @_ (used as an integer) to produce the same value (2) each time. Which would cause $x to change back and forth between 0 and 2. 01:32
Mouq Alas, a spectest vail in S26 :( to be expected
BenGoldberg Aha, parens needed: 01:33
rn: my $x = 0; say sort { ($x ^= 1) ?? Less !! More }, 1 .. 10
TimToady what is the point of a one() junction there?
camelia niecza v24-109-g48a8de3: OUTPUT«Stack overflow: IP: 0x503212, fault addr: (nil)␤Stacktrace:␤ at Niecza.Kernel.UnboxAny<T> (Niecza.P6any) <0x0002b>␤ <...>␤ at Niecza.CtxJunctionBool.Get (Niecza.Variable) <0x00117>␤ at Niecza.CtxJunctionBool.Get (Niecza.Variable) <0x0025…»
..rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«1 3 5 7 9 2 4 6 8 10␤»
pmichaud I was wondering that also. -) 01:34
why do my smileys all seem to start missing characters lately? (/me checks keyboard and fingers)
BenGoldberg Basically, I wanted to write a particular type of japh, but dicovered a bug, instead.
niecza should not have a stack overflow just because someone passes an unstable comparator to sort. 01:35
Mouq r: my $x = 0; say sort { ($x ^= +@_) ?? Less !! More }, 1 .. 10
camelia rakudo-jvm 1bd1ca: OUTPUT«2 4 7 9 6 5 8 10 3 1␤»
..rakudo-moar 1bd1ca: OUTPUT«5 9 4 10 8 1 6 3 7 2␤»
..rakudo-parrot 1bd1ca: OUTPUT«8 7 9 10 1 6 5 3 2 4␤»
Mouq Uh. hm 01:36
r: my $x = 0; say sort { ($x ^= +@_) ?? Less !! More }, 1 .. 10
camelia rakudo-jvm 1bd1ca: OUTPUT«2 4 7 9 6 5 8 10 3 1␤»
..rakudo-moar 1bd1ca: OUTPUT«5 9 4 10 8 1 6 3 7 2␤»
..rakudo-parrot 1bd1ca: OUTPUT«8 7 9 10 1 6 5 3 2 4␤»
TimToady r: my $x = 0; $x ^= 2; say $x.perl
camelia rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«one(0, 2)␤»
Mouq I guess each is allowed to use its own sort algorith
m
TimToady r: my $x = 0; $x ^= 2; $x ^= 2; say $x.perl
camelia rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«one(one(0, 2), 2)␤»
TimToady would be much for efficient to use +^= 01:37
s/for/more/ 01:38
BenGoldberg r: say 1 +^ 3
camelia rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«2␤»
BenGoldberg rn: my $x = 0; say sort { ($x +^= 1) ?? Less !! More }, 1 .. 10
camelia rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«1 3 5 7 9 2 4 6 8 10␤»
..niecza v24-109-g48a8de3: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
BenGoldberg rn: say sort { ((state $) +^= 1) ?? Less !! More }, 1 .. 10 01:39
camelia rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«===SORRY!=== Error while compiling /tmp/tmpfile␤Unsupported use of $) variable; in Perl 6 please use $*EGID␤at /tmp/tmpfile:1␤------> say sort { ((state $)⏏ +^= 1) ?? Less !!…»
..niecza v24-109-g48a8de3: OUTPUT«Use of uninitialized value in numeric context␤ at /home/p6eval/niecza/lib/CORE.setting line 1389 (warn @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 301 (Any.Numeric @ 8) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /tmp/tmpfile line…»
pmichaud not to mention that one(one(one(one(0,2), 2), 2), 2) ends up being false 01:39
oh, maybe not.
hrm.
TimToady well, if you're trying to get it to alternate, that's good
BenGoldberg rn: say sort { ((state $x = 0) +^= 1) ?? Less !! More }, 1 .. 10
camelia niecza v24-109-g48a8de3: OUTPUT«Potential difficulties:␤ $x is declared but not used at /tmp/tmpfile line 1:␤------> say sort { ((state ⏏$x = 0) +^= 1) ?? Less !! More }, 1 .. 1␤␤1 2 3 4 5 6 7 8 9 10␤»
..rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«1 3 5 7 9 2 4 6 8 10␤»
pmichaud r: say ?one(one(one(0,2), 2), 2) 01:40
camelia rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«True␤»
pmichaud okay
yes, alternating, but also ickily inefficient
dalek kudo/pod-code-allow: ef5d59c | Mouq++ | src/Perl6/Actions.nqp:
Missed paragraph_raw -> paragraph_comment
 01:42
BenGoldberg I've always thought that using | and & for superpositions, and +| and +& for bitwise operators, wasn't the best decision... I'm more likely to want bitwise operators than superpositions in my code 01:43
TimToady yer weird
Mouq lol
BenGoldberg likes being weird, thank you very much :)
Mouq weird++ 01:44
pmichaud I think that if ($x == 2 | 3 | 7 | 11) { ... } # a bit more likely 01:45
BenGoldberg Anyone coming to perl6 from... lots and lots of languages (perl, /^c.*/i, /java(?script)/, ...), is going to expect |, &, and ^, to do bitwise operations.
pmichaud ...and they'll be wrong. :)
BenGoldberg wondered what happened to the principle of least surprise. 01:46
Mouq BenGoldberg: I personally like the fact that +|, +&, etc. are clearly numerical
BenGoldberg: And I find this consistency to be the least surprising way to do it, once the rule is learned
BenGoldberg r: say "a" ~| "b" 01:47
camelia rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«c␤»
pmichaud we still have principle of least surprise... but sometimes it's important to not do things "the way they've always been done" if that's clearly counterproductive. Besides, we're not aiming only at existing programmers, we're also aiming at people-who-don't-yet-program
TimToady It's impossible to surprise everyone leastly.
pmichaud TimToady: somehow that doesn't surprise me. :) 01:48
dalek kudo/nom: 0bc8fb7 | Mouq++ | src/Perl6/Grammar.nqp:
Separate =code etc. from =comment in grammar
rakudo/nom: 39308e5 | Mouq++ | src/Perl6/ (3 files):
rakudo/nom: Mostly working Pod code block refactor
Mouq sorry dalek :( 01:48
BenGoldberg r: print 'JSuosrtt ahnaoctkheerr, '.comb.sort( { ((state $ = 0) +^= 1) ?? Less !! More } ) 01:57
camelia rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«Just another Sort hacker,»
BenGoldberg :)
Mouq \o/ cool! 01:58
dalek ecs: 3d80107 | Mouq++ | S26-documentation.pod:
Fix ASCII graph formatting

The Pod parser correctly removed the begining indentation from the outputted graph. Adding '=code' makes it explicit that we don't want this to happen		 02:00
Mouq wonders if 'outputted' is grammatikal 02:01
BenGoldberg Generated?
Mouq BenGoldberg: Probably would have been a wiser choise
-_-
choice
BenGoldberg rn: for ( 1 ) { print (state $ = True) ^= False } 02:04
Nieczabug?
camelia niecza v24-109-g48a8de3: OUTPUT«(timeout)»
..rakudo-parrot 1bd1ca, rakudo-jvm 1bd1ca, rakudo-moar 1bd1ca: OUTPUT«one(Bool::True, Bool::False)»
Mouq I promised someone I would look in to something... 02:25
Oh, tables! good thing I'm already doing that 02:27
irclog.perlgeek.de/perl6/2014-02-19#i_8311340 The problem is that tables are complicated :P 02:28
dalek kudo/nom: 33c19c7 | Mouq++ | docs/ChangeLog:
:allow is now merged, add this back to ChangeLog
 02:43
lue Mouq: is that =code required to fix the graph, or just to preserve indentation? 'Cos requiring explicit code block to de-break the graph still feels Wrong™ to me. 03:10
lue could easily see someone putting a method declaration in an implicit code block, interjecting with para-like paragraphs, and expecting it to Just Work. 03:14
lue Mouq: tables _are_ complicated :) Emacs' org-mode has a separate minor mode devoted to table stuff, for instance 03:23
Mouq lue: Sorry, was AFK 03:25
lue Mouq: no worries 03:27
Mouq lue: When you indent in a normal pod section, it begins a code block. The problem was that the first line sets the indentation level
lue Yeah, I figured that a long time ago. Funny thing is, lesser indent levels seem to change that supposed indent level. 03:28
(As evidenced by the text graph's boxes making subsequent indentation sane again)
It'd be nice if indentation worked more on "least-indented line" than on "first indented line", but I imagine that's very difficult to do. 03:29
lue (And S26 doesn't mention anything specific about how implicit code blocks are to be indented) 03:29
Mouq lue: That's because each time its indented wrong, that line gets re-parsed as a new block of code 03:30
*not re-parsed
but
Mouq it enters the same rule again 03:30
lue yeah, that would make sense 03:31
lue fears the better method of "least-indented line" gets into multi-pass parsing territory :( 03:32
Mouq "A code block may be implicitly specified as one or more lines of text, each of which starts with a whitespace character at the block's virtual left margin. The implicit code block is then terminated by a blank line." 03:33
"An indented Pod block is considered to have a virtual left margin, determined by the indentation of its opening delimiter."
lue: So it is spec. Whether the spec should be changed is, as always, up for debate :) 03:34
lue: And the thing with tables is that they actually are essentially parsed twice 03:35
lue Mouq: that doesn't actually say anything; it only talks about the block containing the implicit code block, and implicit blocks don't have opening delimiters.
Mouq Hm
I took "virtual left margin" to be specifying a very specific thing 03:36
lue There's the possibility of doing it like heredocs, only instead of an ending string, we go until a blank line, and instead of going off the ender's indent, we go off the least-indented line.
Mouq lue: Oh? I thought heredocs used the same rule 03:38
Huh
lue Mouq: heredoc indentation is based off the ending delimiter's indent. So if a q:to/EOF/'s EOF is indented with four space, the "virtual left margin" (so to speak) of the heredoc is 4 spaces in.
Mouq Ohh, that's right 03:39
lue: Well.. I'm not convinced it needs changing, but if you do, the code to change is at src/Perl6/Grammar.nqp:L562 03:41
lue Yeah, I kinda had a feeling I should be the one to fix it :D 03:42
Mouq *and maybe src/Perl6/Actions.nqp:L675 too depending on how it gets done
lue To be fair, the spec doesn't actually specify how to handle implicit code blocks, so any way is equally right atm. I'd like more input on how it should work first :) 03:43
Mouq lue: You do have the advantage that TheDamian apparently thought the graph should be displayed right as it was, though I still think it was formatting bug 03:44
TimToady code blocks are a bit different from heredocs, insofar as most modern languages will have a minimum indentation at the start of a block, but you can't really predict what will be in a heredoc 03:46
lue Cool, didn't know that :) Still would like to make sure from others (esp. TimToady) ... oh hey look!
TimToady: Was curious if you think implicit code blocks should not break indentation like that, or if you think someone with that problem should switch to explicit by that point. 03:47
TimToady and, to be frank, we optimized code blocks more for Perl code than for Fortran :)
one could always throw in a no-break space at the start to force it 03:48
assuming we don't count that as part of the indent 03:49
Mouq Z<>?
lue Mouq: that'd require an :allow, which requires an explicit block :) [i think]
TimToady you'd have to allow it, and it'd throw off the indent
Mouq lue: Yeah
TimToady: Yeah
Yeah.
TimToady :) 03:50
lue TimToady: my idea for a workaround (were I writing a pod doc) involves a lone period on the first non-blank column, one line before the actual code
Mouq goes back to trying to comprehend $<pod_string> and how it could be re-parsed for tables…
lue By the way, I only brought up heredocs as the "most similar" thing to code blocks, at least in terms of how I'd prefer they be handled :) 03:51
Mouq Oh, wait, just don't build it... build its contents by hand....
lue TimToady: "You can also give the option a value of zero, ... For example, to specify a block of code that should appear without its usual nesting:" <-- does this mean all code blocks have an implicit nesting of :nested(«int where * > 0») ? 03:52
TimToady hasn't read the spec there in a while 03:53
TimToady shouldn't do design while he has pneumonia, either 03:54
lue :( 03:55
Mouq This Pod stuff really needs to be done in Perl 6, not NQP… 03:57
Does the perl6 that does syn/ get updated? 04:04
lue Perl5 is doing syn/, I have something designed to eventually replace it though 04:10
github.com/lue/synopsis-generator
Mouq lue: I mean whatever perl6 generates S26 :) 04:10
lue The Pod6 synopses currently have to be hand-generated
Mouq Oh? Ok 04:11
Thank you
lue No-one's generated S15 yet, for instance.
lue &
Mouq nqp: say('' ~~ str) 04:26
camelia nqp-moarvm: OUTPUT«Cannot find method 'ACCEPTS': no method cache and no .^find_method␤ at /tmp/tmpfile:1 (<ephemeral file>::25)␤ from gen/moar/stage2/NQPHLL.nqp:1090 (/home/p6eval/rakudo-inst-1/languages/nqp/lib/NQPHLL.moarvm::94)␤ from gen/moar/stage2/NQPHLL.nqp:1076 (/h…»
..nqp-parrot: OUTPUT«No method cache and no find_method method in meta-object␤current instr.: '' pc 48 ((file unknown):69150712) (/tmp/tmpfile:1)␤»
..nqp-jvm: OUTPUT«java.lang.NullPointerException␤ in (/tmp/tmpfile:1)␤ in (gen/jvm/stage2/NQPHLL.nqp:1099)␤ in eval (gen/jvm/stage2/NQPHLL.nqp:1085)␤ in evalfiles (gen/jvm/stage2/NQPHLL.nqp:1291)␤ in command_eval (gen/jvm/stage2/NQPHLL.nqp:1195)␤ in command_line (g…»
Mouq nqp: say(''.WHAT)
camelia nqp-parrot: OUTPUT«Can only use get_what on a SixModelObject␤current instr.: '' pc 43 ((file unknown):51005000) (/tmp/tmpfile:1)␤»
..nqp-moarvm, nqp-jvm: OUTPUT«␤»
Mouq nqp: say(nqp::isstr('')) #ah-hah 04:27
camelia nqp-moarvm, nqp-jvm, nqp-parrot: OUTPUT«1␤»
Mouq nqp: say(nqp::isstr(nqp::join('',[])))
camelia nqp-moarvm, nqp-jvm, nqp-parrot: OUTPUT«1␤»
dalek ast: 2ff24c0 | Mouq++ | S26-documentation/09-configuration.t:
Add slightly better :begin code :allow test
 04:38
Mouq nqp: say(nqp::isstr([''])) 05:12
camelia nqp-moarvm, nqp-jvm, nqp-parrot: OUTPUT«0␤»
Mouq nqp: my @a = ['a','b','c']; given @a[1] -> $foo { $foo := 'q' }; for @a { say($_) } 05:22
camelia nqp-jvm: OUTPUT«Assignment ("=") not supported in NQP, use ":=" instead at line 2, near " ['a','b',"␤ in panic (gen/jvm/stage2/NQPHLL.nqp:378)␤ in infix:sym<=> (gen/jvm/stage2/NQP.nqp)␤ in !protoregex (gen/jvm/stage2/QRegex.nqp:784)␤ in infix (gen/jvm/stage2/NQPHLL.nqp)…»
..nqp-parrot: OUTPUT«Assignment ("=") not supported in NQP, use ":=" instead at line 2, near " ['a','b',"␤current instr.: 'panic' pc 15934 (gen/parrot/stage2/NQPHLL.pir:5941) (gen/parrot/stage2/NQPHLL.nqp:425)␤»
..nqp-moarvm: OUTPUT«Assignment ("=") not supported in NQP, use ":=" instead at line 2, near " ['a','b',"␤ at gen/moar/stage2/NQPHLL.nqp:369 (/home/p6eval/rakudo-inst-2/languages/nqp/lib/NQPHLL.moarvm:panic:120)␤ from <unknown>:1 (/home/p6eval/rakudo-inst-2/languages/nqp/lib/n…»
Mouq nqp: my @a := ['a','b','c']; given @a[1] -> $foo { $foo := 'q' }; for @a { say($_) }
camelia nqp-parrot: OUTPUT«Confused at line 2, near "given @a[1"␤current instr.: 'panic' pc 15934 (gen/parrot/stage2/NQPHLL.pir:5941) (gen/parrot/stage2/NQPHLL.nqp:425)␤» 05:22
..nqp-jvm: OUTPUT«Confused at line 2, near "given @a[1"␤ in panic (gen/jvm/stage2/NQPHLL.nqp:378)␤ in comp_unit (gen/jvm/stage2/NQP.nqp:922)␤ in TOP (gen/jvm/stage2/NQP.nqp:820)␤ in parse (gen/jvm/stage2/QRegex.nqp:1290)␤ in parse (gen/jvm/stage2/NQPHLL.nqp:1377)␤ in…»
..nqp-moarvm: OUTPUT«Confused at line 2, near "given @a[1"␤ at gen/moar/stage2/NQPHLL.nqp:369 (/home/p6eval/rakudo-inst-2/languages/nqp/lib/NQPHLL.moarvm:panic:120)␤ from gen/moar/stage2/NQP.nqp:917 (/home/p6eval/rakudo-inst-2/languages/nqp/lib/nqp.moarvm:comp_unit:346)␤ fro…»
Mouq I take it NQP doesn't have given 05:23
nqp: my @a := ['a','b','c']; my $foo := @a[1]; $foo := 'q'; for @a { say($_) }
camelia nqp-moarvm, nqp-jvm, nqp-parrot: OUTPUT«a␤b␤c␤»
PerlJam Mouq: nqp does have multi method dispatch, which can be like "given" if you squint :) 05:26
Mouq PerlJam: Heh, I'm just not sure of how I'm limited in NQP 05:32
Mouq nqp: call-reg(/abcd/); sub call-reg($r) { say('hijkabcdefg' ~~ $r) } # for example? 05:33
camelia nqp-moarvm, nqp-jvm, nqp-parrot: OUTPUT«abcd␤»
Mouq Yay!
inahandizha VisitsToMoney.com/index.php?refId=386970 05:58
Mouq nice to meet you too, inahandizha 05:58
lue r: say "A" ~~ /\d?/; say $/.so; say $/.defined; 06:15
camelia rakudo-parrot 33c19c, rakudo-jvm 33c19c, rakudo-moar 33c19c: OUTPUT«｢｣␤␤True␤True␤»
lue I understand why this is, but it feels *so* wrong.
PerlJam * has the same "problem" 06:16
TimToady particularly when scanning for something that starts x* 06:16
and there's an x later in the string 06:17
TimToady but leftmost beats longest 06:17
lue Is there a reason why we can't treat an empty match like an empty string, that is its truthiness is false? 06:18
PerlJam But the match succeeded!
pmichaud you can always do ?~$/ or (~$/).so
TimToady so you think a match of 0 should also be false? :) 06:19
pmichaud r: say "A" ~~ /\d?/; say ?~$/;
camelia rakudo-parrot 33c19c, rakudo-jvm 33c19c, rakudo-moar 33c19c: OUTPUT«｢｣␤␤False␤»
pmichaud or something like that.
TimToady why put the ? if you don't mean it?
lue Yeah, I went with $/.Str.chars
TimToady: POD's =over, and how the number to indent defaults to 4 06:20
pmichaud r: "A" ~~ /\d/; say $/.so;
camelia rakudo-parrot 33c19c, rakudo-jvm 33c19c, rakudo-moar 33c19c: OUTPUT«False␤»
PerlJam lue: where does the truthity come into play? 06:21
lue Originally I tried something akin to $<indent> // 4 , if that helps 06:21
pmichaud camelia: say Nil.defined; 06:24
er
wrong command
pmichaud reads latest S05 updates 06:25
lue PerlJam: My POD converter for the moment blindly converts =over <number> to =begin pod :nested(<number>) , and =over defaults to a <number> of 4. So when I write in the Pod6 replacement, I need to test if I actually got that optional match.
pmichaud new S05 says that the ?-quant produces Nil on a failed match 06:26
and that can be tested using //
r: "A" ~~ /(\d)?/; say $[0].defined; 06:27
camelia rakudo-parrot 33c19c, rakudo-jvm 33c19c, rakudo-moar 33c19c: OUTPUT«===SORRY!=== Error while compiling /tmp/tmpfile␤Unsupported use of $[ variable; in Perl 6 please use user-defined array indices␤at /tmp/tmpfile:1␤------> "A" ~~ /(\d)?/; say ⏏[3…»
pmichaud r: "A" ~~ /(\d)?/; say $0.defined;
camelia rakudo-parrot 33c19c, rakudo-jvm 33c19c, rakudo-moar 33c19c: OUTPUT«False␤»
pmichaud r: "A" ~~ /(\d)?/; say $0 // 4;
camelia rakudo-parrot 33c19c, rakudo-jvm 33c19c, rakudo-moar 33c19c: OUTPUT«4␤»
pmichaud seems correct to me. 06:28
lue r: "A" ~~ /$<indent>=[\d+]?/; say $<indent> // 4; 06:29
camelia rakudo-parrot 33c19c, rakudo-jvm 33c19c, rakudo-moar 33c19c: OUTPUT«｢｣␤␤»
lue r: "A" ~~ /$<indent>=[\d+]?/; say $<indent>[0] // 4; # o.o?
pmichaud r: "A" ~~ /$<indent>=(\d+)?/; say $<indent> // 4;
camelia rakudo-parrot 33c19c, rakudo-jvm 33c19c, rakudo-moar 33c19c: OUTPUT«4␤»
lue I'd've thought () would require that [0], not [] ... 06:31
pmichaud there's a subtle difference between using [...] and (...) for submatches
TimToady I'd'a thought that the $<indent>= erases that distinction 06:32
pmichaud in particular, [...] produces a substring match, which always succeeds
/ $<foo>=[abc]* / creates a single match object containing all of the abc's 06:33
pmichaud the quantifier acts on the brackets before the aliasing... i.e., it's like / $<foo>=([abc]*) / 06:33
moritz \o
lue o/ 06:34
pmichaud directly define in S05... :-)
*defined
lue pmichaud: that feels a bit weird, but admittedly I haven't read S05 :) 06:35
pmichaud From S05: "Note, however, that a set of quantified non-capturing brackets always returns a single Match object which contains only the complete substring that was matched by the full set of repetitions of the brackets ..."
TimToady right, I forgot we put that in 06:37
pmichaud anyway, / $<indent>=(\d+)? / does exactly what you want here
it causes $<indent> to be Nil (undefined) if no digits are found
there's no [0] required because ? doesn't produce an array of matches 06:38
lue pmichaud: for some reason, I expect $<rule>=() to create a $<rule>[0] that I have to check, which is why I always go for =[]
pmichaud it's an alias, so no.
() produces a subrule match, which would go into $0 if it weren't aliased
to get $<rule>[0] you'd do $<rule>=((...)) 06:39
anyway, brackets nearly always means "give me the characters matched, no submatches" 06:40
parens means "createa new match that can have submatches"
in particular, with / $<rule>=[ abc (def) ghi ] / there won't be a $<rule>[0] 06:41
lue pmichaud: OK, I think my problem is seeing $<rule>= as an inline regex rule { }, I'll try to remember it as a change in destination instead :)
pmichaud $<rule>=( ) would be the inline regex rule. 06:41
pmichaud afk, sleep 06:48
lue My POD converter isn't perfect, but it hopefully gets rid of some of the tedium of conversion :) Everything from here involves cleverer stuff. 06:51
jnthn morning o/ 09:16
nwc10 \o 09:27
FROGGS o( 09:38
o/
-.-
moritz )o(
gfldex moritz: thank you for making we whistle www.youtube.com/watch?v=wbWOVfY-rxU 09:44
moritz gfldex: 10 hours long? :-) 09:52
gfldex all 10 hours of it
actually, the shart version is banned in germany
on youtube that is
we are allowed to whistle the short version 09:53
timotimo o/ 10:31
timotimo yawntimo is yawn 10:31
timotimo corecatcher: still not allowed to message you 10:39
tell Mouq gist.github.com/timo/6132249 10:43
how do i do that again? 10:44
timotimo i think it's time to start with the release process 11:04
did we decide against doing release names last time? :P 11:09
well, if nobody has any decent idea, i'll just selfishly call it Karlsruhe.pm
FROGGS +1 11:10
timotimo so the wording for the for loops should be "statementlist level", not "statement-level", right? 11:21
jnthn Probably, yes 11:22
timotimo i'll change that in the changelog, too
jnthn +1 11:23
timotimo builds cpanm 11:26
or rather: sets up
... i don't have a cpanm binary any more o_O 11:28
timotimo tai2utc says This program should next be run in February. 11:33
but it *is* february
tadzik maybe it's not february enough 11:34
timotimo mhh
february is just not evenly distributed this year
dalek kudo/nom: 77ffb02 | (Timo Paulssen)++ | docs/ (2 files):
draft a release announcement
 11:38
timotimo hm, how do i formulate that i've been annoying my local .pm group with perl6 at every meeting so far? :P 11:42
dalek kudo/nom: d10eb8a | (Timo Paulssen)++ | docs/announce/2014.02.md:
short explanation for the codename.
 11:45
timotimo time to cut the nqp release 11:46
jnthn timotimo: You have Geoffrey Broadwell and Geoff Broadwell in the "who contributed" 11:48
timotimo thanks
jnthn timotimo: Also maybe capitalize Rakudo and MoarVM
timotimo done
jnthn And maybe "to MoarVM backend" 11:49
Since that's the bit specific to Rakudo :)
For the debugger could be worth mentioning the consequence of the change (can be used on all backends) 11:50
timotimo ah, right
dalek kudo/nom: 861d38e | (Timo Paulssen)++ | docs/announce/2014.02.md:
announcement fixes, jnthn++
 11:51
timotimo should the changelog of rakudo say that panda now works on moarvm? (the full changelog, that is) 11:56
dalek p: c6a69a1 | (Timo Paulssen)++ | tools/build/ (2 files):
bump moar and parrot
 12:01
p: 82e2b76 | (Timo Paulssen)++ | VERSION:
bump the nqp version
timotimo building an nqp with parrot and moar from a completely clean directory 12:03
timotimo nqp built and tested successfully on all three backends 12:21
testing the tarball now 12:22
timotimo make -j4 install aborts a few times with some ... job pipe read error thingie from make 12:26
make[1]: *** read jobs pipe: Invalid argument. Stop.
timotimo t.h8.lv/nqp-2014.02.tar.gz ← if someone could be so kind 12:44
dalek kudo/nom: 974171e | (Timo Paulssen)++ | tools/build/NQP_REVISION:
[release] bump NQP revision
 12:46
kudo/nom: 2be58d0 | (Timo Paulssen)++ | VERSION:
[release] bump VERSION
timotimo building rakudo in a clean environment
timotimo the stresstest rule has gone from the makefile, may want to reintroduce it to depend on the *-stresstests 13:14
timotimo one potential problem is: how do i properly fudge the jvm spectests? 13:21
i don't really think running every test file one by one manually would be a good idea >_>
moritz why would you want to do that anyway?
you run a spectest with TEST_JOBS=1
then you get a list of failing tests
timotimo i still get "no subtests run" from that 13:22
moritz eeks
timotimo: want me to run them?
timotimo r: my class Priv { method x { self!foo } }; Priv.x 13:23
you could do that
camelia rakudo-jvm 861d38: OUTPUT«(timeout)»
..rakudo-parrot 861d38, rakudo-moar 861d38: OUTPUT«===SORRY!=== Error while compiling /tmp/tmpfile␤No such private method 'foo' for invocant of type 'Priv'␤at /tmp/tmpfile:1␤------> my class Priv { method x { self!⏏foo } }; Priv.x␤»
timotimo the thing is if we're going to make a rakudo star release this month, the spectests ought to be fudged
moritz builds 13:24
timotimo so that a potential "customer" can differentiate between "problem on my machine" and "problem with rakudo-jvm in general"
moritz timotimo: is nativecall good enough for a JVM R*?
timotimo i don't think this really should say '!foo', if it says "private method 'foo'"
i think it ought to
moritz and: who hammers the R* build system in place to support precompilation of modules for perl6-j? 13:25
timotimo ah ...
someone called "(unknown yet)" 13:26
according to the star release guide
moritz you can't expect that from somebody who just signed up als release manager 13:26
timotimo agreed. 13:27
so no rakudo jvm star this month. that's fine then
no need to make extra sure jvm spectests are clean - at least not for the moment
moritz I'm not saying it can't (or won't) be done 13:28
just that an R*-j release isn't automatic
timotimo that's right 13:29
so should we take extra care about jvm's spectests?
masak good afternoon, gentlebots.
timotimo hello masak!
masak ...and, um, humans too, I guess.
tadzik hey ey
jnthn timotimo: Take care of them in the event we manage the r-j release 13:30
timotimo: Much of the work needed will involve changes in places we need to do them for r-m too... 13:31
timotimo so the rakudo release i'm doing right now is basically a parrot-based release?
moritz no 13:32
jnthn timotimo: No, the compiler release supports all 3
moritz exactly
timotimo so what's "the r-j release"? :)
moritz R*-j
not R-j
jnthn timotimo: As in, a Star that also works
With JVM
timotimo ah
that makes more sense
jnthn moritz++ # knows what I meant to say :)
moritz currently, the R* release can build perl6-j, but not the modules precompiled as .jar 13:33
masak oh, today is release day, right?
timotimo we want to have nicer precompilation factoring in the future anyway
masak exciting!
oh, and lurkers -- my offer still stands: someone pick March, and I'll pick April. 13:34
dalek kudo/nom: 00252db | (Timo Paulssen)++ | docs/release_guide.pod:
little fixes to release guide
 13:34
timotimo i has a tarball and i am going to test it 13:35
moritz 's j-spectest is in S05 now
timotimo feel free to test it along with me: t.h8.lv/rakudo-2014.02.tar.gz 13:36
tadzik masak: once in a lifetime deal! I'm in
timotimo release one, get one free! 13:36
(you gotta release one to get one free)
masak tadzik: that's the spirit. 13:39
masak makes it so
moritz builds and tests the tarball too 13:40
dalek kudo/nom: f42ac95 | masak++ | docs/release_guide.pod:
[docs/release_guide.pod] we're signing up!

  tadzik++ gets March, and I get April. nice.
timotimo updates the INSTALLING.txt to reflect current truths 13:41
nwc10 ah yes. "Bananas are marsupials! Cars run on gravy! Salmon live in trees and eat pencils! Reform in South Africa is on the way" 13:42
from a Spitting Image sketch
the previous line being IIRC "my fellow South Africans, the time has come for me to tell you the truth" 13:43
masak nwc10: this was in relation to... what, exactly? :) 13:44
moritz to "current thruths", I'd guess 13:44
nwc10 yes, the lines being read by the puppet for PW Botha
timotimo is one gig of ram still sufficient to do a full rakudo + parrot + nqp build? 13:45
nwc10 this dates it
moritz timotimo: no 13:45
nwc10 timotimo: a better answer probably depends on 32 or 64 bit
masak .oO( the gig is up )
nwc10 however, let's have a test 13:46
moritz timotimo: needed at least 1.5G or more last I tried (on 64bit)
masak at *least* 1.5G *or more*? wow! :P
jnthn :P 13:46
nwc10 I was going to test the answer for MoarVM on a 32 bit system
jnthn Moar is sub-gig even on 64-bit last I checked :)
timotimo is spectest_smolder still a thing? 13:47
moritz maybe my information is outdated 13:48
moritz oh, and it depends on your total memory 13:48
like, if you have 4GB, and set a ulimit of 1.5GB virtual memory, it won't be enough
because parrot uses the total available memory to tune its GC, or something
timotimo ah! 13:49
smolder.parrot.org/app/projects/smoke_reports/5 <- seems to be down 13:50
should i just remove the section about the smolder from INSTALLING.txt?
timotimo puts in a reference to perl6-roast-data instead 13:51
nwc10 OK, I just hit swap building MoarVM with 2 jobs on a 512M machine 13:56
d4l3k_ kudo/nom: c062718 | (Timo Paulssen)++ | / (2 files):
bring README and INSTALL.txt up to date
 13:57
timotimo please review: ^
moritz timotimo: perlpunks.de/paste/show/530609eb.278a.236 my j-spectest output from git (not your tarball), revision Plano-134-g2be58d0 13:58
timotimo well, the TODO passed's i could probably "fix" :) 13:59
timotimo doesn't look terribly bad, to be honest. 13:59
timotimo oh, the smolder is still up 14:01
the last run was on 09/27/13
so ... that's 2009, right?
or is it 2013?
or is it 1927?
fuck dates! >:(
well, it gives internal server errors when clicking on the details button 14:02
nwc10 well, silly ambiguous mixed endian dates.
timotimo so i'll just claim it doesn't exist any more
or rather: not claim it exists
nwc10 it's obviously the 9th day of the 27th month of mumble13
moritz timotimo: why not fudge the failures? 14:06
timotimo the perl6-roast-data repository may want to get a How to Help -> rakudo.moarvm section written by someone ... :) 14:11
huh
i get TTIAR in t/spec/S02-types/baghash.rakudo.parrot
and mixhash.t and sethash.t
seems like there ought to be a : after that sort call
er ...
moritz timotimo: what revision are you on? 14:12
in t/spec/ that is
dalek ast: 0905696 | (Timo Paulssen)++ | S02-types/ (3 files):
fix *hash sorting stuff.
 14:13
ast: 8f634c4 | (Timo Paulssen)++ | S32-exceptions/misc.t:
the private method doesn't need to have ! in its name.
jnthn nwc10: Building *MoarVM* itself?! Or Rakudo on... 14:14
nwc10 MoarVM 14:15
done nqp now without swap
jnthn Whoa
timotimo please verify that this is correct: ^
... wait for it ...
nwc10 actually, wasn't at HEAD, so pulled to HEAD
timotimo i made a commit to spec, seems like dalek doesn't care
not spect
roast.
nwc10 turned swap off and on to get the 28k of swap used down to 0 14:16
and now chugging away
this is an 8 year old laptop with 512K of RAM
jnthn I hope you mean MB? :)
nwc10 er yes
last I tried, I could just get blead to build on 16Mb of RAM and 96Mb of swap 14:17
timotimo "blead"?
FROGGS Perl 5 HEAD
timotimo ah
nwc10 Perl 5 is to blead as Rakudo is to nom 14:17
timotimo 8f634c4cd787a3135726b6b57cc97fa5d71bd2dc - this is my roast commit now, just as dalek showed 14:20
also, why was i the only one for which these tests failed?
nwc10 the setting hits swap 14:24
jnthn Nice that it takes until the setting to do so, though :) 14:25
nwc10 Peak was
r b swpd free buff cache si so bi bo in cs us sy id wa
0 3 130524 5532 124 27552 2130 3421 2262 3422 280 357 1 2 0 97
so 512M RAM plus just over 128M of swap 14:26
(this thing does have an X login session up, but I'm ssh-ing in to it)
nwc10 so I think it's fair to say we need about 640M to build Rakudo for MoarVM on a 32 bit system 14:27
Ulti hmm has this channel grown in size recently? 14:31
jnthn YES AND IN VOLUME 14:31
jnthn hasn't been watching the numbers, but I know backlogging when I disappear fro a while feels like it's getting harder... :) 14:32
masak 194 nicks. 14:32
we've been fluctuating slightly below 200 nicks for years. 14:33
to the point where I'm wondering if there's a hard limit around 200 for some reason.
masak that said, there can certainly be more active users recently. 14:33
Ulti jnthn: yeah I vanished for a while to stop procrastinating from finishing my thesis and it feels a lot bigger around here only after a month 14:33
masak: orly 14:34
Ulti maybe its just I haven't gone and come back to see the numbers 14:34
masak ya rly. 14:35
masak it was ~200 way back in the Pugs era. 14:35
dalek kudo-star-daily: 8246b14 | coke++ | log/ (5 files):
today (automated commit)
 14:48
hoelzro good *, #perl6 14:52
colomon o/ 15:09
[Coke] +1 to making the spectests run clean. We haven't had a single backend with a clean roast in some time. 15:13
ideally, each impl. should only be dirty one day before getting a clean run (or at least new failures) the next day 15:14
[Coke] hoelzro: if you are not able to close p6 tickets in RT, ping me. 15:25
hoelzro [Coke]: trying it now 15:27
is there a way to filter RT tickets by reporter? 15:28
FROGGS there is a custom query thingy 15:29
I think you have to "edit" the current search query
hoelzro I found that, but it wasn't clear on how to filter by reporter to me
FROGGS rt.perl.org/Search/Results.html?Fo...%27%27%2C% 15:30
0A%27%3Csmall%3E__Requestors__%3C%2Fsmall%3E%27%2C%0A%27%3Csmall%3E__CreatedRelative__%3C%2Fsmall%3E%27%2C%0A%27%3Csmall%3E__ToldRelative__%3C%2Fsmall%3E%27%2C%0A%27%3Csmall%3E__LastUpdatedRelative__%3C%2Fsmall%3E%27%2C%0A%27%3Csmall%3E__TimeLeft__%3C%2Fsmall%3E%27&Order=DESC|ASC|ASC|ASC&OrderBy=id|||&Query=Queue%20%3D%20%27perl6%27%20AND%20%28%20%20Status%20%3D%20%27new%27%20OR%20Status%20%3D%20%27open%27%20OR%20Status%20%3D%20%27stalled%27%
20%29%20AND%20Requestor.EmailAddress%20LIKE%20%27froggs%27&RowsPerPage=100&SavedChartSearchId=new&SavedSearchId=new
timotimo … huh
FROGGS ohh /o\
well, I searched for: client email contains "froggs"
moritz enterprise URLs!
hoelzro o_O
FROGGS rt.perl.org/Search/Results.html?Format=' <b><a href[...] <--- I dunno why somebody does that 15:31
hoelzro k, there we go
thanks FROGGS
FROGGS yw
timotimo could someone verify that my changes to readme and install are okay, as well as the changes to roast?
hoelzro ok, I can close tickets! \o/ 15:33
FROGGS timotimo: give me a few minutes 15:34
[Coke] hoelzro++ 15:38
FYI, rakudo is currently at 735 open tickets. Oldest ticket was opened 5 years ago.
s/open/new and open/ 15:39
[Coke] Opened a ticket to add VM as a field to RT for perl6 tickets. 15:45
TimToady_ which VM did you classify that under? :P 15:48
FROGGS timotimo++ # README/INSTALL updates 15:49
nwc10 Is Rakudo now complete enough that Star on JVM is possible? 15:50
moritz nwc10: nobody really knows 15:50
hoelzro should run his moar-support script on perl6-j
I'll do that later today, maybe
maybe I should automate that 15:51
FROGGS timotimo: and the roast changes seem to be sane too
timotimo i'm not 100% convinced on the sorting of the kv 15:52
does it surely not flatten?
nwc10 hoelzro: even a manual run for the JVM would be excellent
hoelzro nwc10: I'll start it in a few minutes 15:52
my battery is low, and I don't want to provoke it =/ 15:53
nwc10 cool, thanks 15:53
nwc10 OK, parrot just hit swap during its tests 15:53
japhb timotimo: I only scanned the README changes, but I did notice one oddness -- an apparently reversed git command at the end.
tadzik NativeCall still fails tests for me on jvm
gist.github.com/tadzik/9115094 15:54
timotimo japhb: ah, should format-patch get origin/nom..HEAD? 15:54
yeah, that does make more sense
japhb yeah 15:55
jnthn Just origin/nom will do, no? 15:55
dalek kudo/nom: 6f8bd8b | (Timo Paulssen)++ | README:
format-patch fix, japhb++
nwc10 I thought that arnsholt thought that it worked 100%. At least on his machine
tadzik yeah, my bug looks linux-specific 15:56
[Coke] I'm happy to make peschwa a bugadmin so they can close their own tickets. 15:59
nwc10 tests on Linux
a big Linux machine. Not the laptop 16:00
timotimo t.h8.lv/rakudo-2014.02.tar.gz updated with a new tarball 16:16
please2test
dalek kudo/nom: 9d0eb58 | (Timo Paulssen)++ | docs/release_guide.pod:
list this month's release at the bottom
 16:28
timotimo [Coke]++ pschwa++ # RT cleanup and tests and stuff 16:31
[Coke] 731 tickets left - 31 of which are marked testneeded, and just need someone who can write (or cut and paste) perl6 code. 16:33
timotimo corecatcher: ^
timotimo i think when my stresstest goes through, i'll tag the version and ask for someone to upload the tarball to rakudo.org 16:34
timotimo Result: PASS 16:42
timotimo tag'd 16:43
hoelzro \o/
timotimo++
timotimo t.h8.lv/rakudo-2014.02.tar.gz <- still the same tarball as half an hour ago. please upload it to rakudo.org, someone 16:44
jnthn concert & 16:49
timotimo tadzik? moritz? 16:51
timotimo PerlJam: or you perhaps? :) 16:54
PerlJam me what?
PerlJam reads
oh, sure. hold on. 16:55
timotimo cool, thanks! :)
37a90cc0edae8035ef44de51146903edf2d6fc75 rakudo-2014.02.tar.gz ← sha1sum to verify
PerlJam timotimo: done. 16:58
timotimo excellent!
i'll send the email
hoelzro I think the JVM * module tests will have to wait 16:59
they didn't finish in time, and I have to get going 17:00
I'll upload them later
timotimo thank you, hoelzro 17:01
hoelzro happy to help =)
TimToady_ rosettacode.org/wiki/Maximum_triang...sum#Perl_6 17:03
TimToady_ wishes [Z+]= worked
hoelzro why is PSGI in rakudo-star/modules twice?
timotimo hm. does last.rotate allow to go from one end of the triangle to the other in "one step"? 17:05
TimToady the Z throws away the extra value
timotimo ah!
helpful!
TimToady the Z+, actually
[Coke] Hey, anyone who has an open grant (hague or otherwise), please consider writing up a status report. 17:09
masak noted.
timotimo what if i have a rant? 17:10
[Coke] Please consider writing up a blog post
timotimo :P
masak [Coke]: short version: aside from $work being overwhelming lately and me wanting to finish up p6cc2012, the macros grant is first in line. my next step, before digging into the D3 milestone of it, is to triage all the now-open RT tickets having to do with macros and quasis, some of which I know how to fix. 17:12
notjack what if I have an ant?
masak TimToady: any compelling language reason [Z+]= shouldn't work?
TimToady it's just a bug
masak niecza: my @a = 1, 2, 3; @a [Z+]= @a; say @a.perl 17:13
camelia niecza v24-109-g48a8de3: OUTPUT«[2, 4, 6].list␤»
masak \o/
rakudo: my @a = 1, 2, 3; @a [Z+]= @a; say @a.perl
camelia rakudo-jvm 6f8bd8: OUTPUT«(timeout)»
..rakudo-parrot 6f8bd8, rakudo-moar 6f8bd8: OUTPUT«===SORRY!=== Error while compiling /tmp/tmpfile␤Preceding context expects a term, but found infix = instead␤at /tmp/tmpfile:1␤------> my @a = 1, 2, 3; @a [Z+]=⏏ @a; say @a.perl␤»
masak I see.
timotimo /o\
masak is that rakudobugged?
timotimo also, could someone power-cycle rakudo-jvm? :\
evalbot control restart
timotimo i can do it! \o/ 17:13
TimToady I've certainly brought it up before
don't recall if it got officiated 17:14
TimToady n: my @a = 1, 2, 3; @a Z+= @a; say @a.perl 17:15
camelia niecza v24-109-g48a8de3: OUTPUT«[2, 4, 6].list␤» 17:15
TimToady n: my @a = 1, 2, 3; @a Z[+=] @a; say @a.perl
camelia niecza v24-109-g48a8de3: OUTPUT«[2, 4, 6].list␤»
TimToady note it works either way :)
the latter form is accepted by rakudo, but doesn't have the semantics correct 17:16
p: my @a = 1, 2, 3; @a Z[+=] @a; say @a.perl
camelia rakudo-parrot 6f8bd8: OUTPUT«Array.new(1, 2, 3)␤»
segomos n: my @a = 1,2,3; my @b = 1,2,3,4; @a Z+= @b; say @a.perl; 17:21
camelia niecza v24-109-g48a8de3: OUTPUT«[2, 4, 6].list␤»
PerlJam perl6 17:29
oops
[Coke] we have tickets open regarding opensolaris. Looks like OS was shut down by oracle in 2010. Can we safely reject these tickets? (if not, can someone with opensolaris volunteer to fix them?) 17:30
ah, bbkr is one step ahead of me, in 2011. 17:32
bbkr++
geekosaur opensolaris /per se/ is dead, illumos is still at least twitching (see also omnios, smartos)
[Coke] I'll close those tickets in a day or two unless someone here objects. 17:33
geekosaur is actually standing up an omnios server here, but it's not likely to be running perl6 (or much of anything else beyond the server functions) 17:34
[Coke] perl6: BEGIN { a() }
camelia rakudo-moar 9d0eb5: OUTPUT«===SORRY!===␤Cannot invoke null object␤» 17:35
..rakudo-jvm 9d0eb5: OUTPUT«===SORRY!===␤java.lang.NullPointerException␤»
..niecza v24-109-g48a8de3: OUTPUT«===SORRY!===␤␤Undeclared routine:␤ 'a' used at line 1␤␤Unhandled exception: System.NullReferenceException: Object reference not set to an instance of an object␤ at Niecza.StashCursor.Core (System.String key, Boolean fina…»
..rakudo-parrot 9d0eb5: OUTPUT«===SORRY!===␤Could not find sub &a␤»
TimToady added an FP version to rosettacode.org/wiki/Maximum_triang...sum#Perl_6 17:37
[Coke] perl6: a()
camelia niecza v24-109-g48a8de3: OUTPUT«===SORRY!===␤␤Undeclared routine:␤ 'a' used at line 1␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1502 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6 line 1147 (P6.comp_uni…»
..rakudo-parrot 9d0eb5, rakudo-jvm 9d0eb5, rakudo-moar 9d0eb5: OUTPUT«===SORRY!=== Error while compiling /tmp/tmpfile␤Undeclared routine:␤ a used at line 1␤␤»
[Coke] niecza++ (RT #81502) 17:38
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=81502
[Coke] p: say ord.Cool #RT #114014 17:40
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=114014
camelia rakudo-parrot 9d0eb5: OUTPUT«===SORRY!=== Error while compiling /tmp/LfGjr1KqtY␤Calling 'ord' requires arguments␤ Expected: :(Cool $s)␤at /tmp/LfGjr1KqtY:1␤------> say ⏏ord.Cool #RT #114014␤»
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=114014
[Coke] synopsebot: you should ignore other bots, maybe.
TimToady it's funny, I have pneumonia, and I'm coughing my lungs out, but I can still program, even if I'm an idiot in all other ways... :) 17:46
[Coke] You might want to check your bias filters, there. 17:48
(the big p sucks. Get well soon. :| )
grondilu TimToady: doesn't "open($filename).lines" make more sense than "slurp($filename).lines"? 18:02
TimToady slurp closes it again :) 18:15
TimToady added a right-assoc solution 18:20
[Coke] TimToady: is there a bug queue for STD? 18:21
TimToady just git issues 18:22
TimToady well, and a Todo file in my directory that I mostly ignore :) 18:22
nwc10 TimToady: tarball works on "my" machine for all 3 backends
[Coke] er, right, better question: is it ok to put std bugs in the github issue queue. :)
k.
TimToady sure
[Coke] There are several RTs that are "rakudo does this wrong and so does STD" 18:23
and I'm sure the rakudo response is "very dangerous - you go first"
TimToady well, I can't really fix anything that depends on having a real compiler 18:24
in that sense, niecza++ is a better example of what STD should be these days
nwc10 er naughty tab key 18:26
^timotimo
timotimo nwc10: thank you :) 18:34
timotimo should i run some benchmarks? hmm. 18:35
i guess i could
nwc10 swap, om nom nom 18:45
hoelzro I have returned 18:57
JVM * module status: gist.github.com/hoelzro/9120694
arnsholt hoelzro: Oh, those NativeCall failures are fixable 19:26
(Sort of)
Getting the command-line to compile on JVM is hard, so we punt on that 19:27
Have to run the tests with Parrot first (that compiles things) then JVM 19:28
nwc10 peak swap building Rakduo on the same machine with parrot: 498460 19:29
er no, sorry, 504488 19:30
so need about 1G for Rakudo with parrot, but 640M for Moar
hoelzro arnsholt: compiles what things? 19:31
arnsholt The libs with the C code of the tests
hoelzro ah ha 19:32
arnsholt Part of the tests are in C (since we're testing calling out to C, after all =)
hoelzro makes sense =)
is it easy to fix so that you don't need to use Parrot first?
arnsholt Not AFAIK 19:33
colomon do it at module build, maybe? 19:34
[Coke] Eventually we'll need to not require parrot to do JVM stuff, I think. 19:36
(even if it means pulling in some compiler detection into rakudo and then using that to drive the backends instead of letting the backends discover things) 19:37
arnsholt Yeah 19:39
The problem is figuring out how to compile things
[Coke] hard problem, your workaround seems fine for now. :) 19:41
rt.perl.org/Ticket/Display.html?id=121071
note that Custmo fields now says: "VM: Parrot"
*Custom
Anyone wants to help bugadmin, adding that to tickets where it's explicitly stated would be great. 19:42
arnsholt Yeah, on Parrot we just pull it out of $*VM
[Coke] right now VM is a select-one field. I can probably have it changed to multiselect if we find we need it 19:44
shinobicl hi all! 20:07
moritz timotimo++ # rakudo release 20:08
hi shinobicl
shinobicl i'm looking for a perl6 example using REST, is there such a thing yet?
masak shinobicl: don't think so.
[Coke] shinobicl: you looking for a client or server side example? 20:12
timotimo moritz: thanks 20:14
i'm benchmarking 2014.01 vs 2014.02 now on moar and parrot 20:15
shinobicl masak: i'm guessing a XML serializer/deserializer is needed first, right? 20:17
i would like to try writing one, if there is no such thing yet
moritz REST works fine with JSON too
masak yeah, XML is not a strict requirement. 20:18
and we already have JSON.
shinobicl JSON::Unmarshall, right?
moritz JSON::Tiny 20:19
timotimo we have exemel, i'm not entirely sure what kinds of stuffs it can do 20:23
dalek ar: 43245f3 | moritz++ | / (17 files):
Remove modules/perl6-psgi

it is a duplicate of modules/PSGI. hoelzro++		 20:24
moritz oh, and that commit also bumped several module versions 20:24
dalek ar: 83d6a65 | moritz++ | modules/Perl6-MIME-Base64:
reset MIME::Base64 revision
 20:26
shinobicl another question: in java there are @Annotations, what could be the equivalent in perl6 for that? 20:27
[Coke] you can use POD6 for custom annotations, I think. You can also add "is deprecated", though I'm not sure if you can add random ones like "is frobnosticable" without pre-defining them somehow. 20:32
hoelzro is annotated('Annotation')? 20:34
you could easily create a module that adds something like that
[Coke] in java, those annotations have some function. I think we have a few others besides "is deprecated" spec'd. 20:38
hoelzro you can't just add adhoc annotations in Java, iirc 20:39
you have to have defined them with an @interface declaration
in Java, @SuppressWarnings is builtin, along with @Override and a few others
I think method traits are the closest thing that Perl 6 offers 20:40
timotimo all you need for an annotation is a sub trait_mod:<annotated>(...)
shinobicl i think that maybe traits is the appropiate construct to add metadata, but it bothers me a little that "is" is used also for inheritance 20:40
(add metadata as @Annotations does in java)
PerlJam shinobicl: you could create an alias for "is" that uses a word that makes you feel better about it :) 20:41
timotimo i don't have a big problem with "is" being for both subclassing and trait applying
[Coke] r: class deprecated { method eek() is deprecated {} } 20:42
camelia rakudo-moar 9d0eb5: OUTPUT«use of uninitialized value of type Any in string context␤Unhandled exception: No exception handler located for warn␤ at <unknown>:1 (/home/p6eval/rakudo-inst-1/languages/perl6/runtime/CORE.setting.moarvm:print_exception:4294967295)␤ from src/gen/m…»
..rakudo-parrot 9d0eb5: OUTPUT«use of uninitialized value of type Any in string context in any at src/Perl6/World.nqp:1661␤␤use of uninitialized value of type Any in string context␤===SORRY!=== Error while compiling /tmp/tmpfile␤Can't use unknown trait 'is …»
..rakudo-jvm 9d0eb5: OUTPUT«use of uninitialized value of type Any in string context␤Unhandled exception: use of uninitialized value of type Any in string context␤ in warn (gen/jvm/CORE.setting:677)␤ in warn (gen/jvm/CORE.setting:673)␤ in Str (gen/jvm/CORE.setting:1023)␤ …»
[Coke] pretty sure I wrote that wrong. 20:44
[Coke] p: class DEPRECATED { method eek() is DEPRECATED {} } 20:45
camelia rakudo-parrot 9d0eb5: OUTPUT«use of uninitialized value of type Any in string context in any at src/Perl6/World.nqp:1661␤␤use of uninitialized value of type Any in string context␤===SORRY!=== Error while compiling /tmp/JwDH2frZ3O␤Can't use unknown trait '…»
[Coke] I think that should probably work, or give a better error message. 20:45
hoelzro has anyone considered shipping a script that uses rlwrap to wrap the real perl6 executable as perl6? 20:47
it's frustating that the arrow keys don't DWIM 20:48
masak hoelzro: it does if you have the right thing installed, IIRC.
PerlJam yep. 20:49
Arrow keys work for me on my laptop
(but not my desktop and I've been too lazy to figure out what I needed to fix it)
moritz for parrot, you need libreadline-dev installed
timotimo it even says so in the INSTALL.txt :P 20:50
moritz but who reads those files anyway?
hoelzro oh, really?
PerlJam has libreadline-dev (but hasn't used a parrot-based rakudo in a while)
hoelzro what about for JVM/moar?
moritz works here in perl6-j 20:51
moritz and iirc moar also has something readliny 20:51
timotimo it has now? 20:52
rlwrap has your back in any case :)
except in gdb
shinobicl in my last java project we had this class declaration: @XmlRootElement @XmlType @Entity @Table public class AtributoMap extends AtributoCollection implements Serializable, Cloneable, Iterable<Atributo> 20:55
i perl6, that would be something like: class AtributoMap is XmlRootElement is XmlType is Entity is Table is AtributoCollection does Serializable does Cloneable does Iterable[Atributo] 20:56
i kind of liked the way in which annotations came before starting to declare the class and its dependencies. 20:57
shinobicl i know, is more about style than functionality, but i just wanted to illustrate my point 20:58
timotimo r: my @pieces = "foo::baz"; my @params = <ver nqp mvm>; my %foo; %foo<test>{@params} = @pieces; say %foo.perl; 22:16
camelia rakudo-jvm 9d0eb5: OUTPUT«("test" => {"nqp" => Any, "ver" => "foo::baz", "mvm" => Any}).hash␤» 22:17
..rakudo-parrot 9d0eb5, rakudo-moar 9d0eb5: OUTPUT«("test" => {"ver" => "foo::baz", "nqp" => Any, "mvm" => Any}).hash␤»
timotimo t.h8.lv/p6bench/2014-02-20-nqp.html 22:20
t.h8.lv/p6bench/2014-02-20-rakudo.html
benched something for yous :)
timotimo r: my @pieces = "foo::baz".split(":"); my @params = <ver nqp mvm>; my %foo; %foo<test>{@params} = @pieces; say %foo.perl; 22:27
camelia rakudo-jvm 9d0eb5: OUTPUT«("test" => {"nqp" => "", "ver" => "foo", "mvm" => "baz"}).hash␤»
..rakudo-parrot 9d0eb5, rakudo-moar 9d0eb5: OUTPUT«("test" => {"ver" => "foo", "nqp" => "", "mvm" => "baz"}).hash␤»
timotimo r: my @pieces = "foo:bar".split(":"); my @params = <ver nqp mvm>; my %foo; %foo<test>{@params} = @pieces; say %foo.perl; 22:28
camelia rakudo-parrot 9d0eb5, rakudo-moar 9d0eb5: OUTPUT«("test" => {"ver" => "foo", "nqp" => "bar", "mvm" => Any}).hash␤»
..rakudo-jvm 9d0eb5: OUTPUT«("test" => {"nqp" => "bar", "ver" => "foo", "mvm" => Any}).hash␤»
timotimo r: my @pieces = "foo".split(":"); my @params = <ver nqp mvm>; my %foo; %foo<test>{@params} = @pieces; say %foo.perl;
camelia rakudo-jvm 9d0eb5: OUTPUT«("test" => {"nqp" => Any, "ver" => "foo", "mvm" => Any}).hash␤»
..rakudo-parrot 9d0eb5, rakudo-moar 9d0eb5: OUTPUT«("test" => {"ver" => "foo", "nqp" => Any, "mvm" => Any}).hash␤»
TimToady Mouq: o/ you off school this week? 22:32
TimToady rotten net weather 22:33
geekosaur some kiddie's probably ddosing freenode again 22:35
Mouq TimToady: o/ Yeah, certainly tonight; I have a ton of stuff I need to do work on tonight in addition to just coming back from the first track practice this year
also
timotimo: \o/ for the grammar for pod tables 22:36
timotimo i hope you can do something at all with it 22:36
it's not really a drop-in replacement ;) 22:37
Mouq timotimo: Does it work?
timotimo: perl6
timotimo it runs exactly like shown ;)
or at least it used to :)
Mouq r: gist.github.com/timo/6132249
camelia ( no output ) 22:38
timotimo oh, huh
Mouq lemme try locally
-_- I need to rebuild 22:39
But it looks good :) 22:40
timotimo thanks :)
BenGoldberg rn: print (state $ = True) ^ False;
camelia rakudo-parrot 9d0eb5, rakudo-jvm 9d0eb5, rakudo-moar 9d0eb5, niecza v24-109-g48a8de3: OUTPUT«one(Bool::True, Bool::False)»
BenGoldberg rn: print (state $ = True) ^= False;
camelia niecza v24-109-g48a8de3: OUTPUT«(timeout)»
..rakudo-parrot 9d0eb5, rakudo-jvm 9d0eb5, rakudo-moar 9d0eb5: OUTPUT«one(Bool::True, Bool::False)»
TimToady rn: print (state $ = True) ?^= False; 22:43
camelia rakudo-parrot 9d0eb5, rakudo-jvm 9d0eb5, rakudo-moar 9d0eb5, niecza v24-109-g48a8de3: OUTPUT«True»
TimToady rn: print (state $ = True) ?^= True;
camelia rakudo-parrot 9d0eb5, rakudo-jvm 9d0eb5, rakudo-moar 9d0eb5, niecza v24-109-g48a8de3: OUTPUT«False»
BenGoldberg It's superposition specific, I think 22:44
rn: print (state $ = True) & False; 22:45
camelia rakudo-parrot 9d0eb5, rakudo-jvm 9d0eb5, rakudo-moar 9d0eb5, niecza v24-109-g48a8de3: OUTPUT«all(Bool::True, Bool::False)»
BenGoldberg rn: print (state $ = True) &= False;
TimToady rn: print (state $ = True) ^^= True;
camelia niecza v24-109-g48a8de3: OUTPUT«(timeout)»
..rakudo-parrot 9d0eb5, rakudo-jvm 9d0eb5, rakudo-moar 9d0eb5: OUTPUT«all(Bool::True, Bool::False)»
niecza v24-109-g48a8de3: OUTPUT«Unhandled exception: System.NullReferenceException: Object reference not set to an instance of an object␤ at Niecza.StashCursor.Core (System.String key, Boolean final, Niecza.StashCursor& sc, Niecza.Variable& v, Niecza.Variable bind_to) [0x00000] i…»
..rakudo-parrot 9d0eb5: OUTPUT«use of uninitialized value of type Any in string context in block at /tmp/tmpfile:1␤␤»
..rakudo-jvm 9d0eb5, rakudo-moar 9d0eb5: OUTPUT«use of uninitialized value of type Any in string context␤»
lue what pod table grammar? 22:47
Mouq lue: gist.github.com/timo/6132249 22:47
lue ooh, wonder if it works on the type of tables I make (see S15) 22:48
Mouq lue: It looks like the current Pod parser doesn't even do your tables right :( 22:52
lue: The one(s?) that's boxed 22:53
lue yeah, that's how I do table usually.
*tables
Mouq Also, I don't think it handles +'s at row-column intersections 22:54
lue I don't think so either.
Mouq *perl6 does, I mean timotimo++'s code
Mouq *perl6 does, from knowing the code, not because I've actually tried it :/ 22:54
Yup, it works fine with perl6 once the borders are removed 22:57
lue Sample from my last generation of S15:
<tr>
<td>|------------+-------+--------+--------|</td>
</tr>
Mouq Yup... I guess detecting if it's boxed shouldn't be too hard, but... 23:00
Mouq There's already so many ways to write tables. The current code to re-parse them is kind of a mess, just because that's kind of the way is has to be with its approach 23:01
lue There are many ways to text tables; I doubt the code considers box-drawing characters :) 23:02
TimToady how hard can it be? 23:05
arnsholt At a guess, PhD-level computer vision project? =)
TimToady we've already earned about 500 PhDs in here, what's one more? 23:06
lue
.oO(Perl 7 will include a new language on the braid: Table)
Mouq lue: You mean Table will include Perl 6 on its language braid 23:07
japhb__ Anyone have any experience using a cloud platform (e.g. AWS, Azure, App Engine) from a not-officially-supported language? 23:10
(I would assume Perl has this distinction for at least some of the cloud platforms, which is why I ask this community ....) 23:12
Mouq *Perl 7 23:16
Mouq -> bed, needs to get functional
timotimo i think my code shouldn't need to care about intersections between cells 23:20

Powered by Cro and the Raku® Programming Language.

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

Thank you!