»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo: / pugs: / std: , or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by moderator on 20 October 2009.
jnthn japhb: Pisa, for the Italian Perl Workshop 00:00
It's only just over an hour's flight from here. :-) 00:01
japhb nice 00:02
OOC, are most international Perl conferences done in English?
jnthn Most - in fact, probably all - are open to English language talks from foreign visitors. 00:03
But usually the majority of talks are in the language of the country it's the workshop for.
There are exceptions. For example, the Nordic Perl Workshop is done in English.
I was told, since the standard of English is very high there, it's easier to do that than to have people speaking in their own language, even though Danish/Swedish/Norwiegan are quite close. 00:04
But the majority (French Perl Workshop, Italian Perl Workshop, German Perl Workshop, etc) have most talks in their own language. 00:05
japhb OK, makes sense. 00:06
jnthn I mostly enjoy going to them to meet more Perl people, speak, visit a new city, and so on.
And can happily hack while talks are in a language I can't understand. :-)
japhb :-) 00:07
00:11 lumi joined 00:12 Aisling joined 00:13 payload joined
cognominal jnthn, but you know well that english speakers are always welcome in french perl workshops :) 00:16
this year we got ingy
00:20 PapaChub joined
jnthn cognominal: Yes, FPW is very welcoming. :-) 00:21
Hopefully I make it next year.
00:25 nihiliad joined
jnthn -> sleep 00:25
00:33 KyleHa joined 00:36 SmokeMachine joined 00:50 szabgab joined 00:52 Eevee joined 00:53 Schwern joined 00:56 lumi joined 00:57 Confield left 01:00 ihrd left 01:08 agentzh joined 01:17 lumi joined
colomon rakudo: say 3/4 < 1 01:17
p6eval rakudo 073d5e: OUTPUT«1␤»
colomon rakudo: say 3/4 cmp 1 01:18
p6eval rakudo 073d5e: OUTPUT«Multiple Dispatch: No suitable candidate found for 'cmp', with signature 'PP->I'␤in Main (file src/gen_setting.pm, line 295)␤»
colomon rakudo: say .5 cmp 1 01:19
p6eval rakudo 073d5e: OUTPUT«-1␤»
01:19 carlin joined, quux joined
carlin quux: say OH HAI 01:19
quux says OH HAI
carlin quux: buy jnthn a beer 01:20
quux buys jnthn a beer
carlin quux: become sentient
quux becomes sentient
nbrown phenny: tell jnthn I tried updating parrot to 41963 (the latest at the moment) and it didn't fix the issue with rakudo being unable to run the test suite 01:21
phenny nbrown: I'll pass that on when jnthn is around.
01:31 quux joined
carlin quux should hopefully now respond to pings, so I can leave it in the channel just for fun :-) 01:32
colomon quux: go home 01:33
quux gos home
colomon :)
carlin quux has bad grammar :-( 01:34
colomon anyone on the channel understand WHICH?
rakudo: class Rat is also { multi method WHICH() { self.Num; }; }; say 3/4 < 1 01:35
p6eval rakudo 073d5e: OUTPUT«1␤» 01:36
colomon rakudo: class Rat is also { multi method WHICH() { self.Num; }; }; say 3/4 cmp 1
p6eval rakudo 073d5e: OUTPUT«Multiple Dispatch: No suitable candidate found for 'cmp', with signature 'PP->I'␤in Main (file src/gen_setting.pm, line 288)␤»
01:38 lumi joined 01:40 orafu joined 01:47 nihiliad joined 01:49 carlin joined 01:54 lumi joined 02:03 szabgab joined
diakopter quux: die? 02:20
quux dies ?
diakopter quux: quux:
quux quuxs :
diakopter quux: 02:21
quux s
diakopter giggles
sjohnson hi 02:29
quux: say 'hi';
quux says 'hi';
sjohnson quux: help
quux helps
sjohnson quux: child
quux childs
sjohnson ... i KNEW it
PerlJam Anyone happen to know an email addres for someone at TO.pm? 02:30
never mind. :) 02:33
diakopter quux: ace wild 02:38
quux aces wild
sjohnson ... n1 02:40
02:41 eternaleye joined 02:42 eternaleye joined 02:43 eternaleye joined
pugs_svn r28867 | kyle++ | [t/spec] more tests for duplicate sub params 02:52
02:56 JimmyZ_ joined
PerlJam moritz++ I finally got some time to sit down and go through the release process and was mildly shocked to see the 2009-10 announcement already there :) 03:05
03:09 vmbrasseur_ joined 03:15 blaze-x joined 03:21 frederico joined
cognominal PerlJam, we don't wait for the pmichaud branch to land? 03:24
PerlJam I'm not releasing *now*, but I am reviewing the process for Thu. 03:28
cognominal nice
03:35 envi^office joined, frederico joined
eternaleye PerlJam: That's the trouble about these darn 'active communities', people just jump in and get things done without any regard for the ceremony of it! 03:36
03:41 lumi joined
dalek kudo: db1e525 | duff++ | build/PARROT_REVISION:
Bump PARROT_REVISION for release
 04:02
04:05 zloyrusskiy joined 04:08 lumi joined
diakopter will PARROT_REVISION work like that? 04:17
41959 1.7.0
PerlJam: ^^ 04:19
PerlJam diakopter: it did for me, and that's what the release guide says 04:23
(and I looked at Configure.pl :) 04:24
04:24 jaldhar joined 04:43 kent\\n joined
diakopter PerlJam: oh :) 04:51
pmichaud use.perl.org/~pmichaud/journal/39779 # new article about rakudo, nqp-rx, etc. 04:59
PerlJam reads before bed 05:12
The hardest part about the rakudo release so far seems to be all the waiting. 05:13
cognominal pmichaud++ 05:26
05:30 quietfanatic joined 05:36 Schwern joined
PerlJam sleep & 05:38
05:41 canhope joined
cognominal A refactoring of SVG.pm gives Null PMC access in type() in Main (file src/gen_setting.pm, line 295) 05:43
gist.github.com/214902
any idea?
quietfanatic cognominal: odd, the error seems to happen between parsing and running. 05:51
05:53 Bzek joined
quietfanatic cognominal: it is the 'Node' parameter in sub 'element'. 05:55
Removing the 'our' before 'subset' stops the error message. 05:58
Perhaps rakudo does not understand that.
I think subsets are 'our' by default anyway though, but I'm not sure.
cognominal thx 06:08
06:13 nihiliad joined
cognominal is there a way to get errors as a stack trace? 06:14
06:19 flip214 joined
flip214 I bow my head before the great perl6 masters ... 06:22
and beg for answer to a few questions.
rakudo: our $a=46; our $b=6; say [*] (1 .. $b).map( ($a - *)/(*) );
p6eval rakudo 073d5e: OUTPUT«8145060␤»
flip214 rakudo: sub X(Int $a, Int $b) { say [*] (1 .. $b).map( ($a - *)/(*) );} X(46,6);
p6eval rakudo 073d5e: ( no output )
flip214 Any ideas why that line doesn't work in a sub? 06:23
And btw, does someone have a vim script for syntax highlightning of perl6?
dalek p-rx: 3de213c | pmichaud++ | src/ (2 files):
Handle :w (quote words) quote modifier .
 06:26
p-rx: 5b85aa3 | pmichaud++ | src/ (4 files):
Simple test of subcalls in NQP.
06:30 akl joined
quietfanatic flip214: the $a in map, I think, finds the our-scoped $a but not the parameter $a. 06:32
06:32 lumi joined, envi^home joined
quietfanatic I am not sure how that would be fixed, though 06:33
06:34 rfordinal joined
quietfanatic rakudo: sub X(Int $a, Int $b) { say [*] (1 .. $b).map( {($a - $_)/$_} );} X(46,6); 06:34
p6eval rakudo 073d5e: ( no output )
diakopter semicolon 06:35
quietfanatic rakudo: sub X(Int $a, Int $b) { say [*] (1 .. $b X $a).map( -> $b, $a {($a - $b)/$b} );} X(46,6);
p6eval rakudo 073d5e: ( no output )
quietfanatic oh yeah
rakudo: sub X(Int $a, Int $b) { say [*] (1 .. $b).map( ($a - *)/(*) );}; X(46,6); 06:36
p6eval rakudo 073d5e: OUTPUT«8145060␤»
diakopter golf clas
claps
quietfanatic flip214: there you go :) 06:37
diakopter rakudo: sub X { say [*] (1 .. $^b).map( ($^a - *)/(*) ) }; X(46,6)
p6eval rakudo 073d5e: OUTPUT«8145060␤» 06:38
diakopter o_O
$^a pulled from sub X and not from the inner....?
flip214 I don't quite see the difference between my line and the one from quietfanatic
quietfanatic It is the semicolon 06:39
diakopter flip214: there's an additional semicolon
flip214 rakudo: sub X(Int $a, Int $b) { say [*] (1 .. $b).map( ($a - *)/(*) );}; X(46,6);
p6eval rakudo 073d5e: OUTPUT«8145060␤»
flip214 Ok ...
diakopter quietfanatic: $^a (above)? 06:40
shouldn't $^a pull from the inner/implied block?
flip214 rakudo: sub X(Int $a, Int $b) returns Int { state %brain; return %brain{"$a $b"} ||= ([*] (1 .. $b).map( ($a - *)/(*) )).Int;}; X(45,6);
p6eval rakudo 073d5e: TIMED_OUT 06:41
quietfanatic diakopter: No, I think WhateverCodes work differently
diakopter ok
quietfanatic infix:<-> gets an Any and Whatever as its arguments and that Any is already declared as a parameter. 06:42
I mean you know what I mean
diakopter is trapped reading that last msg. 06:43
quietfanatic sry ^_^
It parses as a call to &infix:<->($^a, *) 06:44
diakopter wonders how std parses it
quietfanatic If you said minus($^a, *), it makes sense that it would get the right $^a.
Another way of looking at it: those parens in ($^a - *) do not create a block. 06:45
diakopter yeah, it's just...
quietfanatic rakudo: say 3 - *
p6eval rakudo 073d5e: OUTPUT«!whatever_closure␤»
diakopter rakudo: say( (3 - *)() ) 06:46
p6eval rakudo 073d5e: OUTPUT«too few arguments passed (0) - 1 param expected␤in Main (file src/gen_setting.pm, line 295)␤»
diakopter rakudo: say( (3 - *)(*) )
p6eval rakudo 073d5e: OUTPUT«!whatever_closure␤»
quietfanatic Heh
diakopter rakudo: say( (3 - *)(&say) ) 06:47
p6eval rakudo 073d5e: OUTPUT«2␤»
quietfanatic rakudo: say( (3 - *)(*)(*)(*)(*)(*)(*)(*)(3) )
p6eval rakudo 073d5e: OUTPUT«0␤»
flip214 Any ideas why my last line doesn't work?
quietfanatic Oh the TIMED_OUT?
flip214 I used to do a normal factorial/(fact*fact), but that did give wrong results
quietfanatic It took longer than 15 seconds, so the evalbot killed it
diakopter 6 secs I think
flip214 for some calculations ... like 6 over 41.
quietfanatic 6 seconds? 06:48
Last I knew it was 15
rakudo: sleep 14
flip214 I think that was rounding of Num ... so I tried to rewrite it.
p6eval rakudo 073d5e: TIMED_OUT
quietfanatic rakudo: sleep 7
diakopter :)
p6eval rakudo 073d5e: ( no output )
flip214 But the code mostly does the same as before ... Only that a state variable for caching is used
quietfanatic rakudo: sleep 13
flip214 rakudo: sub X(Int $a, Int $b) returns Int { state %brain; return %brain{"$a $b"} ||= ([*] (1 .. $b).map( ($a - *)/(*) )).Int;}; X(45,6);
diakopter lol 06:49
rakudo: for (;;) {}
p6eval rakudo 073d5e: ( no output )
rakudo 073d5e: TIMED_OUT
rakudo 073d5e: ( no output )
diakopter I mean
rakudo: loop (;;) {}
p6eval rakudo 073d5e: TIMED_OUT
flip214 So a "sleep 13" works, but my simple function declaration with a 6-times-loop doesn't?
quietfanatic diakopter: Make sure that won't keep running forever on the server :) 06:50
diakopter oh yeah
flip214 "Da ist etwas faul im Staate Dänemark"
quietfanatic flip214: it's possible Rakudo is doing something wrong and infilooping or somehting
diakopter Load average: 0.81 0.84 0.59
flip214 rakudo: sub X(Int $a, Int $b) returns Int { state %brain; return %brain{"$a $b"} ||= ([*] (1 .. $b).map( ($a - *)/(*) )).Int;}; X(45,6); 06:51
diakopter returns immediately here
p6eval rakudo 073d5e: TIMED_OUT
flip214 rakudo: sub X(Int $a, Int $b) returns Int { ([*] (1 .. $b).map( ($a - *)/(*) )).Int;}; X(45,6);
p6eval rakudo 073d5e: TIMED_OUT
quietfanatic yeah you're right.
flip214 rakudo: sub X(Int $a, Int $b) returns Int { [*] (1 .. $b).map( ($a - *)/(*) ) }; X(45,6); 06:52
06:52 lumi joined
p6eval rakudo 073d5e: TIMED_OUT 06:52
flip214 Am I on some blacklist or something?
diakopter no
p6eval@debian:~/rakudo$ time ./perl6 -e 'sub X(Int $a, Int $b) returns Int { state %brain; return %brain{"$a $b"} ||= ([*] (1 .. $b).map( ($a - *)/(*) )).Int;}; X(45,6);'
real 0m3.596s
user 0m2.190s
sys 0m0.260s
quietfanatic rakudo: sub X(Int $a, Int $b) returns Int { ([*] (1 .. $b).map( ($a - *)/(*) )).Int;}; X(45,6);
flip214 BTW, should a Rat with denominator 1 be returned as an Int? 06:53
diakopter that's on the p6eval vps
p6eval rakudo 073d5e: TIMED_OUT
quietfanatic I don't think it does that for you
diakopter oh you know what
!!
quietfanatic rakudo: say (1/1).perl
diakopter quietfanatic: lol
p6eval rakudo 073d5e: OUTPUT«1/1␤»
diakopter magic exit value! 06:54
quietfanatic Oh?
Is it used for something else?
diakopter oh nm
rakudo: say $^a 06:55
p6eval rakudo 073d5e: OUTPUT«Use of uninitialized value␤␤»
diakopter rakudo: (*,say $^a)(1)
p6eval rakudo 073d5e: OUTPUT«Use of uninitialized value␤␤invoke() not implemented in class 'ResizablePMCArray'␤in Main (file src/gen_setting.pm, line 295)␤» 06:56
diakopter that's a new one
rakudo: (*;say $^a)(1)
flip214 Should a Num to Int conversion do a bit of rounding, so that 123.9999 doesn't get 123 as Int value? I fell into that trap when I had my X function use factorials, for 6 over 41.
p6eval rakudo 073d5e: OUTPUT«Use of uninitialized value␤␤invoke() not implemented in class 'Boolean'␤in Main (file src/gen_setting.pm, line 295)␤»
diakopter rakudo: ({say $^a})(55)
p6eval rakudo 073d5e: OUTPUT«Not enough positional parameters passed; got 0 but expected 1␤in Main (file src/gen_setting.pm, line 295)␤»
diakopter rakudo: {say $^a}(55) 06:57
p6eval rakudo 073d5e: OUTPUT«55␤»
diakopter rakudo: {say "$^a"}(55)
p6eval rakudo 073d5e: OUTPUT«55␤»
diakopter rakudo: {say "$^a"; say $a}(55)
p6eval rakudo 073d5e: OUTPUT«55␤55␤»
06:57 zamolxes joined
diakopter rakudo: sub X { state %brain; return %brain{"^$a ^$b"} ||= ([*] (1 .. $b).map( ($a - *)/(*) ))}; X(45,6) 06:58
p6eval rakudo 073d5e: OUTPUT«Symbol '$a' not predeclared in X (/tmp/N6Ra9IkFr5:2)␤in Main (file <unknown>, line <unknown>)␤»
diakopter yeah, oops.
rakudo: sub X { state %brain; return %brain{"$^a $^b"} ||= ([*] (1 .. $b).map( ($a - *)/(*) ))}; X(45,6)
p6eval rakudo 073d5e: ( no output )
diakopter rakudo: sub X { state %brain; return %brain{"$^a $^b"} ||= ([*] (1 .. $b).map( ($a - *)/(*) ))}; say X(45,6) 06:59
p6eval rakudo 073d5e: OUTPUT«7059052␤»
diakopter rakudo: sub X { state %brain; return %brain{"$^a $^b"} //= ([*] (1 .. $b).map( ($a - *)/(*) ))}; say X(45,6)
p6eval rakudo 073d5e: OUTPUT«7059052␤» 07:00
flip214 Ok, but there's no type definition for the parameters...
diakopter rakudo: sub X(Int $a, Int $b) { state %brain; return %brain{"$^a $^b"} //= ([*] (1 .. $b).map( ($a - *)/(*) ))}; say X(45,6)
p6eval rakudo 073d5e: OUTPUT«Cannot use placeholder var in block with signature. at line 2, near " $^b\\"} //="␤in Main (file src/gen_setting.pm, line 2099)␤»
diakopter rakudo: sub X(Int $a, Int $b) { state %brain; return %brain{$a~" "~$b} //= ([*] (1 .. $b).map( ($a - *)/(*) ))}; say X(45,6) 07:01
p6eval rakudo 073d5e: OUTPUT«7059052␤»
diakopter rakudo: sub X(Int $a, Int $b) { state %brain; return %brain{$a~" "~$b} //= ([*] (1 .. $b).map( ($a - *)/(*) )).Int }; say X(45,6)
p6eval rakudo 073d5e: OUTPUT«7059052␤»
diakopter rakudo: sub X(Int $a, Int $b) returns Int { state %brain; return %brain{$a~" "~$b} //= ([*] (1 .. $b).map( ($a - *)/(*) )).Int }; say X(45,6)
p6eval rakudo 073d5e: TIMED_OUT 07:02
diakopter it doesn't like "returns Int" in p6eval, but from the console it's fine
p6eval@debian:~/rakudo$ ./perl6 -e 'sub X(Int $a, Int $b) returns Int { state %brain; return %brain{"$a $b"} ||= ([*] (1 .. $b).map( ($a - *)/(*) )).Int;}; say X(45,6);'
7059052
hrm 07:03
flip214 Thank you for your efforts ... I'll try to make something of that.
diakopter rakudo: sub X(Int $a, Int $b) returns Int { state %brain; return %brain{$a~" "~$b} //= ([*] (1 .. $b).map( ($a - *)/(*) )).Int }; say X(45,6);
moritz_ good morning
p6eval rakudo 073d5e: TIMED_OUT
diakopter \\0/
flip214 As soon as I put the "%brain{} ||= [*] ...;" line in my script, I get "get_number() not implemented in class 'Sub'."
What does that mean, exactly?
Hello moritz! 07:04
moritz_ it means that something tries to access the * as a number, in a way that rakudo doesn't support yet
lisppaste3 flip214 pasted "Over()" at paste.lisp.org/display/89028 07:07
flip214 diakopter or moritz: would you be so nice and take a short look? 07:08
moritz_ flip214: try it with .map({ ($a+1 - $_) / $_ }); instead 07:09
the { ... } produce a closure, and the * too
so you had a closure which returns a closure, not which returns a number
07:12 Su-Shee joined
Su-Shee good morning 07:12
moritz_ good morning Su-Shee
flip214 moritz: is there some other, nice looking way? 07:15
diakopter quietfanatic: here's one for you
rakudo: {say $^a,$_}(1,a=>2)
p6eval rakudo 073d5e: OUTPUT«Too many positional parameters passed; got 1 but expected 1␤in Main (file src/gen_setting.pm, line 295)␤»
quietfanatic hmyeah?
diakopter got 1 but expected 1
quietfanatic Probably...
what about
moritz_ no 07:16
quietfanatic rakudo: {say $^a,$_}(1,2)
p6eval rakudo 073d5e: OUTPUT«Too many positional parameters passed; got 2 but expected 1␤in Main (file src/gen_setting.pm, line 295)␤»
moritz_ a => 2 bound to $^a
quietfanatic doesn't matter
moritz_ and 1 is the superfluous positional argument
diakopter but the error message says it expected 1 positional
and it also says it got 1 positional
quietfanatic It thought it was getting 1 positional but it was getting 2 07:17
the a=>2 in a named parameter format
but it bound to the position parameter $^a
diakopter rakudo: {say $^a,$_}(1,b=>2)
p6eval rakudo 073d5e: OUTPUT«Unexpected named parameter 'b' passed␤in Main (file src/gen_setting.pm, line 295)␤»
diakopter rakudo: {say $^a,$_}(b=>1,2) 07:18
p6eval rakudo 073d5e: OUTPUT«Unexpected named parameter 'b' passed␤in Main (file src/gen_setting.pm, line 295)␤»
quietfanatic because you can bind to positionals as though they were named.
07:18 payload joined
moritz_ right 07:18
diakopter rakudo: {say $^a,$_}(a=>1,a=>2)
p6eval rakudo 073d5e: OUTPUT«2Use of uninitialized value␤␤»
07:18 lumi joined
diakopter rakudo: {say $^a}(a=>1,a=>2) 07:18
07:18 rfordinal joined
p6eval rakudo 073d5e: OUTPUT«2␤» 07:18
quietfanatic Well I ought to go unconcious for a while. later 07:19
diakopter rakudo: {say $^a}(a=>1,a=>2,a=>a=>3)
p6eval rakudo 073d5e: OUTPUT«a 3␤»
quietfanatic diakopter: now that I didn't expect
oh no it's right never mind
diakopter rakudo: say a=>3
p6eval rakudo 073d5e: OUTPUT«Unexpected named parameter 'a' passed␤in Main (file src/gen_setting.pm, line 295)␤» 07:20
diakopter meh
quietfanatic see, I'm running out of neurons. Good night
moritz_ good night
diakopter 'nite
rakudo: say (a=>3)
p6eval rakudo 073d5e: OUTPUT«a 3␤»
diakopter rakudo: say \\(a=>3)
p6eval rakudo 073d5e: OUTPUT«a 3␤»
diakopter rakudo: say \\\\(a=>3)
p6eval rakudo 073d5e: OUTPUT«a 3␤» 07:21
diakopter rakudo: say \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\(a=>3)
p6eval rakudo 073d5e: OUTPUT«a 3␤»
diakopter rakudo: say \\*
quietfanatic diakopter: good night!
p6eval rakudo 073d5e: OUTPUT«!whatever_closure␤»
moritz_ rakudo's \\ handling is probably more wrong than right
diakopter quietfanatic: nite
rakudo: \\say \\1; \\ 07:22
p6eval rakudo 073d5e: OUTPUT«Syntax error at line 2, near "\\\\"␤in Main (file <unknown>, line <unknown>)␤»
diakopter rakudo: \\say \\1;
p6eval rakudo 073d5e: OUTPUT«1␤»
diakopter cool
rakudo: \\\\\\\\\\say\\ \\1;
quietfanatic I meant that as an suggestion. nm
p6eval rakudo 073d5e: OUTPUT«say requires an argument at line 2, near "\\\\ \\\\1;"␤in Main (file src/gen_setting.pm, line 2565)␤»
diakopter lol
I thought you might have meant that
07:22 mariuz joined
diakopter rakudo: \\\\\\\\\\say\\ \\\\1; 07:22
p6eval rakudo 073d5e: OUTPUT«say requires an argument at line 2, near "\\\\ \\\\\\\\1;"␤in Main (file src/gen_setting.pm, line 2565)␤»
diakopter rakudo: \\\\\\\\\\say\\ .1; 07:23
p6eval rakudo 073d5e: OUTPUT«say requires an argument at line 2, near "\\\\ .1;"␤in Main (file src/gen_setting.pm, line 2565)␤»
diakopter rakudo: \\\\\\\\\\say\\1;
p6eval rakudo 073d5e: OUTPUT«say requires an argument at line 2, near "\\\\1;"␤in Main (file src/gen_setting.pm, line 2565)␤»
diakopter std: \\\\\\\\\\say\\1;
p6eval std 28867: OUTPUT«[31m===[0mSORRY![31m===[0m␤Confused at /tmp/adGI7hW5ZR line 1:␤------> [32m\\\\\\\\\\say\\[33m⏏[31m1;[0m␤ expecting any of:␤ postfix␤ postfix_prefix_meta_operator␤FAILED 00:01 106m␤»
diakopter rakudo: \\
p6eval rakudo 073d5e: OUTPUT«Syntax error at line 2, near "\\\\"␤in Main (file <unknown>, line <unknown>)␤»
diakopter rakudo: \\;
p6eval rakudo 073d5e: OUTPUT«Syntax error at line 2, near "\\\\;"␤in Main (file <unknown>, line <unknown>)␤»
diakopter rakudo: \\* 07:24
p6eval rakudo 073d5e: ( no output )
07:24 donaldh joined
diakopter rakudo: /* */ 07:24
p6eval rakudo 073d5e: OUTPUT«perl6regex parse error: Quantifier follows nothing in regex at offset 12, found ' '␤in Main (file <unknown>, line <unknown>)␤»
diakopter rakudo: /\\/* .*/
p6eval rakudo 073d5e: ( no output )
07:26 masak joined
masak morning, #perl6. 07:26
diakopter masak: hm. :|
moritz_ \\o/ masak 07:27
diakopter masak: hi 07:29
masak moritz_: looking at the edit you made -- it's fine, but the examples say more than the text does. more specifically, the text sees nothing wrong with the last one, :($a, :@a), since it talks about 'a unique name (sigil included)'
we might want to re-word that.
diakopter rakudo: say loop (;;) { }
p6eval rakudo 073d5e: OUTPUT«Confused at line 2, near "{ }"␤in Main (file <unknown>, line <unknown>)␤»
07:30 lumi joined
mathw Morning #perl6 07:30
I have some tuits
mathw wonders what to spend them on
moritz_ masak: aye, we should
masak: I hope you agree that :($a, :@a) can't be filled unambiguosly 07:31
masak oh, sure. 07:32
that's a problem with both named parameters and public attributes.
mathw: I had a dream last night, and you sort of had a small role in it.
mathw Oh?
masak mathw: I was flirting with this girl, and you interrupted us to point out something about hashes in Perl 6.
mathw Oh dear 07:33
Was she interested?
moritz_ lol
masak mathw: in me or in Perl 6?
mathw perl 6
masak I don't know, I sort of woke up.
she was cute, though. 07:34
some kind of clerk.
mathw There were lots of lovely people at the gig I was at last night
07:34 NorwayGeek joined
mathw Some of them were on stage 07:34
One was behind the bar
The rest were in the audience
None of them were really talking to me though :(
07:35 NorwayGeek joined
masak you should try that hashes-in-perl6 opener. 07:36
it might work.
mathw Oh I know what I need to do with my tuits this morning
and it's not Perl 6
server migration
sigh
masak :/
diakopter rotfl at msdn.microsoft.com/en-us/beginner/cc963401.aspx
Learn about how computers work with code by thinking about how people listen to commands. Understanding the basics of commands makes it easier to understand why programs crash and how debuggin 07:37
mathw gone for a bit, migrating my irc client over...
diakopter "and how debuggin" indeed.
07:39 mathw joined
mathw Well, that worked 07:41
hurrah
diakopter rakudo: sub X returns Int { $^a * $^b }; say X(45,6) 07:45
p6eval rakudo 073d5e: TIMED_OUT 07:46
diakopter masak: is that a known bug?
masak don't think so.
diakopter masakbot: addbug
moritz_ rakudo: say 3+5
p6eval rakudo 073d5e: OUTPUT«8␤» 07:47
masak adds bug
diakopter masak: kidding
it works on the console, just not in p6eval for some reason
masak diakopter: too late. :)
I was just about to check that.
diakopter (it's the "returns Int")
moritz_ well, p6eval is nice, but not all reliable 07:48
masak 对, it works fine locally.
diakopter actually 07:49
try it in a file
or stdin 07:50
moritz_ anonymoushash.vmbrasseur.com/2009/1...wbies.html # I already commented, but the comment waits for approval
vmbrasseur++
masak indeed, vmbrasseur++ 07:52
Su-Shee I've just read pmichaud's posting about refactoring over the next weeks and this always makes me wonder how much companies could learn from open source projects... 07:54
diakopter masak: did you try it from stdin? 07:55
masak diakopter: oh, sorry. didn't see that till now.
trying now. 07:56
pugs_svn r28868 | moritz++ | [S06] more consistency on name collisions in signatures, masak++
moritz_ Su-Shee: I think the first thing to learn (if they don't already) is to have an extensive test suite that makes such a refactor possible without too much risk
masak diakopter: both work here.
Su-Shee moritz_: but also: refactoring is needed, so just do it instead of pushing it further and further into the future. 07:57
masak moritz_: no, that's not exactly right either. :/ let me see if I can change it.
moritz_: you see, $a and @a _are_ fine in a signature.
diakopter masak: well, I'm watching rakudo execute that on the server, and it seems to be infinitely looping for 15s 07:58
masak moritz_: you just can't refer to them by name anymore.
moritz_ masak: they are not, if you try to bind them by name
masak moritz_: but that's a later issue.
moritz_ but maybe I was a bit too strict here
masak moritz_: here it's about the signature itself.
masak fixes
moritz_ masak: as long as you still forbid things like :($a, :@a) at compile time, I'm fine 07:59
masak I wasn't planning to.
moritz_ basically $a and @a should only be allowed if both are positional
masak oh.
missed that second colon.
yes, that should be forbidden, too.
moritz_ ok
08:06 am0c joined
masak how about twigils and naming collisions? 08:09
std: class A { has $.x; method foo($!x, $.x) {} }
p6eval std 28867: OUTPUT«ok 00:02 111m␤»
masak std: class A { has $.x; has @!x; method foo(@!x, $.x) {} } 08:10
p6eval std 28867: OUTPUT«ok 00:01 109m␤»
masak std: class A { has $.x; method foo($!x, $!x) {} }
p6eval std 28867: OUTPUT«ok 00:02 109m␤»
masak rakudo: class A { has $.x; method foo($!x, $!x) { say $!x } }; A.new.foo(42, "OH HAI") 08:11
p6eval rakudo 073d5e: TIMED_OUT
masak p6eval is in a mad mood today.
08:12 lumi joined
JimmyZ_ rakudo: 'oh'.say; 08:22
p6eval rakudo 073d5e: OUTPUT«oh␤»
masak std: my $x; { say $x; my $x } 08:26
p6eval std 28868: OUTPUT«[31m===[0mSORRY![31m===[0m␤Lexical symbol $x (from line 1) is already bound to an outer scope implicitly␤ and must therefore be rewritten explicitly as OUTER::<$x> before you can␤ unambiguously declare a new $x in the same scope at /tmp/ZojMGdajnB line 1:␤------> [32mmy $x;
..…
masak what are those ~83 failures? are they due to the 64 bit platform? smolder.plusthree.com/app/public_pr...reports/18 08:37
moritz_ masak: autothreading.t segfaults on amd64 after printing the plan 08:38
masak oh.
moritz_ masak: and there's a Unicode related file that also crashes
and aborts about ~5 tests
masak I'm doing a dry run today. so, under the assumption that I were the real release manager, what would I do with such failures? 08:39
moritz_ you'd temporarily regress on those 08:40
or ask jnthn or pmichaud to fix them
regressing can involve commenting them out from t/spectest.data 08:41
or fudging them
masak nod.
moritz_ if you can't reproduce that on your architecture, you could ask me to fudge where appropriate :-)
masak sounds like something that should be fairly highly stigmatized.
moritz_ or I could give you access to a 64bit server where you can spectest yourself
masak considering that there might be an app somewhere out in the cheese using that particular feature.
moritz_ yes, it's unfortunate 08:42
the segfaulting code will be replaced shortly after the release
when the pcc revamp lands in parrot
masak so, the general rule is to release with all tests passing, even if that means regressing on a few tests? 08:43
moritz_ yes
masak gotcha. 08:45
pugs_svn r28869 | jimmy++ | [zh-cn/syn/S02-bits.pod] updated translations. 08:46
08:47 lumi joined 08:49 kidd` joined
pugs_svn r28870 | masak++ | [S06] third attempt at describing elusive name collision 08:50
moritz_ masak++ 08:51
masak I changed the focus to be on naming collisions, and extrapolated the rest from that. 08:52
moritz_ it works better that way, yes 08:58
sjohnson how's it going masak ?
masak sjohnson: I've been feeling exceptionally well lately. how about you? 08:59
09:05 riffraff joined 09:06 am0c joined 09:15 icwiener joined
moritz_ +token circumfix:sym<ang> { <?[<]> <quote_EXPR: ':q', ':w'> } 09:16
that looks a bit work-aroundish :-) 09:17
sjohnson masak: so so :) not too bad though 09:18
09:23 lumi joined 09:35 jaffa8 joined 09:37 NorwayGeek joined 09:38 NorwayGeek joined 09:39 Schwern joined
masak loliblogged! use.perl.org/~masak/journal/39780 09:40
JimmyZ_ reads
jnthn oh morning
phenny jnthn: 01:21Z <nbrown> tell jnthn I tried updating parrot to 41963 (the latest at the moment) and it didn't fix the issue with rakudo being unable to run the test suite
jnthn yaymasakblogged 09:42
But I gotta sort some silly other things out before I can read it.
09:43 lumi joined 09:46 payload joined
masak lunch & 10:04
10:10 lumi joined 10:31 ihrd joined, ihrd left 10:33 rfordinal joined 10:36 rfordinal3643 joined 10:39 lumi joined 10:45 rfordinal joined
colomon PerlJam: Is TO.pm Toronto's Perl mongers group? 10:49
colomon is very excited about tomorrow's trip to Toronto...
10:54 kentnl_ joined 10:55 rfordinal3643 joined 10:57 NorwayGeek joined 10:59 kidd`` joined 11:00 rfordinal3643asd joined
jnthn finally joins the p6l roles discussion 11:03
masak DrunkGymnast++
jnthn ;-) 11:04
moritz_ indeed
jnthn masak: Maybe tonight you have one of *those* in your dream. ;-) 11:05
At least, until mathw interupts again.
masak I hate it when he does that.
I mean, Perl 6 hashes are interesting and all, but... 11:06
I really had something going with that chick.
jnthn Aww. :(
masak I'll see if she's still around next time.
jnthn Hmm. I just found a way to optimize multiple dispatch for traits
Or for a lot of them anyway.
moritz_ masak: you know, you can influence what you dream
11:06 hanekomu joined
masak moritz_: aye. 11:07
11:07 rfordinal3643 joined
moritz_ it takes a bit of practice, though 11:07
masak nod.
moritz_ and requires that you have "healthy" sleeping habits, not dropping dead into bed at the end of the day
(which pretty much prevents me from doing it :-)
masak most of my creative dreaming takes place right before waking up. 11:08
moritz_ that's quite usual
jnthn Same.
Meh. Unfortunately, said optimization seems to win us little. 11:09
At startup.
0.04s or so.
Ah well.
moritz_ better than nothing
jnthn Well, worth having in place too.
Trait application isn't so rare that it's not worth skipping something that can notably improve its performance that only takes a few lines of code. 11:10
moritz_ wonders how many negations can meaningfully appear in a single statement 11:11
jnthn I'm not sure that we can't have an unlimited number, but I'm equally not unsure that it doesn't make sentences unreadable. 11:13
masak "I won't promise not to refrain from avoiding hurting you." is from Monkey Island I, I think. 11:14
rakudo: role Drinking { method buy_beer() { self.go_to_bar() }; method go_to_bar() { say "going to bar to drink" } }; role Gymnastics { method go_to_bar() { say "going to bar to do gymnastics" } }; class DrunkGymnast does Drinking does Gymnastics { method go_to_bar() { say "resolve conflict" } }; DrunkGymnast.new.buy_beer
p6eval rakudo 073d5e: TIMED_OUT 11:15
masak jnthn: shouldn't that output 'going to bar to drink'?
(locally, it says 'resolve conflict')
moritz_ it should after jnthn's proposal 11:16
but not by current spec
masak oh, I see.
I thought it was established fact already.
jnthn moritz_: Correct.
masak: No, no, I'm suggesting it. :-)
masak it does sound like a terribly good idea, though.
jnthn To resolve the problem Ovid has brought up.
I'm not quite sure why, since it's going to torment an implementor to do it. ;-)
masak haha, masakism! 11:17
jnthn should talk masak into doing it...he already went trawling through p6opaque.pmc yesterday anyway ;-) 11:18
masak I had t/spec/S32-trig/cotanh.t fail in the spectest harness, but it runs fine outside.
jnthn: I did no such thing! :) 11:19
I took a look and got out of there, quickly.
moritz_ masak: there are a few of these heisenbugs that appear in long-running tests
masak could be the Parrot choking on something.
moritz_ I had one in cosh.t or so
jnthn: do you plan to fix the Unicode failure before the release?
jnthn moritz_: I kinda tried already...and failed. :-/ 11:20
11:20 fridim_ joined, NorwayGeek_ joined
jnthn Need to take another look. 11:20
masak as a dry-run release manager, I'm not too amused by heisenbugs. but they're slightly better than Big Fat Fails. 11:21
jnthn Do segvs come between or below those? :-| 11:22
moritz_ I think the heisenbugs are segvs 11:23
masak but they're not consistent.
colomon I could easily shorten the trig tests, or split them up into even more files. 11:24
But that would be just hiding the bug.
moritz_ aye 11:25
colomon went to try the obsolete mega-long trig test, but since he did git pull earlier, has to do a rebuild first. 11:31
jnthn ooh, pmichaud blogged too! 11:34
masak++ # nice post on one of the things you do best :-) 11:37
masak breaking stuff. :P
colomon moritz_, masak: you're having the issues on 32-bit linux? 11:39
masak no.
moritz_ colomon: no, I'm on 64bit 11:40
masak I'm not on Linux.
moritz_ he's on "Unix with a funny hat" (that's how a popular German blogger calls MacOS)
colomon okay, guess that presumption was way off...
masak a lickable hat.
moritz_ lickable? 11:41
11:41 kentnl_ joined
colomon just wondering about the possibility of running valgrind while running the test. 11:41
moritz_ well
jnthn BTW, I recommend pmichaud++'s post, if you want to know what the coming month of Rakudo dev will look like.
moritz_ valgrind is available on that platform
colomon I can test 64-bit linux and OS X here, no problem.
moritz_ but it's way slow
masak moritz_: when Mac OS X arrived, its transparent, shiny GUI controls were often described as 'lickable'. 11:42
colomon moritz_: yes, that's why I have an quad core with 8 gigs of ram that lives headless in the basement.
masak reads pmichaud++'s post
colomon (that's my 64-bit linux platform.)
masak yay! planned and necessary mayhem! \\o/ 11:43
moritz_ colomon: does valgrind parallelize? if not the 4 cores won't help much 11:44
colomon moritz_: it helps in that it doesn't slow down the machine much, so I can still do $work on it. 11:45
jnthn Hmm. "big tasks" B and C in the roadmap can probably be called done in the next few days.
C I'd say already kinda is.
masak jnthn++ 11:46
jnthn (Please don't rip out - would want to make sure pm agrees first.) 11:47
moritz_ heh, under valgrind rakudo still runs about 3 tests per second (of the trig tests) 11:48
jnthn We run more than 3 a second not under valgrind, surely? 11:49
moritz_ yes
jnthn: I meant to imply that I'd expected even worse performance degrading 11:50
jnthn moritz_: oh, I found another way to parse it :-) 11:51
moritz_ yes, I noticed :-)
Woodi hey, where i can find perl6 web server ? 11:58
Woodi mean webserver coded in perl6 11:59
dalek kudo: 4e5877c | jonathan++ | src/pmc/perl6multisub.pmc:
Make trait dispatch that depends on a required named parameter a bit more efficient by allowing candidates that could never apply to fail much faster.
 12:00
masak Woodi: you mean like mberends++' HTTP::Daemon?
Woodi yes
masak Woodi: you can find it on github. 12:01
no, wait. on gitorious.
Woodi k, i will try to find
masak gitorious.org/http-daemon
also available through proto, with a sort of no-money-back guarantee. :) 12:02
in the ChangeLog, should one take down such things as the .push bug being fixed?
Woodi thank you masak++ :) 12:03
eiro hello world
masak eiro! \\o/
eiro masak \\o/
what'sup ?
masak comment ça va?
takadonet morning all
masak takadonet! \\o/ 12:04
eiro masak, bien
et toi ?
moritz_ masak: if you think it's worth mention, put it in
masak eiro: I'm fine. :) aussi bien.
eiro cool!
masak oaui.
er, ouai.
eiro masak: i wrote my iso2709 parser in perl5: it works fine. I try to convert it in a perl6 grammar 12:05
masak ECOLLOQUIALFRENCHFAIL
moritz_ oui?
masak moritz_: ouai:oui::yeah:yes, IIUC.
Woodi btw. on "p6 is users ready" subject: i found few realy scarry-users things...
eiro i have some pb but i guess i'm close to have a MARC::Record port of perl6 :)
masak Woodi: really? let's hear it. 12:06
eiro (working better than the perl5 one)
masak eiro: nice! looking forward to seeing it.
Woodi a) head -one STD.pm gives ALPHA; b) in Rakud roadmap in section "done" are realy strange things 12:07
eiro masak, if i can ask dump questions about perl6 grammars, it can be done quickly i hope
colomon moritz_: what are you valgrind testing?
masak eiro: you can ask any type of questions you like.
eiro masak, i'll be back tonight and will work on this: i'll ask then
moritz_ colomon: because you suggested it :-)
masak splendid. 12:08
Woodi c) somewhere i saw "frozen p6 syntax" document with list of 4 +- any(1,2,3) elems...
colomon moritz_: not why, what. as in what test file? No sense in duplicating effort.
moritz_ erm, sorry
time valgrind ./perl6 t/spec/S32-trig/cosh.t
Woodi c) was realy depressing... it says that only when/case is working :)
masak Woodi: why is (a) scary? it's alpha. 12:09
Woodi: what do you mean by really strange things in (b)?
Woodi: (c) where? URL?
Woodi masak: but look on it as "not-waterfall" project realisation... it should be realy higger for iteration in spiral-type project 12:10
colomon moritz_: okay, I'll check masak's cotanh failure. (and thanks for indirectly pointing out I could just run perl6 to run the test.)
masak Woodi: I'm not following you.
Woodi masak: cannot find it again... maybe pugs, maybe somewhere on the web...
Woodi makes food for customers eq afk 12:11
masak Woodi: unfortunately, your three points boil down to 'some unspecified things are bad'... :/ 12:12
araujo wonders if Woodi is a cook
eiro first question anyway: pastebin.com/f2e2650f3
i don't understand why showThem is called so many times 12:13
moritz_ rule TOP { ^^ <line>* $$ } 12:14
should probably use ^ and $ instead
eiro moritz_, done! it works and i feel dump :) 12:15
moritz_ it seems to backtrack somehow, though I don't see why
eiro thanks
moritz_ eiro: no need to feel dumb, we all start with silly mistakes
I started out with rather simplistic errors too 12:16
eiro you're right moritz_
masak given that you have to make errors, it's much better to do them in rough order of increasing complexity. 12:17
eiro masak, sure: that's why i don't write the iso2709 grammar for now: i want to do simple cases before 12:19
this was lineByLine parser, next step is parsing a posixAccount database
(as /etc/passwd)
masak nice. 12:20
moritz_ well, ISO2709 seems rather (mostly) fixed-width, I don't think a grammar is the best approach anyway
colomon pmichaud: looking at the roadmap, what's untrue about the current hyper/cross/reverse/other metaoperators?
moritz_ something more like unpack would be appropriate
eiro moritz_, that's right for the header 12:21
not for the data
Woodi masak: about b)
masak: you wroted something in Rakudo, right ? 12:22
eiro the length of datafields are found while reading the record headers
masak Woodi: depends on what you mean by 'wroted'. 12:23
eiro s/wroted/written/ ?
Woodi masak: something that is coded in p6 and works somehow :)
masak I guess I qualify for that.
moritz_ masak wrote working code, yes :-)
masak it happens.
moritz_ occasionally, mostly by accident :-) 12:24
s/accident/effort/
masak won't happen again. :P
Woodi masak: so you used if, for, {}, [], (), {} and other things like that
eiro :))
masak Woodi: in a pinch, yes. 12:25
Woodi masak: now look to Rakudo docs pls. all that things are not mention in "DONE" section of ROADMAP...
vi RAODMAP; ^G
jnthn Woodi: The ROADMAP was written in August 2009. The done list is things done since then.
Not all of the things done in Rakudo ever. :-) 12:26
12:26 lumi joined
moritz_ Woodi: you want rakudo.org/status for that, not ROADMAP 12:26
Woodi what i mean: fresh user looking on docs will be do not know what construct can he/she use...
masak Woodi: what moritz_ said. you're likely looking for that status document. 12:27
Woodi moritz_: oo, that much better
12:27 frederico joined
moritz_ Woodi: but you're right, that's a problem in general 12:27
Woodi: we're working on a book that explains those Perl 6 features that rakudo implements
masak maybe the ROADMAP should refer to that status page.
moritz_ masak: feel free to add a link 12:28
Woodi: that way people can just learn from the book, and what they learn works
Woodi not so big now... just: cat "Current status available at * " >> ROADMAP
moritz_ I'd rather put it at the start
masak I'd rather not duplicate information. 12:29
oh, you meant a link. I see.
Woodi moritz_: but it is rakudo... same about mildew too... 12:30
and other implementations.
and about not-waterfall perl6 development. now we know its not cascading. so what it looks like ? just 2 sentences of official description is needed... spiral methodology (plan, coding, testing, review) or agile-like or refactoring like, or chaotic like :) 12:35
12:36 NorwayGeek joined
colomon What's that about array and hash interpolation in strings? I was under the impression that wasn't supposed to work in Perl 6. Is it really just a NYI in Rakudo? 12:36
Woodi is NOT a troll (hi thinks) :)
colomon notes that the spectest graph as not caught up with this week's changes yet... 12:38
moritz_: ZOMG that's a lot of errors in valgrind...
takadonet colomon: you just want to see that huge spike
colomon takadonet: guilty as charged. :) 12:39
takadonet colomon: me too :)
12:47 kentnl_ joined
colomon moritz_: a lot of errors that look like yyparse / yylex must not be properly ported to a 64-bit platform... 12:47
PerlJam colomon: TO.pm == ThousandOaks.pm 12:49
colomon PerlJam: ah, and you're naming the release after them, yes? 12:50
PerlJam yes
moritz_ colomon: moritz.faui2k3.org/files/cosh.t.valgrind.log is "my" valgrind log - no signs of yyparse/yylex here
PerlJam colomon: arrays and hashes interpolate, but you need the brackets to do so
moritz_ PerlJam: btw that release announcement was just a first shot, and it probably needs more additions and refinements 12:51
colomon PerlJam: The status file indicates brackets are a workaround.
moritz_ they are a workaround, but valid Perl 6 :-)
PerlJam colomon: no, "blah blah {@array} blah" is a work-around, "blah blah @array[] blah" is not
colomon Common things that do not work in Rakudo: "interpolations of arrays, hashes, and method calls in strings RT #62198 (Note that the { ... } form is a usable workaround)"
PerlJam: oh. 12:52
moritz_ things with % and @ sigil only interpolate when they end in a postcircumfix
"@a.elems" does not, "@a.elems()" does 12:53
PerlJam colomon: you should also be able to do "List of stuff: @array.uniq.sort()"
moritz_ colomon: can you put your valgrind log up somewhere, and link in in #parrot?
colomon moritz_: one sec
PerlJam is off to help "teach" a GIS class, bbl
12:54 SmokeMachine joined, payload joined
colomon moritz_: where is #parrot? 12:57
moritz_ colomon: on irc.perl.org
12:59 cxreg joined
masak I haven't seen any pmichaud spectest graph data update in quite a while. 13:03
moritz_ 2009-10-13 was the latest 13:04
masak I skipped that part of my dry run where I had to say how many tests we pass, and how many more since last month. 13:05
but PerlJam will have to know tomorrow.
I have a feeling it's yet another impressive increase this month. :) 13:06
dalek kudo: 4c248e5 | masak++ | docs/release_guide.pod:
[docs/release_guide.pod] s/-/_/ in file name
 13:10
13:14 riffraff left 13:15 pmurias joined
colomon masak: I don't have firm numbers, but I'm guessing about 6000 more passing tests, at least 80% coming out of the yellow region there. 13:17
moritz_ runs tools/test_summary.pl 13:18
masak whoa!
that graph is... no longer self-similar. :>
colomon er?
moritz_ self-similar = fractal 13:19
colomon was the graph ever fractal? 13:20
masak there's a large discontinuity in it now.
moritz_ colomon: not really
erm
not at all :-)
masak colomon: it always looked about the same as time went by. 13:21
colomon you mean the early october jump, or was there another one this week?
masak it had no distinguishing features apart from the upwards trend.
I mean the early October jump.
the one that can be seen on rakudo.org/status
moritz_ there was a smaller jump in April too
colomon oh, sure, early October was my fault. 13:22
masak colomon++ # nice fault
colomon though I suppose you can blame moritz_ for aiding and abetting me with the Complex bits of it.
moritz_ the on in April was pmichaud++ and his regex "fault"s :-)
colomon I would think there will be a similar jump this week mostly attributable to jnthn++ 13:23
I added a lot of skipped tests too in the early October thing, and thanks to jnthn's merge, I was able to unskip all of them this week. 13:25
13:27 ihrd joined, abra joined 13:31 slavik joined 13:32 slavik joined
masak add t/spec/S12-introspection/methods.t to the list of heisenbuggish test files. 13:40
13:43 ruoso joined 13:44 ihrd left
moritz_ shows no anomalies in valgrind 13:46
masak strange. 13:47
it died after 31 tests in the harness here.
ladies and gentlefolks! it is with great pride I present to you the dry-run release of Rakudo Europa: the Perl 6 implementation for carbon-based as well as non-carbon-based life-forms! github.com/masak/rakudo/downloads 13:53
release announcement here: github.com/masak/rakudo/raw/master/...ce/2009-10 13:54
moritz_ masak++
14:00 lumi joined
masak (oops, s/Parrot 1\\.6\\.0/Parrot 1.8.0/) 14:01
14:05 slavik joined 14:10 edgar_ joined 14:11 KyleHa joined
colomon I thought we weren't supposed to make contact with Europa.pm until next year? ;) 14:13
masak colomon: we'll keep it under wraps till then. :) 14:15
colomon BTW, latest parrot is 1.7.0 14:16
14:16 mj41_ joined
PerlJam colomon: he's doing *his* release which will be 1.8.0 14:17
colomon ah, I see! That explains a lot.
practicing very early...
PerlJam colomon: better than that last minute :)
colomon Of course, then he needs to change October 2009 to November 2009, too.... :) 14:18
pmichaud good morning, #perl6 14:25
moritz_ o morning 14:26
pugs_svn r28871 | kyle++ | [t/spec] test based on spec r28870
KyleHa masak: Could I impose upon you to look at that to see if it conforms to your expectations?
masak KyleHa: sure. I'll look.
pmichaud 12:20 <colomon> pmichaud: looking at the roadmap, what's untrue about the current hyper/cross/reverse/other metaoperators? 14:28
currently we cheat quite a lot to get those
in particular, they aren't parsed correctly according to STD.pm -- we parse them all as normal infixes or prefixes, not via the various meta rules.
masak KyleHa: that arg ( 'bughunt', 3 ) in the 'is' call args list looks wrong to me. won't it flatten? 14:29
KyleHa: otherwise, fine.
moritz_ in particular we don't support arbitrary depth of meta nesting, I think 14:30
KyleHa masak: Thank you.
moritz_ rakudo: 1 >>+=<< 3
p6eval rakudo 073d5e: OUTPUT«Confused at line 2, near ">>+=<< 3"␤in Main (file <unknown>, line <unknown>)␤»
KyleHa masak: The test passes, so I don't think anything is being flattened.
moritz_ std: 1 >>+=<< 3
p6eval std 28870: OUTPUT«ok 00:02 106m␤»
moritz_ thinks the test is fine 14:32
KyleHa If I deliberately b0rk the test, it reports that argument as being of the form ["bughunt", 3]
dalek ok: abc7de1 | masak++ | src/multi-dispatch.pod:
[MMD] s:g:ii/stone/rock/

familiar to people.		 14:33
ok: e5d44d9 | masak++ | src/multi-dispatch.pod:
[MMD] replaced the classes by an enum decl
colomon pmichaud: Ah, so that's what's going on with generating all the hyper / meta versions of the operators.
pmichaud: I was wondering if there was some way I could help debug them, but it sounds like that's probably best left until after you get nqp-rx merged in? 14:37
pmichaud probably worth waiting, yes. I'm also wondering to what extent we can/should generate the meta operators lazily, instead of initially as we do now 14:38
moritz_ btw nqp-rx isn't going to get "merged in" as a replacement for NQP or PGE 14:39
14:39 lumi joined
moritz_ but rather stay separately 14:39
pmichaud I took it as "rebase rakudo to use nqp-rx"
moritz_ yes, me too, but I just wanted to clarify
14:40 nihiliad joined
colomon pmichaud: and indeed that was what I was trying to say. danke. 14:43
masak pmichaud: where would such meta-op laziness trade off performance? 14:44
pmichaud startup time
also memory
...do we really want to create subroutines for every possible combination of metaops? ;-) 14:45
masak no idea.
moritz_ there are infinitely many 14:46
masak moritz_: are there? I thought they were more of a hierarchy. 14:47
pmichaud I think it's large but finite
moritz_ can't you cross an arbitrary number of times? 14:48
masak my limited knowledge of STD.pm indicates that there's a finite number of them.
moritz_ and then reduce
masak std: 1 X+ 1
p6eval std 28870: OUTPUT«ok 00:01 104m␤»
masak std: 1 XX+ 1
p6eval std 28870: OUTPUT«ok 00:01 104m␤»
masak std: 1 XXXXXXXXXXXXXXXXXXXXXXXXXXXXXX+ 1 14:49
p6eval std 28870: OUTPUT«ok 00:01 105m␤»
masak std: 1 [XXXXXXXXXXXXXXXXXXXXXXXXXXXXXX+] 1
p6eval std 28870: OUTPUT«ok 00:01 105m␤»
masak seems you can.
I'm surprised that that last one survived.
pmichaud well, for Rakudo Star we have "true hyper/meta operators" as "nice to have", so those might not be forthcoming anytime soon (meaning we may still keep a cheating approach for a while) 14:50
masak isn't [XXXXXXXXXXXXXXXXXXXXXXXXXXXXXX+] a listop?
pmichaud I would think so, yes. 14:51
std: 1 [+] 1
masak and don't listops behave a bit like prefix ops?
p6eval std 28871: OUTPUT«ok 00:01 106m␤»
pmichaud yes, they do.
masak so isn't STD.pm a bit too permissive there?
pmichaud looks like it to me
masak TimToady: ^
pmichaud but STD.pm is often a bit permissive
masak but this is up there with 'two terms in a row'. 14:52
14:52 pmurias joined
pmichaud agreed 14:52
masak because, um, that's essentially what it is.
diakopter pmurias: o|
pmichaud let's see how it's parsing
(btw, I'm updating the rakudo spectest graphs in background, should have results in a couple of hours) 14:53
masak pmichaud++
pmurias diakopter: o| is?
pmichaud does svn up, rebuilds STD.pm
pmurias diakopter: hi 14:54
14:55 Psyche^ joined 14:56 alester joined 14:57 Lorn joined
moritz_ total", 32588, 84, 485, 2140, 35297, 0 14:57
that's the result of my tests_summary run 14:58
pmichaud my box needs to catch up on eight days of stuff :)
15:00 justatheory joined 15:01 payload joined
pmichaud oh, "bracketed infix" 15:03
| :dba('bracketed infix') '[' ~ ']' <infixish(1)> { $<O> = $<infixish><O>; $<sym> = $<infixish><sym>; }
masak huh.
so something like 1 [+] 1,1,1 would be 4? 15:04
moritz_ I don't think a bracketed infix is an infix
std: 1 [+] 1
p6eval std 28871: OUTPUT«ok 00:01 104m␤»
moritz_ oh.
pmichaud S03: Any ordinary infix operator may be enclosed in square brackets
with the same meaning. You may therefore use square brackets 15:05
...
moritz_ 'oh' again
so it can be either an infix or a listop, depending on whether an op or a term is expected
pmichaud correct.
moritz_ wicked.
15:06 nihiliad joined
diakopter std: 1 [[+]] 1 15:06
p6eval std 28871: OUTPUT«ok 00:01 104m␤»
jnthn std: my $foo; $foo[[.]]bar 15:13
p6eval std 28871: OUTPUT«[31m===[0mSORRY![31m===[0m␤Obsolete use of . to concatenate strings; in Perl 6 please use ~ instead at /tmp/A6yskiQVfj line 1:␤------> [32mmy $foo; $foo[[.][33m⏏[31m]bar[0m␤FAILED 00:01 109m␤»
jnthn :-)
diakopter std: my $foo; $foo[[~]]bar 15:15
p6eval std 28871: OUTPUT«[31m===[0mSORRY![31m===[0m␤Two terms in a row at /tmp/E6VG17RcUp line 1:␤------> [32mmy $foo; $foo[[~]][33m⏏[31mbar[0m␤ expecting any of:␤ POST␤ bracketed infix␤ infix stopper␤ postfix␤ postfix_prefix_meta_operator␤ standard stopper␤
..statement modifier loop␤ terminator␤ …
diakopter std: my $foo; $foo[[[]]]bar
p6eval std 28871: OUTPUT«[31m===[0mSORRY![31m===[0m␤Two terms in a row at /tmp/P3N9itXjis line 1:␤------> [32mmy $foo; $foo[[[]]][33m⏏[31mbar[0m␤ expecting any of:␤ POST␤ bracketed infix␤ infix stopper␤ postfix␤ postfix_prefix_meta_operator␤ standard stopper␤
..statement modifier loop␤ terminator␤…
diakopter std: my $foo; $foo[[]]bar
p6eval std 28871: OUTPUT«[31m===[0mSORRY![31m===[0m␤Two terms in a row at /tmp/UyubxGPCBe line 1:␤------> [32mmy $foo; $foo[[]][33m⏏[31mbar[0m␤ expecting any of:␤ POST␤ bracketed infix␤ infix stopper␤ postfix␤ postfix_prefix_meta_operator␤ standard stopper␤
..statement modifier loop␤ terminator␤ w…
15:16 pnate joined 15:27 pnate2 joined 15:35 Bzek_ joined 15:45 am0c joined 15:56 payload joined 15:59 justatheory joined 16:00 dalek joined 16:01 lumi joined
colomon pmichaud: It seems to me laziness in meta-operator generating is the only sane approach. Unless many of them could just be executed on the fly using some meta function, which is kind of how I thought it was already done... 16:01
16:10 meppl joined, cdarroch joined 16:12 jaffa8 joined
masak is it true? is Perl 5 on a monthly schedule nowadays? 16:17
KyleHa masak: That appears to be true. This is the second release of a Perl 5.11, one month after the first. 16:22
masak that is extremely good news, by me.
KyleHa I think so too. 16:23
16:24 pmurias joined
takadonet Seems like the perl 5 community really want to start making progress again 16:26
and i'm glad :)
16:27 lumi joined, xinming_ joined
KyleHa has a hard time making progress, preferring instead to order progress for delivery. 16:29
masak :) 16:30
make: *** No rule to make target `progress'. Stop. 16:31
KyleHa make: *** No rule to make target `practice'. Stop.
masak make: *** No rule to make target `whoopie'. Stop. 16:32
16:35 PerlPilot joined
KyleHa Lexicality or lexicalness? 16:37
masak it might depend. are you referring to the phenomenon or the property? :) 16:38
KyleHa I think I'll just avoid such silly words.
16:40 NorwayGeek joined 16:42 dukeleto joined
Tene I prefer lexi-cool. 16:42
PerlPilot Is feather dead?
dukeleto feather is down for me 16:44
diakopter yeah looks like the feather irssi folks timed out 25 min ago 16:49
16:52 cotto joined 16:54 stephenlb joined 16:58 payload joined 17:06 quietfanatic joined, Praveen_ joined
KyleHa (git-svn)++ # keep working when feather is down 17:06
perlpilot git++ you mean; screw that svn business :)
moritz_ #perl6book meeting in 54min 17:07
colomon is slightly depressed to need to write a Perl 5 program for $work. 17:10
17:11 nbrown_ joined 17:17 lumi joined
colomon Oh noes! My script is too slow even in Perl 5! Gack.... 17:30
TimToady hi guys, you're on screen at NetLogic
17:30 lumi joined
masak hi, screen at NetLogic! \\o/ 17:31
stephenlb :D 17:32
moritz_ is at home, not on screen ;-)
17:32 nbrown joined
masak ah, the splintered identity of the modern life! 17:33
KyleHa welcomes our new NetLogic overlookers. 17:42
17:55 tann1 left
masak #perl6book meeting time! 18:02
18:05 Confield joined 18:10 zamolxes joined, lumi joined 18:11 rfordinal joined, perlpilot joined
colomon pmichaud: just remembered my other question, about overloading cmp.... 18:14
18:19 dalek joined 18:20 dalek joined 18:21 hercynium joined, NorwayGeek joined 18:22 Juerd joined, dukeleto_ joined, pmichaud joined
colomon rakudo: say 0.5 cmp 1; 18:23
p6eval rakudo 073d5e: OUTPUT«-1␤»
colomon rakudo: say (1/2) cmp 1;
p6eval rakudo 073d5e: OUTPUT«Multiple Dispatch: No suitable candidate found for 'cmp', with signature 'PP->I'␤in Main (file src/gen_setting.pm, line 295)␤»
KyleHa rakudo: say (sub{} ~~ Callable && !(sub{} !~~ Callable)); 18:24
p6eval rakudo 073d5e: OUTPUT«0␤»
18:25 pugs_svn joined
dalek p-rx: 6cd6c1a | pmichaud++ | src/HLL/Grammar.pm:
Add default <ws> rule.
 18:25
p-rx: 9924922 | pmichaud++ | src/NQP/ (2 files):
Reorganize code a bit, initial term:sym<identifier> (subcall) implementation.
p-rx: b079e0d | pmichaud++ | src/NQP/ (2 files):
Add basic statements and terminators.
p-rx: a4bfb16 | pmichaud++ | build/Makefile.in:
Refactor Makefile in preparation for nqp tests.
p-rx: 4ed10a8 | pmichaud++ | (5 files):
Add goal-matching syntax to P6Regex.
pugs_svn r28872 | kyle++ | [t/spec] Test for RT #69919 18:29
r28873 | kyle++ | [t/spec] Tests for RT #69915
r28874 | kyle++ | [t/spec] Tests for RT #69869
r28875 | kyle++ | [t/spec] Tests for RT #69762
18:38 hugme joined
dalek ok: 026f34f | masak++ | (2 files):
[bin/book-to-latex] script to generate a .tex file

  $ perl bin/book-to-latex > book.tex
  $ pdflatex book.tex
Now, could someone who knows how please make a Makefile that does the above?		 18:43
18:46 jnthn joined 18:53 desertm4x joined 18:55 Tinus_ joined
moritz_ does anybody know where p6eval comes from? 18:55
I am confused.
it seems to live on neither of the usual hosts 18:56
KyleHa It says it comes from 209.9.237.232. 18:57
moritz_ the usual hosts are: timtowtdi.org, 80.237.201.115 18:58
or diakopter++'s server, wait, that's where it's from 18:59
KyleHa That IP appears to be allocated to a company in Sterling, VA.
moritz_ why can't I find the process then?
moritz_ too dumb
I have been looking for half an hour on the wrong host
KyleHa Try 'netstat -p' (as root) and look for connections to freenode.
19:01 lumi joined 19:03 glurks joined 19:07 FCO joined 19:12 NorwayGeek joined 19:15 dalek joined 19:17 Su-Shee left
japhb pmichaud, jnthn: I've got a patch to add mkdir() to Rakudo's src/builtins/io.pir ... how would you like it? 19:24
moritz_ japhb: rakudobug@perl.org (so it goes into RT)
japhb moritz_, thank you, will do
moritz_, as attachment, or inline? 19:25
moritz_ japhb: attachement
(at least that's what I prefer) 19:26
japhb incoming 19:28
jnthn ducks 19:29
yaz op! 19:30
dalek ok: f027e5e | moritz++ | src/ (5 files):
PseudoPod starts at =head0, so decrease the heading level by one everywhere
 19:32
19:34 icwiener joined 19:36 lumi joined
pugs_svn r28876 | moritz++ | [t/spec] test the now non-special delegation to hash and array attributes 19:38
dalek kudo: 0117dd3 | pmichaud++ | docs/spectest-progress.csv:
spectest-progress.csv update: 452 files, 27582 (72.3% of 38174) pass, 265 fail

S02-builtin_data_types/capture.rakudo 16 - unflattened captures can be passed to subs S02-lexical-conventions/unicode.rakudo aborted 5 test(s) S03-junctions/autothreading.rakudo aborted 78 test(s) S06-multi/proto.rakudo passed 2 unplanned test(s) S06-signature/defaults.t aborted 2 test(s) S06-signature/optional.rakudo aborted 3 test(s) S10-packages/basic.rakudo aborted 41 test(s) S12-attributes/delegation.rakudo aborted 16 test(s) S12-attributes/instance.rakudo aborted 88 test(s) S12-methods/parallel-dispatch.t aborted 21 test(s) integration/man-or-boy.t aborted 10 test(s)		 19:42
moritz_ pmichaud: that's quite a few failures which I don't see here
pmichaud: have you tried a git clean -xdf and regeneration 19:43
?
jnthn That looks odd, yeah. 19:52
I've not seen anyone else turn in a failure report like that.
autothreading.rakudo we know to fail on 64-bit, but the rest, no.
19:52 KyleHa left
pugs_svn r28877 | moritz++ | [t/spec] typos 19:53
19:58 spinclad joined
pmichaud I've not tried that -- didn't know it was needed. 19:59
I can re-run from a fresh git checkout. 20:00
moritz_ git-clean -xdf is enough
pmichaud what does that do, exactly?
moritz_ removes all files under version control, even those that are ignored 20:01
TimToady I may be late for phone call
moritz_ so don't do that if you have patches lying around ;-)
pmichaud ....removes all files *not* under version control
moritz_ erm, yes, sorry
moritz_ shouldn't type today
pmichaud ...as it just deleted all of my files I had lying around.
moritz_ oh, I'm so sorry 20:02
pmichaud np, I don't think I had anything too important
it listed all of the files it deleted, and I think I'm safe.
moritz_ anyway, it's not something that should happen
pmichaud I'll start from a fresh checkout.
20:05 Schwern joined 20:06 __ash__ joined
jaffa8 hi 20:07
Do you remember me? 20:09
moritz_ does
jaffa8 I told you I made perl 5 to perl 6 regular expression converter 20:10
the google does not show it.
so much time has been passed...and it does not appear in google
moritz_ does it have links from somewhere else?
jaffa8 yes
it is not like a dangling pointer 20:11
moritz_ does it contain a good title? text that's visible without javascript? allowed by robots.txt 20:12
20:12 nwc10 joined
jaffa8 there is not much text without javascript 20:12
nwc10 TimToady: how late might late be? :-)
TimToady don't know how long this meeting will run 20:13
are you guys done already? :)
nwc10 no, just started. With Allison.
TimToady I think we're wrapping up now
nwc10 chromatic is running it
20:13 fax joined
japhb In the proto projects.list, what is epoxy-resin? There does not seem to be a project of that name on github .... 20:13
nwc10 rah! I don't have to take the minutes. 20:14
20:16 lumi joined 20:17 nihiliad joined
dalek ok: cac7c9f | (Solomon Foster)++ | src/multi-dispatch.pod:
Add missing t to arbitrary. Rewrite first sentence of paragraph to flow more naturally (IMO).
 20:19
20:23 nbrown joined 20:28 justatheory joined 20:31 nwc10 left 20:33 meppl joined 20:40 dukeleto joined
colomon rakudo: my Int $a; say $a.WHAT 20:48
p6eval rakudo 073d5e: OUTPUT«Int()␤»
dalek ok: 437783b | moritz++ | src/multi-dispatch.pod:
[MMD] further clarifications requested by colomon++
 20:55
20:57 Confield left
moritz_ vmbrasseur: welcome ;-) 20:58
vmbrasseur moritz_: Thanks! I've been lurking for a while, where "lurking" == "forgetting I had IRC open". :-) 20:59
phenny vmbrasseur: 15 Oct 07:18Z <moritz_> tell vmbrasseur re contribution info, you might be interested in <www.perlmonks.org/?node_id=771635> and <www.perlmonks.org/?node_id=780001>. Also feel free to ask on #perl6 if you need a pugs commit bit to change perl6.org
20:59 cotto_work joined
moritz_ wow, that's an old message ;-) 21:00
vmbrasseur Dang, yeah it is. 21:01
phenny a bot? 21:02
moritz_ yes
we have quite a few of those here ;-)
vmbrasseur Looks like 8 of 'em, unless my listing is wrong (possible/likely). 21:03
moritz_ plus a few passive ones 21:04
for example is a pure logging bot
vmbrasseur I'd imagine the logs around here get pretty considerable at times. How're they stored? DB or otherwise? 21:05
moritz_ DB 21:06
2M rows so far
vmbrasseur Oh, that's not so bad then. 21:07
moritz_ which also includes #parrot and a few other channels
jnthn 2M? Sheesh, we natter a lot. :-)
Tene BAR BAR BAR BAR BAR 21:08
moritz_ well, audreyt used to chat a lot ;-)
and I imported old logs starting from 2005 21:09
vmbrasseur I'm surprised there's ONLY 2M lines if it goes back from 2005. 21:10
moritz_ rakudo: say 2e6 / (4 * 365) 21:11
p6eval rakudo 073d5e: OUTPUT«1369.86301369863␤»
moritz_ still more than 1k lines per day ;-)
vmbrasseur Judging from my scrollback for the past few days most of those are you and jnthn. ;-) 21:13
moritz_ only if you discount the bots ;-)
actually the largest single contributor is joins/leaves 21:14
which make up about 0.4M lines
vmbrasseur Oh, good point. I hadn't considered that those get logged as well. 21:15
21:17 lumi joined 21:19 Whiteknight joined 21:21 rfordinal left 21:22 namenlos joined
dalek kudo: c4e6d0f | (Geoffrey Broadwell)++ | src/builtins/io.pir:
Implement mkdir
 21:32
jnthn japhb++ 21:33
japhb :-) 21:34
Thanks, moritz_
jnthn (in the setting with inline PIR may have been nice too, but it's still nice to have :-))
japhb jnthn, ah, sorry, just grepped for chdir and put it in the same place in the same style.
jnthn japhb: It's fine, no need to apologize. You just added a feature. :-)
We're still very much in the process of moving things into the setting. 21:35
jaffa8 has rakudo been optimised yet? 21:37
moritz_ a wee bit
pmichaud No.
Not really.
i.e., it's still slow. :)
jaffa8 50 lines /s
?
ok 21:38
DO you know what is the cofee mug incident? 21:39
moritz_ poor mug ;-)
japhb s/mug/mugs 21:41
moritz_ I wasn't there, just read about it :)
jaffa8 WHat did you read? 21:42
vmbrasseur News to me. What's the coffee mugs incident? (link?)
jaffa8 see part about Topaz 21:43
www.perlfoundation.org/perl6/index....mentations
jnthn pmichaud: My fingers are crossed for parsing speed improvements too. :-)
moritz_ books.google.com/books?id=NYgzEwH5t...amp;q=perl 6 mug&f=false
jnthn pmichaud: The nqp part of nqp-rx looks to be coming on well. :-)
21:44 tak11 joined
japhb vmbrasseur, jaffa8: www.spidereyeballs.com/os5/set1/sma..._9705.html 21:44
pmichaud jnthn: it all seems to be coming along well. :) But I'm stuck on a parsing issue at the moment, which means I'm adding some debugging/tracing tools to the regex engine. 21:45
jnthn I guess those will be good ot have in the long run. :-) 21:46
vmbrasseur japhb: Thanks for the link.
moritz_ pmichaud: add goal matching to get better error messages ;-)
pmichaud I already added goal matching :)
moritz_ I've blogged again: perlgeek.de/blog-en/perl-6/perl-6-t....writeback
jaffa8 Jon Orwant, the actor? 21:47
Who is Jon Orwant?
jnthn "(Typically that's masak)" :-D 21:48
japhb jaffa8, google 'perl Jon Orwant' 21:49
jaffa8 ok 21:52
bye
21:54 lumi joined 22:14 ozymandias_ joined 22:21 d4l3k_ joined 22:28 lumi joined 22:35 ihrd joined
dalek p-rx: 31f8ae7 | pmichaud++ | src/NQP/ (2 files):
Add some (non-working) block rules to NQP.
 22:37
p-rx: 9104507 | pmichaud++ | src/ (4 files):
Add some primitive regex debugging capabilities.
p-rx: 9681684 | pmichaud++ | src/NQP/ (2 files):
[nqp]: Add statement_control:sym<if> .
22:42 meteorjay joined 22:44 hercynium_ joined 22:45 Apocal joined 22:47 Aisling joined 22:48 ihrd left 23:19 rfordinal3643 joined 23:21 dalek joined
jnthn -> sleep 23:24
23:25 rfordinal joined 23:27 lumi_ joined 23:38 SmokeMachine joined 23:49 markmont joined 23:50 quietfanatic joined 23:51 payload joined 23:55 lumi joined 23:58 Schwern joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!