»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo:, alpha:, pugs:, std:, or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by lichtkind on 5 March 2010.
CokeBot9000 lue: in the states too or just the UK? 00:03
lue well, it's the season premeire in the US. It happened in the UK April 4th. 00:05
( makes me wish I were in Britain :) )
CokeBot9000 lue++ 00:08
lue I believe the UK always makes better quality versions of our US shows. The entire Syfy channel looks completely uninteresting most of the time 00:09
( except for Ghost Hunters, Destination Truth, and anytime they show a british sci-fi :) ) 00:10
m6locks lol
syfy channel was epic, saw it a couple of days 00:11
taped the first three star trek movies, they showed it
lue :O
m6locks it was some ten years ago... 00:12
maybe 15
lue They are good sci-fi shows, they just come off as being a bit... american (and look at me, an american! That's bad :P)
CokeBot9000 lue: just wait for TorchWood:NY. =-) 00:13
lue :*( I miss ye, torchwood ( I'd rather now see a show on UNIT :) ) 00:14
m6locks www.youtube.com/watch?v=9ezZgAl6aN8
diakopter sorear: that's AWESOME 00:16
circular stack frames
00:21 tewk joined 00:33 pausenclown joined
pausenclown rakudo: grammar G { token TOP { <nums> }; token { \d <nums>? }; }; say G.parse('123abc'); 00:39
p6eval rakudo 78faa0: OUTPUT«Method 'nums' not found for invocant of class 'G'␤current instr.: 'perl6;G;TOP' pc 431 (EVAL_1:183)␤»
pausenclown rakudo: grammar G { token TOP { <nums> }; token nums { \d <nums>? }; }; say G.parse('123abc');
p6eval rakudo 78faa0: OUTPUT«123␤»
pausenclown rakudo: grammar G { token TOP { <nums>; say $/<nums> }; token nums { \d <nums>? }; }; G.parse('123abc'); 00:40
p6eval rakudo 78faa0: OUTPUT«Malformed regex at line 11, near "TOP { <num"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
pausenclown rakudo: grammar G { token TOP { <nums>; { say $/<nums> } }; token nums { \d <nums>? }; }; G.parse('123abc');
p6eval rakudo 78faa0: OUTPUT«Malformed regex at line 11, near "TOP { <num"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
pausenclown rakudo: grammar G { token TOP { <nums> { say $/<nums> } }; token nums { \d <nums>? }; }; G.parse('123abc');
p6eval rakudo 78faa0: OUTPUT«123␤» 00:41
pausenclown rakudo: grammar G { token TOP { <nums> { say $/<nums><nums> } }; token nums { \d <nums>? }; }; G.parse('123abc');
p6eval rakudo 78faa0: OUTPUT«23␤»
pausenclown rakudo: grammar G { token TOP { <nums> { say $/<nums><nums><nums> } }; token nums { \d <nums>? }; }; G.parse('123abc');
p6eval rakudo 78faa0: OUTPUT«23␤»
00:41 plobsing joined
pausenclown rakudo: grammar G { token TOP { <nums> { say $/<nums><nums><nums><nums> } }; token nums { \d <nums>? }; }; G.parse('123abc'); 00:42
p6eval rakudo 78faa0: OUTPUT«3␤»
pausenclown oh. sorry for the noise. i thought i'm in query with p6eval 00:43
rakudo: grammar G { token TOP { <nums> { say $/<nums>[1] } }; token nums { \d <nums>? }; }; G.parse('123abc'); 00:44
p6eval rakudo 78faa0: OUTPUT«Any()␤»
sorear pausenclown: there is a "CPAN6" project, but the level of secrecy together with the level of ambition gives me very little confidence in the project 00:45
lue that's fine with me at least, it's pretty quiet right now :)
sorear pausenclown: in the meantime, we have the very well designed 'proto'
00:45 ruoso joined
pausenclown why do i need for <nums> above to get the "3"? 00:46
sorear: i failed to install that. maybe i should give it a shot again.
00:47 TiMBuS left
lue is writing an emacs highlighting file for P6, and striving to learn as little elisp as possible :) 00:47
pausenclown
.oO( if emacs wouldn't be so ugly )
00:49
i mean, UI wise. 00:50
lue well, I'd write one for Kate, but emacs highlighting is written in an actual *programming* language, which makes it much easier for dynamic highlighting (new keywords, etc.)
it's not that ugly IMO :)
pausenclown i never got warm with it. 00:51
00:52 Trashlord joined
pausenclown i like Kate but i miss search&replace over files 00:52
rakudo: grammar G { my $i = 1; token TOP { <nums> { say '#' ~ $/<nums> } }; token nums { { say $i++ } \d <nums>? }; }; G.parse('123abc'); 00:53
p6eval rakudo 78faa0: OUTPUT«1␤2␤3␤4␤#123␤» 00:54
lue (just don't ask me to use vim. I developed a distaste with it early on, where I had to shut the the terminal because I couldn't exit vim (last I recall, ^C didn't work))
s/shut the/shut down/
pausenclown rakudo: grammar G { my $i = 1; token TOP { <nums> { say '#' ~ $/<nums> } }; token nums { { say $i++, '>', $/ } \d <nums>? }; }; G.parse('123abc'); 00:55
p6eval rakudo 78faa0: OUTPUT«1>␤2>␤3>␤4>␤#123␤»
pausenclown rakudo: grammar G { my $i = 1; token TOP { <nums> { say '#' ~ $/<nums> } }; token nums { \d { say $i++, '>', $/ } <nums>? }; }; G.parse('123abc');
p6eval rakudo 78faa0: OUTPUT«1>1␤2>2␤3>3␤#123␤»
pausenclown mmh. 00:56
sorear lue: How did you discover ^X^C?
emacs is, if anything, harder to exit than vim 00:57
vim at least responds to ^C with Type :quit<Enter> to exit Vim
emacs responds with C-c
pausenclown
.oO( u have just pressed the doorbell button. to call the inhabitants you need to use the gong in the backyard. thanks for your cooperation )
00:58
lue exiting is easy! ( esp. when using graphical emacs :) ) 01:01
But I developed an early hate towards vi and vim when I was installing gentoo(?) or something and had no GUI. I had to restart to exit, because I couldn't figure it out :/
(mind you, my first days at linux! but still...) 01:02
afk (sorry!)
pausenclown i use joe for command line editing. supports the old WordStar™ shortcuts =) 01:04
01:07 holli joined, holli left
CokeBot9000 lue: they're filming Series 4 now. 01:07
chromatic Any objections to pushing a Rakudo branch to work with Parrot immutable strings? 01:10
pausenclown damn router 01:11
jnthn chromatic: None from me. 01:12
chromatic git push origin immutable_strings ? 01:13
jnthn I *think* so. 01:15
jnthn mostly does local branches... 01:16
chromatic Fingers crossed. 01:17
01:18 M_o_C left
chromatic Shows up on GH anyway. That has to be something. 01:21
colomon \o/ 01:24
jnthn: I'm still looking at the mandelbrot failure (slowly) in the background. 01:26
It's dying after outputting the first few lines of the results when I ask it to do a 501 x 501 image. 01:27
but no errors show up in valgrind or gdb. It's just "program exited normally".
jnthn Oddness. 01:28
colomon: You got the program somewhere I can try it here?
01:29 pausenclown left, alester joined
jnthn colomon: I'm going to sleep Real Soon Now, but I can try it under my debugger tomorrow. 01:29
colomon github.com/colomon/mandelbrot/blob/...t-color.pl 01:31
just pushed it so the latest version is up there.
Huh. It's like it's exiting normally. maybe just a bug?
01:32 wknight8111 left
jnthn But does it say "Done!"; ? 01:34
colomon it does.
I just added that to see if it did.
jnthn Ah.
Maybe some more sutble bug then, rather than a crash. 01:35
colomon which is utterly bizarre, because I'm pretty sure subdivide cannot return fewer than 501 elements.
anyway, I'm testing further here at the moment. :)
wow, subdivide is not working at all. 01:38
01:43 alester left 01:47 nihiliad joined
colomon rakudo: sub subdivide($low, $high, $count) { (^$count).map({ $low + ($_ / ($count - 1)) * ($high - $low) }); }; say subdivide(0, 1, 501).elems 01:47
p6eval rakudo 78faa0: ( no output )
colomon :\ 01:48
jnthn Locally, gives 501. 01:50
colomon in the script, gives 6.
ah, 501 in the REPL here, too. maybe just a timeout? 01:51
01:53 alester joined
jnthn -> sleep 01:54
TimToady o/
colomon \o 01:56
01:57 Psyche^ joined
TimToady wonders if it's a return or some other control exception gone awry 02:00
colomon (^$count).map({ $low + ($_ / ($count - 1)) * ($high - $low) }); is the construct 02:01
02:01 Psyche^ is now known as Patterner
TimToady what if $count-1 is 0? 02:01
then div by 0 02:02
well, but you're passing 501
hmm
colomon right 02:03
and the code works beautifully for smaller numbers. 02:05
TimToady yeah, gives 501 here too
but takes a while
so I'm guessing timeout
colomon thing is, in my script it's giving just the first 6 elements. :\ 02:06
TimToady pugs: sub subdivide($low, $high, $count) { (^$count).map({ $low + ($_ / ($count - 1)) * ($high - $low) }); }; say subdivide(0, 1, 501).elems
p6eval pugs: OUTPUT«501␤»
TimToady 6, maybe it's summing the digits :) 02:07
colomon I wish it were so simple. :) 02:08
TimToady it's in an eager context in the script?
colomon yes
say subdivide($upper-right.re, $lower-left.re, $height).elems; 02:09
say subdivide($upper-right.re, $lower-left.re, $height).perl;
TimToady gather/take bug?
colomon 6
(-2, -1.995, -1.99, -1.985, -1.98, -1.975)
nope, it's map, so it's a MapIterator.
could be a map bug.
TimToady I don't suppose map is implemented with gather/take inside? 02:10
colomon Nope, there definitely is a MapIterator.
colomon doesn't trust gather / take either... 02:11
TimToady well, but that doesn't seem to say anything about the innards of MapIterator
lue oh hello again!
colomon MapIterator is all in PIR with no trace of gather that I can see. 02:13
TimToady does it use any global variables? :)
still think maybe it's getting some weird exception propagating out of the map 02:16
colomon umm... no? (my PIR is not that great.)
let me see if I can simplify the trouble script. 02:17
wow, this is a new piece of ugly. 02:21
wow, maybe it really is 5+1 = 6 ... 02:22
sort of. 02:23
rakudo: sub subdivide($low, $high, $count) { (^$count).map({ $low + ($_ / ($count - 1)) * ($high - $low) }); }; say subdivide(-2, 1/2, "501").elems; 02:24
p6eval rakudo 78faa0: OUTPUT«6␤»
colomon rakudo: say (^"501").perl
p6eval rakudo 78faa0: OUTPUT«0..^"501"␤»
colomon rakudo: say (^"501").eager.perl 02:25
p6eval rakudo 78faa0: OUTPUT«(0, 1, 2, 3, 4, 5)␤»
colomon Is that correct behavior (ie bug is in my code) or wrong (ie bug is in Range, which, okay, is also my code)
02:27 drbean_ joined
TimToady rakudo: say '0'..'501' 02:27
p6eval rakudo 78faa0:
..OUTPUT«012345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913…
TimToady rakudo: say '0'..^'501' 02:28
p6eval rakudo 78faa0:
..OUTPUT«012345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913…
TimToady rakudo: say 0..501
p6eval rakudo 78faa0:
..OUTPUT«012345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913…
colomon rakudo: say 0..^'501'
p6eval rakudo 78faa0: OUTPUT«012345␤»
TimToady it should decide before it starts which type it's doing and just do it 02:29
hmm
rakudo: say ^'zz'
p6eval rakudo 78faa0: ( no output )
TimToady rakudo: say 0..^'5' 02:30
p6eval rakudo 78faa0: OUTPUT«01234␤»
TimToady rakudo: say 0..^'50'
p6eval rakudo 78faa0: OUTPUT«012345␤»
TimToady rakudo: say 0..^'5000000000' 02:31
p6eval rakudo 78faa0: OUTPUT«012345␤»
colomon rakudo: say 5 cmp '501'
p6eval rakudo 78faa0: OUTPUT«-1␤»
colomon oh, that's it. 02:32
TimToady rakudo: say 0...^'501'
p6eval rakudo 78faa0: OUTPUT«0␤»
TimToady rakudo: say 0...'501'
p6eval rakudo 78faa0: OUTPUT«012345␤»
02:32 rgrau joined
colomon it's counting up an Int, but the comparison function is comparing strings. 02:32
TimToady yeah, just defaulting to leg
gee, why hasn't someone specced that yet...<looks around> 02:33
certainly would have been friendlier to blow up in this case
colomon yup.
TimToady yet another thing in my list of things to do first.... 02:34
colomon Range iteration is still a bit messed up anyway.
and that's mostly my fault. :)
TimToady well, I need to afk for a while. 02:36
colomon check ques
can we assume that 0..'50' should be the same as 0..50, somehow?
TimToady certainly if one arg is numeric and the other looks-like-num, I think so
colomon okay, I'm writing a spectest for that. :) 02:37
TimToady colomon++
laters &
colomon o/
lue bye o/ 02:41
pugssvn r30399 | colomon++ | [t/spec] Test for the 0..'50' case, properly fudged to work around Rakudo's current issues with it.
colomon Mandelbrot script (patched) is still running, off to bed for me. o/ 02:45
03:00 envi^home joined, rv2733 left, hercynium joined 03:03 meppl joined
diakopter hi 03:08
wazzup #perl6
lue 'ello o/ 03:09
03:09 dalek joined
diakopter perlesque: my int $a = 2; sub int foo (int $b, int $c) { System::Console.WriteLine($b); return $c; }; System::Console.WriteLine(5 + foo(55, 3)); 03:10
p6eval perlesque: OUTPUT«55␤8␤»
diakopter there, now return values of named sub calls actually work in expressions. strongly typed.
lue \o/ 03:11
does anyone know what language (other than P6) has comments like #`(), where three characters open and 1 closes ? 03:12
diakopter I don't know, but I'm glad I don't have to implement the parsing of that. 03:13
lue then I guess it's custom implementation time :) 03:19
03:20 alester left, _jaldhar joined 03:23 alester joined
diakopter I mean, I'm glad it's already implemented in STD.pm6 03:25
lue gah! elisp regex is so annoying. :*( 03:31
is there a command in perl regex where you can match to anything (ascii, non-ascii, etc.), for example to match the comment in # and #`() 03:35
diakopter do you mean . 03:36
or do you mean "anything but a right parens" 03:37
lue for example. If i had #`([text here]), how would I match [text here] so that anything allowed as a comment is matched? 03:38
same with #[text here], where the end of the comment is a \n
diakopter /#\([^)]*\)/ 03:39
and
lue basically I just want to see how much better perl is than elisp :)
diakopter /#[^\n].*(?:\n|$)/
er 03:40
/#[^\n]*(?:\n|$)/
lue figures. while asking, I'm perusing through the elisp reference, and for THE FIRST TIME, I see . , which matches everything except newline :)
diakopter yes. same with perl.
lue *duh*
I just have to remember the \\ for everything, because ( and [ are not special, by default. In elisp regex, you must use \( and \[ to make them do what you want them to do in a regex. 03:41
diakopter but repetitions are greedy (consume as much as they can) by default, so you have to exclude whatever comes next to match the shortest
lue so something involving [^...] then? 03:42
diakopter yes, as above
lue (In addition, regexes are stored as a string, so the backslash in \( must be escaped, resulting in \\( , \\[ , \\] , \\), and so on. :/) 03:43
"#\\[^\n\\]*\n" might work in elisp, for example. 03:44
03:44 meppel joined 03:45 meppl left, meppel is now known as meppl 03:47 dalek left 03:49 dalek joined, _jaldhar left
lue out of curiosity, what can be in a variable name (bar the sigil/twigil)? [looking] 03:49
03:59 _jaldhar joined 04:15 mikehh joined
lue I wonder, if I told elisp to match any opening delimiter, when I tell it to match any closing delimeter, would it make sure they matched? 04:22
04:49 am0c joined 04:50 _jaldhar left 04:59 ReiniUrban joined 05:00 rurban left, ReiniUrban is now known as rurban 05:02 cls_bsd left 05:06 meppl left
lue afk 05:21
05:21 and joined
and Hi all! 05:25
There are those who have used the socket in rakudo?
05:28 test joined, iblechbot joined 05:29 test left 05:31 molaf joined
diakopter got bottom-up type inference working for initializers 05:35
that's a departure from Perl 6, but that's ok.
05:37 nihiliad left 05:49 alester left, snarkyboojum joined 06:03 meppl joined 06:10 slavik1 joined
slavik1 grammars can have methods? 06:10
diakopter perlesque: my $a = 1; sub int foo (int $b, int $c) { say($b); return $c; }; loop (my $x=0;$x<100;$x+=1) {say($a + (foo($x, 2) + $x)); 1;}; 06:31
p6eval perlesque:
..OUTPUT«0␤3␤1␤4␤2␤5␤3␤6␤4␤7␤5␤8␤6␤9␤7␤10␤8␤11␤9␤12␤10␤13␤11␤14␤12␤15␤13␤16␤14␤17␤15␤18␤16␤19␤17␤20␤18␤21␤19␤22␤20␤23␤21␤24␤22␤25␤23␤26␤24␤27␤25␤28␤26␤29␤27␤30␤28␤31␤29␤32␤30␤33␤31␤34␤32␤35␤33␤36␤34␤37␤35␤38␤36␤39␤37␤40␤38␤41␤39␤42␤40␤43␤41␤44␤42␤45␤43␤
06:31 gfx joined
slavik1 diakopter: ? 06:36
diakopter perlesque: say(my $a = 332)
p6eval perlesque: OUTPUT«332␤»
diakopter slavik1: it's a thing I'm working on.
slavik1 do tell :)
diakopter well 06:37
where do I begin.
slavik1 an elevator sales pitch?
diakopter it's a compiler compiler for the CLR (.NET and/or Mono)
slavik1 in p6?
diakopter using it, I'm making this language called "perlesque" which is made up of a tiny subset of perl5 and a tiny subset of perl6 06:38
no it's written in C#
slavik1 ok
diakopter it compiles to CIL
slavik1 right
diakopter I mean, my compiler compiler compiles to CIL
slavik1 through another version of compile and you lost me 06:39
what the compiler you are compiling?
diakopter I was just clarifying, in case by "right" you meant "I know that C# compiles to CIL". I was clarifying that my compiler-compiler generates compilers that emit CIL.
slavik1 I see 06:40
yes
I understand that
diakopter ok :)
slavik1 are you using grammars/regexes?
diakopter yes, and they're p6-ish, but not as expressive. 06:41
but expressive enough to quickly add syntax & actions
slavik1 I think I get it
diakopter all in 1 source file
slavik1 you have your own yacc which produces a compiler not in binary (like gcc) but a compiler in CIL which is then run on whatever CIL runs on.
did I get it right?
diakopter yes.
:D
slavik1 mentioning yacc would've made it easier :) 06:42
hmm
diakopter (Mono 2.6 on Mac, *n*x, or Windows, or msft CLR on windows)
slavik1 technically, if the grammar/etc was standard or something, you could have compiler compilers that targeted different things 06:43
diakopter yes.
06:43 molaf left
slavik1 so then we can have a perl6 compiler in binary, jvm, cil, etc. 06:43
diakopter yeah
slavik1 but why would write it in C# and not Perl? :P 06:44
diakopter I didn't mention yacc b/c I've never used yacc
slavik1 oh
neither have I :)
I just know what it does
diakopter nor any other compiler-compiler
06:44 justatheory joined
diakopter ... in theory. 06:44
but anyways,
I want to use perlesque as an assembly language 06:45
slavik1 definately sounds cool
diakopter to which to target an emitter from STD/viv
slavik1 viv?
diakopter so that TimToady's standard parser/front-end can compile itself to perlesque
viv is short for V->IV
or five->six 06:46
slavik1 ahh
diakopter er
hahahaha
slavik1 wait, are you using just enough of grammars/regexes to get perl6 compiler to compile?
diakopter VI->V
SIX->FIVE
HA
slavik1 heh
must be too late
diakopter slavik1: yea, just enough
slavik1 I mised the 5->4 :P
sounds neat
diakopter to answer your other question 06:47
why not perl
b/c I wanted a strongly-typed language
slavik1 does mono/.net have java like management extensions?
I see
06:47 Trashlord left
diakopter I don't know. what are java's management extensions 06:47
slavik1 then why not java? (had to be asked)
imagine you have a jvm running
diakopter jconsole? 06:48
slavik1 yes
somewhat
if you go to browsing mbeans you'll see what I mean
imagine you have a web server written in java
diakopter ok..
diakopter thinks of tomcat.
slavik1 you can have management beans that would store number of requests served, longest time, shortest, average 06:49
diakopter: exactly
diakopter: although tomcat is a 'container'
for example, in Jboss, you can create a JDBC connection pool
diakopter isn't a Java programmer, but I've used/debugged it here and there 06:50
slavik1 and JMX allows you to see how many connections are already used and how many are available
diakopter ok
slavik1 not only that, but you can stop/start sub systems this way
diakopter broadly, the answer to your question would be "yes"
slavik1 k
will have to look it up
diakopter I doubt they're called "management extensions" though 06:51
slavik1 it will be nice when parrot has this.
diakopter sounds like lingo
well
slavik1 and, via JMX, you can get a dump of all thread stacks :)
diakopter barely able to bite his tongue
06:52 iblechbot left
diakopter I chose clr over jvm b/c I've had good experiences with Mono, and I hate Eclipse so much, and there was a great compiler library that was MIT-licensed that I started with, 06:52
slavik1 lol 06:53
fair enough
in my view eclipse is nice, when you through lots of memory at it
I am also finding that java as a language is beginning to slowly die ... in favor of groovy.
diakopter and b/c I'd already flipped around from JavaScript to C# to Perl and back >10 times over the past 3 years...
meh; it's all the same if they can be decompiled to each other reasonably readably. 06:54
like all the 100+ clr languages
disassembled, I mean.
chromatic I would like to see such management extensions for Parrot as well. 06:55
diakopter yeah, every VM's gotta have its "hot stuff" it feels it needs to hype so that the VM still sounds hip.
slavik1 groovy is dynamic though
06:55 meppl left
diakopter (I meant "hot stuff" to refer to Groovy) 06:55
slavik1 and it doesn't look like Java ... looking at java code makes me want to go to another profession 06:56
diakopter: what do you use for Perl code? 06:57
diakopter perl.exe?
slavik1 editor/ide
diakopter heh
Notepad++
slavik1 on unix?
diakopter no
wynders 06:58
slavik1 you don't code Perl on unix?
diakopter no.
slavik1 :(
poor soul
either way, try geany :)
diakopter geany
slavik1 yes 06:59
it's an awesome editor
diakopter Notepad++ has a very very responsive UI
slavik1 it doesn't run on Linux
diakopter very pretty text rendering
I don't need it to run on Linux.
slavik1 doesn't have a windows system
diakopter I don't need you to have a windows system
slavik1 lol
diakopter :P 07:00
I wasn't trying to convert you
I was stating the reason I like Notepad++
slavik1 I was ...
I failed :(
diakopter I doubt something built on GTK+ can be as responsive as Notepad++
but I'll try it.
and...... sourceforge is down. 07:01
slavik1 FLTK is the way to go?
yea
some upgrade
badmonkey on thinkgeek gets a discount :)
diakopter I don't see a button to add the thinkgeek models to my cart. 07:02
especially since thinkgeek is down as well.
slavik1 the site is coded in php, hence functionality is not possible
07:03 dalek left 07:07 snarkyboojum left 07:08 dalek joined 07:09 am0c left
diakopter hmm 07:09
time for another perlesquesketch; thanks for attending! 07:10
this week: JFDI times 7
next week: JFDI again, pounding out class/field/method declarations, implicit return expressions, 07:12
slavik1 ...
I need a second brain
07:13 slavik1 left 07:17 Levex joined
diakopter hm 07:20
actually
at some point implementing syntax/semantics in perlesque becomes more difficult than writing emitters in viv for the same functionality
07:25 kaare joined, kaare is now known as Guest33828 07:26 snarkyboojum joined 07:27 Guest33828 left 07:28 devsundar1 joined 07:30 Levex_ joined 07:34 Levex left 07:39 Levex_ is now known as Levex, Levex left, Levex joined
moritz_ good morning lambdamoose 07:52
07:54 pmurias joined
pmurias diakopter: if perlesque is meant to be used as an intermidiate form and propably for writing parts of the runtime why do you intend to add syntax sugar such as implicit return expressions? 07:55
07:56 justatheory left 07:57 agentzh joined 08:09 Levex left 08:20 plobsing left 08:33 JimmyZ joined, devsundar1 left, meppl joined 08:41 JimmyZ left 08:45 gfx left 08:54 cls_bsd joined 09:34 sorear joined 09:37 pausenclown joined 09:38 stepnem joined
pausenclown rakudo: Grammar G { token TOP { <nums> { say $/<nums> } }; token nums { \d <nums> }; }; G.parse('123'); 09:39
p6eval rakudo 78faa0: OUTPUT«Confused at line 11, near "Grammar G "␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
pausenclown rakudo: grammar G { token TOP { <nums> { say $/<nums> } }; token nums { \d <nums> }; }; G.parse('123'); 09:40
p6eval rakudo 78faa0: ( no output )
pausenclown rakudo: grammar G { token TOP { <nums> { say $/<nums> } }; token nums { \d <nums>? }; }; G.parse('123');
p6eval rakudo 78faa0: OUTPUT«123␤»
pausenclown rakudo: grammar G { token TOP { <nums> { say $/<nums><nums> } }; token nums { \d <nums>? }; }; G.parse('123'); 09:41
p6eval rakudo 78faa0: OUTPUT«23␤»
pausenclown rakudo: grammar G { token TOP { <nums> { say $/<nums><nums><nums> } }; token nums { \d <nums>? }; }; G.parse('123');
09:41 masak joined
p6eval rakudo 78faa0: OUTPUT«23␤» 09:41
masak oh hai, #perl6!
pausenclown why isnt that 3?
hi masak
rakudo: grammar G { token TOP { <nums> { say $/.perl } }; token nums { \d <nums>? }; }; G.parse('123'); 09:43
p6eval rakudo 78faa0: OUTPUT«Method 'perl' not found for invocant of class 'Regex;Match'␤current instr.: 'perl6;G;_block70' pc 733 (EVAL_1:281)␤»
pausenclown i wish that would work. =(
moritz_ pausenclown: there was a Match.perl implementation in alpha 09:44
but it currently blocks on Match objects being too parroty
and not Perl6y enough
for which in turn I have a partial fix that waits for review from pmichaud
pausenclown moritz++ 09:45
rakudo: grammar G { token TOP { <nums> { say $/<nums><nums>[0] } }; token nums { \d <nums>? }; }; G.parse('123'); 09:47
p6eval rakudo 78faa0: OUTPUT«23␤»
pausenclown rakudo: grammar G { token TOP { <nums> { say $/<nums><nums>[1] } }; token nums { \d <nums>? }; }; G.parse('123');
p6eval rakudo 78faa0: OUTPUT«Any()␤»
dalek ok: b083ea1 | moritz++ | (3 files):
Enable UTF-8 rendering of LaTeX

problems. Also contains small markup fixes
09:48
pausenclown moritz: i am about to add support for ELEMENT declarations, but the recursive way they are declared in the XML grammar leads me to parsing problems. 09:49
masak rakudo: grammar G { token TOP { <nums> { say $/<nums><nums>[0]<nums> } }; token nums { \d <nums>? }; }; G.parse("123") 09:50
p6eval rakudo 78faa0: OUTPUT«3␤»
masak pausenclown: there you go.
there are some treacherous bugs in there, though.
moritz_ oh
masak pausenclown: the ? is a quantifier, and all quantifiers create an array in the match object.
pausenclown i know. 09:51
rakudo: grammar G { token TOP { <seq> }; token seq { <num> ( , <num> ) ? }; token num { \d } }; G.parse('1,2,3'); 09:57
p6eval rakudo 78faa0: OUTPUT«Malformed regex at line 11, near "seq { <num"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
pmurias diakopter: ping 09:59
perlesque: 1+1 10:01
p6eval perlesque: ( no output )
10:02 ejs joined
pmurias moritz_: perlesque appears to be in p6eval how did you build it on mono? 10:02
pausenclown rakudo: grammar G { token TOP { <seq> }; token seq { <num> ( \, <num> ) ? }; }; say '#' ~ G.parse('1,2,3'); 10:05
p6eval rakudo 78faa0: OUTPUT«Method 'num' not found for invocant of class 'G'␤current instr.: 'perl6;G;seq' pc 721 (EVAL_1:276)␤»
pausenclown rakudo: grammar G { token TOP { <seq> }; token seq { <num> ( \, <num> ) ? }; token num { \d }; }; say '#' ~ G.parse('1,2,3');
p6eval rakudo 78faa0: OUTPUT«#1,2␤»
moritz_ pmurias: diakopter did, not me
pmurias ah
pausenclown so why doesnt the line above parse 1,2,3 but 1,2? 10:07
ah. 10:08
moritz_ because it can only match one ,
pausenclown rakudo: grammar G { token TOP { <seq> }; token seq { <num> ( \, <seq> ) ? }; token num { \d }; }; say '#' ~ G.parse('1,2,3');
p6eval rakudo 78faa0: OUTPUT«#1,2,3␤»
pausenclown silly me
rakudo: grammar G { token TOP { <seq> { say $/<seq><num><seq> } }; token seq { <num> ( \, <seq> ) ? }; token num { \d }; }; G.parse('1,2,3'); 10:14
p6eval rakudo 78faa0: OUTPUT«Any()␤»
pausenclown grrrmpf. why isn't that 2,3? 10:15
moritz_ because the ( ) introduce a positional capture
pausenclown so? 10:16
moritz_ rakudo: grammar G { token TOP { <seq> { say $<seq>[0]<num>[0]<seq> } }; token seq { <num> ( \, <seq> ) ? }; token num { \d }; }; G.parse('1,2,3');
p6eval rakudo 78faa0: OUTPUT«Method 'postcircumfix:<{ }>' not found for invocant of class ''␤current instr.: '!postcircumfix:<{ }>' pc 13944 (src/builtins/Code.pir:120)␤»
moritz_ hrmpf
pausenclown: the <seq> capture is inside the positional capture 10:17
rakudo: grammar G { token TOP { <seq> { say $<seq><num><seq> } }; token seq { <num> [ \, <seq> ]? }; token num { \d }; }; G.parse('1,2,3');
p6eval rakudo 78faa0: OUTPUT«Any()␤»
moritz_ oh wait 10:18
$<seq><num> doesn't have any more captures
pausenclown facepalm
of course!
moritz_ rakudo: grammar G { token TOP { <seq> { say $<seq><seq><num> } }; token seq { <num> [ \, <seq> ]? }; token num { \d }; }; G.parse('1,2,3');
p6eval rakudo 78faa0: OUTPUT«2,3␤» 10:19
moritz_ rakudo: grammar G { token TOP { <seq> { say $<seq><seq><seq><num> } }; token seq { <num> [ \, <seq> ]? }; token num { \d }; }; G.parse('1,2,3');
p6eval rakudo 78faa0: OUTPUT«2␤»
moritz_ rakudo: grammar G { token TOP { <seq> { say $<seq><seq><seq><seq><num> } }; token seq { <num> [ \, <seq> ]? }; token num { \d }; }; G.parse('1,2,3');
p6eval rakudo 78faa0: OUTPUT«3␤»
pausenclown funny. you do i right and suddenly it works :-D 10:20
moritz_ :-)
10:38 IllvilJa joined 10:45 pausenclown left 10:50 molaf joined
colomon so, I set mandelbrot-color.pl to work on a 501x501 before I went to bed last night. 8 hours later, it's still working. Using up 2.12 GB of RAM and 3.25 GB virtual memory. So pathetically slow and wasteful, but still miles beyond rakudo alpha.... 10:56
masak aye.
10:57 rurban left 10:58 wknight8111 joined, rurban joined
masak rakudo: say eval "" 11:05
p6eval rakudo 78faa0: OUTPUT«1␤»
masak may I ask... why 1? 11:06
moritz_ rakudo: say (eval "").WHAT
p6eval rakudo 78faa0: OUTPUT«Int()␤»
moritz_ it should return Nil, really
masak submits rakudobug 11:08
11:15 M_o_C joined 11:23 rgrau left 11:26 rurban left 11:27 rurban joined 11:41 masak left 11:45 Maddingue joined 11:51 agentzh left 12:00 ejs left
jnthn o/ 12:06
12:10 M_o_C left 12:20 M_o_C joined, JimmyZ joined 12:27 rv2733 joined
colomon \o 12:32
12:33 snarkyboojum left 12:34 M_o_C left 12:37 M_o_C joined
colomon jnthn: don't know if you've backlogged or not, TimToady++ and I solved the mandelbrot bug last night. 12:42
12:42 M_o_C left 12:46 M_o_C joined
rurban With Installed parrot 2.2: perl Configure.pl --parrot-config=/usr/bin/parrot_config Parrot revision r45386 required (currently r0) Thats lame 12:47
12:52 M_o_C left 12:53 M_o_C joined
jnthn colomon: Not yet. Was it Rakudobug or mandelbug? 12:59
12:59 ReiniUrban joined
colomon kind of both and specbug too. ;) 13:00
The question boiled down to: what should ^"501" do?
jnthn Ah!
Yeah, I can see where this is going... :-) 13:01
13:01 rurban left, ReiniUrban is now known as rurban
colomon rakudo: say ^"501" 13:10
p6eval rakudo 78faa0: OUTPUT«012345␤»
colomon of course, once you know that's the problem, the workaround for the script is trivial. 13:11
thus www.harmonyware.com/pictures/foo.png
I killed that one when it was about 88% complete because it was slogging my MBP. 13:12
From now on I'm running them on my Linux workhorse. :)
jnthn Wow, pretty! 13:14
But yeah, slooow. :-(
colomon it took about 9 hours to generate that much, and it seemed like at that point it was generating another row every half hour or so. :(
jnthn It got slower over time? 13:15
colomon I don't know.
It's hard to say, because the math gets harder as well.
jnthn Ah, OK.
colomon those outer layers go pretty quickly. 13:16
the inner stuff is the slowest.
however, it did consume an enormous amount of memory, so that may be slowing it down as well.
jnthn Yeah, I fear a leak. 13:17
colomon I'm going to run a series of timings on the Linux box to see what the time characteristics are.
but let's not obscure the big news here!
back in Dec, this script crashed most of the time on Rakudo (now alpha). 13:18
I'm not sure it ever went more than ten minutes and ended successfully.
today it went 9 hours, was still producing results, and only ended because I killed it.
\o/
moritz_ .u smiley 13:19
phenny moritz_: Sorry, no results for 'smiley'.
moritz_ .u smile
phenny U+2323 SMILE (⌣)
jnthn colomon: Well yes, that is progress on stability. :-) 13:20
colomon: Are you making heavy use of complex numbers?
colomon yes
jnthn Yeah, we're pretty inefficient on those for the moment.
colomon jnthn: sure. 13:21
jnthn After Rakudo * when I do the native types work, you could be able to have like: 13:22
colomon I still reckon it's a good test for general speed and stability. It's mostly looping and simple math (wrapped in a class) after all.
jnthn class Complex { has float $.re; has float $.im; }
And then you'll have one object with two fields that are floats.
colomon (That is, I don't expect it to have speed like tuned C at the moment.) 13:23
oooooo, cool.
Right now Complex is just two arbitrary types, and all the type resolution stuff has to be done for every operation, so it's not surprising it's not zippy. 13:27
jnthn Aye.
moritz_ "not zippy" is a nice euphemism :-)
13:28 clintongormley joined 13:46 masak joined
jnthn lolitsmasak 13:47
masak lol 13:48
jnthn is glad that masak is back a *lot* before 2038.
masak I realized that the Singularity will happen ~2035, so I wanted to contribute a bit to Perl 6 before that. 13:49
lue: how is it to travel beyond 2035, by the way? would you describe it as mostly utopian or mostly dystopian? 13:50
colomon jnthn: good news on the short front: going from 101 to 201 in mandelbrot size takes pretty much exactly as long as you'd expect (4x longer, that is). no sign at all of any additional slowdown. 13:53
jnthn Ah, OK. 13:54
Just general epic slowness then.
colomon well, memory usage is still a real issue
It got up to 3.x GB virtual on the MBP. 13:55
jnthn Ouch.
Do you keep a lot of values around in memory? 13:56
colomon and this implementation just exports each line of graphics as it is calculated, so memory usage should be quite light.
jnthn Right, so we're leaky.
:-(
colomon sounds like something that chromatic could look at.... 13:57
masak rakkudo: class A { method postcircumfix:<( )>($s) { say $s.WHAT } }; A.new().("OH HAI") 13:58
rakudo: class A { method postcircumfix:<( )>($s) { say $s.WHAT } }; A.new().("OH HAI")
p6eval rakudo 78faa0: OUTPUT«Capture()␤»
masak jnthn: Capture? not Str?
jnthn Dammit, had we not once managed to get to a point where while 1 { 1 } didn't leak? 13:59
Because it seems to be here for me.
masak: Correct.
masak When is Perl 6 going to stop leaking, dammit? :) 14:00
jnthn: and if I want my Str back, what do I do?
jnthn masak: You get one parameter to postcircumfix:<( )> always, and it's a Capture.
masak: 'cus ruoso said so... :-|
masak ok. ISTR you and pmichaud++ talking about this at YAPC::EU.
jnthn masak: Just unpack the capture with a sub-signature should work.
masak kthx 14:01
colomon oh, our monthly release is next week, isn't it?
jnthn masak: Yeah, I'm still not so fond of this. It feels inconsistent with the way postcircumfix:<[ ]> works for example.
masak it does.
jnthn But it was somehow important to the way smop works.
colomon: Think so.
rakudo: class A { method postcircumfix:<( )>( ($s) ) { say $s.WHAT } }; A.new().("OH HAI") 14:02
p6eval rakudo 78faa0: OUTPUT«Str()␤»
jnthn rakudo++
14:02 meppl left
masak oh, ok. 14:02
jnthn omfg, no wonder the Wien - Kiev train takes 35 hours. At one point, according to the timetable, it takes 8 hours to do 40km! 14:03
jnthn is trying to work out workshop travel plans for the next months... 14:04
The French and Ukrainian ones completely overlap. :-(
colomon It's funny: on the one hand, I feel like we've really accomplished something with the ng branch work. Rakudo is more powerful, and more stable, and easier to work with.
And on the other hand, there is so much more to do!
masak who would have thought; new problems lurked around the corner :P 14:05
jnthn colomon: I remember the days when Rakudo was roughly equal to NQP. 14:06
colomon: There was no OO support at all.
colomon: Not even hashes were properly in place.
Let alone grammars and regexes and all that goodness. 14:07
It gives some perspective. :-)
colomon It feels to me like we're right on the verge of having something that can recommended for general work (of things that don't have to be zippy, anyway ;)
but it also feels like we've only implemented 50% of everything. :) 14:08
jnthn Sure. Not everything in the spec is of equal utility or importance.
colomon I don't mean we've only done half the spec, I mean that we've only half-implemented the stuff we've done. 14:09
I know pretty much every single thing I've worked on could still use more work.
14:09 alester joined
jnthn Well, there's that too. :-) 14:09
14:10 ejs joined
cognominal jnthn, do you have submitted your 2 hours of talks for the french perl workshop? 14:10
jnthn cognominal: I'm...rather conflicted now, having been told I'm welcome at both, want to go to both, and now find the two overlap. 14:11
cognominal there is no problem scheduling your talks the first day of the fpw 14:12
but we can't force you to come :) 14:13
but we can remind you about Chartreuse
jnthn cognominal: I hoped something like that would work, but I fear even that isn't workable. FPW is 11th and 12th. UPW is 12th (+ social stuff on 13th rather than talks).
Granted I'm the only person that this affects, most likely, since few others try to do this level of workshop attendence. :-) 14:14
cognominal watches a map of Europe to see the rspective postion of Sweden, Ukraine and France. 14:16
jnthn This is why I need a private jet.
;-)
cognominal hoping that at this time Island will be irrelevant.
Iceland. 14:17
14:17 Alias joined
jnthn cognominal: lol. I'm supposed to be going there in 2 weeks too! 14:17
cognominal I am sure that in Ukraine they can bribe you with strong alcohol too.
jnthn Amongst other things, yes. :-) 14:18
pmurias diakopter: ping 14:20
14:24 molaf left, Trashlord joined 14:25 ejs left
masak how would you describe the semantics of the 'our' initializer? 14:25
jnthn Creates a lexically scoped alias to a thingy in the package. 14:26
Alias say what now?
oh
carry on
} # Descoping myself 14:27
masak alias, theory... good thing there's no such thing in CS as 'a masak' :)
Alias I promise to invent one
jnthn If I ever invent some really cool CS thing, I now know what to call it!
masak oh noes :)
Alias Maybe I can rename the monica filter in your honour
masak jnthn: I don't grok your description. in what package? 14:28
14:30 ejs joined
jnthn The current one. 14:32
rakudo: module Foo { sub lol() { our $x; $x = 42; }; sub rofl() { say $x }; }; Foo::lol; Foo::rofl; 14:33
p6eval rakudo 78faa0: OUTPUT«Symbol '$x' not predeclared in rofl␤current instr.: 'perl6;PCT;HLLCompiler;panic' pc 152 (compilers/pct/src/PCT/HLLCompiler.pir:108)␤»
jnthn rakudo: module Foo { sub lol() { our $x; $x = 42; }; sub rofl() { our $x; say $x }; }; Foo::lol; Foo::rofl;
p6eval rakudo 78faa0: OUTPUT«Can not find sub Foo::lol␤current instr.: 'perl6;Perl6Exception;throw' pc 14818 (src/builtins/Seq.pir:41)␤»
jnthn oh gah
masak :)
'my' subs...
jnthn rakudo: module Foo { our sub lol() { our $x; $x = 42; }; our sub rofl() { our $x; say $x }; }; Foo::lol; Foo::rofl;
Yeah :-)
p6eval rakudo 78faa0: OUTPUT«42␤» 14:34
jnthn Rakudo is winning against me.
masak \o/
jnthn masak: Anyway, what I was getting at is that those two both declare lexical aliases $x to the sme variable, $Foo::x
14:34 and left
masak ok, so 'our' means "put this var not in the lexical scope, but in the surrounding package -- or use the existing one, if one exists" 14:34
14:34 M_o_C left
masak rakudo: our $x = 42; say $x; { my $x = 5; say $x; { our $x; say $x } } 14:36
p6eval rakudo 78faa0: OUTPUT«42␤5␤42␤»
masak \o/
jnthn Right, the storage is in the package.
masak I think I grok it now. thanks.
so 'our' would play well with inheritance, right? but not 'my'? 14:37
jnthn Not sure what you mean. 14:39
That is, I see inheritance and these as pretty orthogonal.
masak oh, ok. 14:41
rakudo: class A { our $x = 42; say $x }; class B is A { our $x; say $x }
p6eval rakudo 78faa0: OUTPUT«42␤Any()␤»
jnthn B is a different package to A. You'd not expect those two to have any relation. 14:42
masak ah ok, A and B are separate packages, even though one inherits from the other.
jnthn Right, but inheritance only means something because the metaclass says it means something.
bacek o hai 14:43
masak bacek! \o/
jnthn о хай
bacek jnthn, привет :)
masak, aloha 14:44
jnthn
.oO( хай looks a little too close to...something else... :-) )
bacek jnthn, will you have time to try rukudo on immutable_strings_part1 branch in parrot?
.oO( Someone learn Russian too fast )
14:45
jnthn Well, only the *bad* bits of Russian. ;-) 14:46
bacek: I did try it out with chromatic last night.
bacek: We got to the bottom of the problem and found a workaround.
bacek: chromatic++ pushed a branch to the Rakudo repo.
(That builds with immutable_strings_part1)
bacek jnthn, ah-ha!
bacek switching from worrying about rakudo to teach jnthn more rude Russian 14:47
14:47 ejs left
bacek jnthn, how was your trip to Russia? 14:49
jnthn Short, but nice.
bacek jnthn, I hope you consumed only chilled vodka! 14:50
jnthn ;-)
I also used the duty free to bring some vodka back with me.
Less than half the price of what I could get it for here in Sweden.
bacek And about twice of the price from buying in Russia? :) 14:51
jnthn Yeah. But I'd gone hand luggage only, and you can't take liquids in that. :-( 14:52
bacek stupid "security"...
jnthn I guess I coulda tried, and offered to drink some of it to show it was really vodka, not explosive. :-) 14:53
masak jnthn: but an 'our' declaration is enough for a variable $x in the class A to be visible to everyone else as $A::x, no?
bacek
.oO( Check-in of 30 kilogrammes of Vodka can be suspicious )
jnthn bacek: heh, it didn't occur to me that I could use my baggage allowance solely to check in vodka!
masak: Aye. 14:54
masak excellent.
bacek why "our" visible outside of Class/Package??? 14:55
jnthn bacek: our variables are stored int he package
use "my" if you want invizible.
bacek jnthn, you have to learn more from Russians :)
jnthn bacek: (rude Russian) I was also told an amusing story about a guy who went to a money changer and asked to минетЬ rather than менять. :-) 14:56
bacek jnthn, гыыы :) 14:57
jnthn Good job there's bankomats, so tourists like me can't make such embarassing mistakes. :-)
14:57 alester left, ejs joined
masak such an awful play on words almost has to be followed up by some etymology... :) 15:01
where do these two verbs come from?
15:04 _jaldhar joined
jnthn Dunno, but менять ain't so far from the Slovak menit'. 15:05
15:06 alester joined
bacek masak, "менять" is a "change", "минетЬ" is way to close to "минет" which is "blowjob" :) 15:07
JimmyZ I guess the last patch was wrong in the immutable_strings, since there is new `replace` op.
15:07 M_o_C joined
masak bacek: I know what they mean -- checked with Google Translate -- but I was interested in why. 15:08
15:08 XaeroOne joined
bacek JimmyZ, "replace" is same as one of "substr" atm. 15:09
masak, hmm... Good question.
15:09 justatheory joined
JimmyZ bacek: github.com/rakudo/rakudo/commit/cde...dc82f3383, Doesn't just use replace instead ? 15:11
bacek masak, Russian version based on French "minette". They just happen to have almost same pronunciation
masak oh, ok. 15:12
15:12 ejs left
jnthn ...great, that's the first new French word I've learned all year. 15:12
:-)
bacek JimmyZ, I introduces "replace" after this patch. But I'll look at it closer.
JimmyZ oh
jnthn -> shops, bbiab 15:15
15:15 kaare joined 15:16 ruoso left, kaare is now known as Guest85744
JimmyZ bacek++ , an mad coder. 15:17
s/an/a/ 15:18
masak rakudo: "foo" ~~ /(oo)/; say $0.substr(0, 1)
p6eval rakudo 78faa0: OUTPUT«Method 'substr' not found for invocant of class 'Regex;Match'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
15:19 and joined
moritz_ that's because Regex::Match isn't Cool 15:20
masak but it should be, right?
15:20 Guest85744 left
moritz_ aye 15:21
but don't even think of submitting a bugreport
we already have a "$/ is not suffciently Perl6y" or so
masak it might please you to know that my first instinct was to look for one in RT :) 15:22
though I must confess I do not find it. 15:23
moritz_: do you?
moritz_ goes a-lookin'
#73462 15:24
pugssvn r30400 | pmurias++ | [mildew] added the optC option to bread 15:25
masak oh, that one.
yes, that's the underlying cause, of course.
(so I agree that no more RT tickets are needed) 15:26
moritz_ rakudo: my $x = Capture.new
p6eval rakudo 78faa0: ( no output )
moritz_ rakudo: my $x = Capture.new; say $x[0]
p6eval rakudo 78faa0: OUTPUT«Null PMC access in type()␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
moritz_ rakudo: my $x = Capture.new; say $x<a>
p6eval rakudo 78faa0: OUTPUT«Null PMC access in type()␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
moritz_ (probably a manifestation of RT #74110) 15:27
masak adds it to that ticket 15:28
cognominal I am compiling the rakudo usenqprx branch and I get Symbol 'self' not predeclared in return_handler_past when compiling Actions.pm 15:29
15:29 pyrimidine joined
moritz_ cognominal: uhm, that's a very old (and abandoned) branch 15:29
cognominal: is there a good reason to try to build it?
cognominal ho
what is the branch that uses nqprx? 15:30
moritz_ master
cognominal ok
moritz_, and it pulls nqprx parrot? 15:32
moritz_ it uses the nqprx that comes with parrot 15:33
cognominal I am confused with all these branches flying around :)
masak uses 'succeed' for the first time instead of 'break' 15:34
15:35 JimmyZ_ joined
moritz_ maybe I should delete some old branches 15:35
masak spring cleaning :) 15:36
moritz_ jnthn: am I right in assuming that the 'llsig' branch can be deleted?
it's from Oct 2009, and I believe it's merged 15:37
15:37 JimmyZ left, JimmyZ_ is now known as JimmyZ
cognominal moritz_, so parrot is now by default nqprx? 15:38
moritz_ cognominal: parrot ships with nqprx, yes
even though the executable is called parrot-nqp 15:39
15:41 justatheory left
jnthn moritz_: Yes. 15:51
moritz_ gone. 15:53
15:55 wknight8111 left 15:56 TiMBuS joined
lue hello! 15:59
masak hellue! 16:03
spock, a specification framework: code.google.com/p/spock/
looks pretty nice. I especially like the 'detailed information' in reason (4) in code.google.com/p/spock/wiki/WhySpock 16:04
lue I'll be in and out today, all day Doctor Who marathon \o/
masak: looks good. 16:06
2. Are you guys Trekkies by any chance? 16:08
Bingo! But watch out for equally subtle references to Battlestar Galactica, Dr. Who, and other fascinating life forms.
from the Spock FAQ :)
masak lue: sounds like your kind of thing :)
lue Alas, I'm still occupied with P6 major mode for emacs :) 16:10
masak no, I meant, the Trekkies/BSG/Dr. Who references :)
16:10 and left
lue I know :) 16:12
masak rakudo: class A { our sub foo() { 42 } }; my &f = &A::foo; say f
p6eval rakudo 78faa0: OUTPUT«42␤»
masak \o/
moritz_ there's *something* that works in Rakudo 16:13
lue The only bad part is that they're showing mainly the Season 4 specials :/ I had to wake up at 3AM if I wanted to see actual episodes.
masak: \o/
moritz_ rakudo: my %a; say %a<foo>; say %a.keys 16:21
p6eval rakudo 78faa0: OUTPUT«␤␤»
lue apparently, your key is a newline (interesting)
moritz_ nope 16:22
rakudo: my %foo; if %foo<a> {}; say %foo.perl
p6eval rakudo 78faa0: OUTPUT«{}␤»
lue ah (wonder why it spit out two newlines then) 16:23
moritz_ because of two say() calls
lue rakudo: my $_var = 3; say $_var;
p6eval rakudo 78faa0: OUTPUT«3␤»
lue oh (duh)
masak rakudo: sub foo(%hash? is copy) {}; foo
p6eval rakudo 78faa0: OUTPUT«Null PMC access in invoke()␤current instr.: 'perl6;Hash[];!STORE' pc 473702 (src/gen/core.pir:67931)␤... call repeated 1 times␤»
masak submits rakudobug
jnthn wtf
masak GGE++ # bugfinder 16:24
jnthn Two is copy bugs in two days. I should smack ma binder up.
masak jnthn: when are you going to get around to that hash slice business, by the way? turns out I was using it in GGE after all. :)
jnthn masak: Got half-done patch here. :-) 16:25
masak yay
colomon rakudo: my @a = 1...10; say @a[7/3]
p6eval rakudo 78faa0: OUTPUT«3␤»
lue ō.o
masak that's fine by me.
lue just wondering what the [7/3] is supposed to do :) 16:26
moritz_ indexing
colomon It's suppose to do [2]
what's weird is I don't understand how it works at the moment.
rakudo: my @a = 1...10; say @a[7/3.Num]
p6eval rakudo 78faa0: OUTPUT«3␤»
jnthn colomon: I'm sorta looking into some of that stuff along with the hash slicing bits. 16:27
colomon: I'm not very happy with how some that lot works.
colomon I've just gotten another of the
No applicable candidates found to dispatch to for 'postcircumfix:<[ ]>'. Available candidates are:
:(Mu : Int $i;; *%_)
:(Mu : Block $b;; *%_)
:(Mu : !whatever_dispatch_helper ;; *%_)
pugssvn r30401 | moritz++ | [t/spec] test for RT #61882, reading a non-existent hash item should not autovivify it
colomon errors, this time because I was passing a Num instead of an Int.
but it just worked fine in the test back there.
masak colomon: @a[2] is 3 in this case.
colomon masak: yes, my worry is that there is no Rat or Num case for [ ] 16:28
on Array, anyway.
masak oh, you're worried that it works :P
colomon yes, exactly.
masak s/that/because/
colomon that or because are both fine. :)
16:28 JimmyZ left
colomon our multi method postcircumfix:<[ ]> (Int $i) seems like the case that is being called, but how / why? 16:29
why does it fail in my code but work on the REPL?
jnthn I think there's something rotten somewhere in that area of the code, fwiw. 16:30
moritz_ is that area of code inside the state of Denmark? :-)
masak 629 bugs in RT. at this rate, we'll hit 650 next week. 16:31
moritz_ updated rakudo.org/status
jnthn masak: Depends on the rate of fixes too.
masak not really :)
or rather, I took that into account in my estimate. 16:32
but sure, consider it a challenge if that helps.
16:32 Alias left
colomon btw, in this case in my code, casting with .Int fixed the problem. 16:33
moritz_: exactly!
moritz_ colomon: any progress in reviewing that lazy assignment patch? 16:35
moritz_ tries polite nagging
colomon moritz_: not in the least, I'm afraid. :(
I'm still running at about half speed here.
moritz_ rakudo: sub foo(:$) {}; say &foo.signature.perl 16:36
p6eval rakudo 78faa0: OUTPUT«:(Any :())␤»
moritz_ it's ok
jnthn :-)
moritz_ uhm, is that correct?
I'd expected :$
jnthn rakudo: sub foo(:$) { say "ok"; }; my %h = '' => 1; foo(|%h)
p6eval rakudo 78faa0: OUTPUT«ok␤»
moritz_ rakudo: sub foo(:()) { say "ok"; }; my %h = '' => 1; foo(|%h) 16:37
p6eval rakudo 78faa0: OUTPUT«ok␤»
moritz_ ok, fine
jnthn moritz_: Signature.perl gets a tad...confused...when there's no name.
It's consistent with the other output it gives, fwiw.
rakudo: sub foo(:$a) {}; say &foo.signature.perl
p6eval rakudo 78faa0: OUTPUT«:(Any :a($a))␤»
moritz_ well, as long as the result works, why not?
jnthn That is, it doesn't collapse it if the variable name matches the parameter name.
Probably the guy who wrote Signature.perl was lazy. 16:38
TimToady std parses it as an anonymous named parameter
lue afk (doctor who)
jnthn TimToady: So does Rakudo.
colomon jnthn: you're saying he was a Perl programmer?
jnthn colomon: Yeah, and he was awesome!
Well, the empty-string-named named parameter anyway. :-) 16:39
masak still likes the term 'named anonymous (yes I know) parameter'
or was it 'anonymous named'? 16:41
jnthn Anonymous yet named can make sense. :-) 16:42
masak it can. it's just etymologically odd.
moritz_ anonymous named 16:43
masak so is 'atom' nowadays, I guess.
16:50 wknight8111 joined, lorenzo_miwi joined
moritz_ rakudo: sub foo(:$) {}; say &foo.signature.perl ~~ / ':' / 16:50
p6eval rakudo 78faa0: OUTPUT«:␤»
moritz_ rakudo: use Test; plan 1; sub foo(:$) {}; ok &foo.signature.perl ~~ / ':' /, 'bar'
p6eval rakudo 78faa0: OUTPUT«1..1␤ok 1 - bar␤»
16:50 lorenzo_miwi left
moritz_ why does this fail in my test file? 16:50
anyway, I have to run :( 16:51
16:55 pyrimidine left 17:04 mberends joined
masak lolitsmberends 17:04
mberends ahoy, #perl6!
phenny mberends: 14 Apr 10:46Z <masak> tell mberends that I'm glad things are happening with proto right now. I think the sanest way I can contribute long-term is to write that 'testable core' thingy. would like to get together some evening and brainstorm on that -- just to make sure I grok the new model.
mberends 3+ days ago! (was distracted teaching PHP Intermediate) 17:05
jnthn ahoj mberends!
17:06 k23z__ joined, k23z__ left, k23z__ joined
mberends ahoj! good thing you flew back just in time, jnthn 17:06
masak ahojte alle Leute :)
jnthn ...you just exploded my parser. 17:07
masak a bit like GGE is all in pieces now, in other words :/ 17:08
17:09 Kyril joined
masak though to be fair, I've gotten some tests passing today. 17:09
and the road ahead seems clear enough, just thorny.
jnthn Ah, *that's* why we'll hit 650 bugs. :-)
masak++ # becoming a master
er, master guy 17:10
masak yes, that's why.
jnthn :-)
masak I expect to keep doing this with the rest of the app cheese as much as I can this month.
jnthn \o/ 17:14
Hopefully as you discover bugs and they get fixed, it'll get less painful as you go along. 17:15
colomon what's the syntax to open a file for writing in p6? 17:16
jnthn open("file", :w)
colomon jnthn++
masak well, one good piece of news is that some of the things I'm fixing are errors in the app cheese itself, things that slipped through because Rakudo alpha was insufficiently strict.
17:18 envi^home left
diakopter pmurias: hi 17:20
after you've installed mono 2.6, you just type xbuild -p:Configuration=Release Sprixel.sln (actually the -p:Configuration=Release is optional) 17:21
17:23 Kyril left
diakopter pmurias: to answer your other question, yes. :) 17:26
(I shouldn't be considering adding "cloying" syntax such as implicit return expressions) 17:28
diakopter continues class declarations 17:29
17:32 ruoso joined
diakopter masak: the app cheese. 17:37
masak aye. 17:38
diakopter mm
is it a smelly cheese or a processed cheese
masak it's one with holes.
diakopter so a holy cheese.
masak exceedingly so. 17:39
jnthn switzerland++ 17:40
masak which, by an amazing coincidence, borders Austria. 17:41
and which, speaking of holes, isn't in the European Union, even though all its neighbours are. 17:42
rakudo: say $*IN.get 17:43
p6eval rakudo 78faa0: OUTPUT«Land der Berge, Land am Strome,␤»
jnthn masak: Not true.
masak I may be wrongly informed.
jnthn Liechtenstein is its neighbor, and is not in the EU either.
masak ah. 17:44
diakopter rakudo: while($_ = $*IN.get) { .say }
p6eval rakudo 78faa0: OUTPUT«Land der Berge, Land am Strome,␤Land der Äcker, Land der Dome,␤Land der Hämmer, zukunftsreich!␤Heimat bist du großer Söhne,␤Volk, begnadet für das Schöne,␤vielgerühmtes Österreich,␤vielgerühmtes Österreich!␤»
masak so it is. it's very lichtle and easy to miss :)
diakopter shouldn't rakudo require a space between while and ( 17:45
masak yes.
but I do believe there's an RT ticket for that.
rakudo: $*IN.get(3)[2].subst(/ä/, 'u').say
p6eval rakudo 78faa0: OUTPUT«No candidates found to invoke␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
masak rakudo: $*IN.lines(3)[2].subst(/ä/, 'u').say 17:46
p6eval rakudo 78faa0: OUTPUT«Land der Hummer, zukunftsreich!␤»
masak hee hee
diakopter hm. currently perlesque creates blocks with names '_Closure_1', '_Closure_2', etc 17:48
should I make it match nqp/rakudo's '_blockN' instead? :)
or is that something from parrot 17:49
okily dokily
jnthn diakopter: It's just an internal name. 17:50
diakopter: Generated by PAST, iirc.
Rather than anything Parrot-y.
diakopter ok.
jnthn You could call it _Llama_1 and it'd be fine
diakopter but Llamae might feel offended. 17:51
jnthn Who cares, they smell. 17:52
colomon has perl 6 writing out IGES files now, but cannot get his own software to recognize them. :( 17:53
17:58 alester left
diakopter colomon: missing BOM maybe? 18:00
colomon bill of materials?
diakopter byte order mark?
colomon IGES is ascii
80 columns, even.
diakopter o 18:01
colomon oh, I'm an idiot.
think I've been successfully writing out points, and thinking I was failing to write out lines. 18:03
okay, now I'm definitely failing to write out lines rather than writing points. :\ 18:09
18:10 hercynium left
colomon hey, that's 90% there... 18:12
diakopter TimToady: should I assume every postcircumfix:<{ }> node in STD's AST of itself is a Perl 5 hash lookup? 18:13
18:13 mberends left 18:17 rv2733 left 18:18 mberends joined
pmurias diakopter: installing mono 18:19
diakopter pmurias: :) 18:21
2.63?
2.6.3?
pmurias 2.6.3 18:22
colomon Argh. Finally got my own software to recognize the IGES files I'm writing, but now Rhino 3D won't recognize my "improved" versions of these files. Why can't anything ever be easy? 18:24
rurban I studied the chmod 755 dll/so problem. perl5 File::Copy never preserved file attributes, even not with syscopy. Will be ahrd to persuadae them that Extutils::Commands is broken. 18:25
Just a few platforms have special rules, like mac, Win32, mepix
diakopter rurban: as long as it's not the linux or *bsd ones that are broken. ;) 18:27
rurban I think it's better to let perl5 be broken, and fix that entirely in parrot configure hints (for hpux, cygwin already done)
diakopter o wait, mac is *bsd...
arnsholt Probably Mac OS 9 and below 18:28
pmurias rurban: is a failing test not enough to demonstrate the bug? 18:29
rurban we will see. 18:30
masak rakudo: my $x = Nil; say defined $x 18:34
p6eval rakudo 78faa0: OUTPUT«1␤»
masak rakudo: say defined Nil
p6eval rakudo 78faa0: OUTPUT«0␤»
masak should the definedness of Nil change when it's stored in a scalar variable? 18:35
diakopter rakudo: say defined Nil but Nil
p6eval rakudo 78faa0: OUTPUT«No applicable candidates found to dispatch to for 'infix:<does>'. Available candidates are:␤:(Mu $do-it-to-me, Role $r)␤:(Mu $do-it-to-me, ConcreteRole $r)␤␤current instr.: 'infix:<but>' pc 300187 (src/gen/core.pir:3700)␤»
diakopter erm. that's what she said.
colomon masak: I thought that Nil switched to undefined variable of the appropriate type as a scalar?
masak colomon: so the above is wrong? 18:36
colomon easier test:
rakudo: my Int $x = Nil; say $x.defined; say $x.WHAT 18:37
p6eval rakudo 78faa0: OUTPUT«Type check failed for assignment␤current instr.: '&die' pc 17293 (src/builtins/Junction.pir:404)␤»
masak I think that one's in RT already.
colomon if I'm right, this is just NYI
masak I think you are. S02 says that assigning or binding Nil should restore a scalar variable to its pristine state. 18:38
colomon \o/
so my $x = Nil should be Any()
masak aye.
baking & 18:39
18:40 masak left
colomon baking what? :) 18:40
lue hello! (oh! just missed you masak)
jnthn maybe baking @s
colomon woah, he's gone
lue he's baking blue police-box shaped cookies I hope :)
colomon mmmmm, @s 18:41
lue eh, he should bake scalars. Those would get him some money :)
18:42 alester joined
jnthn Hope he's not baking hash. ;-) 18:42
lue bake them in a 2D array at least :) 18:43
And scope them properly!
colomon jnthn: btw, on my big linux box, there is no unexpected slowdown going from mandelbrot 101 to 401 18:47
(that is to say 401 takes 16x as long)
of course, that machine has 8GB ram... 18:48
mberends colomon: is the trend in memory size similar to the trend in run time? 18:49
colomon mberends: I'm afraid it may be.
which would be bad, as memory usage for this should go up by N, not N^2 18:50
do you have a good way of testing that on linux or os x?
mberends no, I sometimes watch it using 'top', but the numbers are very volatile 18:51
colomon drat 18:52
mberends I'm going to look for a better tool
colomon I vaguely remember finding a memory monitoring tool for os x, but don't remember what it was now. 18:53
valgrind can do it, but it's very slow
mberends of course, any scripting language that cat read /proc could do it. Perl 5, for example 18:54
*can read
18:57 Kyril joined 18:58 iblechbot joined 19:02 tri1 joined 19:06 justatheory joined 19:11 alester left
lue rakudo: say `hello` 19:18
p6eval rakudo 78faa0: OUTPUT«Confused at line 11, near "say `hello"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
lue rakudo: say `hello'
p6eval rakudo 78faa0: OUTPUT«Confused at line 11, near "say `hello"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
lue ` is not a quoting construct, is it?
colomon ` was in p5 19:19
mberends only in Perl 5, not in Perl 6. You may be able to use qx{}. 19:20
colomon rakudo: say `ls`
p6eval rakudo 78faa0: OUTPUT«Confused at line 11, near "say `ls`"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
19:20 k23z__ left
lue yeah, I'm looking at the p5 syntax table (emacs), and was wondering if it still applies 19:20
colomon std: say `ls`
p6eval std 30401: OUTPUT«===SORRY!===␤Confused at /tmp/IbSY2WsKco line 1:␤------> say ⏏`ls`␤ expecting any of:␤ bracketed infix␤ infix or meta-infix␤ statement modifier loop␤Other potential difficulties:␤ Unsupported use of bare 'say'; in Perl 6 please use
...say if y…
lue here are all the symbols the perl5 major mode has to define in the syntax table (not counting sigils/comments): *+-/<=>\| 19:23
methinks I'll end up defining a lot more (imagine all the parentheses I might have to define o.o) 19:24
19:24 meppl joined 19:26 tri1 left 19:28 mantovani left, rv2733 joined
mberends colomon: you've got me trying to write a memory monitor utility now ;) 19:29
colomon \o/
cognominal www.appleinsider.com/articles/10/04...pairs.html "acts of God or other external causes" 19:32
sacrés (sic) avocats américain!!
lue sacré bleu! 19:33
cognominal: est-ce que tu parles français? 19:34
cognominal oops, I wanted to paste on #perlfr 19:35
sacrebleu!
lue ah. well, I'm only beginning to learn French. Right now we're at adjectives, and just did verbs ending in 'er'. :) 19:36
cognominal come to #perlfr on irc.perl.org if you want to perfect your french
lue I'll consider it once I finish French 1, I can barely say anything :) 19:38
19:40 tri1 joined 19:42 M_o_C left
diakopter perlesque: my $sub_a = sub int foo (int $arg0) { say(my $b = $arg0 * 2); return $b; }; $sub_a($sub_a($sub_a(5))); 19:47
p6eval perlesque: OUTPUT«10␤20␤40␤»
19:48 tri1 left
diakopter type inference for closures-as-expressions, and hence closures-as-variables 19:48
rakudo: my $sub_a = sub foo ($arg0) { say(my $b = $arg0 * 2); return $b; }; $sub_a($sub_a($sub_a(5))); 19:49
p6eval rakudo 78faa0: OUTPUT«10␤20␤40␤»
diakopter oh good
19:50 mantovani joined
diakopter perlesque: my $sub_a = sub int foo (int $arg0) { say(my $b = $arg0 + 4); return $b; }; my $last = 5; loop (my $x=0;$x<100000;$x+=1) { $last = $sub_a($last) } 19:53
p6eval perlesque:
..OUTPUT«9␤13␤17␤21␤25␤29␤33␤37␤41␤45␤49␤53␤57␤61␤65␤69␤73␤77␤81␤85␤89␤93␤97␤101␤105␤109␤113␤117␤121␤125␤129␤133␤137␤141␤145␤149␤153␤157␤161␤165␤169␤173␤177␤181␤185␤189␤193␤197␤201␤205␤209␤213␤217␤221␤225␤229␤233␤237␤241␤245␤249␤253␤257␤261␤265␤269␤273␤277␤281␤285␤289␤293␤297␤301􏿽xE2
diakopter rakudo: my $sub_a = sub foo ($arg0) { say(my $b = $arg0 + 4); return $b; }; my $last = 5; loop (my $x=0;$x<100000;$x+=1) { $last = $sub_a($last) }
p6eval rakudo 78faa0: 19:54
..OUTPUT«9␤13␤17␤21␤25␤29␤33␤37␤41␤45␤49␤53␤57␤61␤65␤69␤73␤77␤81␤85␤89␤93␤97␤101␤105␤109␤113␤117␤121␤125␤129␤133␤137␤141␤145␤149␤153␤157␤161␤165␤169␤173␤177␤181␤185␤189␤193␤197␤201␤205␤209␤213␤217␤221␤225␤229␤233␤237␤241␤245␤249␤253␤257␤261␤265␤269␤273␤277␤281␤285␤289␤293␤297␤301􏿽xE2
19:54 XaeroOne left
lue oooh 19:55
afk 20:02
20:30 TiMBuS left
diakopter 3.1s for perlesque to do 100million iterations of that on my machine 20:31
oh, with the say() removed. 20:32
heh
otherwise it's obviously IO bound
20:53 synth joined 20:59 ReiniUrban joined 21:01 rurban left, ReiniUrban is now known as rurban 21:04 clintongormley left 21:25 iblechbot left 21:36 s1n joined
moritz_ rakudo: sub foo($v, $w?, $x?, $y?){ say $v~"|"~$w~"|"~$x~"|"~$y}; foo(1,,3,); 21:44
p6eval rakudo 78faa0: OUTPUT«1|3|Any()|Any()␤»
21:45 dual joined
lue hello again 21:46
s1n ya know, no matter how slow rakudo is, it just can't be worse than the jvm :/ 21:48
so, when rakudo just takes a second or two to start, i'm okay with that
lue me too :) 21:49
mberends moritz_: did 3 move from the $x to the $w ? That's wrong, ain't it?
lue I actually don't get why people are concerned with things like "my computer takes 30 seconds to boot up! Let's reduce it to 5!". At least it boots in a reasonable amount of time
moritz_ mberends: it should be a parse error (and there's a ticket for that) 21:50
jnthn std: sub foo($v, $w?, $x?, $y?){ say $v~"|"~$w~"|"~$x~"|"~$y}; foo(1,,3,);
moritz_ std: 1,,3,
p6eval std 30401: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix , instead at /tmp/GgxdASfVG1 line 1:␤------> ?){ say $v~"|"~$w~"|"~$x~"|"~$y}; foo(1,⏏,3,);␤Parse failed␤FAILED 00:01 112m␤»
std 30401: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix , instead at /tmp/T8mmvUXZEd line 1:␤------> 1,⏏,3,␤Parse failed␤FAILED 00:01 108m␤»
moritz_ RT #54804 FWIW 21:51
21:53 _jaldhar left 21:57 cledu joined
pugssvn r30402 | moritz++ | [t/spec] test for RT #69492 22:04
moritz_ btw this test fails on a parrot version newer than the recommended one 22:05
jnthn :-( 22:06
lue Any particular reason why? 22:09
moritz_ because of a regression
lue ah 22:11
jnthn Given the Parrot 2.3 release would happen next week iirc, and we'll need to ship #28 on that, maybe worth reporting?
(If not already done)
lue rakudo: say 3 add 5 22:12
p6eval rakudo 78faa0: OUTPUT«Confused at line 11, near "say 3 add "␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
lue rakudo: say 3 plus 5
p6eval rakudo 78faa0: OUTPUT«Confused at line 11, near "say 3 plus"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
moritz_ jnthn: I wanted to bisec first, but it seems I'm too tired for that right now
lue 3rd tuesday right? In which case, the 20th, 3 days away. R* would release the 22nd 22:13
moritz_ nope 22:14
lue oh :/ 22:15
22:20 M_o_C joined
pmurias diakopter: after i build sprixel how should i run it? 22:24
22:25 justatheory left, justatheory joined
colomon lue: why didn't you tell me the dude from Life on Mars was on Doctor Who? 22:35
22:36 huf left
lue I never saw Life on Mars :) 22:39
mberends colomon: first cut of a process monitor to report maximum memory size: gist.github.com/369869 22:40
colomon mberends++ 22:44
mberends colomon: Linux only (/proc dependent) and only measures the immediate child, so no make processes etc (...yet) 22:45
colomon ...will that work with ./perl6?
mberends probably. must try it ;) 22:46
22:46 huf joined
mberends colomon: yes, 187MB process size for rakudo -v 22:47
colomon lue: Life on Mars (BBC) was fantastic, and starred the dude who plays the Master.
mberends++
jnthn mberends: What do you get for Rakudo just opening and sitting at the command line? 22:48
colomon I'll give it a whirl as soon as I'm doing making chocolate pudding for my wife, now watching her first episode of Dr. Who. :)
lue her first episode is the End of Time?
great one :)
mberends colomon: 187207680(bytes) # amd64 ubuntu 10.4 22:49
colomon lue: ah, I see she just switched over to curling, which was what she meant to turn on. I'm sure she'll be planning on watching those episodes of Who at a later date. :) 22:51
mberends watched "Victory of the Daleks" a few hours ago www.bbc.co.uk/programmes/b00s56d2 22:52
lue Ah well, I figure Part 2 of the End of Time (when it's split apart; right now is the streamlined 1-part "movie" style) is coming up soon :) 22:53
22:53 crythias joined
lue mberends: is that from the old series? I would love to get my hands on so 20th century DW 22:53
s/so/some/
mberends no, this is current series with Matt Smith as the 11th regenerated Doctor 22:55
colomon lue: I think he's gloating about seeing them ahead of us.... ;) 22:57
mberends gloats some more...
find archived episodes at tvunderground.org.ru/index.php?show...p;sid=3405 22:59
lue mberends: i can haz free tiket tu lundin? 23:00
( not really of course :) ) 23:01
mberends lue: the airport's closed, otherwise it would be a fun idea
lue oh yeah ( stupid volcano. can't believe that monster from Pompeii is back ;) ) 23:02
mberends every 1/2 day the authorities are rolling the grounding of planes forward another 1/2 day or more 23:03
lue just wondering; are the bees still disappearing? 23:04
mberends: hopefully that doesn't affect you in any way
pmurias diakopter: perlesque works here
lue *Ba-da-da* 1 hour until the 1-hour special detailing all about the Doctor Who-niverse. 2 hours until the season premeire 23:05
cognominal rakudo: say 'a' ~~ m/ <? a > a/ 23:06
p6eval rakudo 78faa0: OUTPUT«Confused at line 11, near "say 'a' ~"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
jnthn mberends: It's kinda scary and amazing to watch it.
lue jnthn: where is you in the situation?
mberends lue: fortunately there are no health risks, just the risk to aircraft. The trains, ferries and buses are full.
jnthn lue: I made it home from Moscow within 4-5 hours before the point things all shut down here.
lue: So I got lucky.
lue good :) 23:07
jnthn Otherwise I'd be stuck in Russia now.
cognominal STD: say 'a' ~~ m/ <? a > a/
lue
.oO(trains are awesome; there ought to be more in the US)
cognominal std: say 'a' ~~ m/ <? a > a/
p6eval std 30402: OUTPUT«===SORRY!===␤Unrecognized regex assertion at /tmp/IPIMYlizKk line 1:␤------> say 'a' ~~ m/ <?⏏ a > a/␤ expecting any of:␤ assertion␤ name␤Parse failed␤FAILED 00:01 111m␤»
jnthn Though my visa would last through to the 20th, so I'd probably have made the most of my stuckness. :-)
mberends jnthn: yeah, you had incredible timing this week.
jnthn mberends: Aye, getting back without flying woulda turned into a slight adventure. 23:08
mberends: Train to Petersburg, train from there to Helsinki, boat.
I actually have done that in reverse before now. 23:09
cognominal What is the syntax for looking-ahead a sting?
jnthn Though they didn't have the super-fast Moscow - Petersburg trains last time.
lue Spirit Tracks and Phantom's Hourglass in one trip ( oh wait, probably not that fun in real life ;) )
cognominal *string
jnthn cognominal: <?before foo>
colomon jnthn: the good news on mandelbrot: the expected (proper) ratios for timing hold up out to the 501x501 case on my Linux box.
jnthn colomon: OK, so we're just slow.
23:09 orafu left
jnthn But not worseningly slow. 23:09
23:09 orafu joined
jnthn That's good to know. Thanks for doing the analysis. colomon++ 23:09
cognominal thx
colomon and mberends++ wrote the code to let me test memory characteristics. 23:10
which is what I'll tackle next.
I don't expect such good news there, alas...
23:10 pmurias left
jnthn oops, I just vi a-20gb-file 23:13
lue O.o /o\
jnthn: if ye can open that in good time, ye can run rakudo without pause. Good Luck!
.oO(words from the East)
23:14
jnthn: what is this 20gb file? the linux kernel? 23:15
23:16 rgrau joined
mberends it's the todo list for rakudo * ;) 23:16
lue what is on the todo list for R*? I'm missing something :/ 23:17
jnthn lue: Wikipedia dump.
lue: See docs/ROADMAP in Rakudo git repo. 23:18
lue hopefully not all of Wikipedia is 20gb. ( that means it has to get bigger! that's way too small :) ) 23:19
jnthn lue: Current revisions only. 23:20
And text only, no images etc. 23:21
lue where is the REPL implemented? I've been curious about that.
23:24 nihiliad joined
lue ~ 30 minutes until an explanatory special on DW, 1 hour and 30 minutes until the premiere. afk! 23:27
23:29 dual_ joined 23:31 dual left 23:32 dual joined
colomon mberends: \o/ 23:34
23:35 dual_ left 23:46 LionMadeOfLions joined, c1sung joined
diakopter \o\ 23:54
//o/ 23:55
er
\/o/
er
/o/
mberends /o\ 23:56
23:58 M_o_C left