pugscode.org/ | Brave souls needed to fix 'make Install'! Non-brave souls should avoid it for now | sial.org/pbot/perl6
Set by mncharity on 4 June 2007.
Coke what's the name of the thing that's pointed at in for @joe -> $bag_o_donuts {stuff} ? 03:33
Aankhen`` Whatcha mean?
Oh, $bag_o_donuts?
I'm not sure. Isn't it just a loop variable? 03:34
Tene a block parameter 03:35
?
Aankhen`` shrugs. 03:36
Coke "loop variable" it is. lame, but servicable. 03:40
Jmax "Enterprise temporary reliable iterator variable container" 03:42
... "With aliasing*" "* only in Pugs Enterprise Edition" 03:43
Coke I'm trying to the 07-for.t sanity test passing on parrot. 03:44
er, to *get* 03:45
gaal it's indeed a block parameter, and more generally, -> SIG BLOCK 04:22
TimToady a functional programmer should probably be calling it a lambda parameter... 04:25
Aankhen`` wonders if you could pass other types of Routine objects. 04:26
(Routine is what all of them inherit from, right?)
TimToady sure, as long as you call it via statement_control:<for>(@list, &routine) 04:28
Aankhen`` Awww.
Tene Okay, I've completely rebuilt pugs from scratch, and I'm still getting a segfault immediately on startup. 05:04
pasteling "tene" at 24.2.71.129 pasted "pugs segfault" (27 lines, 2.5K) at sial.org/pbot/25385 05:05
TimToady hmm, a SEGV at 0 often means that something tried to jump through a null pointer 05:16
but I'd guess the problem is likelier to be in libperl.so than in pugs 05:17
I wonder if it's possibly related to the problems I've been having with looping and slow tests on both FC6 and F7. When was your last update? 05:20
Tene A couple of days ago. 05:22
On FC
Well, now on F, I suppose
Coke nopaste.snit.ch:8001/10315 05:28
TimToady it's claiming that all 6 versions of Perl suck, I guess... :) 05:31
Tene: I wonder if it's finding the wrong libperl.so. I have a sneaking suspicion they may have rearrange the library search order to put site_perl and vendor_perl before the top-level lib, and maybe there's an old libperl.so out there? 05:38
when I built my own perl and put it into /usr/local/bin and /usr/local/lib it ran fine 05:39
Caelum strace it and see where it dies 05:40
Tene Caelum: I posted an strace
Caelum oh duh
Tene Heh. :)
TimToady on my machine all the previous versions of libperl.so are symlinks to the 5.8.8 version 05:41
Caelum I noticed that there's an inconsistency in the pugs build process, if you do /usr/bin/perl Makefile.PL it uses that perl to link with, but in other parts it does just "perl Makefile.PL" using the perl in your path instead of the one it built with 05:42
TimToady but that hasn't changed since Jan
Tene also, I only have one perl
Caelum maybe I should fix this...
TimToady I don't think that can be this problem, since I'm having trouble on an F7 machine with only one perl 05:44
and I had no trouble when I did /usr/local/bin/perl Makefile.PL on my other machine 05:45
Caelum no, different problem...was just making an observation
it builds fine, but some of the extensions are built with "perl Makefile.PL" 05:46
TimToady as long as they don't try to link in a different perl... 05:47
could introduce some other subtle differences, I suppose
Caelum they do 05:48
Caelum figured out how to build pugs without totally freezing his computer... turns out if I have folding@home running at the same time, the pugs build can't coexist with it 05:56
Aankhen`` Oh, I'd forgotten all about Folding@Home. 06:01
I could probably speed up my builds if I'd just remember to turn it off.
TimToady it's just really weird that my FC6 machine slowed down at exactly the same time as my F7 machine...makes me think it isn't really fedora. my yum update log doesn't show any thing perly...
though of course the F7 upgrade touched all kinds of stuff 06:02
amnesiac TimToady, FC is evil :(
FC is the root of all evil :)
TimToady of all the files pugs has open while running the slow test (propcharset takes 40 minutes), none of them has been updated recently except pugs. 06:29
TimToady so I think it's either pugs or the linux kernel itself, which did get an update a few days ago 06:34
Jmax should i test on openbsd? 06:58
TimToady nobody can stop you. :) 07:00
Jmax ok :) 07:01
Aankhen`` Jmax: Come on, no one sane tries to do anything on OpenBSD except route packets. ;-) 07:06
thoughtpolice i use openbsd :( 07:19
OpenBSD stormwind.youareinferior.net 4.1 GENERIC#104 i386
unobe just did a 'make install' for r16634 and everything installed correctly into /Library/Perl/5.8.6/Perl6 and /Library/Perl6 07:23
Tene audreyt++ 07:27
Jmax warning: implicit declaration of function `perl5_newSViv' 07:37
Jmax from compiling Pugs.Parser.Charnames and Pugs.AST.Internals 07:37
there's a few of those for different functions 07:38
TimToady ignore
torz ello
Tene Hi, torz! 07:39
torz hey Tene :D
whats up? 07:40
Jmax failed a few, seems to be stuck on t/examples/99problems/problem90 08:22
Jmax oh, there it goes 08:22
Caelum yay, I made my first commit :) 08:27
Tene :D
thoughtpolice Caelum: congrats :) 09:44
Aankhen`` Caelum: Welcome aboard. :-) Now please complete Perl 6.0.0, kthx. 09:48
moritz Caelum: did you add your name to the AUTHORS file? if not, please do it! 09:52
thoughtpolice where's svnbot6? 10:00
moritz thoughtpolice: gone with the wind :(
thoughtpolice :( 10:01
moritz thoughtpolice: same with specbot and evalbot
thoughtpolice a sad day
i guess I must have missed the requiem
at least lambdabot is still with us 10:03
Aankhen`` What good is it without a P6 eval?
moritz thoughtpolice: you missed it indeed: moritz.faui2k3.org/irclog/out.pl?ch...05#id_l204
lambdabot Title: IRC log for #perl6, 2007-06-05, tinyurl.com/yo9vc5
Aankhen`` LOL. 10:04
Aankhen`` takes a bow.
Aankhen`` collapses and falls asleep.
thoughtpolice @remember Aankhen`` What good is it without a P6 eval? 10:08
lambdabot Done.
thoughtpolice :)
Aankhen`` o_O
What does that do?
moritz wow, I want to be able to type english text while I'm sleeping too ;) 10:10
thoughtpolice @quote Aankhen``
lambdabot Aankhen`` says: What good is it without a P6 eval?
thoughtpolice that's what it does
Aankhen`` Heh.
Neat.
thoughtpolice lambdabot is pretty fun :)
www.cse.unsw.edu.au/~dons/code/lamb...t/COMMANDS 10:11
there're some pretty useful commands in there
moritz @slap thoughtpolice "Just a test" 10:14
lambdabot beats up thoughtpolice "Just a test"
moritz nice one ;)
thoughtpolice :'[ 10:15
@tell moritz leave me alone 10:16
lambdabot Consider it noted.
thoughtpolice the useful thing about that is people can leave you memo's pretty easily :)
moritz thoughtpolice: that's what we use most of the time 10:17
lambdabot moritz: You have 1 new message. '/msg lambdabot @messages' to read it.
moritz @massages 10:18
lambdabot thoughtpolice said 1m 59s ago: leave me alone
thoughtpolice eval is one of the funnest though 10:19
Aankhen`` heads to bed.
thoughtpolice there're some that're weird though 10:21
@unlambda
lambdabot Done.
thoughtpolice :/
@elite
lambdabot Say again?
thoughtpolice @elite test
lambdabot tes+
thoughtpolice hah.
moritz I'm look at examples/mandel.pl 10:43
loop ($L = 0;; $L++) { last() if $L > 0;
$L is not modified in the body of the loop 10:44
wtf is that?
masak moritz: looks like a fairly silly way of doing a one-pass loop 10:50
maybe it grew to look that way
moritz masak: seems like 10:51
gaal fun project: since Pugs is cabalized, write a lambdabot plugin to it 14:16
shouldn't be vry difficult even, just make sure to have the plug set PUGS_SAFEMODE. 14:17
lumi What does cabalization have to do with \b plugindom? 14:25
gaal lumi: easy to link against pugs 15:45
just -package pugs, import Pugs, call its -e or whatever 15:46
iirc plugs, \bot's run dynamically moosed thing thing, handles limits 15:47
TimToady attempting to restart evalbot on feather 16:05
moritz TimToady++
TimToady but it's gonna have to recompile the world 16:06
avar 1
moritz "the world is not enough"
spinclad ?eval say "I'm baaack" 16:26
evalbot_r16637 OUTPUT[I'm baaack␤] Bool::True
moritz YaY
TimToady someone who is an operator needs to invite svnbot6 16:41
I don't know how to be an operator... 16:44
spinclad rather short on operators here... 16:49
sub prefix:<@> { .can('invite') } 16:50
[particle] i know audrey can op herself
fglock I'm trying to understand method &.(Match $m) {...} in Perl-6.0.0-STD:132 16:54
TimToady well, svnbot6 isn't gonna work in an svk directory in any case... 17:03
wolverian wants a gitbot 17:04
fglock I can't tell what &.() is doing - is it declaring a method named from the match contents? (doesn't look so)
wolverian I'd assume it's declaring a method named ".", the & being there to make the . a code name. but TimToady probably knows better... 17:05
TimToady see S13:144
fglock looking
fglock aha 17:06
gaal AFAIK only audreyt can op herself; I think we need her help if we want to, uh, viopdoers 17:16
b_jonas hi, 19:54
moritz hi
b_jonas I'm asking a question about junctions that integral asked on #perl and we're disputing:
in this code, how many times is the function called: sub f ($x) { $x == 3 ?? "hello" :: "goodbye" } f(2 | 3); 19:55
the relevant passage from S09 is this:
"In particular, if a junction is used as an argument to any routine (operator, closure, method, etc.), and the scalar parameter you are attempting to bind the argument to is inconsistent with the Junction type, that routine is "autothreaded", ..."
but I don't know what "inconsistent" means there, that is, whether that simple declaration of the argument is inconsistent with junctions or not 19:56
[particle] seems to me that $x is untyped, so it can accept a Junction
b_jonas right, so it's called only once 19:57
thanks
integral curiously I str the answer being the other way around originally: you have to explictly declare the type to include a junction to get a junction; so it'd be called twice. I guess it's changed.
moritz I agree
PerlJam moritz: with whom?
[particle] i think my $x is like saying my Any $x
moritz PerlJam: with [particle] 19:58
[particle] where Any is the base type
moritz PerlJam: nobody else answered the question ;)
PerlJam sorry, I was threading #perl and #perl6 :)
moritz no problem ;) 20:00
PerlJam aha!
S02: Any Perl 6 object (default parameter type, excludes Junction)
[particle] hrmm
b_jonas ooh
PerlJam b_jonas: so, again. My answer is f() gets called twice :) 20:01
[particle] perljam: you sank my battleship!
b_jonas wow
integral so I remembered right. I guess my brain's not that tired yet
b_jonas so what about sub f () { $_[0] == 3 ?? "hello" :: "goodbye" } f(2 | 3);
um
so what about sub f () { @_[0] == 3 ?? "hello" :: "goodbye" } f(2 | 3);
I mean
[particle] huh? 20:02
PerlJam b_jonas: that's an error. You've declared that your subroutine takes no parameters, but you've given it one
moritz what about sub f { $^val == 3 ?? "foo" :: "bar" } ?
b_jonas oh
yeah, I'm stupid
PerlJam moritz: that's another Any type.
moritz PerlJam: are $^ variables always Any? 20:03
PerlJam moritz: AFAIK 20:04
bpalmer` so how would you declare that a variable could be a junction or not, you don't care? Or do you just declare it a Junction and accept a junction of one ? 20:05
PerlJam bpalmer`: eh?
sub foo (Junction|Any) { ... } # there's some irony there :) 20:06
integral sub foo (Any|Junction $foo) { ... }
bpalmer` ah
[particle] Whatever
actually, it's my Object $foo 20:07
PerlJam [particle]: if you mean that it's typed as Whatever, I don't think so.
right.
[particle] but Whatever sounds cleverer
ererer
b_jonas PerlJam: thanks for cleating it up then 20:08
PerlJam b_jonas: thanks for making me look at the docs :-) I need to do that every once in a while 20:09
dduncan svnbot appears to be down 20:48
moritz dduncan++ # adding ext/Muldis-DB/ 21:01
dduncan better name, no?
moritz better than QDIDONTKNOWHOWTOSPELLIT 21:02
dduncan indeed 21:10
the previous name was bad on purpose and meant to be temporary, the new name is meant to be quasi-permanent
Coke I have a p6-like rule (PGE) for a pointy_block: 23:49
token pblock { <?ws> [<'->'> <?ws> <signature> <?ws>]? <block>
}
... is there a way in PGE/rules to make the signature read as if it were a literal $_ if that optional []? doesn't match at all? 23:50
(got the explicit -> version parsing and passing in parrot now. Trying to sneak the default as $_ in.) 23:51
TimToady usually such annotation would be done in semantic analysis rather than faking it up syntactically. That's what STD.pm assumes with '->' 23:55
so first chance is in the closure attached to the rule as its "action"
Limbic_Region salutations all 23:57
Coke yes, but TGE has me boxed in a corner here, and if I could cheat, I was gonna. =-)
TimToady to do it within the rule you'd have to find some way of $<signature> := something that isn't there 23:58
Coke right!
TimToady and that current requires a closure
Coke but I don't think that works. =-)
TimToady *currently
Coke at least not in PGE. 23:59
[particle] sure, you can do closures in pge
Coke Ok, back to the drawing board.
[particle] {{ pir here }}
there's examples in the test suite
Coke ... I don't want to run PIR - what would i run?

Powered by Cro and the Raku® Programming Language.

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

Thank you!