pugscode.org | Beware of `make install` and Pugs, it will do bad things to your system
Set by Aankhen`` on 22 May 2007.
ingy seen audreyt 00:13
thoughtpolice ?seen audreyt 00:15
lambdabot audreyt is in #haskell and #perl6. I last heard audreyt speak 3h 58m 41s ago.
thoughtpolice :)
ingy :) 00:23
cj ingy! 03:29
I was just thinking about you, as seen on #perl
ingy hi cj 03:41
audreyt time dd if=1024mb.bin of=/dev/null 04:00
oops, ww :)
zzzz ?eval constant $N = 10; $N =5 07:16
evalbot_r16507 Error: Can't modify constant item: VInt 10 07:17
Aankhen`` Is the default level for regexes graphemes or characters?
dduncan afaik, the default abstraction level for everything is the highest possible, which is generically called characters 07:37
and that often equals graphemes in practice
Aankhen`` Okay. I'm trying to figure out whether this line from porting_howto is correct: \X -> <.> # single grapheme 07:40
renormalist ping moritz 09:13
moritz pong ;)
renormalist kannst du nebenan im talk gucken? 09:14
moritz wo?
query?
renormalist ja, ich weiss nicht, wie das roh geht, ich hab hier ne KDE gui :-) 09:15
siehst du meinen (query?) text? 09:16
moritz ja 09:17
don't you see mine?
renormalist erm, no 09:18
moritz damned, I'm not registered at freenode
renormalist :-)
moritz so I can't query you :(
so, the answer is "yes"
i can read through it right now
are you on irc.perl.org as well?
renormalist yes
eg in catalyst
rindolf Hi moritz, renormalist 09:24
moritz hi rindolf ;)
renormalist hi rindolf 09:34
rindolf moritz: what's up? 09:35
moritz could anybody with a feather account please start run.pugscode.org 09:44
svnbot6 r16508 | moritz++ | t/operators/precedence/self-defined.t: added prefix <-> postfix tests, 10:06
r16508 | moritz++ | corrected ok() to is(), moritz--
Juerd moritz: Do you use runpugs often? 10:48
moritz: If so, you're probably much better off using feather directly 10:49
moritz Juerd: no, I don't ... I just noticed that is was down 10:50
I have pugs locally, so I don't really have a reason to use it
Juerd Okay
andara maintains runpugs and should start it 10:51
(Should have made it start at bootup)
I don't know if any manual environment setting up is needed
moritz re 13:28
fglock re lexical scope: methods and structure members cannot be lexically redefined, right? 13:37
like: my @a[10]
like: my X::meth {...} 13:38
fglock oops - I meant like: my method X::meth {...} 15:28
TimToady you can't use "my" on anything that is package qualified 15:35
and my @x[10] is declaring an array of 10 elements, not the 11th element of @x 15:36
fglock TimToady: thanks - I was afraid I had to "fix" the implementation :) 15:40
fglock I was supposed to be writing a talk on neuromuscular disease, but I find myself reading CS papers instead 15:49
TimToady "creating procrastination" 15:51
*creative rahter 15:52
*rather
TimToady shouldn't type standing up
[particle] "creating procrastination" -- sounds like a book title 15:53
args "in ten days" 15:54
fglock I think I'm most creative when I'm working on things I was not supposed to
Limbic_Region salutations all 16:01
moritz hi Limbic_Region ;)
Limbic_Region is most creative when he is slightly off his rocker 16:02
rindolf Hi all. 16:03
Hi Limbic_Region, moritz , fglock
Limbic_Region: what's up?
Limbic_Region unfortunately, being slightly off my rocker means being more than a little unstable which is a non-option for me (have family to think about) 16:04
fglock Limbic_Region: yes, maybe the problem is with "too serious" work
Limbic_Region not much rindolf - work ATM, you?
rindolf Limbic_Region: oh I forgot it's not Saturday.
Limbic_Region: chillin'.
Limbic_Region: I've updated my FAQ.
I have another update.
And did a backup.
Limbic_Region the FAQ regarding copyrighted digital media or something else? 16:05
moritz I just tried to use ~~ in bleadperl
Feature "~~" is not supported by Perl 5.9.5 16:06
it says when I 'use feature qw(~~)'
args certainly, it's on by default.
we judged it wasn't a syntax hazard enough to warrant mandating "use feature"
and you can't turn it off actually. (like //) 16:07
"use feature" is now only required for new keywords. which may clash (and do) with existing function names.
also, "use 5.10" will do an implicit "use feature ':5.10'"
moritz args: thanks
rindolf Limbic_Region: no, www.shlomifish.org/meta/FAQ/ 16:12
lambdabot Title: Shlomi Fish' Frequently Asked Questions (FAQ) List
rindolf Limbic_Region: it's my personal FAQ.
args do you get frequent questions about yourself ? 16:13
nobody asks me anything :-( 16:14
moritz args: ? 16:15
args: now somebody asked you anything ;)
rindolf args: a/s/l? 16:20
;-)
args I'm a married man 16:21
Limbic_Region . o O ( on planet earth ) 16:22
jjore-w can never remember whether it is "use features ':5.10'" or "use feature ':5.10'". :-/ 16:26
args just "use 5.10"
Jmax heh, i think he just meant the feature/features part 16:28
moritz Jmax: so he can be lazy and omit it ;)
[particle] use feature :creep; 16:29
that's on by default in all my $work projects :/
japhb [particle]: heck, that's on by default in half of my personal projects! 16:33
TimToady hmm :5.10 is a bit problematic from a p6 point of view 16:41
[particle] language interop at its finest :) 16:43
amnesiac TimToady: why? 16:56
riffraff hi everyone 17:25
offby1 glances around nervously.
spinclad TimToady: r14399's "group of" is a clear improvement, but could i suggest "collection" rather than "group" for us math people (though "group" is fine on a general english level)? 18:01
rindolf Hi riffraff
spinclad ((answer: too late, i've already gone and suggested it)) 18:02
cj anybody know if ingy's back in Seattle? 18:22
offby1 I'll go look for him. 18:37
offby1 pretends to go look
obra cj: yes 18:40
cj obra: well then... I should invite him to lunch. you have his number? 18:53
obra I don't 18:55
cj d'oh. his socialtext vm says he's "not available" 18:56
they call him 'i-n-g-y d-o-t- n-e-t'. ha.
obra cj: his PASSPORT calls him ingy-dot-net 18:57
cj it should be ö of course
obra: you're kidding
that's pretty crazy 18:58
his kwiki.org whois data says he lives in Oregon 19:00
but it sounds like him on the vm that answers the phone number there...
svnbot6 r16509 | tumulus++ | Fill in some missing "how to make this non-ASCII operator in vi" entries 19:14
dduncan TimToady, I think that synopsis r14401 is a vast improvement, so congrats there 19:43
wolverian I'm having trouble with feather 19:48
no route to host now and then
tumulus hello - just want to check some Perl 6 before I commit a test for a bug 19:50
my $l = 5; say "$l {$l} {my $l=123} $l";
outputs "5 5 123 5" as I'd expect
is there any reason that 19:51
my $l = 5; say "$l {$l} {my $l=123} {$l}";
shouldn't do exactly the same?
(difference is $l vs {$l} at the end)
moritz I don't know... do multiple {} interpolations compose the same closure? 19:52
TimToady certainly not 19:53
rindolf Hi wolverian, TimToady
We're discussing COBOL on #perlcafe
amnesiac isn't pugs backend written in cobol?
tumulus so the test should expect the same output in both cases?
TimToady yes
tumulus super - thanks :-) 19:54
moritz amnesiac: certainly not ;)
amnesiac ;)
dduncan TimToady, the last synopsis improvement was big ... I also liked the new names for the roles that the sigils represent ... better than having roles like the type names but not like the types 19:55
TimToady thought about those a long time...
dduncan not surprised 19:56
TimToady kinda hard to describe Essence of Sigil :) 19:57
rhandom I had often wondered at the usefulness of the % and @ sigils (similar to thoughts in recent threads) - but the new synopsis really helps lay out the beauty of the sigils. 20:06
It is a sort of nirvana
well maybe not nirvana - but I really like the roles. 20:07
it sort of opens the doors for all sorts of wonderful magic
PerlJam checks his email to see if he missed an important update to S02 20:08
cool 20:09
TimToady++
moritz yes, it clears up some things 20:10
svnbot6 r16510 | tumulus++ | add some extra tests for string interpolation with closures,
r16510 | tumulus++ | particularly the currently breaking:
r16510 | tumulus++ | my $l=1; say "{my $l=2} {$l}";
PerlJam Though "set with additional behaviors" is an unsatisfying description ofa Junction. (But I don't have anything better either) 20:11
moritz: yeah, just the paragraph on scalars should be enough for those people who are trying to get rid of the other sigils. 20:12
moritz long live the sigil++! 20:13
@karma sigil
lambdabot sigil has a karma of 0
tumulus ?eval my $x=1; { my $x=2;}; say $x; 20:27
evalbot_r16510 OUTPUT[1␤] Bool::True 20:27
tumulus ?eval my $l = 5; say "{my $l=123} {$l}"; 20:28
evalbot_r16510 OUTPUT[123 5␤] Bool::True 20:29
tumulus discovers that it's his own local build that is broken
moritz tumulus: I hope that you write the test cases anyway ;) 20:33
Aankhen`` Couple of questions.
tumulus indeed - already committed :-)
Aankhen`` 1. regex foo { { #{{ CLOSURE 1 }} } ... { #{{ CLOSURE 2 }} } } # is there a way to declare a variable within CLOSURE 1 that's scoped to the regex, so that it can be viewed from CLOSURE 2?
TimToady $<foo> 20:34
Aankhen`` 2. Is the spelling of the implicitly parallelized loop `hyperfor` or `hyper for`? If it's the later, may I ask how it reaches into the innards of the `for`? :-S
TimToady automatically scoped to the current $/
Aankhen`` But that sticks it onto the result object, which means it's visible beyond the regex, doesn't it?
Oh.
Hmm.
Right.
Thanks. ^_^ 20:35
TimToady it's hyper for
yes, you may ask
:)
Aankhen`` Consider it asked!
moritz ;)
TimToady same way eager vs lazy propagates context inward 20:36
Aankhen`` Ahh.
TimToady hyper is just an eager variant
Aankhen`` Okay.
I started thinking about that last night before I went to sleep and it was driving me crazy. :-P
moritz I don't like the "hyper" key word, because it tells the compiler what to do... I'd better like "pure", because it describes your own code 20:40
but they are not always identical, so it wouldn't be a good idea to use that either 20:41
TimToady we could change lazy, eager, and hyper to slow, medium, and fast. :) 20:42
Aankhen`` LOL. 20:43
TimToady or maybe lazy, impatient, and hubristic
in the sense of "I deserve all these CPUs running at once..."
moritz lol
Aankhen`` warpspeed for @foo -> $a { say "You only THINK there's just one parameter here." }
wolverian it's more like "do more work" than "be faster"
TimToady actually, one could argue that the faster variants do less work 20:44
eager doesn't have to keep track of the iterators and continuations 20:45
moritz with multiple CPUs faster != less work
TimToady and hyper doesn't have to keep the processes in sequence
wolverian it depends on how much of the result you use
TimToady eager and hyper both promise that you really want it all, more or less
wolverian but it's not explicit in the name 20:46
TimToady eager says "burn memory" and hyper says "burn CPUs"
wolverian (not saying it should be, just pointing out they're different to lazy in that sense) 20:47
moritz and eval says "burn code" ;)
wolverian TimToady, I like that definition. .)
s/\.)/:)/
TimToady if we renamed "hyper" we'd also have to rename hyperoperators
moritz I'm not really suggesting to rename "hyper", I was just pondering idly ;) 20:48
TimToady we could name them neurotypical, autistic, and ADHDish :) 20:50
moritz or maybe the names of some perl developers 20:55
larry, randal, damian
perhaps in a different order ;) 20:56
rindolf moritz: I want to be an operator in Perl! 21:01
moritz rindolf: since you can define operators for yourself, that's no problem
sub prefix:<rindolf> { say "foo" }; my $a = rindolf"42" 21:02
or something ;)
Aankhen`` multi *postfix:<rindolf> (Str $operand) { $operand ~= " rindolf'd"; }
Well, `Str $operand is rw`.
rindolf Aankhen``: what does ~= do?
Aankhen``: wasn't it supposed to be ~~? 21:03
Aankhen`` ~ is the string concatenation operator.
rindolf Aankhen``: oh, OK.
Aankhen`` So that function appends " rindolf'd" to the string.
rindolf Aankhen``: OK.
Aankhen`` Hmm.
multi *postfix:<rindolf> (Str $operand) { "<rindolf>$operand</rindolf>" }
?eval multi *postfix:<rindolf> (Str $operand) { "<rindolf>$operand</rindolf>" }; "abcd" rindolf 21:04
evalbot_r16510 Error: Odd number of elements found where hash expected: VStr "abcd"
Aankhen`` :-( 21:05
moritz Aankhen``: I think posfix ops can't have a whitespace in front of them 21:06
?eval multi *postfix:<rindolf> (Str $operand) { "<rindolf>$operand</rindolf>" }; "abcd"rindolf
evalbot_r16510 Error: Odd number of elements found where hash expected: VStr "abcd" 21:07
Tene I don't see any tests in t/ for defining circumfix operators. Is that accurate, or do I not know what I'm looking for? 21:37
moritz in unspecced/sub_application.t there are a few method postcircumfix:<( )> 21:39
Tene Yes, but that's postcircumfix, not circumfix.
moritz oh, sorry
operators/operator_overloading.t: two circumfix macros 21:40
but there's need for more 21:41
should circumfix:<[[ ]]> or some such work? 21:42
Tene That's what I'm seeing in the spec. 21:43
meppl good night 21:56
moritz Tene: are you writing a test? 21:57
Tene moritz: I wanted to check if there was a reason that it wasn't already tested where other operator definitions are tested.
moritz can't think of any 21:58
rhandom i know this isn't on topic - how do you call a builtin from XS 22:00
offby1 very carefully 22:01
rhandom :)
PerlJam rhandom: Perl 6 doesn't have XS 22:26
rhandom i know - hence the not on topic 22:27
moritz PerlJam: not yet ;)
rhandom there are just many bright perl 5 people here to
offby1 presumably, integration with C will be lovely in p6
rhandom I think NCI will be great
offby1 Lawd knows it's skeery in p5
PerlJam moritz: not ever! 22:28
Aankhen`` NCI?
rhandom In parrot - not sure how it ties in at Perl 6 yet. Native Call Interface (i think)
moritz PerlJam: actually I never used it ;) (happy /me I guess)
Aankhen`` Ah.
PerlJam rhandom: The plan is for Perl to have some form of NCI-like thingy, but it is yet unspecced. 22:29

Powered by Cro and the Raku® Programming Language.

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

Thank you!