pugscode.org | temporary feather address: perl6.wall.org | sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse (show (scanl (*) 1 [1..] !! 4))
Set by allbery_b on 24 March 2007.
svnbot6 r15820 | audreyt++ | * Pugs.Prim: Add &stat(Str) and &lstat(Str) for Dan Kogai's 02:15
r15820 | audreyt++ | YAPC::Asia presentation. (It merely re-exports perl5:File::stat.)
svnbot6 r15821 | masak++ | [util/version_h.pl] 05:44
r15821 | masak++ | * typo
svnbot6 r15822 | unobe++ | * updated all legacy smartlinks syntax in test suite 06:17
r15822 | unobe++ | * updated util/smartlinks.pl and its tests to no longer support legacy syntax
pasteling "evalbot_r15821" at 194.145.200.126 pasted "Pugs build failure" (25 lines, 1.1K) at sial.org/pbot/23963 06:20
gaal unobe++, whoever you are, please add yourself to AUTHORS 06:59
nothingmuch gaal: unobe == the guy next to you at my birthday party 08:02
Tene gaal: he's alread in AUTHORS 08:03
line 67
nothingmuch ah yes, romano
nothingmuch couldn't find the last name =(
Tene found it! 08:04
nothingmuch Tene++
lichtkind Leibsle: moin 09:07
DarkWolf84 I found thread bug 09:12
dduncan does anyone know if the next Pugs release is expected to happen this week, or whether it would definitely be next week or later? 09:14
lambdabot dduncan: You have 1 new message. '/msg lambdabot @messages' to read it.
dduncan okay lambdabot
DarkWolf84 dduncan, I don't now I use subversion 09:15
dduncan I mean, the release on CPAN 09:16
DarkWolf84 CPAN modules O_O
dduncan last ETA was march, but that didn't happen
DarkWolf84 that's great
dduncan anyway, if if the next release is no earlier than next week, then I anticipate making a significant contribution first 09:17
it will probably be done near the end of this week
my contrib that is 09:18
Tene dduncan: what are you working on? 09:31
Leibsle lichtkind, moin :) 09:40
lichtkind aaaah
dduncan in 1-2 days I expect to make the first CPAN release of the Perl 5 version of QDRDBMS, my new database engine 09:45
after that, I plan to to s/// the necessary parts to make it valid Perl 6 code, and check that in under ext/ of Pugs
I would then delete ext/Rosetta/ , which it replaces 09:46
unlike Rosetta, the QDRDBMS rewrite can actually be used for something
dduncan I won't go into further details now, better afterwards 09:46
Tene, that's what I was referring to 09:47
on a tangent, I'm following mugwump's example and now recommending that people try Git as their version control system rather than SVN or whatever ... Pugs et al may benefit from switching to it 09:51
see utsl.gen.nz/talks/git-svn/intro.html for one of his doc's on the subject 09:52
lambdabot Title: An introduction to git-svn for Subversion/SVK users and deserters
Tene kk 09:59
gaal your birthday party huh? you mean I got drunk with a guy I don't remember? how approporiate 10:03
Leibsle gaal, you should not drink too much, so you remember at least who was at the party :P 10:21
Juerd Huh, why is there a temporary feather address in the topic? 10:45
Is this a very roundabout DNS spoofing attack? :P
moritz Juerd: no, it's because of the ICANN (or something) having trouble with registerfly... 10:46
Juerd moritz: So? feather is on a .nl address
That has nothing to do with registerfly, or...? 10:47
moritz Juerd: hm...
Juerd feather.perl6.nl
And afaik, it works well
moritz Juerd: I know that pugscode.org had problems resolving
Juerd pugscode.org is not feather's primary address. It's just some address that happens to point to the same machine :)
moritz Juerd: you're right, that makes no sense
Juerd Feather isn't configured to receive email on perl6.wall.org 10:48
moritz Juerd: ask allbery_b (set the topic)
Juerd allbery_b: ping
moritz does anybody know a decent open source blog system in perl? 10:57
Juerd Depends on your definitions for decent, open source, blog, and system.
And perhaps also perl.
:)
gaal livejournal's pretty decent. :)
Juerd Mostly "decent" is a bit vague 10:58
moritz Juerd: blog = something like a CMS with online editing facillities that provides RSS Feeds, track/pingbacks and comments... 11:01
Juerd: Open Source = OSI-Approved (or better: DFSG) 11:02
Juerd: perl = perl 5 (perhaps mod_perl 2)
gaal: do you have a link for me? googling for livejournal only brings up livejournal.com, but that doesn't look like open source 11:03
lumi moritz: www.livejournal.com/code/ 11:05
lambdabot Title: Code
moritz lumi: thanks 11:07
lichtkind audreyt: ping? 11:15
avar didn't p6 have some feature to assign to captures thus modifying the matched string? 11:42
kolibrie avar: dev.perl.org/perl6/doc/design/syn/S...#Modifiers 'The new :rw modifier' 12:39
lambdabot Title: Synopsis 5: Regexes and Rules - perl6:
avar $ perl5.9.5 -Mblib -E 'use re::engine::Plugin (exec => sub { my ($re, $str) = @_; $re->captures( sub { sub { $$str = shift } } ); 1 }); my $str = "foo"; if (\$str =~ /eek/) { $1->("ook"); warn $str }' 12:56
ook at -e line 1.
moritz ? 12:57
"ook" is a nice message ;-)
avar that'll allow you to implement :rw in p5 space 13:05
moritz how cool ;-) 13:09
avar I'm bugging the core people for a better interface to numbered and named matches:) 13:26
fglock for people coming to YAPC::SA - fisl.softwarelivre.org/8.0/www/?q=en/arena (perl5/6 and haskell is probably ok) 17:14
lambdabot Title: Programming Arena | 8th International Free Software Forum
wolverian no internet use? is there a local CPAN mirror? :) 17:17
fglock hmm - right :)
I can ask them to preinstall CPAN modules 17:19
though it's hard to predict what would be useful :)
fglock how much of perl6-in-parrot is implemented? I wonder if I can rewrite mp6 to bootstrap with what we already have 17:41
moritz fglock: perhaps you should ask in #parrot ;-) 17:43
fglock hmm - no, it's a perl6 question :) 17:44
moritz fglock: ok, you're right ;)
[particle] fglock: shouldn't be too much effort to get there 17:45
fglock [particle]: this could be a task for the YAPC::SA hackathon 17:48
next week
[particle] sounds great to me! 17:50
fglock ok - then I've got a few days for planning 17:51
fglock re kp6-in-parrot: the simplest implementation would be through a perl6-in-parrot emitter 20:00
although it apparently doesn't make sense, but it would make it possible to evolve the p6regex-in-perl6 implementation 20:01
which is a possible path to STD compilation
pmurias fglock: hi 20:06
fglock pmurias: hey 20:08
pmurias why doing kp6-in-parrot thought a perl6-in-parrot emitter dosn't make sense? 20:14
you mean emitting perl6 code which compiles under the parrot perl6 compiler? 20:15
fglock pmurias: yes, because in the end, it would be a perl6-to-perl6 transformation
which really smells like duplicate/throwaway work 20:17
pmurias is emitting PAST much harder? 20:19
fglock yes - the perl6 compiler does a lot of transformations already 20:20
I think what I'm really trying to do is to avoid duplicate work, by using pugs, p6-on-parrot, and p6-on-perl5 as virtual machines for perl6-on-perl6 20:23
6-on-6 makes less sense, as the bottom-up implementations evolve 20:25
__Ace__ why dont I get some error msg by doing: 20:26
perl_eval_pv("use strict; $a = 3;", TRUE);
fglock <__Ace__>: maybe the message is in $@ - you need to print it? 20:27
__Ace__ I get complaints if I add eg: use adsfsdfk; 20:28
fglock <__Ace__>: maybe try a perl5 list or perlmonks.org? 20:30
pmurias fglock: every time i look pir code 6-on-6 makes a *lot* sense to me 20:35
s/look/look on/ #I wouldn't want to look like pir code for anything :) 20:36
pmurias a big advantage of perl6-on-parrot(for me) would be that it's compile time is a fraction of pugs 20:40
devbot6 planet6: Stevan Little: Moose is turning 1yr old soon <use.perl.org/~Stevan/journal/32891?from=rss> 20:43
fglock pmurias: that's the compile time for 6-on-parrot; 6-on-6 should be much slower 20:48
even on parrot
svnbot6 r15823 | kudra++ | Completed the week 21:34
devbot6 planet6: Audrey Tang: Weekly Perl 6 mailing list summary for 11-17 March, 2007 <pugs.blogs.com/pugs/2007/04/weekly_...m.html> 21:43
svnbot6 r15824 | kudra++ | Placeholders 21:52
perlmonkey2 what is the <<== operator do? 22:15
s/is/does/
[particle] a pipe that appends? i'd have to check S03 22:16
S03? 22:17
hrmm. thought there might be a factoid for thta
yep, it's a feed operator that appends 22:18
perlmonkey2 So this:@w <<== $i; is equiv to push(@w,$i); ? 22:19
moritz that looks... unfamiliar ;-)
[particle] yep
lichtkind audreyt: are you awake?
perlmonkey2 cool
moritz ?eval my @a = 1 .. 2; @a <<== 3, 4 22:20
evalbot_r15824 Error: ␤Unexpected "<<=="␤expecting operator 22:20
[particle] moritz, it's a ENOIMPL pugsbug
moritz [particle]: I can see that, <<== is not recognized as an op ;-) 22:21
perlmonkey2 "Note, as of today, append feed is now <<== rather than <==. "
Limbic_Region .seen stevan__ 22:23
?seen stevan__
lambdabot stevan__ is in #perl6. I don't know when stevan__ last spoke.
overkordbaever Seems like stevan__ is online at least 22:24
Limbic_Region stevan ping 22:27
or rather
stevan__ ping
[particle] it's somewhat unfortunate that @a ==>> $s is not equivalent to pop 22:28
perlmonkey2 is it equivalent to shift? 22:45
[particle] it depends what's in $s 22:47
perlmonkey2 it would be weird having the assignment go from left to right.
[particle] if $s does array, it would append every item from @a to $s
if $s is scalar only, it may fail because it can't append, but i'm not sure 22:48
perlmonkey2 you can assign basic type roles to your classes?
[particle] my int $dog; 22:49
types, or roles?
perlmonkey2 you said "does" array.
package MyAPP{ does array;}
probably not correct syntax. 22:50
[particle] i'm not 100% sure what the base role names will be
class Dog is Mammal does Pet { ... }
check out spec.pugscode.org/ 22:51
Limbic_Region particle, I don't care if your Dog is a Mammal - it should, in no way, be doing my Pet
lambdabot Title: Official Perl 6 Documentation
perlmonkey2 so: class Dog is Mammal does array{ ... } and array's operators would have to be overloaded?
[particle] dogs will be dawgs
perlmonkey2 thinks Multiple Dispatch is groovy. 22:56
perlmonkey2 Ahah, you don't say "does array", you do Type Casting. 22:58
Limbic_Region perlmonkey2 - indeed, but keep in mind that multi-method dispatch goes by different names in different groups
perlmonkey2 CharLingua? Wow, Perl6 really will become the tool of choice...this stuff is like magic fairy dust for people working the web. 23:03
lichtkind Leibsle: ? 23:04
Leibsle lichtkind, guguck :) 23:06
diotalevi wondered that of class Dog is Mammal does Pet { ... }, not all Dogs do Pet. 23:09
moritz diotalevi: just name it PetDog ;-) 23:11
perlmonkey2 S29 is like reading a wish list for every shortcut you've ever wanted. 23:15
moritz perlmonkey2: actually the changes in the builtins are small compared to most other changes ;-) 23:18
perlmonkey2: just read S05 ;-) 23:19
perlmonkey2 If the regexps are upgraded too much, won't p6 allow anyone who cares to, to take over the entire world? 23:21
moritz perlmonkey2: when _everybody_ takes over the world, that's communism ;-) 23:22
OSS developers are all freaking communists, didn't you know? *bg*
"World domination, fast" -- Linus Torvalds
perlmonkey2 Reading these synopsis takes me back about a decade to when I read the introduction to the Camel book and realized I was about to see something cooler than I'd ever seen before. 23:27
perlmonkey2 m:c??? come on? how cool is that? if that exists in p5 and I don't know about it, I'll cry. 23:29
dduncan on the contrary, OSS development is the closest software model to true capitalism 23:31
there, everything is driven by supply and demand, rather than copyright laws that artificially change the balance of things
moritz dduncan: I know that, I'm paraphrasing trolls ;-)
Aankhen`` You know what would be a great idea? 23:38
Porting Params::Validate to Perl 6!
Aankhen`` ducks.
dduncan I have an opinion question. If I were to declare a generalized Str subtype, say NEStr, that includes all Str values except for the empty string, then what might be the best default value of such type, since the empty string is the default value of Str? 23:41
geoffb space? 23:42
dduncan Also assume that Undef is also not allowed, or alternately, that Undef when used as a string would default to the type's default value, but that can't be the empty string here.
geoffb, that was actually my first thought, that space could work well
but I was wondering if some people thought other alternatives, such as perhaps the digit zero, may be better? 23:43
I figure, 1 of those 2 options, unless a third is suggested
geoffb space has the advantage of non-jarring interpolation
moritz dduncan: smells like a design - a non-empty type that needs a default (just an opinion ;-)
dduncan sounds reasonable
moritz s/a/ill/ 23:44
dduncan I'm designing a type system that has no concept of undefined ... every variable of that type always contains a valid value of the type
Perl 6 sort of does this by that Undef can pretend to be a default value for a type, eg, zero-but-undef or some such
the type system has a normal Str equivalent, but I thought it would be useful to also include a subtype that couldn't be an empty string, since that's what lots of people use 23:45
for the normal type, the empty string is the default of course
the distinction of these 2 types is similar to having numeric types that can either be only all numbers, vs a subtype that can only be >= 0, and another that can only be > 0 23:46
all 3 of those are common scenarios 23:47
anyway, I'll go with [space] as I was before, unless I'm convinced otherwise
for NEStr, and the empty string for Str as usual
with the 3 numeric types mentioned, the default values would be 0, 0, 1 respectively 23:48
polettix throws a x{00}
dduncan thank you for your feedback 23:49
moritz throws a string with negative length and quickly ducks bevor it detonates ;-) 23:49
dduncan I think in Perl 5 that is actually possible
eg, hash keys that are str with a utf8 flag on are indicated as such with negative lengths, or something? 23:50
polettix wonders what would happen doing ${moritz' string} x -1
dduncan that way, 2 different keys that otherwise have the same bit pattern, but 1 is text and the other binary, don't collide 23:51
moritz polettix: it turns upside down
polettix then, I wonder what would happen feeding it to an inside-out object
[particle] @eval sqrt $negative_string 23:52
er, wrong evalbot, but you get the point :)
moritz polettix: it vomits ;-)
polettix $i * $something
:)
dduncan ?eval sqrt $negative_string 23:53
evalbot_r15824 0.0 23:53
dduncan there you go particle 23:54
moritz ?eval sqrt -1i
evalbot_r15824 0.7071067811865476 + -0.7071067811865475i
dduncan dood!
polettix ?eval sqrt -1
evalbot_r15824 NaN
moritz polettix: that's on purpose..
[particle] ?eval NaN ** NaN
evalbot_r15824 NaN
moritz polettix: not to frighten mathematically non-inclined programmers ;-) 23:55
polettix ?eval sqrt (-1 + 0i)
evalbot_r15824 0.0 + 1.0i
polettix :)
is there a less work-aroundish mechanism to trigger complex computing?
moritz ?eval sqrt(-1 + 0i)**2 23:56
evalbot_r15824 0/1
dduncan cast arguments as Complex ?
?eval sqrt Complex(-1,0)
evalbot_r15824 Error: No such subroutine: &Complex
dduncan I would expect something like that to work
polettix what does 0/1 mean? 23:57
dduncan it means zero
dividing anything by 1 is a trivial situation
?eval 0/1i 23:58
evalbot_r15824 0.0 + 0.0i
polettix ah, it's a division sign. so the result is wrong?
moritz I assume, yes
polettix ?eval 1i**2
evalbot_r15824 0/1
polettix ?eval (1i + 0)**2 23:59
evalbot_r15824 0/1
polettix ok, complex computations are still a bit too "complex" 23:59
dduncan mind you, I can't assert whether a number of those results were correct or not, since I don't know too much about complex math

Powered by Cro and the Raku® Programming Language.

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

Thank you!