pugscode.org/ | sial.org/pbot/perl6
Set by agentzh on 7 June 2007.
00:01 sfullenwider joined
sfullenwider how far away is perl6 to production? 00:04
moritz one, two years perhaps 00:05
depends on your notioin of "production"
Tene I've used it to produce things several months ago. 00:06
moritz evalbot and svnbot are "producing" things as well ;) 00:07
Tene ?eval "a small red dog" 00:09
evalbot_r16653 "a small red dog"
Tene evalbot just produced a small red dog!
00:09 larsen_ joined 00:10 knewt joined
offby1 can we keep it? Can we? 00:13
TimToady you'll have to take better care of it than you did the camel... 00:19
00:31 sfullenwider left 00:33 deq` joined 00:37 weinigLap joined 00:45 nipotaway is now known as nipotan
Tene TimToady: I reinstalled F7 and pugs runs fine now, btw. 01:08
TimToady yeah, saw that earlier, but too busy writing a talk to respond then... 01:20
but I don't really want to trash my current install...
I keep thinking there's some subset I can delete and reinstall 01:21
Tene nods.
However, after I installed, I added a bunch of packages with the 'Add/Remove Software' tool instead of just yum, and it failed to install dependencies for many of the packages installed, so beware. 01:22
TimToady thanks
silug Tene: did it just fail to download the packages? usually you can just tell it to try again... 01:23
<- fedora developer
TimToady silug: I upgraded my laptop to F7, and now gnome always hangs coming up, so limping along on kde. also, if I click on Languages on the greeter it immediately crashes. any ideas? 01:25
Tene silug: It claimed to have completed successfully, so I didn't bother to check. Later, when trying to use applications, they would complain about missing deps.
after rpm -e and isntall via yum, everything worked fine. 01:26
silug other people have told me about the gnome hangs. the best guess i've heard is that they are dns-related.
i'm a kde user, so i'm the wrong one to ask. :)
<- not terribly helpful fedora developer
Tene: weird. pirut is just a front-end to yum, so the behavior should have been identical. 01:27
TimToady seems related to scim to me, maybe.
can't bring up scim under kde either.
01:28 weinigLap joined
TimToady which is why I haven't been typing much Japanese here lately... 01:28
01:29 weinigLap joined
TimToady so I'm guessing that F7's overall internationalization is less than robust under upgrade, at the moment. quite possibly a complete reinstall would work, but I only have one major partition on my laptop. 01:29
silug bugzilla.redhat.com/bugzilla/show_...?id=243234 may be the problem. 01:30
imho f7 is one of the rougher fedora releases so far... the core/extras merge was way more painful than ever anticipated. 01:31
TimToady heh, the page with the fix is entirely in Japanese. I guess I deserve that... :) 01:33
silug and, for the record, since i'm extremely monolingual, i know almost nothing about scim 01:34
TimToady but from the rest of the bug report, it definitely seems like the problem doesn't affect fresh installs, only upgrades
SamB scim is also usable for maths input... 01:35
or would be if unicode characters were all it took
silug "rm -rf /tmp/gconfd-* /tmp/scim-panel-socket-*" isn't the fix? 01:36
TimToady oh, I didn't see the translation, just followed the link 01:39
silug babelfish does a reasonable job translating that page. i'm impressed. 01:41
TimToady well, definitely looks like it got further, but will have to restart X
01:45 dolmans joined
silug i wonder if bugzilla.redhat.com/bugzilla/show_...?id=243563 is related at all 01:46
01:49 sunnavy joined
silug wanders off to find some dinner 01:49
01:56 Eidolos_ joined 02:00 Eidolos_ is now known as Eidolos 02:07 japhb joined
TimToady やった！ありがとうございます。 02:13
but my Language button in the greeter still crashes X. 02:14
silug: but yes, the other one was related too.
02:14 offby1 joined
TimToady I'm definitely in the category of person who has an english locale but wants a asian input method. 02:18
*an
amnesiac TimToady, Watashi wa tamago-desu!!! 02:21
eewww.. no.. egg no :P
TimToady hmm, gnome hung again... :( 02:25
02:50 mako132_ joined 02:51 obvio171_ joined 03:00 jdv79 joined 03:17 unobe_away is now known as unobe 03:38 mako132_ joined 03:42 obvio172 joined 03:43 obvio172 is now known as obvio171 03:44 stevan_ joined 04:11 deq` joined 04:35 iblechbot joined 04:54 Mahmoud joined 05:04 BooK_ joined 05:19 baest joined 05:23 devogon joined 05:46 japhb joined 06:21 LeCamarade joined 06:24 Averell joined 06:26 weinigLap joined 06:27 weinigLap joined 06:35 weinigLap_ joined 06:36 weinigLap joined 06:48 chaoslawful joined 06:57 weinigLap_ joined 07:01 justatheory joined 07:02 justatheory joined 07:19 jisom joined, [particle1 joined 07:24 buetow joined
meppl good morning 07:27
07:29 deq` joined 07:34 LeCamarade is now known as [LeCamarade] 07:45 justatheory joined 07:55 torz joined 08:10 chylli joined, r0bby__ joined 08:11 r0bby joined 08:29 jisom_ joined 08:35 offby1` joined
dduncan well, I've basically just proven that a lot of web servers out there are refusing traffic just because the client's user agent says it is LWP ... but self-identifying as something else works fine 08:52
08:52 unobe is now known as unobe_away
dduncan this adds a bother if I want to test now and then that links in a web site list are still valid, automatically 08:53
08:53 unobe_away is now known as unobe
dduncan at least, it means I can't use LWP::Simple 08:53
Tene dduncan: wow. 08:57
gaal dduncan: a quick peek suggets you can override _init_ua and get around that. 08:58
dduncan mind you, I got other benefits from switching to the slightly more complicated LWP::UserAgent
yes, I already worked out the solution 08:59
just set $ua->agent() to something else
and it worked
Tene Set it to "Ell Double-You Pee"
dduncan oh, your suggestion was how to get LWP::Simple itself to work?
gaal It would have been great if LWP::Simple had called "$ua->agent( $class->get_default_agent )" or something similar
dduncan well, no matter, I like the additional benefits of switching
gaal but the shotgun method is *{"LWP::Simple::_init_ua"} = sub {...} 09:00
dduncan one moment ...
Aankhen`` $ua->agent('gaal'); 09:02
Mwahaahahahahaahaaha.
dduncan the init-ua thing wouldn't have helped in my situation 09:03
but just using UA instead did
gaal $ua->agent('sekkrit') 09:04
Tene $ua->agent('møøse');
dduncan see, the way I was using it before, LWP::Simple::get wouldn't have been using an actual $ua behind the scenes, so the init_ua thing would never be invoked
gaal actually you should do agentname/version
'mse/1.øø'
Tene laughs.
dduncan it was using _trivial_http_get , which hardcodes everything 09:05
gaal ouch
::Simple modules are
dduncan the downside of using the full UserAgent is extra CPAN prereqs ... HTML-Parser and HTMl-Tagset
gaal { package DWIM::Simple; sub AUTOLOAD { die "not supported" } }
dduncan which I didn't have to install before
the LWP package included both Simple and UserAgent, so both were already installed anyway 09:06
Tene was going to make a joke about that statement being produced by gaal::Simple while waiting for the continuation.
dduncan that is, installing simple installed the other
gaal good thing cpan.org accepts LWP::Simple UAs... 09:08
dduncan well, the majority of urls in my database that I tried did accept LWP::Simple ... just a few hundred out of about 20-thousand didn't 09:09
so maybe 1-2% or so are the problem
still, that's a large number 09:10
Tene dduncan: what % by host instead of url? What market segment do these hosts represent? Etc.
dduncan the web addrs are businesses and/or other organizations that provide amenities to tourists 09:11
its a tourism related website 09:12
that I'm working on for my $job
the database was seeded a few years ago by other people
Tene 'kay
dduncan so for the most part, every url is a different host 09:13
maybe that's what you wanted to know
Tene It is.
dduncan on a tangent, today is probably the first time I ever used alarm() in a program ... in this case, to timeout a page fetch ... done before the move to UA ... some pages simply hung Simple ... not sure if this is still necessary, as I also give UA a timeout value 09:17
currently, I give it 30s, and have my own alarm for 40s
on an unrelated matter, I note an interesting coincidence that both the GPLv3 (date announced 2 weeks ago) and the iPhone (date announced today) are both planned to be released on the same day, June 29th 09:23
gaal is disappointed to find you can't tie package symtable hashes in Perl 5 09:28
Aankhen`` Heh.
sub FETCH { if (rand(10) % 2) { return "Sorry, no symbols for you today."; } } 09:29
Er, int(rand(10))
gaal doesn't work :(
doesn't work either :)
Aankhen`` Pah. 09:30
A mere trifle.
dduncan so it seems the end result was about 600 web addrs out of 23000 or so were rejecting traffic based on 'lwp' identifier 09:31
09:31 buetow joined
dduncan so about 2-3% 09:31
09:50 riffraff joined 09:51 dduncan left 09:58 blaze-x_ joined 10:00 unobe is now known as unobe_away 10:13 jedai joined 10:24 aukjan is now known as aukjan|gone 10:31 buetow joined 10:46 blaze-x_ joined 11:08 ruoso joined 11:13 blaze-x joined 11:29 Apple joined
Apple 􏿽xCA􏿽xC7􏿽xD1􏿽xCA􏿽xB4􏿽xD5 11:29
Tene Hello! 11:30
11:31 buetow joined 11:33 blaze-x joined 11:43 blaze-x joined 11:45 nipotan is now known as nipotaway 11:55 iblechbot joined 12:08 kolibrie joined 12:26 chris2 joined 12:30 blaze-x joined 12:31 |Jedai| joined 12:50 blaze-x joined 12:52 mico_ joined, HiTech69 joined 12:53 mico_ left 13:03 blaze-x joined 13:05 perlygatekeeper joined 13:18 kanru joined 13:20 blaze-x joined 13:21 buubot joined 13:23 buubot is now known as mutebot, kanru joined 13:28 mutebot is now known as bonkbot, blaze-x joined 13:30 bonkbot is now known as buubot 13:31 buubot joined 13:33 buubot joined 13:36 offby1` is now known as offby1 13:45 buetow joined 13:53 IllvilJa joined 13:56 vel joined 14:26 fglock joined 14:46 Limbic_Region joined 14:51 rindolf joined 14:58 vel joined 15:02 daxim joined 15:09 rlpeacock joined 15:12 amnesiac joined 15:29 f0rth joined 15:47 vel joined 15:51 f0rth joined 16:07 schmalbe joined 16:28 xinming joined, weinigLap joined 16:29 chylli left 16:45 coke joined
coke Is the p6 PM lying about? 16:46
wolverian PM? 16:48
16:48 rintaro_ joined 16:49 stevan__ joined 16:52 coke is now known as Coke
Coke PM is project manager 16:52
wolverian we have one? :)
[particle1 obra: ping 16:57
16:57 [particle1 is now known as [particle], fglock joined
wolverian (also, which p6?) 16:59
fglock [particle]: the problem pmichaud was talking about was: 17:01
lambdabot fglock: You have 1 new message. '/msg lambdabot @messages' to read it.
fglock token xyz { <foo> | <bar> }; eval " token foo { abc }; /<xyz>/ "
right?
[particle] i think so
fglock I meant to compile that to:
token xyz { <foo> | <bar> }; eval " my token xyz { <foo> | <bar> }; token foo { abc }; /<xyz>/ " 17:02
using an AST transformation, so that it can be figured out statically
where the 'my' happens at the appropriate OUTER level 17:03
fixing: 17:04
token xyz { <foo> | <bar> }; eval " my token foo { abc }; /<xyz>/ "
token xyz { <foo> | <bar> }; eval " my token xyz { <foo> | <bar> }; my token foo { abc }; /<xyz>/ "
17:04 Psyche^ joined
[particle] fglock: okay, so you need lexical subroutines 17:05
or, anyway, all rules would be stored in a set of pads 17:06
fglock not sure if methods can be lexical in Perl 6, but I think they can 17:07
[particle] yeah, methods. right.
fglock non-lexical regexes are static, already 17:08
17:09 weinigLap_ joined
fglock in the case of: eval " token foo { abc } " 17:10
the global 'xyz' is redefined 17:11
which probably requires that you keep the AST around at runtime
or use tables instead
17:18 weinigLap_ is now known as weinigLap 17:19 rintaro_ is now known as rintaro 17:20 Psyche^ is now known as Patterner
obra pong 17:23
17:27 crashmatrix joined 17:28 ofer joined
[particle] obra: coke was looking for you 17:33
Coke obra;hey. 17:39
17:40 fglock_ joined
obra Coke: what's up? 17:41
17:42 obvio171 joined 17:43 Jmax joined, schmalbe is now known as barney 17:46 barney is now known as schmalbe 17:48 schmalbe is now known as barney 17:50 rindolf joined, justatheory joined 17:54 justatheory joined 18:24 fglock joined 18:27 jisom joined 18:31 mj41 joined 18:37 Hamish joined 19:05 devogon_ joined, drupek12 joined 19:06 devogon joined 19:13 dduncan joined 19:16 GabrielVieira joined 19:20 weinigLap_ joined 19:22 mj41 joined
Coke obra: Wondering if you consider parrot a subset of the total perl6 plan. or adjunct. or something. how does parrot fit into your plans, since you're a potential consumer. 19:29
This will help me firm up *my* plans. 19:30
obra Perl 6 on parrot is certainly part of "the perl 6 project" 19:39
Parrot is both part of the perl6 project and a bird of another feather.
*waves hands*
how will it impact you? ;)
Coke well, I have a presentation to give in 2 weeks..
Other than that? =-) 19:40
Can you include me on any p6 project management stuff as a cc?
obra It's mostly phone calls right now. 19:43
What are you actually looking for?
19:49 larsen_ joined
Coke I need to know what you need from parrot. you're one of the main customers. 19:50
obra It's what Patrick needs from Parrot
Coke so I should just go through patrick then? ... ok.
obra Patrick is the main consumer of Parrot for Perl 6.
I believe he's about to mail the lists about DFA and regexps 19:51
Coke k. I'll bug him. 19:52
dduncan I may be missing context, but if that exchange wasn't a joke, it didn't make much sense 19:55
dduncan web backlog ...
Coke it only had to make sense to me and obra. 19:56
dduncan nothing else in prior backlog
Coke which part doesn't make sense?
dduncan something about a person being the main customer 19:57
but then I may have misread it
obra dduncan: made perfect sense to me and coke
dduncan well never mind
Coke well, parrot's "customers" are the language developers. one of the languages is going to be perl6 - need to make sure parrot supports perl6 - do I talk to the perl6 PM for those requirements? (most of which are there and/or obvious) Turns out, no, I talk to the one guy who's in charge of perl6 specifically on parrot, patrick. 19:58
this is all part of my trying to prepare for parrot 1.0...an dmore importantly, my talk about same in 2 weeks. 19:59
dduncan yup
[particle] do you think it'll work? 20:00
Coke do I think what will work?
No, i think we're all doomed. =-)
[particle] *phew*
moritz Coke: did you speak to fglock? iirc his miniperl6 project has a PIR emitter, maybe he has some good suggestions 20:01
Coke He's on #parrot. haven't had a full discussion with him yet, though, I think he's coordinating with patrick. 20:02
p6 on parrot, btw, is reaaaaly close to running Test.pm; hopefully that'll open some floodgates for more development. 20:03
Should I complain that there are things in Test.pm that aren't tested in sanity?
[particle] in pugs current sanity tests?
yes.
in parrot's old sanity tests? maybe not.
Coke ok. we are parsing all the 01- tests. however, 'use Test' fail with a syntax error. So, some syntax isn't being testing in 01-sanity. 20:04
[particle]: : I've already resolved that issue to at least last week's version of pugs tests.
[particle] oh? fab!
Coke in fact, 'make spectest' in .../perl6 will svn export the pugs tests right there and run them, in case I missed something. 20:05
[particle] coke: it may be the difference between namespace and class
20:08 Jmax joined 20:10 fglock_ joined 20:13 IllvilJa joined
svnbot6 r16654 | lwall++ | 1st attempt to whack in block comments 20:23
moritz YaY, svnbot6 is back ;) 20:25
TimToady curtsies 20:26
Yaakov TimToady: How gentile!
dduncan svnbot is back 20:27
moritz TimToady++
Yaakov Err... genteel!
TimToady that too 20:28
Yaakov Well, you are a gentle genteel gentile.
Generally.
20:29 jjore-w joined
spinclad one of the gentry 20:29
Yaakov He is also know for genorosity. 20:30
spinclad true dat
20:31 obvio171_ joined 20:33 the_dormant joined
spinclad very lazy 20:34
unlike me
TimToady wonders why goy, gaijin and gringo all start with 'g'... 20:35
[particle] honky is a notable exception to the rule
TimToady off by 1
well, it's somewhat after 1, but I must still be off. 20:36
commuting &
20:37 obvio171_ joined 20:43 weinigLap joined 20:44 penk joined 21:00 pbuetow joined 21:01 polettix joined 21:05 justatheory joined 21:18 marmic joined
japhb TimToady: I suspect the reason for them all starting with 'g' has something to do with why of the 'w' question words, 'how' has it at the end ... 21:23
;-)
riffraff wasn't it the "wh" words? 21:31
21:43 ron__ joined
ron__ ?eval $::a = 1 21:45
21:45 evalbot_r16653 is now known as evalbot_r16654
evalbot_r16654 \1 21:45
ron__ ?eval $a=1
evalbot_r16654 \1
ron__ ?eval class Bozzle {has $.a = 2;has $!a = 4;} my $c = Bozzle.new; say $c.a; 21:46
evalbot_r16654 OUTPUT[4␤] Bool::True
meppl good night 21:47
TimToady ?eval class Bozzle {has $!a = 2; has $.a = 4}; my $c = Bozzle.new; say $c.a; 21:49
evalbot_r16654 OUTPUT[4␤] Bool::True
TimToady ?eval my $x = 2; my $x = 4; say $x 21:50
evalbot_r16654 OUTPUT[4␤] Bool::True
TimToady ?eval class Bozzle {has $.a = 2; has $.a = 4}; my $c = Bozzle.new; say $c.a; 21:52
evalbot_r16654 OUTPUT[4␤] Bool::True
moritz ?eval class B {has $.a = 'a'; has @.a=<b c>;}; my $c = B.new; $c.a 21:54
evalbot_r16654 ["b", "c"]
21:58 ron__ joined 22:04 drowned joined 22:09 drowned joined 22:11 weinigLap joined
drowned ?eval class Grozzle {has $a;method set_a { $a = 1 }} 22:12
evalbot_r16654 undef
drowned ?eval class Grozzle {has $a;method set_a { $a = 1 }}; my $b = Grozzle.new; $b.set_a; 22:13
evalbot_r16654 \1
TimToady ?eval class Grozzle {has $!a; method set_a { $!a = 1 }}; Grozzle.new.set_a 22:14
evalbot_r16654 \1
TimToady ?eval class Grozzle {has $!a; method set_a { $a = 1 }}; Grozzle.new.set_a
evalbot_r16654 \1 22:15
TimToady ?eval class Grozzle {has $.a; method set_a { $!a = 1 }}; Grozzle.new.set_a
evalbot_r16654 \1
TimToady ?eval class Grozzle {has $.a; method set_a { $!a = 1 }}; $b = Grozzle.new; $b.set_a; say $b.a 22:16
evalbot_r16654 OUTPUT[1␤] Bool::True
TimToady ?eval class Grozzle {has $.a; method set_a { $a = 1 }}; $b = Grozzle.new; $b.set_a; say $b.a
evalbot_r16654 OUTPUT[␤] Bool::True
TimToady ?eval class Grozzle {has $.a; method set_a { $a = 1 }}; $b = Grozzle.new; $b.set_a; say $Grozzle::a 22:17
evalbot_r16654 OUTPUT[1␤] Bool::True
TimToady (no strict in eval...)
22:46 Limbic_Region joined 22:59 Aankh|Clone joined 23:17 TreyHarr1s is now known as TreyHarris 23:20 Aankh|Clone is now known as Aankhen`` 23:22 weinigLap joined
TimToady hmm, I think svnbot6 and specbot6 need to be reinvited... :/ 23:22
Limbic_Region how does that work 23:24
23:27 TimToady joined
TimToady how does which work? inviting? I don't know, except it apparently takes a channel operator, and doesn't seem to be sticky. that's all I've intuited so far... 23:30
Tene /invite svnbot6 23:31
23:32 obvio171_ is now known as obvio171
Tene sends a message to the client saying "you've been invited to suchandsuch channel" and also on many servers overrides a ban, or the channel could be set to invite-only. 23:32
Limbic_Region oh, chan op required - makes sense 23:39

Powered by Cro and the Raku® Programming Language.

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

Thank you!