Camelia, the Perl 6 bug

IRC log for #mojo, 2011-07-29

| Channels | #mojo index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:27 j3nnn1 joined #mojo
00:49 preflex joined #mojo
00:53 preflex_ left #mojo
00:58 xaka left #mojo
01:19 stephen left #mojo
01:20 stephen joined #mojo
01:29 Alias_ joined #mojo
02:01 blabos joined #mojo
02:07 vel joined #mojo
02:25 SmokeMachine left #mojo
02:45 SmokeMachine joined #mojo
02:51 crab hi.
02:59 xaka joined #mojo
03:01 alnewkirk is now known as alnewkirk-hiring
03:01 alnewkirk-hiring hi
03:13 j3nnn1 left #mojo
03:23 ispy_ joined #mojo
03:23 ispy_ Mojolicious::Plugin::ShareHelpers works pretty nicely so far...
03:26 kaare joined #mojo
03:29 sri anyone know if return in a BEGIN blog is a feature or a quirk?
03:29 sri s/blog/block/
03:32 sri FOO=1 perl -e'BEGIN {return if $ENV{FOO}; print "BAR!"} print "BAZ!"'
03:39 SmokeMachine left #mojo
03:45 mateu it's a subroutine, why would return be quirky?
04:02 shonorio_ left #mojo
04:10 jwang joined #mojo
04:22 GitHub75 joined #mojo
04:22 GitHub75 [mojo] kraih pushed 1 new commit to master: https://github.com/kraih/mojo/commit/9e​e90bc3573c68620d032e0ff2f3a63e3882103e
04:22 GitHub75 [mojo/master] simplified resolver a bit - Sebastian Riedel
04:22 GitHub75 left #mojo
04:28 sri if you're using an outdated perl you surely won't need ipv6 ;p
04:39 ispy_ sri: I'm pretty sure it's a feature... because I've used it before and I recall reading it somewhere (noteworthy) ... I'll try and double check to be sure though...
04:39 sri ispy_: ?
04:39 sri oh, return
04:39 ispy_ sri: You're questions about return in a BEGIN block.
04:39 ispy_ yeah
04:40 sri well, it looks weird, so i'll prolly not use it anyway :)
04:41 ispy_ sri: Yeah I suppose. :) hehe...
04:41 ispy_ sri: In practice I wouldn't need something like, because if I'm using a BEGIN block and something fails, I don't want to continue...
04:42 ispy_ sri: I guess with that logic, using a return value to continue proceeding is sort of a wierd way to use BEGIN in the first place.
04:42 ispy_ weird*
04:42 sri it could be useful for hackish stuff like what i removed in the commit above
04:42 sri return if $ENV{MOJO_NO_IPV6}
04:44 ispy_ Yeah, checking it out now...
04:45 ispy_ Perl and its flexibility...
04:45 ispy_ One of the things I like about it.
04:48 yko left #mojo
04:54 D4RK-PH0ENiX joined #mojo
04:57 d4rkie left #mojo
05:13 perlrocks Twitter: ""@tomaszmiklas: $self->res->headers->header('X-Bender' => 'Bite my shiny metal ass!'); <- from #mojolicious docs" that's going in my code :)" --hvcco http://twitter.com/hvcco/status/96810583407595520
05:15 crab people are easily amused.
05:17 d4rkie joined #mojo
05:20 D4RK-PH0ENiX left #mojo
06:07 marcus left #mojo
06:07 stephen left #mojo
06:07 Kulag left #mojo
06:07 bzero left #mojo
06:07 bzero_ joined #mojo
06:07 Kulag joined #mojo
06:07 tempire_ joined #mojo
06:07 mattp_ left #mojo
06:07 newbie joined #mojo
06:07 tempire left #mojo
06:07 bzero_ is now known as bzero
06:07 kaare left #mojo
06:07 marcus joined #mojo
06:07 kaare joined #mojo
06:07 mattp_ joined #mojo
06:07 stephen joined #mojo
06:07 xaka left #mojo
06:08 vel left #mojo
06:19 GitHub8 joined #mojo
06:19 GitHub8 [mojo] kraih pushed 1 new commit to master: https://github.com/kraih/mojo/commit/ea​a85ff257ddace5e098e36ba8033d4ff9af56e4
06:19 GitHub8 [mojo/master] fixed typo - Sebastian Riedel
06:19 GitHub8 left #mojo
06:21 Foxcool joined #mojo
06:52 blabos left #mojo
07:05 Alias_ left #mojo
07:14 AmeliePoulain joined #mojo
07:18 perlrocks Twitter: "Mojolicious-Plugin-PlackMiddleware 0.19 by JAMADAM - http://metacpan.org/release/JAMADAM/Mo​jolicious-Plugin-PlackMiddleware-0.19/" --cpan_new http://twitter.com/cpan_ne​w/status/96841894209925120
07:25 spleenjack joined #mojo
07:26 sugar joined #mojo
07:38 newbie left #mojo
07:43 D4RK-PH0ENiX joined #mojo
07:48 d4rkie left #mojo
08:06 sugar hello
08:06 sugar where can i get up of remote client&
08:07 sugar * s/up/ip/
08:07 sugar &
08:07 sugar ?
08:13 sugar is anybody else alive?
08:18 sugar many thanks: https://metacpan.org/module/Mo​jo::Transaction#remote_address
08:49 Alias left #mojo
09:48 D4RK-PH0ENiX left #mojo
09:59 tabbi joined #mojo
09:59 tabbi left #mojo
10:46 fmerges joined #mojo
11:26 sugar left #mojo
11:28 sugar joined #mojo
11:33 alnewkirk-hiring left #mojo
11:41 alnewkirk-hiring joined #mojo
11:50 dekimsey joined #mojo
12:24 mattastrophe joined #mojo
12:31 Akron joined #mojo
12:44 d4rkie joined #mojo
13:28 tabbi joined #mojo
13:44 tabbi left #mojo
14:12 AmeliePoulain left #mojo
14:12 mattastrophe left #mojo
14:15 Foxcool left #mojo
14:22 amoore joined #mojo
14:37 perlrocks Twitter: "Going to put #Mojolicious in production for the first time soon... Let's see if it's so awesome as everyone says :)" --cstrep http://twitter.com/cstrep/status/96952367928442880
14:44 sugar left #mojo
15:24 minimalist ping marcus
15:33 spleenjack left #mojo
16:00 xaka joined #mojo
16:05 crab shouldn't you have decided if it's awesome _before_ you put it into production?
16:06 stephan48 why?
16:07 stephan48 its much better to directly test your dev apps in production.. annoying users & customers.... you know?
16:09 crab oh, sorry. you're right, of course.
16:12 fmerges left #mojo
16:13 amoore left #mojo
16:15 nickl left #mojo
16:29 mmcleric left #mojo
16:30 mmcleric_ joined #mojo
16:31 stephen left #mojo
16:37 stephen joined #mojo
16:43 ajgb left #mojo
17:03 GabrielVieira joined #mojo
17:06 jwang left #mojo
17:17 mattastrophe joined #mojo
17:18 Akron left #mojo
17:21 Drossel joined #mojo
17:21 Kulag left #mojo
17:28 perlrocks Twitter: "The MacBook Air now, ended up in an environment most often used Perl. ~ The following JavaScript and Python and Java. WAF is to use something in Perl, that would not do what now? That would be the Mojolicious, does not it go up with the version of the angry waves." (ja) --apke99 http://twitter.com/apke99/status/96995436400873473
17:47 tempire_ "That would be the Mojolicious, does not it go up with the version of the angry waves?"
17:47 tempire_ Sounds like a biblical proverb
17:47 sri heh
18:05 tempire_ Behold, out of the eastern city there shall be a westward wind, and it shall be called, revolucion.
18:09 tempire_ left #mojo
18:10 tempire joined #mojo
18:27 GitHub9 joined #mojo
18:27 GitHub9 [mojo] kraih pushed 1 new commit to master: https://github.com/kraih/mojo/commit/fa​ce2d1da74bf018edf6a008e98cc5d7b8c1390e
18:27 GitHub9 [mojo/master] made resolver a little more resilient - Sebastian Riedel
18:27 GitHub9 left #mojo
18:36 GitHub159 joined #mojo
18:36 GitHub159 [mojo] kraih pushed 1 new commit to master: https://github.com/kraih/mojo/commit/57​447fa4d2054f81c9a9d66e201b6addd159bf3b
18:36 GitHub159 [mojo/master] made tests a little more resilient - Sebastian Riedel
18:36 GitHub159 left #mojo
18:40 perlrocks Twitter: "Mojolicious 1.68 by SRI - http://metacpan.org/release/SRI/Mojolicious-1.68/" (ms) --cpan_new http://twitter.com/cpan_ne​w/status/97013568532250624
18:52 * alnewkirk-hiring is dreaming of a cool new plugin M::P::Blacklisted that redirects ips to a blacklisted page
19:01 xantus[] is hypnotoad useful for general ioloop servers?
19:01 xantus[] I'm not using any mojolicious features in this specific server
19:02 sri interesting question
19:02 * purl scratches her chin...
19:04 sri maybe, but you'll have problems with EV
19:04 xantus[] true
19:04 sri events that are defined on startup don't survive the fork
19:05 sri once you get past that it could be fine
19:06 sri sadly even timers don't survive
19:08 sri it doesn't make much sense, maybe there's a trick
19:08 sri not listening on any ports shouldn't be such a big deal
19:09 sri the EV examples i gisted recently all work in hypnotoad btw
19:10 xantus[] I can probably pull together a preforking server lib out of it
19:11 xantus[] I need to prefork proxy
19:11 SmokeMachine_ joined #mojo
19:13 sri well, if you can get past the event setup problem hypnotoad itself could work well
19:13 sri s/well//
19:14 sri maybe there's really a trick to make timers survive fork
19:14 xantus[] hypnotoad is using timers internally
19:14 xantus[] sri: I think there is
19:15 sri yes, but only after fork
19:15 sri and only a recurring for heartbeats, to see if the event loop is still alive
19:16 SmokeMachine_ is now known as SmokeMachine
19:43 kaare left #mojo
19:45 amoore joined #mojo
19:46 arthas left #mojo
19:49 vel joined #mojo
19:50 arthas joined #mojo
20:03 dekimsey left #mojo
20:24 perlite_ joined #mojo
20:28 perlite left #mojo
20:28 perlite_ is now known as perlite
20:28 minimalist left #mojo
20:38 abra left #mojo
20:49 abra joined #mojo
21:03 arthas left #mojo
21:03 arthas joined #mojo
21:07 xantus[] sri: I'm not getting on_close when I drop() a connection
21:08 xantus[] sri: is that by design?
21:10 minimal joined #mojo
21:10 minimal left #mojo
21:13 sri xantus[]: you do
21:14 sri but only when the connection gets actually dropped
21:18 sri half of our websocket tests depend on on_close working
21:21 xantus[] I'm def not getting them using ev
21:23 minimalist joined #mojo
21:27 sri ev has msot likely still quite a few bugs
21:27 sri but i can't do anything without unit tests
21:29 sri xantus[]: in fact, you know EV better than me ;p
21:30 sri no close event hints at a leak though
21:30 sri since it is tied to DESTROY
21:32 sri actually...
21:32 vel left #mojo
21:32 sri it doesn't even have to be a bug
21:32 sri maybe your code just leaks
21:34 amoore left #mojo
21:39 xantus[] lol
21:41 sri i take that as a yes ;p
21:42 xantus[] well, now that you remind me that its called on destroy
21:42 sri it's so damn easy to leak in perl…
21:43 sri wonder which event loop handles that problem best
21:52 GitHub157 joined #mojo
21:52 GitHub157 [mojo] kraih pushed 1 new commit to master: https://github.com/kraih/mojo/commit/ad​32b8ecc60d6a9d35b11b12a0641461748c1544
21:52 GitHub157 [mojo/master] more Mojo::DOM documentation - Sebastian Riedel
21:52 GitHub157 left #mojo
22:04 SmokeMachine left #mojo
22:05 xantus[] I can't spend much time on it now, but it works ok if I don't drop the connection myself
22:17 sri just played a bit with a raw ioloop app and it seems to work flawless, with and without ev
22:21 xantus[] i'm using a drain cb to drop
22:21 xantus[] then I never get a close
22:21 xantus[] prob a leak, but I'm weakening everything I use in the sub
22:25 sri which version?
22:25 purl rumour has it which version is that
22:25 sri i fixed a drain leak recently
22:39 meraxes left #mojo
22:55 xaka how can i watch (wait) for events (io, timer) for N seconds?
22:55 stephen left #mojo
22:59 xaka ok, got it, one_tick
23:01 xaka but what i see is it sleeps for exactly $timeout so if i have recurring every 2 seconds and run one_tick(5), my recurring won't be called twice, only once
23:08 j3nnn1 joined #mojo
23:24 sri because that's not how it works
23:26 xaka yes, i understand, just a thought
23:26 stephen joined #mojo
23:27 sri btw. the answer to your question is a N second timer that stops the loop
23:32 sri hmm
23:32 * sri might have an idea for the fork problem
23:33 sri Mojo::IOLoop->deferred(sub { … });
23:34 sri it would not initialize the iowatcher and run the callbacks on the next/first reactor tick
23:44 GabrielVieira left #mojo
23:54 xaka it'd be nice to have an access to ioloop (and iowatcher) via Test::More ($t->ioloop or $t->app->ioloop) because i'm too lazy do import of Mojo::IOLoop and write long-line Mojo::IOLoop->singleton->..
23:55 stephen left #mojo

| Channels | #mojo index | Today | | Search | Google Search | Plain-Text | summary