Camelia, the Perl 6 bug

IRC log for #mojo, 2013-04-25

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

All times shown according to UTC.

Time Nick Message
00:02 moltar joined #mojo
00:04 tempire was a 4.0 unicode character decided on?
00:07 tempire I like alien monster
00:07 tempire but if it's already used, then top hat.
00:11 sri not yet
00:34 jb360 joined #mojo
00:41 shmuel joined #mojo
00:56 btyler joined #mojo
01:15 ka2u joined #mojo
01:24 jb360 joined #mojo
01:29 Kripton joined #mojo
01:45 moltar joined #mojo
01:57 Meiermann joined #mojo
02:01 d4rkie joined #mojo
02:32 jberger joined #mojo
03:17 rem_lex| joined #mojo
03:19 good_news_everyone joined #mojo
03:19 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/qRJw6A
03:19 good_news_everyone mojo/master 464562e Sebastian Riedel: made WebSocket tests a little more reliable
03:19 good_news_everyone left #mojo
03:27 good_news_everyone joined #mojo
03:27 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/3Bc0Ew
03:27 good_news_everyone mojo/master 9a0f9db Sebastian Riedel: better description for finished_ok test
03:27 good_news_everyone left #mojo
03:30 powerman left #mojo
03:45 xaka joined #mojo
04:14 Adura joined #mojo
04:27 zacts joined #mojo
04:38 egopro joined #mojo
04:55 egopro_ joined #mojo
05:03 ka2u joined #mojo
05:21 BonSequitur joined #mojo
05:23 BonSequitur So I'm teaching myself Mojolicious and I just got the most bizarre bug: I'm wrapping a little text-generating script in Mojolicious::Lite, and I noticed (after having the script run several times in the test server) that it was getting not initialized errors.
05:24 BonSequitur It took me like five minutes to realize that the closures that module creates to iterate over content were running out of content because, of course, what was separate instances before is now one process calling the same code over again. tl;dr I'm not too clever...
05:30 yakudza joined #mojo
05:41 ka2u joined #mojo
05:56 egopro joined #mojo
06:01 d4rkie joined #mojo
06:03 egopro joined #mojo
06:06 alnewkirk joined #mojo
06:07 egopro_ joined #mojo
06:20 lukep joined #mojo
06:23 Mike-PerlRecruiter_ joined #mojo
06:27 dpetrov_ joined #mojo
06:42 ver joined #mojo
06:48 Vandal joined #mojo
06:56 suy joined #mojo
07:03 dod joined #mojo
07:12 entro joined #mojo
07:13 denisboyun joined #mojo
07:17 entro joined #mojo
07:21 ObseLeTe joined #mojo
07:28 denisboyun_ joined #mojo
07:28 dod joined #mojo
07:34 egopro joined #mojo
08:07 yakudza joined #mojo
08:12 rem_lex joined #mojo
08:13 Meiermann joined #mojo
08:13 jzawodn joined #mojo
08:16 egopro joined #mojo
08:17 dod joined #mojo
08:32 mire joined #mojo
08:39 Vandal joined #mojo
08:47 egopro joined #mojo
08:55 egopro joined #mojo
08:55 maxhq joined #mojo
08:59 denisboyun joined #mojo
09:02 nelio joined #mojo
09:13 egopro joined #mojo
09:13 entro joined #mojo
09:44 denisboyun_ joined #mojo
09:44 amirite_ joined #mojo
10:00 xaka joined #mojo
10:06 bowtie joined #mojo
10:40 Britzel joined #mojo
10:51 egopro joined #mojo
10:55 mrphilov joined #mojo
11:07 mire joined #mojo
11:22 egopro joined #mojo
11:29 Slatibart Hi, is anyone familiar with the setup of apache/mod_proxy for mojo ? http://mojolicio.us/perldoc/Mojolici​ous/Guides/Cookbook#Apachemod_proxy If I understand correctly, the example would map everything to the mojo server. But since other contents and cgi scripts live on the server I only can map port of the url. For example , www.example.com/mojo should go through mod_proxy
11:57 yakudza joined #mojo
11:57 egopro joined #mojo
12:13 hrupp Slatibart: we use mod_rewrite in similar case
12:13 hrupp http://pastie.org/7716408
12:14 Slatibart hrupp This is a different approach, right ?
12:15 Slatibart I'm not so deep into apache, unfortunately
12:16 hrupp yes different, but the same goal
12:18 Slatibart so basically everything not plain html goes to the "mojo" server
12:19 hrupp everything from our old cgi envoirement is kept
12:19 bluescreen joined #mojo
12:20 hrupp cgi-bin and the static content
12:20 hrupp everything else is rewritten to reach hypnotoad
12:20 Slatibart ah got it now
12:21 Slatibart thanks for the help.
12:21 hrupp its a neogated rewrite
12:21 hrupp no prob
12:58 egopro joined #mojo
13:06 moltar joined #mojo
13:13 mire joined #mojo
13:17 dvinciguerra joined #mojo
13:38 denisboyun_ joined #mojo
13:40 btyler joined #mojo
13:46 Kripton joined #mojo
13:48 nelio joined #mojo
13:58 asarch joined #mojo
13:58 egopro joined #mojo
14:03 gryphon joined #mojo
14:22 rihegher joined #mojo
14:24 SmokeMac_ joined #mojo
14:28 kvorg_ joined #mojo
14:28 sh4 joined #mojo
14:29 kvorg joined #mojo
14:47 good_news_everyone joined #mojo
14:47 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/43cpsQ
14:47 good_news_everyone mojo/master bc4ac25 Sebastian Riedel: added diagnostics information for finished_ok test
14:47 good_news_everyone left #mojo
14:51 whitebook joined #mojo
14:54 labrown joined #mojo
14:59 egopro joined #mojo
15:02 good_news_everyone joined #mojo
15:02 good_news_everyone [mojo] kraih created log_diagnostics (+1 new commit): http://git.io/wDWk1w
15:02 good_news_everyone mojo/log_diagnostics 072a3d5 Sebastian Riedel: added testing attribute to Mojo::Log
15:02 good_news_everyone left #mojo
15:02 sri jberger, marcus, tempire, crab: what do you think?
15:04 travis-ci joined #mojo
15:04 travis-ci [travis-ci] kraih/mojo#523 (log_diagnostics - 072a3d5 : Sebastian Riedel): The build passed.
15:04 travis-ci [travis-ci] Change view : https://github.com/kraih/mojo/commit/072a3d5dda49
15:04 travis-ci [travis-ci] Build details : http://travis-ci.org/kraih/mojo/builds/6632955
15:04 travis-ci left #mojo
15:05 jberger I suppose that could be useful
15:07 sri the point is that log messages don't interfere with the test harness
15:18 jberger yeah, I get that
15:18 Kripton joined #mojo
15:24 jberger my wife thinks that \o/ looks like a drowning guy :-o
15:25 tempire that's better than what most people think it is originally
15:26 ka2u joined #mojo
15:27 sri |\      \o/
15:30 sri hmmm
15:30 sri i suppose log output interfering with the test harness is a non issue... since it goes to STDERR by default
15:32 * sri rejects his proposal
15:32 good_news_everyone joined #mojo
15:32 good_news_everyone [mojo] kraih deleted log_diagnostics at 072a3d5: http://git.io/qlYbuw
15:32 good_news_everyone left #mojo
15:37 good_news_everyone joined #mojo
15:37 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/hj8eeA
15:37 good_news_everyone mojo/master 6d522b1 Sebastian Riedel: better diagnostics message for finished_ok test
15:37 good_news_everyone left #mojo
15:46 good_news_everyone joined #mojo
15:46 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/1mO0MA
15:46 good_news_everyone mojo/master d8c7fd1 Sebastian Riedel: better test descriptions
15:46 good_news_everyone left #mojo
15:47 sri kind of dislike that the diagnostics message is the same sentence as the description... but can't think of anything better :/ https://github.com/kraih/mojo/commit/d8c7f​d158628d0908f188f5146960f1bfd57e5a9#L0R114
15:53 d4rkie joined #mojo
15:59 egopro joined #mojo
16:03 xaka joined #mojo
16:03 amirite joined #mojo
16:08 amirite_ joined #mojo
16:16 basiliscos joined #mojo
16:18 mrphilov apologies if this is newbish, but i've been going in circles through the documentation with no luck… i've got a regular application, not lite, and i'm doing too much concurrency now such that i'm running it with hypnotoad. everything works, but i cannot for the life of me figure out how to get logging to go anywhere.
16:18 mrphilov clues?
16:19 whitebook joined #mojo
16:19 mrphilov (doing $self->app->log->info("…") which may be incorrect)
16:19 marty joined #mojo
16:21 marty_ joined #mojo
16:25 whitebook1 joined #mojo
16:26 sri mrphilov: did you skip the lite tutorial?
16:27 mrphilov I did not, no, but I'll take that as a hint and go review it now.
16:29 whitebook joined #mojo
16:29 mrphilov So the only reference to logging I can find in that tutorial is this: "http://mojolicio.us/perldoc/Mojo/Log messages will be automatically written to STDERR or a log/$mode.log file if a log directory exists." - STDERR on a pre-forking webserver goes… where? Not to my terminal STDERR, certainly. I created a "log" directory in the PWD where I was running the server, but that got nothing in it.
16:30 mrphilov Logging works fine when I'm running single-threaded. It goes to STDERR.
16:31 mrphilov But that's not really gonna work long-term.
16:31 whitebook1 joined #mojo
16:35 sri so, if it goes to STDERR with morbo, obviously your log directory is in the wrong place
16:36 sri also, nothing in mojolicious uses threads
16:38 whitebook1 joined #mojo
16:40 sri worst thing about postgresql is the zealots that keep bashing all other databases indiscriminately
16:40 sri it's almost as bad as python
16:43 sri every time i open a mysql article i know in advance that there's gonna be an ad hominem attack in the comments
16:46 whitebook joined #mojo
16:50 ka2u joined #mojo
16:52 alester joined #mojo
17:00 egopro joined #mojo
17:01 yakudza joined #mojo
17:08 marty joined #mojo
17:09 rihegher left #mojo
17:11 Mikey joined #mojo
17:19 sri wwdc tickets sold out in under 2 minutes :O
17:22 Mikey they should have a conference for real developers, too
17:29 alester rrroowwww
18:00 egopro joined #mojo
18:18 sh4|2 joined #mojo
18:23 Mike-PerlRecruiter_ joined #mojo
18:29 amirite_ joined #mojo
18:31 egopro joined #mojo
18:41 sh3 joined #mojo
18:42 basiliscos joined #mojo
18:50 komodo1 joined #mojo
18:58 nelio joined #mojo
19:04 mire joined #mojo
19:29 yakudza joined #mojo
19:30 * sri wonders if there's a "correct" way to handle unicode characters in test descriptions
19:31 sri been browsing through TAP documentation, but couldn't find anything about encoding
19:31 egopro joined #mojo
19:34 sh4 joined #mojo
19:41 tianon joined #mojo
19:43 kvorg joined #mojo
19:43 kvorg_ joined #mojo
19:46 btyler_ joined #mojo
20:14 mattp_ joined #mojo
20:19 sh3 joined #mojo
20:21 ObseLeTe joined #mojo
20:32 egopro joined #mojo
20:50 BonSequitur joined #mojo
20:50 BonSequitur Hey is there somewhere I can find clean copies of the Mojolicious logo to link back to mojolici.us in an app?
20:54 GabrielVieira joined #mojo
20:55 GabrielVieira how to send a signal to hypnotoad reload the app?
21:03 egopro joined #mojo
21:03 chansen sri: the TAP specification is flawed, it says "Any text after the test number but before a # is the description of the test point", it's not possible to interchange "any text" without an encoding
21:03 sri chansen: yea
21:06 chansen I would stick with ASCII and escape any Unicode characters using the U+XXXX(?:XX) notation
21:16 buu GabrielVieira: Don't you just rerun the hypnotoad script and that reloads it?
21:17 buu And or USR2?
21:19 GabrielVieira how can I send a signal to hypnotoad to reload the app OR share a hash between workers (I set an attr and updated its value, but the other workers didn't update)?
21:19 GabrielVieira buu: I can try system("hypnotoad /path/to/app.pl"); but I wan't this to be the last solution
21:19 rl joined #mojo
21:20 sri chansen: maybe, it's strange Test::Simple has no solution for it yet
21:21 chansen sri: true, why not bring it up in #toolchain ?
21:21 rl Hi folks, I just experimented with Mango and Mojolicious::Lite
21:21 gryphon joined #mojo
21:22 buu GabrielVieira: If you want to share data between workers you need to store it outside the worker in a shared environment
21:22 rl And found that the URL mongodb://dev1/testdb was not parsed correctly
21:22 buu AKA memcache
21:22 btyler_ redis also works well
21:23 rl Just added some tests for t/mojo/url.t to test that scenario and they fail there (so it's not a problem in Mango, but in Mojo::URL)
21:24 rl Did I get something wrong, I thought that Mojo::URL should parse any URI like stuff
21:24 sri rl: oh yea, my bad, i broke it with the last Mojolicious release
21:25 sri new Mango fixing it should be out soon
21:26 rl Im pretty much on master, so that's probably something your cooking right now?
21:26 sri yea
21:26 rl Ok great, so no tests needed for Mojo::URL?
21:27 sri nope
21:27 sri i'm rewriting the Mango core atm
21:27 rl Ah ok, sorry to interrupt ;-)
21:31 sri question is if it makes sense to add Mojo::URL::schemes, to make supported schemes configurable
21:31 sri Mojo::URL->new->schemes([qw(http https ws wss mongodb)])->parse(...)
21:32 sri or if i should just use what we have now to rewrite the mongodb url
21:32 sri $url = Mojo::URL->new($url->scheme('http'));
21:33 rl Just digged into the RFCs a bit
21:33 rl Mojo::URI (with an i not L) would allow authority for all protocols
21:33 rl as we call it URL it should be limited to supported schemes
21:33 sri we don't do URIs
21:34 egopro joined #mojo
21:34 sri that RFC is mostly irrelevant for us
21:34 rl yepp, so acutally mongodb:// would be an unsupported scheme anyway
21:35 sri the way browsers handle URLs is a big mess, and we follow browsers more closely than the specs
21:35 sri http://url.spec.whatwg.org # the closest to a spec you will get
21:35 rl if the RFC is irrelevant, why not parse for all types by default than
21:35 lukep joined #mojo
21:35 rl (not limit to http and ws)
21:36 sri because we follow the whatwg spec
21:37 sri all schemes can be parsed, but unsupported ones have only two parts, scheme and scheme data
21:38 rl ok, understood, thanks for the hint
21:39 rl To make it do the right thing you would need to parse a function to correctly resolve the parts
21:40 rl but that sound a little to funky to be usable
21:41 ask joined #mojo
21:44 sri mojolicious master has everything needed to make it work already
21:45 sri only change in Mango::new needed is the addition of "$url = Mojo::URL->new($url->scheme('http'));"
21:48 good_news_everyone joined #mojo
21:48 good_news_everyone [mojo] kraih tagged v3.97 at d0e4ede: http://git.io/fWP6jA
21:48 good_news_everyone left #mojo
21:49 sri simple solution and quick mojolicious release it is then ;p
21:51 njlg nice
21:57 whitebook joined #mojo
22:05 egopro joined #mojo
22:06 amirite joined #mojo
22:11 hrupp joined #mojo
22:13 rem_lex|pivo joined #mojo
22:35 egopro joined #mojo
22:41 whitebook joined #mojo
23:06 tianon sri: glad to see you came up with the same workaround we ended up with for our FTP parsing - makes me feel like I'm doing it the "right way" when you use the same solution :)
23:07 egopro joined #mojo
23:07 sri :)
23:08 sri Mojo::URL->new->schemes([qw(http https ws wss mongodb)])->parse(...) is still an option, depending on feedback
23:09 sri the scheme change trick is a feature and will keep working though
23:10 sri ohoh, github having problems again :S
23:15 sivoais joined #mojo
23:37 egopro joined #mojo
23:50 mire joined #mojo

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