Camelia, the Perl 6 bug

IRC log for #mojo, 2010-11-23

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

All times shown according to UTC.

Time Nick Message
00:18 xantus[] sri: http://www.kevs3d.co.uk/dev/asteroids/
00:28 jamesw especially now there are Mouse and Moo
00:30 jamesw if you're following the modern perl playbook and using Moose, shouldn't you be using POE too? heh
00:31 MojoGuest874 left #mojo
00:48 MojoGuest994 joined #mojo
00:48 MojoGuest994 From: http://www.google.co.uk/search?aq=f&source​id=chrome&ie=UTF-8&q=html+5+irc+client (1 hits)
00:53 MojoGuest994 left #mojo
01:27 jdixon joined #mojo
01:28 jdixon is it possible to have optional placeholders in a uri?
01:28 jdixon e.g. to identify api and version?
01:29 jdixon just to clarify, I'm referring to Mojolicious::Lite
01:38 Alias joined #mojo
01:43 achromic left #mojo
01:43 jdixon sri
02:36 fhelmber_ joined #mojo
02:42 stephen left #mojo
02:44 Akron joined #mojo
02:53 Akron left #mojo
03:13 trcjr joined #mojo
04:03 tl joined #mojo
04:05 tholen42 left #mojo
04:20 MojoGuest777 joined #mojo
04:22 MojoGuest777 left #mojo
04:22 MojoGuest183 joined #mojo
04:22 MojoGuest183 Hello there
04:25 MojoGuest183 left #mojo
04:48 achromic joined #mojo
05:05 yakudzo left #mojo
05:08 yakudzo joined #mojo
05:14 su-bzero joined #mojo
05:18 trcjr left #mojo
06:00 kaare joined #mojo
06:31 und3f[A] is now known as und3f
06:45 ltriant left #mojo
07:00 sri jdixon: i think thats mentioned in the routing guide
07:02 und3f is now known as und3f[A]
07:08 ysyrota joined #mojo
07:50 * xantus is hacking on a riak mojo module
07:52 koban joined #mojo
08:11 y0j left #mojo
08:26 y0j joined #mojo
08:38 ysyrota what does mean an option 'default' in M::Plugin::I18n?
08:48 otaviof joined #mojo
08:48 otaviof left #mojo
09:18 arthas joined #mojo
09:20 tholen42 joined #mojo
09:22 tl left #mojo
09:41 sri ysyrota: default language?
09:47 ysyrota sri: I thought in the same way, actually it is a language used in string Ids
09:48 sri ?
09:48 ysyrota sri: is there a way to set default language? I want to use some specific language is there is no Accept-Language header in request
09:49 sri umm, thats what i just said
09:49 sri you are very hard to understand :S
09:49 ysyrota sri: it is hard to explain :)
09:52 ysyrota sri: if you set default language then i18n plugin will generate the following %Lexicon for this language: (_ AUTO => 1)
09:53 ysyrota sri: so the plugin won't do translation for this language
09:53 ysyrota sri: but I want to setup default translation
09:54 sri not sure you can
09:56 sri i would accept a patch if it had really good tests
10:03 arthas left #mojo
10:03 ysyrota sri: i see
10:10 perlrocks Twitter: "kansai.pm upload videos that we've taught from friends lapis25 Mojolicious:: Lite touch from now on I'm gonna! !" (ja) --AzureStone http://twitter.com/AzureSto​ne/status/7013201552080896
10:11 sri and of course you would have to find a nice solution :)
10:14 sri something like default => 'en', pass_through => 0
10:14 sri pod updates would also be nice
10:14 spleenjack joined #mojo
10:15 crab these translations from japanese are insane
10:17 sri they are much better than nothing ;)
10:37 snearch joined #mojo
11:08 sri xantus: found the javascript glitch btw?
11:10 MojoGuest909 joined #mojo
11:11 MojoGuest909 left #mojo
11:12 sri xantus: ngrep just tells me that nothing leaves the browser
11:12 sri no outgoing network traffic at all
11:13 sri so it has to be something around your websocket write code
11:15 sri nothing in the javascript console either
11:18 und3f[A] is now known as und3f
11:20 snearch left #mojo
11:47 Akron joined #mojo
11:50 arthas joined #mojo
11:53 arthas left #mojo
11:53 arthas joined #mojo
12:20 MojoGuest935 joined #mojo
12:20 MojoGuest935 From: http://news.ycombinator.com/item?id=1277067 (57 hits)
12:20 MojoGuest935 left #mojo
12:24 s0va sri: i'm getting Event "error" failed for connection "a18f55c": Can't call method "error" without a package or object reference at /home/siolboxadmin/siol_box/c​ontrib/Mojo/Server/Daemon.pm line 183.
12:24 s0va ... just as logging object would dissappear.
12:25 s0va well, problem is that daemon stops completely.
12:25 sri whatever you are doing, don't do that :)
12:25 s0va process still lives but connection to daemon is not accepted
12:25 s0va i didn't do anything.
12:26 s0va :) just plain lite app
12:26 sri do tests pass?
12:26 s0va mojo is behind nginx proxy
12:26 s0va yep.
12:26 sri well, daemon has like a million tests
12:26 sri it is something you do
12:26 s0va fucking weird... and this happens after server minutes of running
12:27 sri then you are leaking descriptors or so
12:30 s0va hm...
12:31 s0va the only "illegal thing", that i'm doing is: ioloop->timer(10, sub { config_reload })
12:31 s0va config_reload() just re-enqueues itself into timer
12:32 sri do you have a test case?
12:32 sri does reload config open a file each time?
12:33 s0va only if it changed.
12:33 s0va file is read using IO::File->new()
12:33 s0va file's fd should be closed automatically when io::file object is destroyed
12:34 s0va ... since reloading is done in sub file should be closed asap as sub finishes.
12:34 sri https://github.com/kraih/mojo/blob/master/lib/Mojo​licious/Guides/FAQ.pod#I_think_i_have_found_a_bug,_what_should_i_do_now?
12:38 s0va hm, maybe it has to do something with installed perl version
12:38 s0va i'm developing on 5.12
12:38 s0va prod box has 5.10
12:40 sri unlikely
12:41 s0va it's fedora install
12:41 sri still unlikely
12:41 s0va ... even rsync doesn't works *always* as advertised on mentioned host.
12:41 sri if it was a perl bug it should have been reported before and possibly showed up in tests
12:42 s0va fuck man
12:42 s0va make test fails on problematic host!
12:42 sri hmm
12:43 s0va #   Failed test '500 Internal Server Error'
12:43 s0va #   at t/mojolicious/lite_app.t line 1293.
12:43 s0va #          got: '404'
12:43 s0va #     expected: '500'
12:43 s0va git master, pulled few minutes ago
12:43 sri now a perl bug becomes a possibility again
12:45 s0va shit man.
12:45 s0va i'm gonna rerun tests on my laptop...
12:46 jdixon sri: regarding the placeholder I mentioned above...
12:46 jdixon the only thing I see in the routing guide mentions optional placeholders as the last part of the uri
12:46 jdixon I don't see any examples where it would be at the root of the uri (or somewhere in the middle)
12:46 sri jdixon: you can also control the regex compilation process
12:47 jdixon is there a better way of combining "normal" and api requests in a single route based on the uri?
12:48 sri jdixon: ?
12:48 sri jdixon: see "More Restrictive Placeholders" for regex features
12:50 sri it's rather uncommon to do that though
12:50 jdixon hmm
12:51 s0va sri: tests fail on my laptop too (perl 5.12.1)
12:51 s0va sri: http://pastebin.com/KYTQGecr
12:51 sri s0va: weird
12:52 sri s0va: only lite_app.t?
12:52 s0va perl Makefile.PL; make; make test
12:52 sri is that the whole output?
12:52 s0va jep.
12:52 s0va make test hangs.
12:52 sri so all previous tests pass?
12:53 s0va yup
12:53 sri very weird
12:53 s0va i'll also check on ubuntu lucid.
12:54 sri that doesn't make any sense at all, there is nothing special about that test
12:55 s0va sri: ubuntu lucid 10.04 LTS passes test!
12:55 s0va perl 5.10.1
12:55 s0va let's see on 32bit ubuntu lucid...
12:58 s0va 32bit ubuntu lucid passes the test.
13:00 jdixon sri: is there any access http headers within Lite?
13:00 jdixon e.g. I just want to dump out everything to see what referer I'm getting
13:01 sri $self->req has everything
13:01 jdixon thx
13:02 s0va sri: on which platform do you develop?
13:02 sri s0va: os x
13:02 sri which is very BSDish
13:02 s0va hmm... sri: tests on master pass on 32/64bit ubuntu (perl 5.10.1) with and without io::epoll installed
13:03 daviddelikat left #mojo
13:03 sri oh, broken epoll is also a possibility
13:03 sri or rather broken IO::Epoll
13:03 s0va failed: 32bit opensuse 11.3 perl 5.12.1, 32bit fedora 10, perl 5.10.0
13:04 s0va tests also fail if i remove io:epoll
13:04 s0va wtf.
13:04 sri right, ignore that
13:04 sri epoll is deactivated for tests
13:06 chansen left #mojo
13:06 s0va just removed io::epoll from my opensuse completely, tests still fail...
13:06 s0va omfg...
13:07 sri have you tried 0.999941?
13:07 koban left #mojo
13:07 koban joined #mojo
13:07 sri would be helpful to compare the results with master
13:15 GitHub20 joined #mojo
13:15 GitHub20 mojo: master Sebastian Riedel * fdd42b5 (1 files in 1 dirs): improved Mojolicious::Lite tutorial - http://bit.ly/hQGrix
13:15 GitHub20 left #mojo
13:20 chansen joined #mojo
13:26 s0va sri: all tests were done on fresh git cloned repository.
13:26 s0va therefore: on master branch
13:27 sri yea, but a test run with 41 would help a lot
13:27 jdixon sri: re the "Mojo server" post, it's not a mojo issue.  sorry for the noise.
13:28 jdixon stupid fontdeck server is passing back the wrong referer, wtf.
13:29 sri jdixon: expected something like that, but your report was too cryptic
13:31 jdixon sorry
13:33 s0va sri: just a sec.
13:48 MojoGuest851 joined #mojo
13:48 MojoGuest851 From: http://news.ycombinator.com/item?id=1277067 (58 hits)
13:48 MojoGuest851 left #mojo
13:57 kaare left #mojo
14:05 daviddelikat joined #mojo
14:16 vel joined #mojo
14:32 MojoGuest185 joined #mojo
14:32 MojoGuest185 left #mojo
14:42 yko joined #mojo
14:43 sri s0va: results?
14:44 und3f hello
14:44 * sri waves
14:51 und3f sri, we got tcp connect check
14:51 sri und3f: you make no sense
14:51 und3f why?
14:52 und3f on check-host.net
14:52 sri because i don't understand what you're trying to say
14:52 sri ah, context helps ;)
14:53 und3f sri, i am about tcp connection ping
14:54 sri now i get it :)
14:58 sri und3f: seems very close to normal ping
14:59 sri tcp connect is more interesting though
14:59 und3f why?
15:00 sri because ping doesn't matter like at all, blocking it doesn't matter
15:01 sri connect actually gives me a useful value
15:01 sri i know how quickly a tcp connection can be established from many places around the world
15:02 sri at least for web servers thats more interesting
15:03 und3f so we rock )
15:04 sri und3f: maybe more information about the test servers would be interesting
15:05 sri like data center and backbones they are connected to
15:05 sri in case you have that information
15:06 und3f i have, but we don't want to post it in public, at least now
15:07 und3f Can send you it in private
15:08 sri why not?
15:10 sri it would be nice as a tooltip
15:10 sri a list of the backbones every server is connected to, in case something is blocked or so
15:11 sri just a thought
15:12 und3f hm, actually backbones information we don't have, possible we can request it from providers...
15:12 und3f but not now
15:12 und3f Will create such ticket, possible will resolve it some day :)
15:15 sri my primary hosting provider has a public list
15:23 kvorg joined #mojo
15:26 s0va sri: git checkout v0.999941
15:27 sri 41 is also the latest cpan release
15:28 s0va sri: opensuse 11.3, 32bit, perl 5.12.1, fail at: t/mojolicious/lite_app.t ...................... 3/664
15:29 sri good, then your box is screwed :)
15:29 sri (good for me)
15:29 sri http://www.cpantesters.org​/distro/M/Mojolicious.html # because of this
15:29 spleenjack1 joined #mojo
15:31 s0va sri: fedora 10, 32bit, perl 5.10.0, fail at: t/mojolicious/lite_app.t ...................... 1/664 We won't be doing any rendezvous publishing, please install a Net::Rendezvous::Publish::Backend:: module
15:31 s0va :)
15:32 sri yea, enough seen, now i can lay back and order some pizza :D
15:32 ashleydev_logbot left #mojo
15:32 lee7 left #mojo
15:32 spleenjack left #mojo
15:32 lee7 joined #mojo
15:32 Akron left #mojo
15:33 Alias left #mojo
15:33 ashleydev left #mojo
15:34 * s0va is fucking pissed off...
15:35 sri pizza helps
15:36 s0va this is weird.
15:36 s0va master/head tests run ok on my ubuntu boxes...
15:36 s0va v0.999941 tests fail at lite_app.t 3/664
15:37 perlrocks Twitter: "Mojolicious made about the tutorial, this time writing roundup" (ja) --shiba_yu36 http://twitter.com/shiba_yu​36/status/7095363399847936
15:38 achromic left #mojo
15:43 achromic joined #mojo
15:51 achromic left #mojo
15:53 achromic joined #mojo
16:19 ashleydev_logbot joined #mojo
16:22 arthas left #mojo
16:28 perlrocks Twitter: "Video http://bit.ly/fLvSXX: Kansai.pm 13th meeting in Kyoto - Perl Study Memory - Hatena:: Group:: Perl Perl | 21:40 | Hello, Mojolicious:: Lite in learning A.." (ja) --recenturl http://twitter.com/recentu​rl/status/7108165032222720
17:27 ysyrota left #mojo
17:29 spleenjack1 left #mojo
17:29 spleenjack joined #mojo
17:31 koban left #mojo
17:42 yko :D
17:45 und3f yko, you got the power)
17:51 yko right you are, young padawan
17:52 * yko likes to use power
17:57 und3f lhm
18:08 ashleydev joined #mojo
18:32 sri heh
18:38 stephen joined #mojo
18:41 stephen left #mojo
18:42 stephen joined #mojo
18:45 Akron joined #mojo
18:53 spleenjack left #mojo
19:00 * und3f on dark side of power
19:00 und3f yko, ^_^
19:03 jdixon left #mojo
19:08 yko -_-
19:17 GitHub173 joined #mojo
19:17 GitHub173 mojo: master Sebastian Riedel * 7d63452 (4 files in 4 dirs): cleanup - http://bit.ly/fecG3s
19:17 GitHub173 left #mojo
19:30 GitHub60 joined #mojo
19:30 GitHub60 mojo: master Sebastian Riedel * 731216f (1 files in 1 dirs): improved tick and idle events - http://bit.ly/i2jau1
19:30 GitHub60 left #mojo
19:32 arpadszasz joined #mojo
19:40 marcus joined #mojo
19:41 perlrocks Twitter: "Mojolicious:: Lite to make chat with at WebSocket http://bit.ly/h2ifz4" (ja) --MorningMae http://twitter.com/MorningM​ae/status/7156915402444800
19:50 marcus left #mojo
19:58 marcus joined #mojo
20:29 tempire joined #mojo
20:31 tempire I would love to see more client javascript frameworks.
20:31 tempire There's a massive need for such things.
20:32 tempire JavascriptMVC has too much scaffolding.  A lot of typing.
20:33 tempire Sammy looks interesting, though it doesn't seem to allow for progressive enhancement without writing wrapper functions.
20:34 tempire I may have to create my own, based on Nicholas Zakas' talks.
20:35 Akron Well it allows for helpers - and you've got the complete jQuery enhancement stuff.
20:35 tempire I'm hoping Sammy will do the trick for my next app.
20:37 Akron I was thinking of a simple Mojolicious plugin that creates Sammy routes from Mojolicious - to allow one-site app as well as No-JS-App with not much overhead.
20:47 janus left #mojo
20:48 sri i would like a perl to js compiler :)
20:48 sri something like Pyjamas
20:52 und3f i think i could add measuring of connection time in http_request
20:53 und3f (for check-host)
20:54 und3f that seems to be useful
20:58 sri it does
21:02 ltriant joined #mojo
21:09 und3f and also resolve time
21:14 und3f sri, i want to create connection by myself and then pass it to $tx, but the problem is in callback
21:14 sri there is no problem
21:14 sri just put the id into $tx->connection
21:14 und3f seems like now i must to use hacks and set connection callbacks to Mojo::Client private method
21:15 und3f sri, i found that part
21:15 und3f sri, but what about callback of connection? There are no problems with them
21:15 sri ?
21:15 und3f on_connect, on_read
21:16 und3f actually on_read
21:16 sri hmm
21:16 sri thats fixable, but i need a test case
21:18 und3f i don't want write anything right now
21:18 und3f i am gonna sleep soon
21:18 sri ok
21:19 fhelmber_ left #mojo
21:33 MojoGuest828 joined #mojo
21:33 MojoGuest828 From: http://www.reddit.com/r/html​5/comments/bt76k/irc_client/ (15 hits)
21:34 MojoGuest828 left #mojo
21:34 Akron Is there a possibility to deactivate "strict" in templates?
21:36 GitHub105 joined #mojo
21:36 GitHub105 mojo: master Sebastian Riedel * 3f5e2cc (2 files in 2 dirs): improved Mojo::Client slightly and added tests - http://bit.ly/e4PZ10
21:36 GitHub105 left #mojo
21:37 sri und3f: it's a good test case, so i made it work anyway
21:39 chansen marcus: where will you stay during LPW?
21:39 chansen sri: will you attend LPW?
21:39 sri i've not attended any perl conference yet :S
21:40 chansen neither have I, but perhaps it's time we old monkeys show some world order? ;P
21:40 sri :D
21:41 und3f thanks )
21:42 sri und3f: the test is broken though, needs more fixing
21:42 und3f but even because of this commit i will not start to code now. i need to sleep
21:42 und3f i am ill a little bit..
21:43 sri sounds like you need more vodka
21:44 und3f i don't drink alcohol at all
21:44 sri wait, you are not russian?
21:45 und3f But vodka with onion and black pepper is one of effective healing method here (tried it sometimes)
21:45 und3f sri, Ukrainian
21:45 GitHub111 joined #mojo
21:45 GitHub111 mojo: master Sebastian Riedel * df168eb (1 files in 1 dirs): fixed test - http://bit.ly/hOcGeg
21:45 GitHub111 left #mojo
21:45 sri close enough
21:46 und3f нуфр
21:46 und3f yeah
21:48 GitHub84 joined #mojo
21:48 GitHub84 mojo: master Sebastian Riedel * d5a3364 (1 files in 1 dirs): upated changes - http://bit.ly/gP0Pw8
21:48 GitHub84 left #mojo
21:53 tempire If you folks show up at a perl conference, I'll make that one my first one as well.
21:53 tempire Onward!  To the Mojolicous!
21:54 und3f good night
21:54 spleenjack joined #mojo
21:54 sri ALL GLORY TO THE HYPNOTOAD!
21:54 sri nn
21:55 tempire left #mojo
21:55 sri this weeks futurama was a little disappointing
21:55 sri very few new quotes
21:59 xantus[] :(
22:04 yko left #mojo
22:07 daviddelikat left #mojo
22:07 daviddelikat joined #mojo
22:08 daviddelikat left #mojo
22:08 arpadszasz left #mojo
22:17 Akron sri: Regarding autovivication - how about $prepend .= 'no strict "vars";' if ...; in line 70 of Mojolicious::Plugin::EpRenderer.pm and making this activable?
22:18 sri Akron: i don't like encouraging bad code
22:19 Akron sri: It's pretty usefull for <% if ($var) { %> ... <% } %>!
22:19 sri % no strict 'vars'; should be fine too, no?
22:19 Akron Yes ... but this is more ugly, I think.
22:20 Akron It's internal hacking in a template.
22:24 sri plugin ep_renderer => {name => 'foo', prepend => 'no strict "vars";'};
22:24 sri that maybe, but i'm against encouraging bad style
22:26 sri the whole if ($var) thing is not good though
22:26 Akron Hm ... looks too internal to me. How would you do this thing?
22:27 sri i'd rather suggest using if (stash 'var') or set a default in app->defaults(var => undef);
22:27 Akron It - how ever - works with stash('var') of course.
22:27 Akron ;)
22:27 sri or even the route
22:28 sri get '/foo' => {var => undef} => sub {...};
22:29 Akron But than I have to trace all variables used in my templates ... this is not user friendly.
22:29 sri there are always tradeoffs
22:29 sri the stash helper is prolly the best option for optional stash keys
22:30 sri no strict 'vars' has this php feel to it
22:31 memememomo_ is now known as memememomo
22:31 Akron Yes - I know. But I really think it's user friendly. Mixing stash helper values and $-vars in a template doesn't seem right.
22:31 memememomo is now known as memememomo_
22:32 Akron But a "prepend"-Parameter is too internal ...
22:32 sri % no strict 'vars'; has the advantage that the non standard behavior becomes obvious
22:32 Akron How about a "bad style" => "yes"? ;)
22:33 sri bad_style => "yes seriously i'm absolutely really sure i want perl to explode"
22:34 Akron That's a better she bang for my apps.
22:34 und3f is now known as und3f[A]
22:34 sri but i really think % no strict 'vars'; is the best option
22:35 sri keeps everything together
22:35 sri no surprises
22:35 Akron Okay. Thank you!
22:35 sri what did i do? :)
22:35 vel left #mojo
22:37 Akron So Mojolicious::Plugin::BadStyle with a "bad_style" helper is a good thing?
22:38 sri what would the helper do?
22:38 Akron Warn for valid code.
22:38 Akron Encourage using PHP instead.
22:38 sri Oo
22:39 sri Mojolicious::Plugin::PHPishRenderer? :)
22:40 Akron "Execution failure - too many comments. And stop using indentation - you are no Python Hacker!"
22:41 sri sounds sensible
22:46 sri Zzz
22:47 sri btw. with the prepend argument you could wrap everything in a separate renderer and make it look like "plugin bad_style_renderer" or so
22:47 sri zzZ (rly)
22:49 daviddelikat joined #mojo
22:51 chansen sri: Will you attend?
22:52 chansen we could drink wise beer if that will make you more comfortable ;P
22:57 MojoGuest423 joined #mojo
22:57 MojoGuest423 hello
23:03 MojoGuest423 left #mojo
23:39 fhelmberger left #mojo
23:39 fhelmberger joined #mojo

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