Camelia, the Perl 6 bug

IRC log for #mojo, 2011-11-24

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

All times shown according to UTC.

Time Nick Message
00:12 mojimbo joined #mojo
00:15 elb0w_ joined #mojo
00:24 MojoGuest129 joined #mojo
00:24 MojoGuest129 From: http://www.google.com/url?sa=t&rct=j&am​p;q=&esrc=s&source=web&cd=7&amp​;ved=0CGkQFjAG&url=http%3A%2F%2Fdev.xan​tus.org%2F&ei=oY7NTsW4GK2rsAL2pozaDg&am​p;usg=AFQjCNGw6nZv0wO42uzUnHhWcD0vcKNULg (1 hits)
01:52 mire joined #mojo
02:17 metaperl joined #mojo
02:20 gshank joined #mojo
02:56 lammel2 joined #mojo
02:58 lammel2 left #mojo
03:04 Eugene joined #mojo
04:48 plu morning!
04:48 purl Mornings are great. Every time you experience morning, you're not dead yet!
04:49 plu purl++
04:49 plu purl: have a cookie!
04:49 purl Yum, yum!
05:03 sri botsnack!
05:03 purl thanks sri :)
05:10 tempire :(
05:32 ispy_ joined #mojo
05:40 Foxcool joined #mojo
05:59 Eugene joined #mojo
06:33 arpadszasz joined #mojo
06:40 batman joined #mojo
06:58 marcus Moin
07:04 sri ALL GLORY TO THE HYPNOTOAD!
07:06 koban joined #mojo
07:06 crab 3.5 units of glory to the hypnotoad
07:07 crab sometimes i'm very thankful that i have such detailed log messages
07:07 crab e.g. when a client says "i don't know what happened but YOUR APP has EATEN MY DATA!" and i can say "no, you exported it at 10:50:06, here's the relevant section of the log"
07:30 Vandal joined #mojo
07:38 Shaeto joined #mojo
07:39 Shaeto is easy way to trace/profile hypnotoad processes ? i still trying to understand why application in idle mode eats all cpu on fedora 16
07:40 Shaeto think to add trace messages to ioloop but mb exists more convinient tool
07:52 sugar joined #mojo
08:08 batman joined #mojo
08:18 foomator joined #mojo
08:25 sri hmm
08:27 sri i wonder if making tests a little more complex is ok if it allows us to use default values that perform better in the real world
08:28 Shaeto how to catch messages from _sandbox ?:)
08:48 berov joined #mojo
08:49 cosmincx joined #mojo
08:50 Shaeto well, per profiler all time eats EV::run
09:04 kjeldahl joined #mojo
09:06 sri Shaeto: you can't yet, i'll change that
09:07 noganex joined #mojo
09:08 Shaeto i have changed sandbox to print event name, "Timer xxxx" spawns crazy fast
09:09 GitHub197 joined #mojo
09:09 GitHub197 [mojo] kraih pushed 1 new commit to master: http://git.io/JKy8Jw
09:09 GitHub197 [mojo/master] added experimental error event to Mojo::IOWatcher and improved Mojo::IOLoop performance by increasing the default cleanup interval from 0 to 0.025 seconds - Sebastian Riedel
09:09 GitHub197 left #mojo
09:09 sri that should do the trick
09:09 Shaeto thank you, really was a showstopper for me :)
09:10 * sri loves adding new events
09:10 sri still a little scared about the interval change, since it creates some randomness around connection cleanups
09:11 sri which tends to mess up tests
09:12 Shaeto for tests where you check that daemon did cleanup all connections ?
09:12 sri kinda, yes
09:13 smpb joined #mojo
09:13 sri but those tests are updated for the new default now, i'm more scared about those that didn't break
09:14 sri random failures that may happen once every few thousand test runs
09:15 Shaeto confirm this patch fixed problem with cpu
09:15 sri that's not a problem
09:15 sri EV was just too fast
09:16 sri if you had said you were using EV i could have told you to change cleanup_interval
09:17 Shaeto yes i have changed cleanup_interval directly in IOLoop.pm
09:18 Shaeto there was a good side, i met nytprof
09:22 Foxcool joined #mojo
09:58 noganex_ joined #mojo
10:08 sugar hello
10:08 purl hello, sugar.
10:08 sugar https://gist.github.com/1391022
10:08 sugar bug or not?
10:09 sugar on 2.11 and 2.31 does not work
10:09 sugar on 1.81 worked fine
10:09 sugar or how to send utf8 via mojo_ua?
10:15 sri doesn't look like a bug
10:16 sri i'm sure there are tests
10:16 sugar i don't see any utf test
10:17 sri the whole test looks broken
10:17 sugar ok
10:18 sri don't even know where to start
10:18 sri (yours)
10:18 sugar ?
10:19 sugar hmm
10:19 sugar interface of tests are changed?
10:21 sri there is no way that test ever worked with 1.81, what you're saying is just plain wrong
10:22 sri maybe someone with more patience can talk you through it
10:23 sugar no i am wrong, test is working fine
10:23 sugar should be post_form not post
10:33 sugar https://gist.github.com/1391022
10:33 sugar i updated a gist
10:34 sugar now it looks like a bug
10:35 sugar we need to post some xml with utfs, and Mojo::UserAgent fails
10:36 sri no it doesn't you're feeding it chars when it expects bytes
10:36 sri this is all perfectly fine
10:36 sugar ok
10:37 sugar it worked in 1.81
10:37 sugar how now i need to pass utfs to Mojo::UserAgent
10:37 spleenjack joined #mojo
10:37 sugar &
10:37 sugar ?
10:38 sugar i about last example
10:39 noganex joined #mojo
10:40 sugar decode_utf8 and encode_utf8 dont work, both
10:45 random joined #mojo
10:51 sugar sri: but encode_utf8 generate bytes, and if you see in test 6, it does not work fine
10:57 mire joined #mojo
11:01 sugar i update gist https://gist.github.com/1391022
11:02 sugar how i should pass utf via Mojo::UserAgent?
11:02 sugar <sri> there is no way that test ever worked with 1.81, what you're saying is just plain wrong
11:02 sugar sorry
11:02 sugar <sri> no it doesn't you're feeding it chars when it expects bytes
11:03 sugar now passes only bytes
11:03 sugar but test fails
11:03 purl test fails are fixed in master
11:03 sugar what i shoul do? _utf8_on
11:04 sugar sri: what do you think?
11:04 purl I think sugar should try flossing more often!
11:11 sugar hey, anybody knows how to pass utf via Mojo::UserAgent?
11:11 sugar or it's bullshit is from 90th's?
11:11 crab shaeto: how are you profiling your mojo app with nytprof?
11:15 Shaeto sec
11:15 sugar hmm
11:16 sugar if i in render will add decode_utf8( text ), it will work fine
11:16 Shaeto perl -d:NYTProf ./script/MyService daemon --listen http://localhost:2600
11:17 sugar sri: mb add it to tests?
11:17 Shaeto actually not uner hypnotoad
11:17 sugar sri: https://gist.github.com/1391022 - passes
11:17 Shaeto then stopped process by kill
11:17 Shaeto and decoded result by nytprofhtml --open
11:18 sugar because before write tests and post it here i read tests from mojo's distro
11:22 crab Shaeto; cool, i thought you were doing it with hypnotoad.
11:22 crab SHaeto: still, you should add that to the wiki
11:23 Shaeto i think it is possible with hypnotoad there is DB::enable_profile(file) / disable_profile
11:34 random joined #mojo
11:37 random joined #mojo
11:43 spleenjack1 joined #mojo
11:48 aptituz joined #mojo
11:56 sugar http://showmetheco.de/articles/2010/10/how-t​o-avoid-unicode-pitfalls-in-mojolicious.html
11:56 sugar good article about working with utf in mojolicous
12:03 Quai wtf8
12:12 Psyche^ joined #mojo
12:18 Debolaz joined #mojo
12:21 Foxcool joined #mojo
12:26 abra joined #mojo
13:03 mire joined #mojo
13:23 MojoGuest47 joined #mojo
13:23 MojoGuest47 From: http://blog.kraih.com/perl-is-ready-for-html5 (31 hits)
13:59 MojoGuest782 joined #mojo
13:59 MojoGuest782 From: http://www.google.com/url?sa=t&amp;rct=j&am​p;q=&amp;esrc=s&amp;source=web&amp;cd=4&amp​;ved=0CD4QFjAD&amp;url=http%3A%2F%2Fdev.xan​tus.org%2F&amp;ei=m03OTs-1JcOBgAeyrfC9DQ&am​p;usg=AFQjCNGw6nZv0wO42uzUnHhWcD0vcKNULg (1 hits)
13:59 g0n hi
14:00 MojoGuest217 joined #mojo
14:03 sugar joined #mojo
14:42 Shaeto joined #mojo
14:49 koban left #mojo
14:54 elb0w_ joined #mojo
15:03 sugar new mojo exceptions are suxx
15:03 sugar =(
15:17 sugar how i can get normal exception (module and line) but not this crap http://pb.rbfh.de/32MDgtADoI7JG
15:17 sugar anybody knows?
15:18 arpadszasz joined #mojo
15:18 Shaeto i check modules by "perl XXXX.pm" before application (re) start :)
15:19 Shaeto i think it is because Loader and eval
15:20 batman sugar: do you have a 00-load.t test or something?
15:21 batman https://metacpan.org/module/Test::Compile#SYNOPSIS <-- maybe something like this can help you to check if all the modules can load
15:21 Shaeto thats fine for syntax etc check but what about runtime exceptions
15:22 batman ah. sorry. didn't read the paste carefully enough :(
15:24 sugar batman, Shaeto, thanks guys, those are good suggestions, i'll try it
15:24 batman nah. mine suck :P
15:24 batman sugar: is this in dev or prod?
15:24 sugar dev
15:24 batman so you don't see the full stacktrace in the browser?
15:25 sugar one of my co-workers fault dev and went home
15:25 sugar no
15:25 batman ok
15:25 batman you could try to do something evil like this: $ENV{'__DIE__'} = \&Carp::confess;
15:25 sugar white paper
15:25 purl white paper is, like, 127 lines of pod so far.  8sigh*
15:25 batman i sometimes toss that in, hoping to get more information (it works most of the time)
15:26 ispy_ joined #mojo
15:27 Shaeto left #mojo
15:27 sugar batman: trying it, i has [error] Unknown error
15:28 sugar not informational...
15:37 MojoGuest451 joined #mojo
15:37 MojoGuest451 From: http://www.google.com/url?sa=t&amp;rct=j&amp;q​=websocket%20demo&amp;source=web&amp;cd=9&amp;​sqi=2&amp;ved=0CG8QFjAI&amp;url=http%3A%2F%2Fd​ev.xantus.org%2F&amp;ei=WmLOTrHsG_CGsAKUkdWyDg​&amp;usg=AFQjCNGw6nZv0wO42uzUnHhWcD0vcKNULg (1 hits)
15:37 MojoGuest620 joined #mojo
15:37 MojoGuest620 From: http://www.google.com/url?sa=t&amp;rct=j&amp;q​=websocket%20demo&amp;source=web&amp;cd=9&amp;​sqi=2&amp;ved=0CG8QFjAI&amp;url=http%3A%2F%2Fd​ev.xantus.org%2F&amp;ei=WmLOTrHsG_CGsAKUkdWyDg​&amp;usg=AFQjCNGw6nZv0wO42uzUnHhWcD0vcKNULg (2 hits)
15:40 sugar_ joined #mojo
15:49 gshank joined #mojo
16:00 sugar joined #mojo
16:01 mire joined #mojo
16:16 sugar joined #mojo
16:20 elb0w_ joined #mojo
16:31 elb0w_ joined #mojo
16:32 mattastrophe joined #mojo
16:50 kaare joined #mojo
17:02 elb0w_ joined #mojo
17:05 smpb joined #mojo
17:19 MojoGuest939 joined #mojo
17:19 MojoGuest939 From: http://news.ycombinator.com/item?id=1277067 (142 hits)
17:20 MojoGuest939 joined #mojo
17:20 MojoGuest939 From: http://news.ycombinator.com/item?id=1277067 (143 hits)
17:21 MojoGuest939 left #mojo
17:25 cosimo joined #mojo
17:35 cosimo joined #mojo
17:36 mattastrophe joined #mojo
17:40 ereslibre joined #mojo
18:04 mire joined #mojo
18:06 grim_fandango joined #mojo
18:37 chorny joined #mojo
18:44 batman joined #mojo
20:10 perlite_ joined #mojo
20:28 MojoGuest481 joined #mojo
20:28 MojoGuest481 From: http://news.ycombinator.com/item?id=1277067 (144 hits)
20:36 arpadszasz joined #mojo
20:42 Freddy joined #mojo
20:42 Freddy Hey there.
20:43 Freddy I was reading your documentation, and I have a bit of a doubt about it
20:43 Freddy could someone help me?
20:44 Hinrik we won't know until you ask your question
20:44 Freddy oh. right.
20:44 Freddy I was reading your guide in Routing
20:44 Freddy (actually, translating it for the Google Code-In)
20:44 Freddy and it says
20:44 Freddy "Nameless routes get an automatically generated one assigned that is simply equal to the route itself without non-word characters."
20:45 Freddy by non-word, do you mean non-alphanumeric?
20:45 Freddy or just non-alphabetic?
21:12 mire joined #mojo
21:50 marty Freddy:  I think it's non-alphanumeric.  I Just tested it and  a generated route for /foo1/bar2  works as $self->url_for('foo1bar2');
21:51 Freddy Thanks!
21:53 Freddy left #mojo
21:59 MojoGuest406 joined #mojo
22:01 MojoGuest406 joined #mojo
22:04 Strengths gnivigsknaht yppah
22:07 Strengths left #mojo
22:56 jwang joined #mojo
23:23 diogo joined #mojo

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