Camelia, the Perl 6 bug

IRC log for #mojo, 2010-10-01

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

All times shown according to UTC.

Time Nick Message
00:09 ltriant left #mojo
00:09 ltriant joined #mojo
00:46 su-bzero_ joined #mojo
00:48 fod joined #mojo
00:50 su-bzero left #mojo
00:59 MojoGuest9 joined #mojo
00:59 MojoGuest9 From: http://news.ycombinator.com/item?id=1277067 (8 hits)
01:00 MojoGuest9 left #mojo
01:11 stephen left #mojo
01:38 alnewkirk joined #mojo
01:43 lafille joined #mojo
01:45 lafille left #mojo
02:09 GitHub139 joined #mojo
02:09 GitHub139 mojo: master Sebastian Riedel * 80c6def (2 files in 2 dirs): fixed typo - http://bit.ly/9seDIH
02:09 GitHub139 left #mojo
02:35 janus left #mojo
02:35 janus joined #mojo
02:41 perlrocks Twitter: "alnewkirk closed pull it alnewkirk request 1 / Mojolicious-Command-Package: Fixed ... # # github http://bit.ly/9DSJ1t alnewkirk" (fr) --alnewkirkassoc http://twitter.com/alnewki​rkassoc/status/26041458017
02:52 GitHub126 joined #mojo
02:52 GitHub126 mojo: master Sebastian Riedel * 0c33eeb (2 files in 2 dirs): added more pseudo classes to Mojo::DOM - http://bit.ly/ctT44c
02:52 GitHub126 left #mojo
03:35 perlrocks Twitter: "RT @vtivti Just donated $100 to the awesome #Mojolicious project. Sebastian++! Let's support the best #Perl web framework!" --xomaa http://twitter.com/xomaa/status/26046237724
03:54 und3f[A] left #mojo
03:54 doubi left #mojo
04:04 tholen42 joined #mojo
04:05 tl left #mojo
04:37 tempire I think I'm going to start using Mouse a bit
04:45 alnewkirk package Rat; use base 'Mouse'; sub add_gangsta { return shift->transform; } 1;
05:01 marcus_ D     BGVG    SSSSSSSSSSSH
05:01 marcus_ ]\]=[=
05:32 marcus_ (that was Eva saying good morning)
05:33 koban joined #mojo
05:33 marcus_ sri: How would you unit test a web socket?
05:35 marcus_ sri: another thing, I was thinking that Test::Mojo could be made to test any other web component through as well through Mojo::Client
05:44 ash_ left #mojo
06:02 kvorg left #mojo
06:03 koban left #mojo
06:04 koban joined #mojo
06:05 koban left #mojo
06:05 koban joined #mojo
06:12 fod left #mojo
06:18 arpadszasz joined #mojo
06:24 kvorg joined #mojo
06:49 Rhaen hola!
06:56 marcus_ HOHA
06:56 Rhaen :)
06:56 * marcus_ is listening to the music from lion king and finish Mojo::Guides::Testing
06:57 marcus_ o/' And can you feel the love tonight / It is where we are o/'
06:57 marcus_ !tweet o/' And can you feel the love tonight / It is where we are o/'
06:57 perlrocks marcus_: http://twitter.com/perlrocks/status/26057542407
06:58 Rhaen marcus_: that's awesome! Another guide. That's really nice!
06:59 Rhaen guys! This has nothing to do with mojo - but one thing which is horrible is non UTF-8 alike systems!
06:59 marcus_ Rhaen: http://gist.github.com/605856 is what I have so far. Trying to finish the web socket section atm
07:00 marcus_ Rhaen: encoding is pain
07:00 marcus_ Rhaen: but it's no better in most other languages.
07:00 koban left #mojo
07:00 Rhaen huuuh! Cool! Great job!
07:00 marcus_ ruby is much worse, for one
07:00 Rhaen marcus_++
07:00 Rhaen right, I'll get iso-8859-1
07:01 Rhaen awesome so now let's convert it to UTF-8
07:02 Rhaen hm, is utf8::upgrade capable for such tasks?
07:03 marcus_ I think you need utf8::decode
07:03 marcus_ I could be wrong tho
07:03 marcus_ but I think that'll turn your iso-8859-1 content into internal perl encoding.
07:04 Rhaen k, I'll check. Never had a problem like this before :)
07:04 Rhaen so - rtfm and google :)
07:04 marcus_ there was a really good presentation about encoding on slideshare recently
07:04 marcus_ http://www.slideshare.net/Penfold/perl-and-unicode there we go
07:05 Rhaen oh, cool - thanks!
07:16 Rhaen marcus_: the slideshow is cool :)
07:16 ysyrota joined #mojo
07:17 Rhaen if the speaker was good, too - that must have been a great presentation
07:18 marcus_ Rhaen: I think he's pretty good, he works for Yahoo iirc. Unfortunately it's not on http://www.presentingperl.org/ye2010/ yet
07:21 koban joined #mojo
07:22 Foxcool joined #mojo
07:22 koban left #mojo
07:22 koban joined #mojo
07:28 ltriant left #mojo
07:41 koban left #mojo
07:41 koban joined #mojo
07:54 marcus_ heh
07:56 Kulag left #mojo
07:56 Kulag joined #mojo
08:03 kvorg hi all
08:03 kvorg another day, another confusing question:
08:04 kvorg say i want to sanitize the parameters i get from my form, making sure they make sense before using them
08:04 kvorg and i also want to display the form on the result page, of course taking advantage of the tag helpers
08:05 koban left #mojo
08:05 koban joined #mojo
08:05 kvorg so while params are by default non-mutable, the param helper actually cecks the mojo.params from the stash first:
08:06 kvorg so should i just sanitize mojo.params and everything will just work?
08:06 kvorg or will parameters that are set in params but not in mojo.params still get trough?
08:06 koban left #mojo
08:07 koban joined #mojo
08:07 kvorg i am asking to understand how things should work, but also testing to see how they actually do. i believe this is a case common enough it might be worth me writing an addition for the guides or smth.
08:08 koban left #mojo
08:08 kvorg sri:?
08:08 kvorg vti:?
08:08 kvorg i am all alone ...
08:08 koban joined #mojo
08:09 Kulag left #mojo
08:09 Kulag joined #mojo
08:18 Kulag left #mojo
08:21 Kulag joined #mojo
08:42 Kulag left #mojo
08:43 esskar__ joined #mojo
08:43 Kulag joined #mojo
08:44 Rhaen no.
08:44 Rhaen we'll watch you and query behind you :D
08:44 Rhaen no, we don't!
08:44 Rhaen I am alone here...
08:47 esskar left #mojo
08:47 esskar__ is now known as esskar
08:48 Drossel joined #mojo
08:50 Kulag left #mojo
08:54 kvorg Rhaen: we are all alone here ...
08:54 Kulag joined #mojo
08:54 kvorg what happened to your newbie docs?
08:54 kvorg working on an update to this: http://github.com/kvorg/mojo-handbook
08:54 kvorg would you like to contribute?
08:55 kvorg i have had several promises, but no actual contributions beyond typo fixes so far
08:55 Drossel left #mojo
08:55 kvorg and it is a bit outdated, but i am moving to update it again
08:55 kvorg lots of neat info from this channel
09:04 Kulag left #mojo
09:06 arpadszasz left #mojo
09:11 vti kvorg: !
09:11 kvorg vti: aaah, have i done it again?
09:12 vti kvorg: you've highlighted me :)
09:12 * kvorg notices Rhaen is all alone in his channel.
09:12 kvorg i was spamming the silent channel with some questions about parameters ... and hoped you would be a fountain of wisdom on the matter, what with your validator and all
09:13 Kulag joined #mojo
09:15 vti hm
09:16 kvorg mh
09:20 Kulag left #mojo
09:22 spleenjack joined #mojo
09:26 vti i don't understand :)
09:28 marcus_ vti: did you ever test a web socket?
09:28 marcus_ been trying to look at you examples, but none of them have tests afaict.
09:30 kvorg neither do i
09:30 kvorg i will try to ask properly: imagine I want to sanitize the GET parameters, so that i present the templates and tag helpers with the sanitized version
09:31 kvorg i.e., i want to set some defaults if they were not passed, and want the tag helpers to just fill them in
09:31 kvorg what is the recommended way to do that?
09:34 Kulag joined #mojo
09:39 marcus_ kvorg: this is one of the eternal questions of perl web development.
09:39 marcus_ 'Which form processor should I use?'
09:40 marcus_ I know vti have written some mojo specific ones in his Mojo-common package.
09:40 marcus_ mojox-common even
09:41 Kulag left #mojo
09:43 kvorg marcus_: i am not expectin to find a processor done for me (but his validator is nice) ... I want to sneak in different values from what the request actually received to sanitize the application
09:43 vti marcus_: let me think
09:43 kvorg i could, of course, copy the parameters in a hash and manipulate them, but then i loose all the tag helper goodness
09:43 vti there is mojox::validator on cpan
09:44 kvorg and creating a new parameter object and replacing the one in request seems a bit brutal, no?
09:44 vti marcus_: http://github.com/vti/mojolicious-plugi​n-websocket-chat/blob/master/t/plugin.t
09:46 kvorg or should i do just exactly that and just feel dirty afterwards?
10:09 marcus_ vti: sweet
10:21 kvorg no answers no mojo no sri
10:21 marcus_ sri was probably up late
10:21 marcus_ at least he answered me around 12.30 last night
10:21 * marcus_ slept 5.5 hours :(
10:29 stereonaut joined #mojo
10:30 baton8 left #mojo
10:30 baton8 joined #mojo
10:36 fhelmberger joined #mojo
10:44 baton8 left #mojo
10:50 baton8 joined #mojo
10:52 kvorg hmmm. the only way i find to overrule the actual parameters is by replacing the $self->req->{_params}, which is scary messing in internal representation.
10:53 kvorg two weeks ago there was some discussion that the param helper touches the stash in some way ... but i can't seem to find that in the code at all.
10:58 arpadszasz joined #mojo
11:00 perlrocks Twitter: "Mojolicious, 'framework' minimalist web http://bit.ly/cFvImt http://bit.ly/cS2n4f # softwarelibre Perl" (es) --besmarthinkfree http://twitter.com/besmarth​inkfree/status/26068970611
11:04 kvorg it really buggs me when people talk about Mojlicious as a minimalistic web framework
11:05 kvorg it is minimalistic in many senses, such as efficient, uncluttered, no dependencies
11:05 kvorg but it is Lite that is a microframework, Mojolicious is a real framework done right
11:05 kvorg we have to spin that
11:15 * kvorg is spinning in place.
11:17 * kvorg is idling.
11:23 und3f joined #mojo
11:33 yko joined #mojo
11:39 MojoGuest617 joined #mojo
11:42 MojoGuest617 left #mojo
11:42 MojoGuest459 joined #mojo
11:42 MojoGuest459 left #mojo
11:42 MojoGuest798 joined #mojo
11:51 arpadszasz left #mojo
11:53 esskar__ joined #mojo
11:57 MojoGuest798 left #mojo
11:57 esskar left #mojo
11:57 esskar__ is now known as esskar
12:10 sri moin
12:10 sri kvorg: don't ever touch mojo.* stash values
12:11 sri if you want to sanitize params just clone your own and put them in the stash
12:12 sri $self->req->params->clone or so
12:13 sri getting called a minimalistic framework is ok, over time people will just realize that there's a lot more, we just need more blog posts showing what can be done
12:14 sri marcus: maybe look at the websocket tests in t? :)
12:14 marcus_ sri: I did. vti's example is easier tho
12:15 sri ah ok
12:15 und3f good day
12:15 sri yea, the embedded https proxy is not a good example to get started :D
12:16 und3f Could someone help me with uploading files - i got problem when run my application with apache (with FastCgiServer) - file just cuts
12:16 und3f and it is cutted on different values - from 300 bytes up to 3kb
12:16 sri latest github version?
12:17 und3f sri, yes
12:17 sri weird
12:17 und3f oh, wait
12:17 und3f i didn't installed it
12:17 sri ah
12:17 und3f just pulled :P
12:17 sri then it's likely fixed
12:17 und3f i'll try
12:18 sri actually, it should also be fixed in the cpan version
12:19 und3f sri, no, it doesn't work still
12:20 und3f sri, with latest mojo
12:20 sri wait FastCGIServer is external server isn't it?
12:21 und3f sri, it is part of apache :|
12:21 und3f sri, just directive in apache vhost
12:21 sri ah, it's actually what i use in the unit test
12:21 sri never seen or heard of such a bug
12:22 sri no idea
12:22 und3f sri, what is test name?
12:22 und3f i can try to run it here
12:22 sri it is os x specific
12:22 und3f will it run on linux?
12:23 sri 14:22] sri: it is os x specific
12:25 sri we've has so many success stories for the fastcgi binding, and it didn't change in months
12:25 sri i somehow doubt that there's a bug
12:26 sri could be a broken perl or apache too
12:27 cognominal left #mojo
12:29 sri und3f: are you 100% sure there is no IO::Poll in your installed version of Mojo::Server::FastCGI?
12:30 sri that would be the broken older version
12:30 und3f i'll check now
12:31 und3f sri, there is no io::poll
12:32 sri then i'm out of ideas
12:33 und3f i can check apache with simple upload script
12:36 Foxcool left #mojo
12:40 und3f sri, cgi script uploads files right
12:56 marcus_ sri: best way to have a custom json config for your test suite with Plugin::JsonConfig?
12:56 sri ?
12:57 marcus_ lets say I have a database config in my app, and I want to set it to a test-db for the unit tests.
12:57 marcus_ am thinking of setting ext to '_test.json'
12:59 sri you could also use a "test" mode
12:59 marcus_ does the mode affect the config file?
12:59 sri not automatically
12:59 sri but you could make a decision based on it
13:00 GitHub146 joined #mojo
13:00 GitHub146 mojo: master Sebastian Riedel * 3c75519 (3 files in 2 dirs): added fastcgi huge message test - http://bit.ly/aaZ1vr
13:00 GitHub146 left #mojo
13:01 sri actually, we could habe it check myapp_$runmode.json by default too
13:01 marcus_ that would be nice.
13:02 sri patches welcome :)
13:02 vti myapp.$runmode.json :)
13:02 Kulag joined #mojo
13:03 sri und3f: as far as i can see it works flawless
13:03 sri und3f: i'm afraid there is nothing more i could do for you
13:05 und3f thanks, i'll keep looking for a problem
13:12 fhelmber_ joined #mojo
13:15 cognominal joined #mojo
13:21 ash__ joined #mojo
13:36 yko left #mojo
13:42 perlrocks Twitter: "Retrying to implement MojoX::CouchDB: http://github.com/esskar/mojox-couchdb wish me luck and stay tuned. #mojolicious #couchdb" --esskar http://twitter.com/esskar/status/26080486844
14:02 esskar__ joined #mojo
14:05 esskar left #mojo
14:05 esskar__ is now known as esskar
14:34 und3f Got problem with runnin as psgi: MOJO_APP=Catalogue MOJO_HOME=/var/www/catalogue/ plackup -Ilib ./script/catalogue
14:34 und3f Fri Oct  1 17:33:26 2010 debug Mojolicious::Plugin::JsonConfig:53 [18915]: Config file "/var/www/catalogue/plackup.json" missing, using default config.
14:34 und3f it ignores Mojo_app :|
14:34 und3f during config loading
14:36 sri likely plackup related
14:36 sri it might change $0 and stuff, who knows
14:36 und3f sri, so it is not possible to run with plack?
14:36 sri ?
14:36 sri of course it is
14:36 sri we are psgi compliant
14:37 und3f but not plack compliant? :)
14:37 sri there is no plack spec
14:37 sri every server does other wonky stuff
14:37 und3f yeah
14:40 sri member:und3f: i guess you could try making t/mojo/apache_fastcgi.t work on linux and create a real test case i can replicate
14:40 * sri slaps colloquy
14:42 sri you might only have to change a few paths
14:42 und3f sri, very possible, but now i need to make things work
14:42 sri well, if there's a bug and you can replicate it i can fix it in a few mins ;p
14:42 und3f Ok
14:42 alnewkirk the site will go live this evening, still have two articles to write, check out this article though, ..., http://ana.im/press/2010/09/mojoli​cious-the-web-in-a-box-with-a-bow/
14:42 und3f i'll try to fix a test
14:43 sri ok, there is still the risk that it's a mod_fastcgi bug or so
14:48 sri alnewkirk: did i say "designed by people that don’t love web development"? a better word might be "developed", since it's a pun at the current state of catalyst :)
14:49 alnewkirk okay chaing now
14:49 alnewkirk changing*
14:49 sri the whole sentence sounds a bit weird
14:50 alnewkirk what about
14:50 alnewkirk many web frameworks are developed by people that don�t have a passion for web development
14:50 sri the ending "while Mojolicious is very close to an awesome 1.0"
14:50 sri it seems unrelated
14:51 und3f sri, actually different linuxes got different pathes for libraries :|
14:51 alnewkirk sri, has been refactored
14:51 sri und3f: i don't want a new test file, just a test case i can add to the current os x version ;)
14:52 und3f sri, but gentoo and debian has same
14:52 und3f sri, i understand you
14:52 tempire_ joined #mojo
14:52 tempire_ ack!  A 13k line cgi script!
14:54 sri alnewkirk: the part still sounds a little weird to me, but otherwise i like it
14:54 alnewkirk okay let me take another look
15:00 sri "hopefully the second ajax revolution" refers to realtime web, but in your version it seems to belong to mojolicious
15:01 sri i think the next evolutionary step for the web will be persistent client/server connections, while ajax is based on a pull model
15:02 sri maybe i didn't elaborate enough on that
15:02 und3f sri, i make apache_fastcgi work, but it fails on tests 3 and  4 (mojo returns 500 )
15:02 sri und3f: interesting
15:02 und3f sri, post modifications?
15:03 sri und3f: the test passes flawless here
15:04 alnewkirk cool, makesd sense
15:04 sri 3 and 4? does that mean one of the two request tests passes?
15:05 und3f sri, and now test passed :|
15:05 und3f dunno why
15:05 und3f :|
15:05 sri ah, maybe increase the timeout
15:05 sri your test box might be slower
15:05 sri no wait
15:06 und3f sri, checkout pull request
15:06 sri add a sleep 2 after the sleep 1 loop
15:06 koban left #mojo
15:06 sri your apache might need a little more startup time
15:07 und3f sri, now it works ok, without any problems, dunno why it didn't before
15:07 und3f sri, but i received 500 from Mojo
15:07 und3f sri, in tests 3 and 4
15:07 und3f *status code 500
15:07 alnewkirk Mojolicious is very close to an awesome 1.0 release, as real-time web communication (persistent client/server connections) are currently taking shape, "hopefully the second ajax revolution",  Mojolicious is in a very good position to supply the latest demands.
15:07 und3f sri, i received it again :)
15:07 alnewkirk stills feels a bit fragmented, what do you think
15:08 sri alnewkirk: much better :)
15:08 und3f Seems i got it 1 time to 5 test
15:08 sri alnewkirk: yea maybe
15:08 und3f sri, http://pastebin.ca/1952639
15:09 sri und3f: harmless, it's just apache needing a few more microseconds to start
15:10 sri a sleep 2 before the first request test will make it work consistently
15:12 sri those live tests are not very reliable in general
15:12 sri too many random factors
15:12 und3f sri, so, now add uploading to this test?
15:12 sri ye
15:13 sri the second test should have actually been enough
15:13 sri it is kinda huge
15:14 sri sends a big urlencoded body and triggers an even bigger chunked response
15:14 und3f sri, second test? Where? In test_apache just 1 request to server and it is small
15:14 sri huh
15:15 sri latest github
15:15 sri i added one earlier
15:15 und3f see now
15:16 und3f sri, tests are passed but i got 100% cpu loading during tests
15:17 sri it's a stress test, not unexpected
15:18 sri now add a test that fails :)
15:21 und3f :|
15:21 und3f :(
15:21 und3f ok 0, 'that fails :P'
15:22 sri the test is parsing 100000 urlencoded parameters, and turns each one into a response chunk, it's actually a really neat stress test
15:27 und3f sri, but data transfered to apache not checked
15:27 su-bzero_ left #mojo
15:28 sri ?
15:28 su-bzero joined #mojo
15:28 und3f oh, it is checked
15:47 kvorg left #mojo
15:52 ysyrota left #mojo
15:53 sri und3f: that test passing is a very bad sign for you though, it pretty much rules out the mojo io layer
15:54 und3f sri, i am trying to simulate situation
15:54 esskar__ joined #mojo
15:58 esskar left #mojo
15:58 esskar__ is now known as esskar
16:11 sri http://blog.fupps.com/2010/04/26/resty/ # oh this is fun, someone should build a mojo command like that :)
16:14 sri actually put/post/delete commands might be even better
16:14 sri ./myapp get /foo is close enough
16:15 sri ./myapp put /inbox < foo/bar/test.txt could be fun
16:15 kvorg joined #mojo
16:25 und3f sri, i just made clean Mojolicious::lite app and connected to Apache
16:25 und3f uploading doesn't work ok
16:26 alnewkirk gotta go, i'll see y'all later
16:26 sri laters
16:26 alnewkirk left #mojo
16:26 sri und3f: test case?
16:27 und3f sri, possibly i should include all configuration files of apache to test? :P
16:27 sri i don't care about your apache configuration
16:27 sri if a minimal test case works it's not my problem ;p
16:27 und3f sri, it is clean configuration
16:27 und3f sri, just installed
16:28 sri please just add it to the test
16:28 sri i really don't want to spend an hour setting up some weird apache configuration just to find out it is configured wrong
16:30 sri you can even turn the generated fcgi script into a lite app if you want
16:36 sri rewrite the whole test if you have to, but thats the format i need
16:43 perlrocks Twitter: "Looks like we'll soon be able to run our #mojolicious apps inside @couchdb, wow! http://bit.ly/ajiTvT #perl" --kraih http://twitter.com/kraih/status/26096753547
16:47 sri that could be interesting
16:48 und3f sri, seem like i made it
16:49 perlrocks Twitter: "RT @kraih: Looks like we'll soon be able to run our #mojolicious apps inside @couchdb, wow! http://bit.ly/ajiTvT #perl" --CouchDB http://twitter.com/CouchDB/status/26097218553
16:50 und3f sri, btw, does $client->post( $url, file => {content => $content } ) uploads a file?
16:50 sri that request looks wrong
16:50 sri where in the pod did you see that syntax?
16:51 sri there is no way that produces a sensible request
16:51 und3f my $tx= $client->post_form(
16:51 und3f 'http://kraih.com/foo',
16:51 und3f {file => {content => 'lalala'}}
16:51 und3f );
16:51 sri ...
16:52 und3f oh right
16:52 und3f i missed just 2 brackets
16:52 sri and the method
16:52 und3f sri, let me rephrase question
16:53 sri if you want to know what i actually produces try print $client->build_form_tx(...)->req;
16:53 sri takes the same arguments
16:53 und3f thanks
16:57 perlrocks Twitter: "RT @kraih: Looks like we'll soon be able to run our #mojolicious apps inside @couchdb, wow! http://bit.ly/ajiTvT #perl" --florinmatinca http://twitter.com/florinm​atinca/status/26097857901
17:11 sri und3f: do you have a test case? because i'll be afk for a bit soon
17:11 und3f sri, i got something like
17:11 und3f but i have a small problem:
17:12 und3f i made standalone apache configuration that  returns not all data of sended file
17:12 und3f (test it with browser)
17:12 und3f But for Mojo::Client request i got 500
17:12 und3f sri, can you check it?
17:13 sri depends
17:13 sri nopaste
17:13 und3f sri, can post everything in my repo
17:14 und3f sri http://github.com/und3f/mojo/commit/52​c3a08b50f49779226fafacb409bf88f325290f
17:15 sri why <Directory "<%= $dir %>/">?
17:16 und3f sri, in my config i also got there Order, All from all directives
17:17 sri but why is it important for the test?
17:17 und3f sri, but access modules seems not to be included
17:17 und3f sri, i tried to make my configuration in test
17:17 sri so you just put it there for fun?
17:17 und3f em..
17:18 sri does it change the behavior?
17:18 und3f I'll check now
17:18 sri like i said before, i don't want to debug your apache configuration
17:18 sri my interest is finding and fixing mojolicious bugs
17:20 und3f sri, right, <Directory is just for fun  it is
17:20 und3f sri, it doesn't change anything
17:21 sri thats very important information
17:21 und3f sri, also i am having fun with mojo upload arond 4-5 hours
17:21 und3f :P
17:21 und3f *around
17:21 und3f sri, your previous apache configuration works same
17:23 sri ok, now i can take over and debug
17:24 und3f sri, so there is a bug, right?
17:24 sri unknown
17:24 sri first step is to make the test case sane and compare daemon behavior
17:24 und3f sri, but you received cut file from apache?
17:25 sri not yet
17:25 und3f sri, try to star apache manually and send file to MOjo::HeloWorld
17:26 kvorg left #mojo
17:41 sri ok, confirmed
17:41 sri cgi and daemon are not affected
17:42 sri and size doesn't seem to matter
17:42 sri 1 ..10 has the same effect as 1 .. 100000
17:44 tempire_ left #mojo
17:44 kvorg joined #mojo
17:52 und3f sri, yay :)
17:52 und3f Now i can take my harmonica and play some mojo blue
17:52 und3f *blues
17:55 spleenjack left #mojo
17:59 sri http://pastie.org/1193905 # ok, now i have diagnostics, seems the content arrives, time to dig deeper
18:00 sri (in case someone wants to join the fun)
18:04 sri really nothing spectacular as far as i can see
18:09 und3f sri, i've had so much fun today :P
18:09 und3f too much fun for a day
18:10 stereonaut left #mojo
18:12 perlrocks Twitter: "RT @ kraih: Looks like we'll Soon Be Able to run apps inside Our # mojolicious @ CouchDB, wow! perl http://bit.ly/ajiTvT" (es) --elsanto_wa http://twitter.com/elsanto_wa/status/26103613143
18:13 und3f i also have to debug one more my mojo app
18:14 sri i have a feeling this bug will be very hard to track down but easy to fix
18:40 sri hmm
18:40 sri looks like i can even replicate it in a message.t test
18:40 sri thats a very good sign :)
18:41 und3f :)
18:56 spleenjack joined #mojo
19:22 ilbot2 joined #mojo
19:22 Topic for #mojo is now ♥ Join the revolution! ♥ | http://mojolicio.us | http://irclog.perlgeek.de/mojo/today
19:28 fhelmber_ left #mojo
19:38 GitHub142 joined #mojo
19:38 GitHub142 mojo: master Sebastian Riedel * 08a8370 (7 files in 4 dirs): fixed a multipart parser bug affecting mostly file uploads - http://bit.ly/aqODMY
19:38 GitHub142 left #mojo
19:38 sri und3f: fixed
19:39 und3f sri, checking
19:39 sri it was a side effect of the refactored state handling that was resetting the multipart parser at times
19:40 sri explains your described problems perfectly
19:40 esskar_ has anyone retrieved {"error":"bad_request","reason":"invalid UTF-8 JSON"} as an error from couchdb
19:40 esskar_ i post
19:40 esskar_ {"foo":"bar"}
19:40 esskar_ to it
19:40 esskar_ so nothing special here
19:41 esskar_ content-length is 13
19:41 esskar_ in the couchdb log i see
19:42 esskar_ [debug] [<0.2912.0>] Invalid JSON: <<"{\"foo\":\"bar\"}">>
19:42 esskar_ not sure if it always prints it like that (with the \)
19:42 Rhaen_ joined #mojo
19:44 und3f sri, how to check that upload limit were exceeded?
19:45 sri req->error
19:47 Rhaen left #mojo
19:47 Rhaen_ is now known as Rhaen
19:53 MojoGuest165 joined #mojo
19:54 MojoGuest165 left #mojo
20:00 esskar_ left #mojo
20:28 und3f sri, uploading work just find, thanks :)
20:28 esskar_ joined #mojo
20:28 sri \o/
20:35 esskar_ i think its my couchdb
20:35 esskar_ futon returns the same error
20:48 esskar_ left #mojo
21:00 GitHub83 joined #mojo
21:00 GitHub83 mojo: master Sebastian Riedel * 30032c7 (4 files in 3 dirs): added some experimental methods to Mojo::Message - http://bit.ly/bPVYIV
21:00 GitHub83 left #mojo
21:05 perlite left #mojo
21:08 perlite joined #mojo
21:21 und3f sri, that will help much :)
21:21 sri still unsure about the method name
21:22 sri but can't think of anything better than is_limit_exceeded
21:23 und3f for me it is ok
21:24 und3f but that means that there is only one limit for Message
21:24 sri you can always check req->error
21:25 und3f sri, it is not clear for developer how to check if limit exceed
21:25 sri ?
21:25 und3f sri, before you added is_limit_exceeded
21:26 sri thats why i added it
21:26 und3f sri, also i found how to increase limit just in tests
21:26 sri doc patches are welcome
21:26 und3f sri, so your latest commit will help much to developers
21:27 und3f sri, i have a lot of my mojo app to debug :)
21:27 sri you are really hard to understand :S
21:27 und3f I know :)
21:28 und3f need to improve my english, but i am so lazy
21:36 und3f sri, seems like <%= $a %> doesn't produce warning message if $a is undef
21:36 und3f sri, can remove all <%= $a || '' %> :P
22:19 vti traveling to berlin tomorrow
22:26 kvorg back to munging parameters if we want tag helpers to just work: would the recommended way then be to redirect to a new request with corrected parameters?
22:27 kvorg that should be highly kosher
22:28 kvorg (munging = i was trying to discuss parameter sanitization with an empty channel in the morning)
22:31 kvorg vti: i missed giving a talk in berlin this tuesday. we so missed each other.
22:38 kvorg left #mojo
22:40 perlrocks left #mojo
22:49 kvorg joined #mojo
23:00 perlrocks joined #mojo
23:02 sri kvorg: that question doesn't make any sense
23:31 und3f is now known as und3f[A]
23:45 * kvorg reformats and turns off.
23:46 * kvorg after uploading http://github.com/kvorg/mojolic​ious-plugin-tag_helpers_extra/
23:54 Alias left #mojo

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