Camelia, the Perl 6 bug

IRC log for #mojo, 2010-10-19

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

All times shown according to UTC.

Time Nick Message
01:19 kimoto left #mojo
02:05 vel left #mojo
02:35 janus left #mojo
02:35 janus joined #mojo
04:03 tholen42 joined #mojo
04:05 tholen left #mojo
05:16 Kulag left #mojo
05:18 ysyrota joined #mojo
05:19 Kulag joined #mojo
05:26 Kulag left #mojo
05:29 Kulag joined #mojo
05:36 tempire_ joined #mojo
05:36 tempire_ left #mojo
05:37 koban joined #mojo
05:37 Kulag left #mojo
05:42 Kulag joined #mojo
05:43 alnewkirk|com joined #mojo
05:44 Drossel joined #mojo
05:44 marcus goood morning
05:47 alnewkirk left #mojo
05:47 alnewkirk|com is now known as alnewkirk
05:47 tempire_ joined #mojo
05:48 koban hi!
05:48 kvorg left #mojo
05:48 * tempire_ wants to mojo deploy ec2
05:48 Kulag left #mojo
05:51 tempire_ left #mojo
05:56 kvorg joined #mojo
05:59 ltriant left #mojo
06:50 stephan48 http://abc123.pastebin.com/vTKZqEDq <- any clue on this?
06:58 und3f[A] is now known as und3f
06:58 kvorg git submodule init ; git submodule update
06:58 kvorg or just
06:58 kvorg git submodule update --init
06:59 kvorg wait, it dies at the init already? hmm, can you do a git cone of Mojolicious from the same comp at this time?
06:59 stephan48 kvorg: i dont it...
06:59 stephan48 what?
07:01 stephan48 http://abc123.pastebin.com/mAzb119x - this made it  to work
07:02 kvorg interesting. i had a git url and it worked with no problem
07:03 stephan48 it was confused about the double mojo part
07:04 kvorg aaah, only now do i understand. i falsely assumed you were just setting up bootylicious ... sorry.
07:04 * kvorg stops pretending to actually understand anything at this time of day (or any, for that matter).
07:07 stephan48 :)
07:13 marcus sri: Please please please release
07:13 omega sri: please please release
07:13 marcus peace, release...
07:13 omega a release pretty please?
07:14 marcus HOT BEVERAGE
07:15 rhaen moin!
07:15 rhaen sri: please please please
07:15 rhaen whazzup?
07:16 Drossel left #mojo
07:16 rhaen can I haz a release?
07:16 Kulag joined #mojo
07:49 alnewkirk left #mojo
07:49 alnewkirk|com joined #mojo
07:49 alnewkirk|com is now known as alnewkirk
07:49 vti stephan48: sorry about that :)
08:02 und3f is now known as und3f[A]
08:23 dotan joined #mojo
08:26 kvorg left #mojo
08:26 kvorg joined #mojo
08:38 dotan left #mojo
08:43 omega Json::Config + plackup is not working very well?
08:43 omega at least I have to specify MOJO_HOME
08:43 omega and the config-file name
09:03 dotan joined #mojo
09:04 stephan48 vti: no problem :)
09:04 dotan rhaen: (or anyone)  - are there links to presentations in the github wiki? If not I'm adding some.
09:10 rhaen dotan: please go for it!
09:10 rhaen that's an awesome point - on slideshare there are a lot!
09:11 rhaen Maybe that's worth a topic on the homepage of the wiki
09:11 rhaen cool!
09:11 dotan rhaen: doing it :) - is there a reason the front page of the wiki is in textile? not that I'm complaining, just wondering.
09:18 koban left #mojo
09:18 koban joined #mojo
09:22 marcus I updated the testing wiki page a bit.
09:32 koban how can i use flash messages?
09:34 koban $c->flash(msg => 'Hi!') in controller and <%= flahs('msg') %> in template seems don't work
09:36 viking-3 maybe <%= flash('msg') %=> will do? ;)
09:40 koban what does '%=>' mean?
09:42 viking-3 sorry, i meant flash instead of flahs
09:42 viking-3 that = got there by accident
09:46 spleenjack joined #mojo
09:57 koban ah, it's my typo
10:12 koban <%= flash('msg') %> also does not work
10:32 y0j left #mojo
10:34 sri moin
10:35 sri marcus: just use github master ;p
10:35 sri cpan release won't make it any better
10:47 y0j joined #mojo
11:01 koban it works!! :)
11:02 rhaen dotan: uhm, yes and no.
11:03 rhaen dotan: the trick is easy, it's more difficult to wiki link the pods :)
11:03 rhaen dotan: one day we'll assemble an offline version of the wiki - but this requires more documentation, this will be shipped entirely in pod
11:09 jawnsy left #mojo
11:16 jawnsy joined #mojo
11:20 rhaen well, here is something strange
11:20 rhaen I can perform tests against webpages with Test::Mojo.
11:21 rhaen is_status(200) is perfectly working, however as I want to test existing webapplication which are not implemented in Mojo
11:21 rhaen I have different status codes like 200 and 302
11:22 rhaen is_status(200 || 302) is not working, any other suggestions on that?
11:24 sri testing for variable results is always bad
11:24 rhaen yep :) It is. But I want Mojo even for dirty things!
11:25 rhaen but hej, sri?
11:25 sri do it on your own with ->tx
11:25 marcus sri: most people that come in here with trouble has been trying to use the cpan version.
11:25 marcus sri: and my client's build system is built around cpan.
11:25 marcus why not keep it updated?
11:26 marcus might as well remove it if you're not going to maintain it on cpan.
11:26 rhaen sri: even after a few weeks now - Mojo is still cool. Really, it's starting to start a small revolution in my company here
11:26 sri marcus: was thinking about that yesterday :)
11:26 rhaen marcus: agreed. Being a maintainer for NetBSD I know the problems.
11:27 rhaen marcus: we build our packages from CPAN - and we don't use github versions
11:27 rhaen I will commit a mojolicious-devel version soon which points to the github repo
11:28 marcus sri: it's also an issue for other people who release modules on cpan and want to depend on your modules.
11:28 rhaen sri: so - come on! Push a new version to cpan and increase to .**30
11:28 sri what should i do? make a major release without hypnotoad?
11:29 rhaen no - a minor minor minor - with .30 :)
11:29 rhaen we had soooo many changes and fixes
11:29 marcus sri: yeah, push .30
11:29 rhaen more docs, etc!
11:29 sri daemon prefork will be deprecated without a replacement
11:29 rhaen there are alot of modules which reside on CPAN which don't have such a thing like Changes
11:30 marcus sri: it's deprecated in trunk?
11:30 marcus as in don't work?
11:30 marcus I'm using prefork on a couple of web services :/
11:30 rhaen then just remove it :)
11:30 sri it might work, but untested and with less features
11:31 sri it will go away
11:31 sri if you care so much you can adopt it
11:31 rhaen hm, no problem with it :) just remove it :)
11:32 sri marcus: maybe now you understand why i'm waiting
11:33 sri it's not for fun
11:39 marcus sri: I think you just removed the pid file support from it, so I wouldn't be able to use it anyways.
11:39 marcus sri: Finish Hypnotoad! :p
11:40 ilbot2 joined #mojo
11:40 Topic for #mojo is now ♥ Join the revolution! ♥ | http://mojolicio.us | http://irclog.perlgeek.de/mojo/today
11:42 sri marcus: the main problem of the prefork daemon is that it was designed for http
11:42 marcus sri: I like http.
11:42 sri it works reasonably well for that, but fails completely at websockets
11:43 sri like i said, you can have the namespace if you like :)
11:43 marcus sri: btw, I had to use nginx rewrite directive to proxy thefeed.no/api to prefork daemon.
11:43 marcus sri: I'll be happy to use hypnotoad.
11:44 marcus sri: does it only exist in your local branch atm?
11:44 sri thats correct
11:44 sri been testing it at a client too, but there are still problems
11:45 sri at least they sponsor the bug fixes :)
11:45 spleenjack is it expected that url_for() helper adds format (and other values from captures() also) of current route to the generating url? if i call url_for() from html-template and want to generate an url to the other route that doesn't support html-format, i have to call url_for() with { format => '' } explicitly.
11:47 sri spleenjack: that behavior is pretty much undefined atm
11:48 sri does make sense though
11:49 sri someone would have to investigate, make a list of cases where it doesn't work so well, and write tests to show how it should work
11:51 vti i reported it like years ago
11:51 sri vti: got evidence?
11:51 sri where's the test case?
11:52 spleenjack my application have no utilize explicit format (like .html suffix) in path for www-routes
11:52 spleenjack and url_for() do unexpected work with adding format-suffix to my url
11:52 sri test cases!
11:52 vti i always fight with it in bootylicious
11:53 spleenjack sri: one moment
11:53 marcus mhmph
11:55 sri ok i have a fix
11:55 sri will commit if i get a test ;p
11:55 marcus I'm listening to uxebu talking about html5 based apps.
11:56 marcus friggin' dojo consultants
11:56 marcus giving html5 a bad name :p
11:56 marcus all his examples are butt ugly as well.
11:57 sri so far i've not seen many examples the beat the mozilla video
11:57 sri http://www.youtube.com/watch?v=-BsB0HpS768
11:57 sri that really shows what html5 is about
11:59 spleenjack http://pastebin.com/ZpgcqaMi
12:00 sri argh
12:01 stephan48 what alternatives are there to prefork? or is deamon also ok?
12:01 sri spleenjack: the case is ok, but a lite_app.t patch would have been nicer
12:01 sri stephan48: there will be hypnotoad
12:01 stephan48 "hypnotoad"?
12:02 sri prefork done right
12:02 sri new server
12:03 stephan48 okeeeeeyyy...
12:03 spleenjack sri: but i don't know what is wrong exactly, is it euther inclusion the whole captures() of current route or captures()->{format} only?
12:03 sri marcus: now you are confusing our users...
12:03 sri this is what i was trying to avoid...
12:05 stephan48 *runs back to cat* atleast i know there whats evil!
12:05 marcus sri: me?
12:05 sri YOU!
12:06 marcus sri: I'm not the one deprecating stuff without providing a replacement :p
12:06 sri ...
12:06 marcus seeing as you're telling everyone 'use github'
12:06 marcus it now doesn't have a working http daemon anymore.
12:06 sri what should i have done instead?
12:06 marcus deprecated it in the hypnotoad branch
12:06 arthas joined #mojo
12:06 sri the normal daemon works flawless
12:07 sri actually much better than before
12:07 sri marcus: umm....
12:07 sri marcus: you do realize prefork has serious bugs?
12:07 sri as in if you throw websockets at it things will go horribly worng
12:08 sri i don't have the resources to fix those
12:08 omega can I with Test::Mojo somehow get the content? I only find the _like, _is methods etc?
12:09 marcus omega: you can get to it through $t->tx
12:09 vti sri: where's perlrocks?
12:09 marcus wow, ths guy is telling us jquery mobile probably won't be enough, and you should use dojo.
12:09 sri ok once again, if anyone here does have the resources, please adopt the prefork daemon!!!
12:09 sri otherwise, shut up!
12:10 marcus sri: grumpy today?
12:10 sri marcus: only because of you
12:11 sri you are acusing me of making mistakes, but refuse to tell me which
12:12 sri prefork is broken by design, there is no easy fix
12:12 sri websockets kill it
12:12 marcus sri: I don't care about web sockets.
12:12 marcus I never wrote an app that use it.
12:12 marcus but your trunk changes broke my apps anyways.
12:12 marcus without giving me any laternatives.
12:12 marcus alternatives.
12:12 omega plackup?
12:13 sri ^^^
12:14 sri marcus: i think you have co-maint on daemon prefork
12:14 sri go ahead!
12:15 marcus sri: go  ahead and do what?
12:15 sri maintain it in a separate package
12:16 marcus why? You won't release a new version to cpan before hypnotoad is done anyways?
12:16 sri if you split out daemon prefork i can
12:17 sri because i don't have to worry about loss of features anymore
12:18 sri for example i removed pid support to address a bug in the normal daemon
12:19 sri since hypnotoad has its own pid file handling
12:20 sri daemon prefork could just as well get its own pid file handling
12:20 marcus yepp
12:20 sri someone just has to invest time
12:20 sri i can't
12:20 sri so if you turn it into a new distribution you would take pressure off me
12:20 marcus I'll look into it this afternoon.
12:20 marcus Am at a conference now, and have to pick up in the kindergarden soon.
12:21 omega why not just use plack and its multitude of servers/daemons?
12:21 marcus that is a pretty good question, I suppose.
12:21 Akron joined #mojo
12:21 sri for http thats a valid option
12:22 omega but not for websockets?
12:22 sri nope
12:22 stephan48 so with websockets what to use?
12:23 sri normal daemon only atm
12:24 sri hmm, i could remove daemon prefork for now and point to PSGI
12:24 sri that would allow a quick release
12:24 marcus yeah, that'd be fine.
12:24 marcus imo
12:25 sri vti: whats your vote?
12:28 omega If I was to add more convenience methods to Test::Mojo, what are the odds of that getting merged in?
12:28 sri omega: depends on popularity
12:30 marcus omega: I want your json partials method.
12:31 omega where are the tests for Test::Mojo?
12:31 sri all over the place
12:37 meraxes left #mojo
12:41 dotan marcus: json partials? link?
12:43 vti sri: i am with you
12:44 sri allright then
12:44 GitHub47 joined #mojo
12:44 GitHub47 mojo: master Sebastian Riedel * a911542 (4 files in 4 dirs): fixed url_for format handling - http://bit.ly/byoNOK
12:44 GitHub47 left #mojo
12:45 sri spleenjack: that should solve your problem
12:45 sri format handling should be very intuitive now
12:47 sri is this a major or minor release?
12:48 sri marcus, omega, vti?
12:51 vti 1.0? :D
12:51 omega I don't know what the difference is
12:51 * sri slaps vti
12:51 sri hot beverage or not hot beverage
12:52 und3f[A] left #mojo
12:53 spleenjack sri: thanks!
12:54 vti i think ye, you'vechanged a lot
12:54 vti and remove prefork
12:54 vti *removed
12:55 GabrielVieira2 joined #mojo
12:58 GitHub65 joined #mojo
12:58 GitHub65 mojo: master Sebastian Riedel * 42c1768 (2 files in 2 dirs): removed Mojo::Server::Daemon::Prefork due to unfixable design flaws regarding WebSocket support, please use a PSGI server instead for HTTP production setups - http://bit.ly/cd0Zhg
12:58 GitHub65 left #mojo
12:59 sri it is done
12:59 sri please review and let me know if i can push a release
13:00 GabrielVieira left #mojo
13:02 perlrocks joined #mojo
13:05 Akron Sorry for interrupting your release chat ... ;)
13:05 Akron Always when I do something like $client->get('https://www.google.com')->res is HTTP/1.1 200 OK\nDate: ...\nContent-Length: 0. No debug info in the Log ... how can I debug? Do I have to say explicitely, that I want to SSL handshake?
13:05 omega dotan: I'm not sure "partials" is a good explantion, but here are an example test from my current test-suite: https://gist.github.com/d25085cde48261f37e9b
13:06 sri Akron: do you have IO::Socket::SSL installed?
13:07 sri because perl -Ilib -Mojo -e 'print g("https://www.google.com")' works flawless for me
13:07 Akron Yes - latest version - Net::SSLeay as well.
13:07 omega marcus: https://gist.github.com/d25085cde48261f37e9b <-- example test :)
13:08 * sri pokes vti and marcus
13:08 Akron sri: And I'm wondering, that there is really no problem mentioned in the log ...
13:09 sri Akron: no idea, works flawless here
13:10 omega sri: marcus is fetching his daughter at the kindergarten atm
13:10 Akron sri: Hm ... okay. So it seems, that I#m doing it right ... then I have to test IO::Socket::SSL separately. Thanks!
13:10 vti sri: works ok with bootylicious :)
13:15 perlrocks Twitter: "We are going to take an interim step with the next #mojolicious release and delay the addition of the new server. http://bit.ly/bMoZU3 #perl" --kraih http://twitter.com/kraih/status/27831687619
13:18 perlrocks Twitter: "I'm not happy with this solution, but hypnotoad will take some more time and we need to get those bugfixes out there. #mojolicious #perl" --kraih http://twitter.com/kraih/status/27831959072
13:19 fhelmber_ joined #mojo
13:21 perlrocks Twitter: "Mojo::Server::Daemon::Prefork is up for adoption if anyone really depends on keeping it maintained. #mojolicious #perl" --kraih http://twitter.com/kraih/status/27832211844
13:34 sri vti: good
13:34 sri guess i can release then
13:42 rhaen sri: cool! Thank you!
13:42 sri rhaen: oh, what did i do this time? :)
13:43 rhaen I think that's a step in the right direction.
13:43 rhaen releasing it to cpan for a newer version
13:43 rhaen :)
13:43 sri well, that release comes with a cost
13:43 rhaen it makes sense for everyone coming to mojo!
13:43 sri no prefork for now
13:44 rhaen I don't think so. Prefork is a part - however, all those people which are interested in the client stuff are not using it
13:44 rhaen It's just one way to deliver mojolicious content, but it's not the only way.
13:45 rhaen Yes, it breaks working installations but who cares? It's not 1.0 - no need to keep everything in shape if serious problems occur.
13:45 sri oh, perl people are very picky when it comes to backwards compatibility
13:45 rhaen The Test::Mojo stuff is cool and solves some of the problems of Test::WWW::Mechanize
13:46 rhaen The Mojo::Client stuff is awesome, the Mojo::DOM stuff - this is just awesome.
13:46 rhaen And for delivery - there are other options like PSGI or reverse apaches to the standard daemon.
13:46 GitHub114 joined #mojo
13:46 GitHub114 mojo: master Sebastian Riedel * 5e6bd42 (4 files in 4 dirs): cleanup - http://bit.ly/9X05yw
13:46 GitHub114 left #mojo
13:47 sri still makes me feel a little bad
13:47 rhaen I haven't looked in the commits, have you removed the documentation from it?
13:47 sri yes
13:47 sri removing last traces rigth now
13:47 rhaen sri: yes. It's bad, but it leads to a better future without keeping the broken stuff which doesn't get fixed in time.
13:49 sri weird how these loss/loss decisions are so much easier if decided by committee
13:49 sri if it was just me i couldn't have decided this
13:49 rhaen that's why we are here :)
13:49 sri \o/
13:50 rhaen I am at a conference tomorrow (IT-SA in Nürnberg) and I am meeting the guys from the freeX
13:50 rhaen I think I won't be able to write an article about Mojo in a whole as it is way too much stuff
13:50 sri make it a series :D
13:51 rhaen I would like to focus on just a few aspects like: using Test::Mojo to test your website
13:51 rhaen Or, deploy options for Mojolicious and other perl alike web daemons
13:51 sri simple lite app might be a good start
13:52 rhaen I really would like to contribute the stuff I wrote to the project instead of keeping it to a small audience only
13:53 rhaen sri: hm, yup. You are right. Think I will show how to use Mojo::Client and Test::Mojo as a replacement for WWW::Mechanize or Test:WWW.*
13:53 rhaen everything with one additional module - keep your perl stuff clean :)
13:53 sri proxying is usually a lot of fun too
13:53 rhaen right. I'll talk to them :)
13:53 rhaen ok. next meeting for me - cya later
13:54 sri laters
13:54 sri which was the best pause upload tool again?
13:55 koban left #mojo
13:56 stephan48 pause?
13:56 sri pause.perl.org
13:57 stephan48 theres alink on pause page howto give a github tag tarball link
13:57 stephan48 als download source
13:57 sri ooh
13:57 stephan48 i know pause :)
13:57 stephan48 atleast i think so that it was there some weeks ago
13:57 stephan48 look for upload module and then around the uri field
13:58 forwardever joined #mojo
13:58 sri anyway, new release is on its way to cpan
13:58 stephan48 found it?
13:58 sri nope, uploaded old school again
13:59 sri don't think we have a manifest on github anyway
13:59 stephan48 http://blog.usarundbrief.com/?p=36
14:01 vti sri: there is a tool written by yko
14:02 vti http://github.com/yko/mojox-cpan-uploader
14:02 Akron left #mojo
14:03 forwardever sri: "perl script/myapp routes" shows a list of routes, however, the request method (get,post) is missing
14:03 perlrocks Twitter: "Released #mojolicious 0.999930 (Hot Beverage) with countless new features and bug fixes, enjoy! #perl" --kraih http://twitter.com/kraih/status/27835807566
14:05 sri forwardever: patches welcome
14:06 forwardever I'm just an end user :)
14:08 perlrocks Twitter: "Recently I was anxious to hear about the really good mojolicious" (ja) --TokyoIncidents http://twitter.com/TokyoIn​cidents/status/27836326234
14:13 kvorg left #mojo
14:18 sri omega: whats this json query stuff in your example?
14:18 hoelzro joined #mojo
14:19 sri jsonpath?
14:20 sri whenever i see jsonpath it feels weird, since we can just represent everything in json with native perl structures
14:29 omega sri: its just something I made now
14:30 omega only allows simple things
14:30 sri hmm
14:33 omega dunno if the whole jsonpath spec makes sense in this matter either
14:33 omega I just wanted a simple way to target parts of the json
14:33 sri i don't like jsonpath at all
14:34 sri a) it makes no sense in a language like perl b) it looks just as ugly as xpath
14:34 omega mhm
14:34 sri i can see a simple structure notation make sense for tests
14:35 sri quite unsure about it though
14:36 omega mhm
14:37 omega what is the best practice way of specifying invalid request in ::lite?
14:37 dotan omega: does json_query do partial de-serialization of the json string, or just convert it to a data structure and then look up the path spec?
14:38 sri don't think there is a best practice for that yet
14:38 omega dotan: it uses the ->json method on the response, so it de-serializes everything, then just tries to walk the path spec
14:38 sri we expect the lower levels to take care of that usually
14:39 omega but I'm building a webservice, and want to reply correctly if someone requests something they shouldn't
14:39 sri would be a normal response then i guess
14:40 sri ->render('invalid', status => '40x')
14:40 omega mhm, ok
14:45 perlrocks Twitter: "RT @kraih Released #mojolicious 0.999930 (Hot Beverage) with countless new features and bug fixes, enjoy! #perl" --xomaa http://twitter.com/xomaa/status/27839633317
14:46 spleenjack left #mojo
14:46 omega sri: if you want something akin to json_query in Test::Mojo, I'm going to need som direction, if not I'll just keep in a sub-class as now
14:47 sri i kind of like the idea, but for now i think it might be better of in its own distribution
14:47 omega ok
14:47 sri if it gets popular it can always be promoted
14:48 omega mhm
14:48 sri the question for something like that has come up a few times at least
14:49 fish_ my $data = $self->req->param('data'); <- should this work for more complect data structures in a post? i just fed jquery.post a 'hash' with two keys were one value are another two keys with values.. but in mojo its just undef..
14:50 sri fish_: that does not sound like a urlencoded form post
14:50 perlrocks Twitter: "RT @kraih: Released #mojolicious 0.999930 (Hot Beverage) with countless new features and bug fixes, enjoy! #perl" --rabenfeder http://twitter.com/rabenfeder/status/27840129887
14:53 fish_ sri: i probably should have a look what jquery.post really does
14:53 sri fish_: that migth be a very good idea :)
14:53 fish_ fdata%5B0%5D%5Baction%5D=hello&data%5B0%5D%​5Bparam%5D%5Blat%5D=52.50933&data%5B0%5D%5B​param%5D%5Blng%5D=13.4560725&data%5B%5D=-15
14:54 omega marcus: your testing of ::lite spews out the HELP for the script each time if I run prove -v?
14:54 fish_ s/^//
14:55 sri omega: all apps do that
14:56 sri we are having a hard time detecting test usage
14:56 sri it's too similar to a normal command line invocation
14:57 omega ok
14:57 omega HARNES_ACTIVE?
14:57 omega or perhaps HARNESS
14:57 sri perl t/myapp.t
14:58 sri i was thinking about setting something in Test::Mojo
14:58 omega well, if people aren't running their tests under a harness, I doubt they care about the output that much, and it would fix it for the majority
15:00 sri whats the most common env var for the test harness?
15:00 rhaen foo?
15:01 omega HARNESS_ACTIVE
15:03 omega whats the easiest way to "fake" a response into Test::Mojo?
15:03 sri guess it's easiest not to fake it :)
15:03 omega can I somehow inline a small app in the test-script?
15:04 sri $t->tx($tx) might work
15:04 dotan omega: I think you should find a way to generalize the json path stuff to paths in perl data structures. It could be useful to perl developers everywhere.
15:04 sri omega: yes, inlining a lite app is absolutely trivial
15:04 sri omega: http://github.com/kraih/mojo/blob/mas​ter/t/mojolicious/charset_lite_app.t
15:05 omega dotan: perl datastructures are infinetly more complex than json-structures
15:05 sri no setup needed
15:05 sri you just inline the lite app and Test::Mojo will just work
15:06 GitHub33 joined #mojo
15:06 GitHub33 mojo: master Sebastian Riedel * 8bf5ce5 (3 files in 3 dirs): hide command overview from prove - http://bit.ly/a00NS9
15:06 GitHub33 left #mojo
15:07 omega ok
15:08 omega well, heading out. Will try to clean up and doc Test::Mojo::JSON later today or tomorrow
15:09 sri laters
15:11 perlrocks Twitter: "RT @kraih: Released #mojolicious 0.999930 (Hot Beverage) with countless new features and bug fixes, enjoy! #perl" --skaurus http://twitter.com/skaurus/status/27842052725
15:22 rhaen hm. I should write a Nagios::Plugin::Mojo::Client module for Nagios probing
15:22 sri you should
15:22 spleenjack joined #mojo
15:23 rhaen sri: would be cool as it just pulls one dependency into the system - Mojo.
15:23 rhaen That's the only thing you need.
15:30 GitHub120 joined #mojo
15:30 GitHub120 mojo: master Sebastian Riedel * 293e653 (1 files in 1 dirs): cleanup - http://bit.ly/9jsy56
15:30 GitHub120 left #mojo
16:21 dotan sri: what's the big issue with websockets and the preforking daemon?
16:22 sri dotan: there are multiple deadlocks
16:23 sri a malicious person could easily block your server
16:23 sri i'm not going into detail ;p
16:23 sri but there are multiple such problems
16:25 sri prefork was simply not designed for long lasting connections
16:28 ysyrota left #mojo
16:55 dotan left #mojo
16:57 sri behind nginx everything is fine though
16:57 sri so no need to panic
17:04 chansen left #mojo
17:08 kvorg joined #mojo
17:11 forwardever sri: improved my resourceful_routes plugin: http://github.com/forwardever/mojo​licious-plugin-resourceful_routes , added scaffolding of templates and controller, now have to write some tests, still not sure how to do this exactly
17:11 chansen joined #mojo
17:13 sri forwardever: you need to blog it, and possibly post to mailing list
17:14 forwardever sri: would like to add tests first
17:17 forwardever e.g., would like to test whether routes are generated in a correct way through resources command
17:18 spleenjack left #mojo
17:19 perlrocks Twitter: "Mojolicious (PSGI) Starman deploy the app - Getting Started with Perl code samples http://htn.to/PA9dF3" (ja) --yusukebe http://twitter.com/yusukebe/status/27852719499
17:20 perlrocks Twitter: "RT! Yusukebe: Mojolicious (PSGI) Starman deploy the app - Getting Started with Perl code samples http://htn.to/PA9dF3" (ja) --perlism http://twitter.com/perlism/status/27852762102
17:29 marcus sigh
17:32 sri ohoh
17:36 stephan48 good or bad?
17:38 sri i don't know, just wanted to sound like i knew what marcus meant
17:57 stephan48 k
18:14 vti me sighs
18:17 GabrielVieira2 left #mojo
18:22 spleenjack joined #mojo
18:51 kvorg ohoh, I missed some exciting events today
18:51 kvorg sri: congrats on the new release
18:52 kvorg not that it would count, but there are other deployment options, i think you really want to make websockets work and all our bugging was just holding you back from the hypnotoad ... so it must have been a good decision
18:52 kvorg (it also seems that your work on the hypnotoad must have had something to do with all the new features in the last two months ...)
18:52 marcus sri++ # release
18:53 marcus everybody should buy a kindle, btw.
18:53 marcus it's awesome.
18:53 sri somehow i'm not convinced
18:53 marcus I just got a new one
18:53 marcus ok, correction, everyone who reads should get a kindle ;)
18:56 sri perl -Ilib -Mojo -e'g("search.cpan.org/dist/Mojolici​ous")->dom("td.version")->each(sub { b($_->text)->say if $_->text })'
18:56 sri latest mojolicious version
18:57 * sri always waits for the movies
18:57 marcus sri: then don't get a kindle
18:57 sri epaper bad for movies?
18:57 marcus I think it would suck for movies.
18:57 marcus mmm
18:58 marcus unless you like 2 fps
18:58 sri mmmm
18:58 * kvorg has several tons of paper books and can read e-books with a nice laptop (or netbook when in bed/bath/toilet mode).
18:58 * kvorg climbs on top of his books and grins.
18:58 marcus kvorg: screw reading on a laptop screen
18:59 marcus epaper is the same quality as reading paper
18:59 marcus I have 15 boxes with books under my stairway.
18:59 marcus Going to give them away the next time I move
19:02 kvorg marcus: language, content, decade of plubishing and approximate length when on the shelf - and perhaps I give you my snailmail address (postage on delivery, of course). :-P
19:03 marcus kvorg: sounds like too much work.
19:03 * kvorg wants his laptops to have dual-mode screens.
19:03 * marcus wants his ipad to have a dual mode screen.
19:03 marcus but for now I'm pretty happy with having an ipad, a kindle and a laptop
19:04 Htbaa joined #mojo
19:05 kvorg Kristiansand? A bit out of the way, I might get to Norway on a conference in spring. Anything in English?
19:06 marcus huh?
19:06 marcus I am in Oslo
19:06 marcus sure, I have lots of english books
19:07 marcus a lot of sci-fi and fantasy stuff, but also some stuff like bukowski and murakami.
19:07 kvorg ah, so just your email sever ext.unicore.no is in Kristiansand? Or maybe geolocation is broken?
19:07 GabrielVieira2 joined #mojo
19:08 marcus ext.unicore.no is in oslo too, but this is a irssi running on a screen
19:08 * kvorg will just send a truck, no worries.
19:08 marcus which conference are you going to?
19:08 kvorg Nordugrid/NDGF, it is more like a meeting, I am not sure where the yearly conference will be.
19:09 marcus neat
19:09 kvorg marcus: Is kindle now ok if you have pdf, or pdf on a local server?
19:10 marcus kvorg: for pdfs, I would get the kindle dx. The normal kindle screen is on the small side for most technical stuff
19:10 marcus with the latest firmware, it can render pdfs fine tho.
19:12 perlrocks Twitter: "RT @kraih: Released #mojolicious 0.999930 (Hot Beverage) with countless new features and bug fixes, enjoy! #perl" --marcusramberg http://twitter.com/marcusr​amberg/status/27860580558
19:12 marcus kvorg: one interesting feature of the new kindle; The cover has a built in led lamp that takes electricity from the kindle
19:13 snearch joined #mojo
19:13 kvorg nice, i have that on the laptop and the phone (but not the mojo server, damn sri, where is the hypnotoad, it sure has bioluminence built in)
19:14 kvorg incidentaly, anyone find a good loking Unicode HOT BEVERAGE in some font?
19:19 perlrocks Twitter: "RT @kraih: Released #mojolicious 0.999930 (Hot Beverage) with countless new features and bug fixes, enjoy! #perl" --sharifulin http://twitter.com/sharifulin/status/27861058503
19:19 marcus kvorg: what about on your paper books? ;)
19:20 kvorg lol, I have some standing next to my paper book, thank you very much, and it is pretty nice and in colour (but getting cold).
19:21 marcus mm
19:21 stereonaut left #mojo
19:21 marcus I like this one a lot because it only lights up the kindle screen
19:21 marcus since I have daughter and wife in the bed as well
19:22 forwardever marcus: tried $t->app->log->level('error'); to get my tests silent as outlined in your guide, but doesn't seem to work, any ideas? still get debug messages, using prove -l t/*
19:22 yulia joined #mojo
19:22 kvorg Something like that: http://www.shutterstock.com/pic-9908137/sto​ck-vector-hot-beverage-in-a-cup-symbol.html
19:22 kvorg For the news on the release. But it would be better if it came from an actual font, of course.
19:23 marcus forwardever: uhm, that's how sri does it in the mojolicious test suite.
19:23 GitHub103 joined #mojo
19:23 GitHub103 mojo: master Sebastian Riedel * 850776a (1 files in 1 dirs): improved version command - http://bit.ly/91y86Z
19:23 GitHub103 left #mojo
19:23 sri :D
19:23 kvorg And for cases like: "Mojolicious ☕ has been released to CPAN."
19:23 kvorg sri++
19:24 kvorg i like it when I am ahead with github
19:24 stereonaut joined #mojo
19:26 sri !tweet ☕!!!
19:26 perlrocks sri: http://twitter.com/perlrocks/status/27861572613
19:27 und3f joined #mojo
19:27 jamesw now with all the emoji in unicode 6.0 tweets are going to be much more concise
19:27 jamesw heh
19:34 kvorg Btw: I just noticed that HOT BEVERAGE can be used as an indication to wait. Which should be the default response if anyone nags sri about hypnotoad.
19:34 sri HOT BEVERAGE
19:35 marcus sri: so did you come up with a code name for the hypnotoad release yet?
19:35 fhelmber_ left #mojo
19:35 sri marcus: no, that will be a huge problem
19:35 marcus yeah, it should be a good one
19:36 sri maybe snowflake
19:36 sri little unrelated though
19:36 marcus better than hot turd or what it was called
19:36 sri are there toads in unicode 6.0?
19:36 kvorg hmm, damn, we could have used SNOWMAN WITHOUT SNOW for this one
19:37 sri TOAD WITH HEARTYEYES or so maybe?
19:37 marcus that would have been awesome
19:37 * kvorg hides in a FUNERAL URN (U+26B1)
19:38 sri that would be the last release
19:39 marcus FACE WITH LOOK OF TRIUMPH
19:39 kvorg or HISTORIC SITE for 1.0.
19:39 sri :D
19:39 kvorg (U+26EC)
19:40 kvorg (but you get U+26F B⛻ JAPANESE BANK SYMBOL if you purchase commercial support)
19:40 marcus CAT FACE WITH TEARS OF JOY
19:41 marcus 1F47D is pretty good
19:41 marcus (EXTRATERRESTRIAL ALIEN)
19:42 marcus 1F438 FROG FACE
19:42 marcus 1F42B BACTRIAN CAMEL
19:42 sri frog face is actually cute
19:43 sri PIG FACE
19:43 marcus pig face is cute too
19:43 sri 1F437
19:43 marcus but PANDA FACE
19:43 marcus cuteness overload
19:43 marcus I like PAW PRINTS too
19:43 sri awwww
19:44 DaTa ist there a FACEPALM? :)
19:44 marcus 1F433 fail whale
19:44 sri go find out!
19:44 sri marcus: i do not believe you
19:44 sri oh, close enough
19:45 marcus 1F46F WOMAN WITH BUNNY EARS
19:45 marcus playboy is now in unicode
19:45 sri JAPANESE OGRE
19:45 marcus 👻 GHOST
19:45 marcus HUNDRED POINTS SYMBOL = perfect score
19:46 sri DANCER 1F483
19:46 marcus lol
19:46 marcus it's elvis
19:46 sri 💆FACE MASSAGE1F486
19:46 kvorg 1F438 🐸 FROG FACE
19:47 kvorg ah, too late ... u have done it all while i was looking at the standard
19:47 marcus 1F419 OCTOPUS
19:47 kvorg 1F46F?? marcus? really?
19:47 marcus kvorg: it's pretty awesome.
19:48 marcus sri: I think you should use 1F459
19:48 marcus Mojolicious Bikini
19:48 kvorg sri: starting a flamewar? poor dancer people will get confused.
19:48 marcus sri would never do that. He loves dancers
19:48 kvorg Mojolicious Bikini Release sounds attractive.
19:49 sri Mojolicious Bikini Issue?
19:50 kvorg well, we cant do Mojolicious 1F46A 👪 FAMILY release before 1.0, and these are more for experienced public.
19:51 sri think we should do an emoticon next
19:51 sri FACE WITH LOOK OF TRIUMPH
19:51 sri or so
19:51 marcus FACE WITHOUT MOUTH
19:52 marcus FACE WITH OPEN MOUTH AND COLD SWEAT would be relevant for hypnotoad
19:53 kvorg sorry to close it back in a loop, but FACE WITH LOOK OF TRIUMPH (marcus at 9:39:30) is definitely appropriate for the hypnotoad release, while  FACE WITH OPEN MOUTH AND COLD SWEAT is more of a hypnotoad branch symbol.
19:53 kvorg ok, we have to nudge Unicode to get HYPNOTOAD in Unicode 7.
19:53 marcus or SEE-NO-EVIL MONKEY
19:54 kvorg nope, that is 1.0 RC1
19:54 sri FACE WITH STUCK-OUT TONGUE AND WINKING EYE
19:54 marcus don't panic?
19:55 kvorg 1F3AF 🎯 DIRECT HIT and 1F3B0 🎰 SLOT MACHINE will get handy too ...
19:55 * kvorg calms himself and closes Unicode 6.
19:56 sri SEE-NO-EVIL MONKEY
19:56 sri wow, total win
19:56 marcus its nice
19:56 sri next 3 releases
19:56 sri HEAR-NO-EVIL MONKEY
19:57 sri SPEAK-NO-EVIL MONKEY
19:57 marcus is that hynotoad, rc1 and 1.0? ;)
19:57 sri :D
19:57 yulia left #mojo
19:59 forwardever marcus: for none lite apps, it seems to be $self->log->level('fatal'); in the startup method
20:00 marcus forwardever: app->log->level('fatal') then?
20:00 marcus for lite apps
20:01 forwardever just tested my non lite plugin, and it works, not sure about lite apps
20:01 forwardever maybe, it's different in lite
20:02 forwardever $t->app->log->level('fatal'); did NOT work
20:02 marcus drop $t
20:02 forwardever no, it's also non lite
20:05 forwardever the lite app tests indeed use just app->log->level('error'); (lite_app.t)
20:08 MojoGuest444 joined #mojo
20:08 MojoGuest444 From: http://news.ycombinator.com/item?id=1277067 (7 hits)
20:09 sri think i'm goig to change something regarding log output
20:09 sri *+n
20:09 sri and just default to fatal
20:11 MojoGuest444 left #mojo
20:14 perlrocks Twitter: "The Web In A Box, With A Bow – Al Newkirk & Associates – Internet Business Consultants http://t.co/5ovuljX via @AddThis" --icjr http://twitter.com/icjr/status/27864790179
20:14 Htbaa left #mojo
20:15 forwardever sri: how can I test whether the correct route name has been set, can I get access to the name of the current route via $c?
20:15 sri $c->match->endpoint->name
20:16 forwardever is something like this also possible via $t (test object)
20:17 sri that does not make sense
20:18 forwardever probably not, so I will try your first suggestion, render the name, and than test
20:20 sri our tests use an actual http server and client
20:20 sri there is no direct connection
20:21 omega http://github.com/omega/test-mojo-json
20:22 omega for those that might be interested
20:22 omega feedback welcome
20:22 omega etc
20:22 omega :)
20:22 sri why inline pod?
20:22 omega because dzil rewrites it for me
20:23 omega this way I can more easily check that docs and code are in coherence
20:23 sri pod after __END__ is soooooo much nicer
20:23 omega but thats how it will end up
20:23 omega on CPAN
20:23 sri hmm
20:23 sri i couldn't work like that anymore :)
20:24 omega :)
20:27 GitHub163 joined #mojo
20:27 GitHub163 mojo: master Sebastian Riedel * 9e52f8e (18 files in 6 dirs): default to silent tests in Test::Mojo - http://bit.ly/aqsGml
20:27 GitHub163 left #mojo
20:28 sri app->log->level('fatal') has become a convention
20:28 omega for tests?
20:28 sri ye
20:28 omega then the testing-guide and example-test you showed me both need updating :p
20:29 omega they both say error
20:29 sri already updated ;p
20:29 omega :)
20:30 omega updated :)
20:33 sri elasticsearch does look interesting
21:01 perlite left #mojo
21:01 ltriant joined #mojo
21:01 perlite joined #mojo
21:01 und3f is now known as und3f[A]
21:11 forwardever sri: how to render a template from the C<DATA> section while changing tag_start/tag_end ? Not sure at which point to change the tags?
21:12 sri thats quite advanced stuff
21:12 sri there's an example in twinkle_lite_app.t or so
21:13 forwardever want to create templates through the "generate resource" command, which is a problem as I cannot escape the tags
21:14 sri umm, thats a whole different context....
21:14 sri you might want to mention that next time
21:14 sri look at the current generator commands
21:14 sri some change the tags
21:16 forwardever think I have found it in the lite app generator, thanks
21:45 snearch left #mojo
22:10 hoelzro left #mojo
22:18 arthas left #mojo
22:19 GitHub178 joined #mojo
22:19 GitHub178 mojo: master Sebastian Riedel * 9fac394 (1 files in 1 dirs): dead simple installation - http://bit.ly/9PgkYg
22:19 GitHub178 left #mojo
22:19 sri :)
22:26 DaTa miyagawa++ :)
22:29 forwardever left #mojo
22:30 tempire_ joined #mojo
22:32 GitHub144 joined #mojo
22:32 GitHub144 mojo: master Sebastian Riedel * 88b389a (1 files in 1 dirs): made installation a bit cooler - http://bit.ly/bO2QaO
22:32 GitHub144 left #mojo
22:32 sri :D
22:44 kvorg cool! and nifty as hell
22:44 kvorg see? releasing makes you smuger (and I hope this is a good thing)
22:45 kvorg sri++
22:45 * kvorg observes sri overflow his integers and goes to bed.
22:45 sri it surely took away some pressure
22:46 sri nn
22:49 DaTa vti: http://data.rbfh.de/p/2010-10-20vEZW0fLH for MojoX::Validator, $Mojo::VERSION is gone
22:50 sri no cookie for vti!
22:53 tempire_ v .30!
22:53 * tempire_ dances
22:54 * tempire_ releases updated basicauth module
22:54 * tempire_ goes back to work on the upcoming high profile site based mojolicious
22:55 tempire_ anyone seen the gnome shell, in gnome 3.0?
22:55 tempire_ almost makes me want to switch away from osx
22:58 forwardever joined #mojo
23:00 forwardever sri: <%= text_field 'foo', value => '' %> creates <input name="foo" value="" />, isnt there a type="text" attr missing
23:05 sri forwardever: no
23:06 sri tempire_: seen a video, wasn't that impressed tbh
23:06 tempire_ really?  That side bar is killer to me, with seemless keyboard navigation.
23:07 tempire_ osx keyboard nav is still weak.
23:07 forwardever sri: TagHelpers.pm says there should be type="text"
23:07 sri forwardever: typo then
23:10 tempire_ I can see you being ok with current keyboard, nav, though.  I launch textmate once every couple of days just to look at the pretty twilight theme.
23:10 sri haha
23:10 tempire_ but then I go back to vim and leave the mouse to itself
23:10 GitHub144 joined #mojo
23:10 GitHub144 mojo: master Sebastian Riedel * 395115b (1 files in 1 dirs): fixed typos - http://bit.ly/98A3MP
23:10 GitHub144 left #mojo
23:12 sri maybe i use the mouse more because i draw a lot too
23:13 forwardever nn
23:13 sri nn
23:13 forwardever left #mojo
23:24 skaurus joined #mojo
23:29 skaurus sri: it's strange but when I start mojo with that script: http://gist.github.com/635395 I doesn't get any warns into log besides "server listening". If I remove --daemonize option, warns are warned to the console and not to the log; of course I can write > log/development.log 2>&1 in start line. But also sock file now created with 755 rights and root owner, so nginx doesn't work with that. And that last problem I couldn't solve... Any suggestions how
23:29 skaurus to fix that?
23:30 sri daemon_prefork is not part of mojolicious anymore
23:31 skaurus yeah I know...
23:32 sri in that context your question seems rather weird
23:35 spleenjack left #mojo
23:38 sri domain socket problem could have been a bug
23:39 skaurus trying to move to PSGI right now...
23:40 sri wish hypnotoad was ready to make this more smooth :/
23:41 skaurus me too :p
23:42 sri anyway, time to sleep
23:42 sri nn
23:42 skaurus bye

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