Camelia, the Perl 6 bug

IRC log for #mojo, 2010-08-23

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

All times shown according to UTC.

Time Nick Message
00:50 ka2u left #mojo
00:50 ka2u joined #mojo
02:35 janus left #mojo
02:41 janus joined #mojo
02:50 Foxcool joined #mojo
04:04 tl joined #mojo
04:06 tholen42 left #mojo
05:16 yakudzo joined #mojo
05:33 ysyrota joined #mojo
05:57 koban joined #mojo
06:43 Foxcool_ joined #mojo
06:43 Foxcool left #mojo
06:43 Foxcool_ is now known as Foxcool
07:03 perlrocks Twitter: "hatebu Mojo + DBIx:: Skinny + Test:: mysqld such that - perl Study Notes: 6 users http://bit.ly/bV4r6q" (ja) --tk0451 http://twitter.com/tk0451/status/21894017575
07:03 perlrocks Twitter: "RT @ tk0451: hatebu Mojo + DBIx:: Skinny + Test:: mysqld such that - perl Study Notes: 6 users http://bit.ly/bV4r6q" (ja) --AE_PerlIsm http://twitter.com/AE_PerlIsm/status/21894017826
07:03 [1]awnstudio joined #mojo
07:14 skaurus left #mojo
07:31 ltriant left #mojo
07:41 tl left #mojo
07:48 tholen joined #mojo
07:58 jest joined #mojo
08:03 marcus goood morning vietlamb
08:05 marcus DBIx::Skinny looks awesome
08:05 marcus I'm going to use that on my next project.
08:16 fod left #mojo
08:32 perlrocks Twitter: "learning #mojolicious" --skaurus http://twitter.com/skaurus/status/21897776383
08:34 cristi8 joined #mojo
08:54 geekus left #mojo
08:57 fish_ ohh.. that twitterbot is nice :D
08:57 fish_ but i could right away ask here :)
08:59 koban left #mojo
09:00 marcus fish_: yepp. :)
09:17 [1]awnstudio is now known as alnewkirk
09:20 yko joined #mojo
09:32 dotan joined #mojo
09:45 koban joined #mojo
10:40 pid joined #mojo
10:58 dotan left #mojo
11:13 * sri yawns
11:14 sri daemon_prefork is getting a rewrite, wonder if i should give it a catchy new name
11:16 yakudzo left #mojo
11:24 sri yea, i think i'll give the prefork daemon a new cutesy name :)
11:31 yko left #mojo
11:33 fish_ marcus: so, do you plan to write a mojo powered wiki?
11:33 fish_ i had a look at mojowka, but i looked more like a simple cms than a wiki..
11:34 fish_ -> so revisions, which is - for me - mandentory to call it a wiki
11:34 marcus fish_: I have been pondering doing mojomojo2 with mojolicious and couchdb
11:35 marcus using couch would be so much easier than sql for a tree based wiki, and you even get revisions for free.
11:35 yko joined #mojo
11:35 marcus but there are issues, like the fact that couchdb can delete older revisions as it feels like it.
11:36 fish_ marcus: sounds good. is there already some code somewhere? :)
11:38 marcus fish_: no. There is only the catalyst based mojomojo. ( github.com/marcusramberg/mojomojo/ )
11:39 fish_ well okay, guess than i'll stick to oddmuse for now
11:39 marcus sure.
11:40 fish_ btw: git as wiki backend would be great ;)
11:43 janus sri: daemon + prefork = dork? *hide*
11:45 janus actually that sounds more like a name for the java/fop daemon over here :>
11:47 marcus fish_: yeah, git might be fast enough
11:48 marcus fish_: I've been seeing wikis with rcs and cvs backends, and they pretty much sucked.
12:00 * sri slaps janus
12:00 sri something along the lines of tornado and unicorn ;p
12:03 marcus robot unicorn tornado
12:04 marcus http://games.adultswim.com/robot-uni​corn-attack-twitchy-online-game.html
12:14 sri aaaaaaaaaaaaaaaaaaaaaaah *runs screaming around the channel*
12:19 yko Oo
12:26 arthas joined #mojo
12:31 sri interesting, the first oreilly html5 book is out
12:40 janus mojo is the knife, each process is a fork but users are still nutured with a spoon - based on that it could be named cutlery or something :)
12:41 sri lol
12:53 MojoGuest751 joined #mojo
12:53 MojoGuest751 From: http://xantus.tumblr.com/ (1 hits)
12:56 MojoGuest751 ddd
12:56 sri eee
12:56 MojoGuest751 :)
12:57 MojoGuest751 left #mojo
12:58 perlrocks Twitter: "Thinking about giving the #mojolicious prefork daemon a new cutesy name after the cleanup, suggestions? :) #perl" --kraih http://twitter.com/kraih/status/21911690121
13:14 skaurus joined #mojo
13:15 skaurus \wave
13:15 * skaurus waves
13:17 alnewkirk left #mojo
13:27 * marcus shaves
13:37 vel joined #mojo
13:37 vel left #mojo
13:38 vel joined #mojo
13:40 dotan joined #mojo
14:00 vel left #mojo
14:27 sri hmm, i'll make the worker pool less dynamic, and just go with unicorn like signals for spawning more/less
14:33 DaTa the content is transfered, but the response does not finish: http://pastie.org/1109990 what am i missing?
14:35 kimoto joined #mojo
14:40 kimoto I want to get route name in controller.
14:40 kimoto $c->match->endpoint->{_name} is not good.
14:41 sri c->match->endpoint->name
14:41 sri not {_name}
14:41 sri DaTa: that doesn't look right
14:41 sri DaTa: to end a chunekd stream you simply do ->write_chunk('')
14:41 sri thats all
14:42 kimoto sri: thanks
14:42 sri and the callback is only invoked once
14:42 sri the following ->write_chunk call needs its own callback
14:45 MojoGuest154 joined #mojo
14:45 MojoGuest154 From: http://www.google.com.hk/search?q=html5+we​bsocket+demo&hl=en&newwindow=1&amp​;safe=strict&client=firefox-a&hs=W​ct&rls=org.mozilla:fr:official&ei=​P4lyTKaRGcSecfXjjK0N&start=20&sa=N (1 hits)
14:46 MojoGuest154 tesr
14:46 MojoGuest154 iii
14:46 janus fail :P
14:46 MojoGuest154 impressive
14:47 janus sounds mojolicious
14:47 MojoGuest154 left #mojo
14:47 DaTa sri: ok, so i just loop inside the call back
14:48 sri DaTa: no, you build a callback chain and let mojolicious pull chunks
14:48 sri normal event based programming
14:51 sri i think there is an example in the tests, let me look it up
14:52 sri DaTa: http://github.com/kraih/mojo/blob/​master/lib/Mojo/HelloWorld.pm#L50
14:53 sri ->finish is not neccessary though
14:54 sri it's a lowe level example, but the pattern is the same
14:54 sri *lower
14:55 DaTa ah, ok, thank you
14:57 flu joined #mojo
15:04 DaTa yay, it works :)
15:04 sri :)
15:08 sri hah, i might rename daemon_prefork to bender :D
15:09 koban left #mojo
15:14 sri or maybe not
15:19 skaurus bender and forks?
15:20 skaurus i have one little Bender... He is receiving messages from IM and posts it to one of Russian blog services)
15:21 skaurus it make no sense, yeah - bender and blogs...
15:22 skaurus sri: can you add any new name for daemon_prefork not as replacement but as synonym please?
15:23 sri skaurus: what difference would it make?
15:23 janus in relation to futurama and forking i have to think of the slurm worms, they seem to fork pretty nicely at least in one episode ;)
15:23 sri pretty much all command line options from the new prefork daemon are gone anyway
15:24 sri all configuration will depend on a config file
15:24 sri there will be --daemonize and --config thats it
15:25 sri thats why i think a name change would be appropriate
15:25 skaurus sri: ok, it don't mean much for me now, but would't that change break compatability with existing setups?
15:25 skaurus *wouldn't
15:25 sri it might
15:26 chansen_ joined #mojo
15:26 sri if you see an alternative please share it
15:27 skaurus sri: add new mode?
15:27 sri not possible, daemon internals need to change to support hot deployment for tcp ports
15:27 sri prefork will break because of daemon changes
15:28 skaurus sri: can be old behaviour be emulated with new code?
15:28 sri maybe if i had a few months
15:28 kimoto left #mojo
15:29 sri daemon would basically need to be forked and some of the ioloop code
15:29 chansen left #mojo
15:30 sri prefork would never be able to provide any form of hot deployment
15:30 sri it is broken by design
15:30 sri fyi
15:30 sri it's not like i'm sacrificing a perfectly working module
15:30 sri it is fundamentally broken
15:31 sri feel free to hack in hot deployment if you see a better way though :)
15:32 skaurus sri: so in general i can quite safely rely upon existing features without fear they will be replaced?
15:32 sri if something can be deprecated before being replaced, we will
15:33 skaurus ok, i think it good enough)
15:33 sri right now i'm in the process of eliminating design flaws though
15:33 sri which often go very deeply
15:34 sri since we are close to 1.0 :)
15:34 skaurus preparing for 1.0? :) yeah :)
15:34 skaurus i see ^)
15:36 sri also the reason i keep asking people to point out flaws ;)
15:36 sri now is the time to eliminate them
15:38 koban joined #mojo
15:38 sri that said, pretty much nobody even notices the stuff i consider a design flaw xD
15:38 sri i'm focused on enabling future optimizations
15:42 sri in recent months i've really only broken the way dynamic content generation worked i think
15:43 sri from ->write_cb(sub { return 'lalala' }) to ->write('lalala')
15:44 Foxcool left #mojo
15:45 Akron joined #mojo
15:45 DaTa note to self: always localize $!
15:46 sri i'm not even sure if command line options qualify as breaking backwards compatibility, does it?
15:47 sri *+changing
15:47 sri the command system is supposed to be used reactively, not automated
15:47 skaurus sri: glad to hear that :) i'm a little paranoid at start of getting to know local subculture )
15:48 skaurus sri: nah, if it will die anyway, who cares how to call it...
15:50 sri the new config files qualify though for deprecation policy
15:50 vel joined #mojo
15:50 jest left #mojo
15:54 sri hmm, Mojo::Server::Hypnotoad? :)
15:54 janus lol
16:10 sri hypnotoad > unicorn
16:13 robinsmidsrod is there any gotchas when using DBI together with Mojolicious::Lite?
16:13 sri not really, besides the usual
16:14 sri (the usual being that DBI blocks and sql sucks) :)
16:14 robinsmidsrod I have an object which has a DBI handle in it which is initialized at the start of the script and then I do $obj->something_that_uses_dbi() inside get handlers - should this work out of the box?
16:14 ysyrota left #mojo
16:14 sri you should use DBIx::Connector or so
16:15 robinsmidsrod the problem I have is that a query that takes 144ms on the console seems to "hang" for around 40sec before it runs when I use it through the webapp
16:15 robinsmidsrod github.com/robinsmidsrod/twitter_stream (web_app.pl) in there
16:16 sri maybe the dbi experts here can help
16:16 * sri pokes vti
16:16 * robinsmidsrod is just wondering if he does something completely basic wrong
16:17 * robinsmidsrod uses "daemon" to run the web app
16:17 sri well, you are not using the prefork daemon right?
16:17 sri so it shouldn't be a multi processing problem
16:17 robinsmidsrod how does "daemon" handle concurrency?
16:17 sri non blocking io
16:18 sri so dbi blocks the world
16:18 sri which shouldn't be a problem unless your query takes forever
16:19 robinsmidsrod no, I've got it down to 144ms in the common case now
16:19 robinsmidsrod when I run it from the console - but not from within the web app
16:20 sri people have been using dbi with mojolicious like forever, not heard something like that before
16:22 sri how are you running anyevent in the same process?
16:23 robinsmidsrod I'm not - it's not used from the webapp
16:23 sri ah
16:23 robinsmidsrod perl -MTwitterStream -Ilib -e 'print join("\n", @{ TwitterStream->new->get_links​({precision=>"day",limit=>25, offset=>0, age => 0 }) });' from the console seems to be slow too
16:23 robinsmidsrod but psql < TwitterStream::get_links query is fast
16:24 robinsmidsrod anyevent is just used by the twitter_stream.pl process
16:25 sri ok
16:25 sri no idea then
16:26 sri the global object should only get you in trouble if multiple processes are involved
16:26 robinsmidsrod $ time psql twitter_stream < q.sql  >/dev/null
16:26 robinsmidsrod real    0m0.174s, user    0m0.050s, sys     0m0.020s
16:26 sri not in this case
16:26 robinsmidsrod "global object"? who's global object?
16:27 sri my $ts = ...
16:27 robinsmidsrod you're referring to my $ts object?
16:27 sri DBIx::Connector would fix that i guess
16:27 sri but thats another problem
16:28 esskar_ joined #mojo
16:28 robinsmidsrod I guess I could change my sub _build_dbh {} to use DBIx::Connector, but would it really matter?
16:29 sri not with the daemon
16:29 esskar_ moin
16:30 esskar_ why is "ladder sub { return; }; get '/admin' => 'admin';" still the admin page?
16:30 esskar_ *showing
16:32 sri esskar: if thats the case it would be a bug, can you prepare a unit test? i'll fix it then
16:33 esskar_ i'll try
16:34 robinsmidsrod OMG! when I used placeholders for LIMIT and OFFSET it is wickedly slow, but when I build the sql string with that data inline it is super fast (as expected)
16:34 robinsmidsrod WTF is DBD::Pg doing?
16:36 sri esskar: but i just checked, looks like we already have tests for it in t/mojolicious/lite_app.t
16:37 sri maybe a side effect from something else
16:38 robinsmidsrod sri: not really, afaik - the only thing I'm changing is to use $dbh->prepare() with placeholders for limit/offset and then using $sth->execute($limit,$offset) and then with the $limit/$offset vars inline in the sql string
16:39 sri robinsmidsrod: oh, i meant esskar :) your problem seems to be obviously a bug ;p
16:39 robinsmidsrod maybe my sql statement is just good at pushing the problem, because my sql statement without the limit/offset takes a ridiculous time to execute
16:40 robinsmidsrod which must be what is happening
16:40 robinsmidsrod I guess I'll go and talk to the guys over at #dbi
16:40 * sri nods
16:46 esskar_ sri: http://perlpunks.de/paste/show/4c72a5ea.2926.38
16:46 esskar_ on windows
16:47 esskar_ with Mojolicious 0.999926
16:47 sri umm
16:47 sri thats ancient
16:49 sri of course already fixed
16:49 sri :)
16:49 esskar_ let me check
16:50 sri we don't do 401 out of the box, but it generates 404
16:52 esskar_ http://perlpunks.de/paste/show/4c72a749.36b3.a3
16:53 sri i just ran your test and it worked
16:53 sri with s/401/404/
16:53 esskar_ maybe it's a windows thing
16:53 sri flawless
16:53 sri unlikely, tests seem to pass on windows
16:54 esskar_ i had this behaviour on (the ladder) on a new maschine running strawberry and vista
16:54 esskar_ and on my current maschine runing activestate
16:55 sri ladder is deprecated btw. new name is under
16:55 esskar_ now i use 0.999930
16:55 esskar_ same behaviour
16:55 esskar_ changed 401 to 404
16:55 esskar_ strange
16:56 esskar_ what about => http://perlpunks.de/paste/show/4c72a749.36b3.a3 ?
16:56 sri that is a windows bug
16:56 sri we've been unable to fix it yet
16:57 esskar_ ok
16:57 sri maybe
16:57 sri actually i'm not sure if thats the windows bug
16:58 sri could also be an old bug in an old version of mojolicious
16:58 sri if it is the windows bug t/mojo/client.t would also fail
17:00 sri well, i don't have a windows box, so i depend on volunteers
17:00 esskar_ well, i just got the version from github
17:00 esskar_ and run the tests
17:00 esskar_ then i posted the paste above
17:01 esskar_ sri: where is under implemented?
17:01 sri Mojolicious::Lite
17:02 dotan left #mojo
17:03 perlrocks Twitter: "Would be nice if someone with a windows box could investigate this issue. http://bit.ly/9jfXJ3 #mojolicious #perl" --kraih http://twitter.com/kraih/status/21928544699
17:05 sri i'm pretty much helpless here
17:07 sri it seems to be some kind of read/write timing issue, but those are impossible to track down without a windows box
17:09 sri if nobody steps up i'll just move the tests and disable them on windows ;p
17:10 skaurus maybe I later today
17:10 skaurus for me "today" ends in three hours
17:13 sri http://www.cpantesters.org​/distro/M/Mojolicious.html # obviously a general windows problem
17:17 sri http://github.com/kraih/mojo/blo​b/master/lib/Mojo/IOLoop.pm#L560 # i suspect it could be this line
17:18 sri windows networking code might be too fragile for it
17:19 pid left #mojo
17:20 pid joined #mojo
17:20 esskar_ sri: i got my test to run now.
17:21 sri esskar_: what caused the problem?
17:21 esskar_ sorry
17:21 esskar_ still now
17:21 esskar_ wait
17:24 BlueT_ joined #mojo
17:25 esskar_ well, i changed something on Lite.pm to get it to work
17:26 esskar_ sri, Lite.pm line 82
17:26 esskar_ s/->to($defaults)->name($name)/->via($​methods)->to($defaults)->name($name)/
17:26 esskar_ don't ask me why
17:26 pid left #mojo
17:28 sri that doesn't even make sense
17:28 sri under has no methods associated with it
17:28 sri ever
17:29 esskar_ right
17:29 esskar_ return 1 in ladder makes is buggy then :)
17:29 esskar_ no clue
17:30 sri under is actually quite well tested, this doesn't make sense
17:31 sri maybe broken perl that optimizes away constant closures or whatever...
17:31 sri no idea
17:33 skaurus left #mojo
17:33 esskar_ the ladder/under sub is called (i put a warn "ladder"; before the return)
17:34 koban left #mojo
17:36 esskar_ let me make a linux test
17:48 yko left #mojo
17:52 pid joined #mojo
17:54 esskar_ sri: http://perlpunks.de/paste/show/4c72b59f.5309.183
17:55 sri esskar_: does it fail?
17:55 esskar_ nope
17:55 yko joined #mojo
17:55 sri harmless
17:55 sri we have exception tests too you know ;p
17:56 Htbaa joined #mojo
18:10 esskar_ sri: http://perlpunks.de/paste/show/4c72b982.70d0.d6
18:10 esskar_ i'm on pandol
18:10 esskar_ tested it on linux
18:10 xantus yawn
18:10 esskar_ downloaded newest version from githib
18:12 sri esskar_: what exactly is the bug?
18:12 esskar_ shouldn get / return 200?
18:12 sri no
18:12 sri bad template, look at the log
18:14 esskar_ i dont see it, maybe it's the panadol
18:15 esskar_ fuck
18:16 esskar_ well i changed .epl to .ep
18:16 esskar_ still no go
18:18 esskar_ i give up
18:19 sri ok, which version are you really using?
18:19 sri i get a whole different result with latest github
18:20 esskar_ perl -e 'use Mojolicious; print $Mojolicious::VERSION' => 0.999930
18:21 esskar_ justed installed new 10 minutes ago
18:22 sri for me template detection kicked in and .epl just worked too
18:25 Akron left #mojo
18:26 esskar_ sri: ok
18:26 esskar_ it was line endings bug
18:27 esskar_ but know, i get 200 again for /admin
18:27 esskar_ but i expect 404
18:27 esskar_ i'm running cricles here
18:27 esskar_ s/know/now/
18:30 esskar_ i dont get it
18:31 esskar_ sri: thats the result http://perlpunks.de/paste/show/4c72be81.2270.3a9
18:33 sri esskar_: found a way to replicate it
18:35 arthas left #mojo
18:40 sri and i know why :)
18:43 sri it's actually more of a feature
18:44 sri auto rendering kicks in because you did not render anything
18:44 sri it uses the name of the actual endpoint
18:44 sri which is the action that never ran
18:46 esskar_ sri: what does it mean?
18:47 sri ?
18:47 esskar_ exactly
18:47 esskar_ :)
18:47 sri what i just said :/
18:47 esskar_ what endpoint?
18:48 sri the action that actually matched the route
18:48 sri the under is just something in between
18:49 esskar_ http://perlpunks.de/paste/show/4c72c28e.47a1.2ea
18:49 esskar_ thats the script
18:49 sri i'll just look for a solution, no point explaining in detail
18:49 esskar_ ok
18:49 esskar_ thanks for that
19:05 sri it's quite complicated actually
19:05 sri we bascially don't want bridges to be able to trigger auto rendering
19:08 sri esskar_: should be fixed in http://github.com/kraih/mojo/commit/fa​c3b1edd5bf06b15cde30216f3fc947dc2bfad0
19:09 GitHub8 joined #mojo
19:09 GitHub8 mojo: master Sebastian Riedel * 52dbf4a (6 files in 3 dirs): test cleanup
19:09 GitHub8 mojo: master Sebastian Riedel * 7bf05e3 (2 files in 1 dirs): cleanup
19:09 GitHub8 mojo: master Sebastian Riedel * 52e57b0 (10 files in 5 dirs): cleanup
19:09 GitHub8 mojo: master Sebastian Riedel * 9e3690b (3 files in 2 dirs): cleanup
19:09 GitHub8 mojo: master Sebastian Riedel * 86f24d1 (1 files in 1 dirs): cleanup
19:09 GitHub8 left #mojo
19:12 GitHub141 joined #mojo
19:12 GitHub141 mojo: master Sebastian Riedel * f8743b4 (1 files in 1 dirs): cleanup - http://bit.ly/bq4DJC
19:12 GitHub141 left #mojo
19:22 esskar_ does not look to good
19:23 yko left #mojo
19:25 sri ?
19:25 esskar_ tons of errors now
19:26 sri paste
19:28 esskar_ sri: http://pastebin.com/DN5aDe83
19:29 sri umm
19:29 sri i have not the slightest idea
19:31 sri oh well, maybe it will force people to find the windows bug which :D
19:32 sri *-which
19:33 sri everything exploding is rather unlikely though
19:34 sri it seems to affect all network tests
19:34 sri maybe it's your machine
19:36 pid left #mojo
19:37 Htbaa sri: building on strawberry perl 5.12.1 also gives me failed tests. Where do you want me to put the result?
19:38 Htbaa only 2 failed tests though
19:38 sri some paste site please
19:38 sri 2 failed tests seems like the usual windows bug
19:39 sri client.t and lite_app.t i guess
19:39 Htbaa http://pastebin.com/98qLhFVK
19:39 sri ah that one
19:39 Htbaa lite_app.t only
19:39 sri exactly what i expected
19:40 sri esskar_: just throw your windows box out the window ;p
19:40 Htbaa too bad company protocol demands me developing on Windows, but target Linux. How stupid is that :-)
19:40 sri i doubt anyone will be able to tackle the failing lite_app.t test, so i'll just make it optional
19:41 Htbaa how important are they? (haven't taken a look at it yet though)
19:41 sri it's testing an exception
19:42 Htbaa which isn't being thrown in Windows?
19:42 sri on deployment platforms i consider it not unimportant
19:42 sri but on windows i don't really care ;p
19:42 sri it's a rather harsh test, it's a wonder windows passes the rest
19:43 sri i think it has something to do with the windows kernel send buffer and possibly timing
19:43 sri i even suspect a specific line
19:43 Htbaa who would deploy on Windows anyway huh?
19:43 sri but since i don't have a windows box no debugging
19:44 sri http://github.com/kraih/mojo/blo​b/master/lib/Mojo/IOLoop.pm#L560 # i suspect if that line was commented out it might pass
19:44 janus Htbaa: you don't ask them, they ask you ;P
19:45 * janus has been a greencard for that hell
19:45 sri and possibly the line before that
19:45 janus err, s/been//
19:45 Htbaa I can try that sri
19:45 Htbaa janus: :-)
19:45 sri \o/
19:47 janus but actually, it's easier with mojo than with most (any?) other frameworks
19:48 janus try cat on win... pukadelic!
19:48 sri and i basically didn't touch a windows box during the whole development process :)
19:48 Htbaa I've no issues running Catalyst on Windows. Or most other libraries. It's usually the plugins and some other libraries that have external bindings that have issues
19:49 Htbaa ok now this is funny. I manually downloaded mojo from CPAN and now client.t also gives me a few fails
19:49 sri as long as the github version passes
19:49 sri thats all that matters :)
19:49 janus does here, just pulled :)
19:50 Htbaa but I'm testing 0.999929  from cpan :P, Earlier cpanm fetched this one as well
19:50 sri i would really like to know if commenting out the two lines helps
19:50 Htbaa I'll do a git clone and test it for you
19:50 sri 560 in Mojo::IOLoop
19:50 janus Htbaa: well, it's been over a year ago and was painful for me (iis/fcgi was the biggy btw)
19:51 Htbaa ah you mean deployment. I don't even dare to touch that
19:53 Htbaa sri: lite_app.t passes if I comment out line 560 in Mojo::IOLoop. Even better, the whole test suite passes
19:53 janus especially deployment... the point in time where the project is about to bring in the cash ;)
19:53 Htbaa although I wonder if that actually is 'better' :-)
19:53 sri it is much slower
19:54 sri but as long as it works
19:54 Htbaa I solely deploy Perl apps on Linux. Aside from one wxWidgets PAR based application :-)
19:54 Htbaa All tests successful.
19:54 Htbaa Files=50, Tests=4129, 48 wallclock secs ( 0.89 usr +  0.25 sys =  1.14 CPU)
19:54 Htbaa Result: PASS
19:54 sri \o/
19:54 sri thanks, that helped a lot
19:55 Htbaa no problem
20:02 GitHub61 joined #mojo
20:02 GitHub61 mojo: master Sebastian Riedel * c361b6f (2 files in 2 dirs): fixed Mojo::IOLoop Windows support - http://bit.ly/b3nD7A
20:02 GitHub61 left #mojo
20:02 Htbaa lol @ http://cpanratings.perl.org/dist/Mojolicious - It also makes you pie. :-)
20:02 sri :)
20:02 sri we need more ratings btw!
20:03 Alias left #mojo
20:03 Htbaa haven't used it enough yet to do any ratings :-)
20:10 marcus Where's my pie? :/
20:11 * xantus hits marcus in the face with a pie
20:11 * marcus licks cream
20:13 xantus wtf, 1 of 3 found that rating helpful
20:13 * xantus clicks yes
20:13 * marcus clicked yes too
20:13 marcus so now it should be 3 of 5
20:23 xantus http://cpanratings.perl.org/dist/Mojolicious
20:23 xantus added mine
20:24 sri \o/
20:24 sri xantus++
20:32 xantus maybe I should have put a gist link in there
20:36 skaurus joined #mojo
20:36 skaurus sri: btw, did you read my yesterday message about strange behaviour with templates? (tt and ep1)
20:37 sri skaurus: no
20:39 skaurus sri: i don't sure if its bug, but it's unexpected behaviour. Starting position: default mojolicious. Then I add TT renderer, .tt template and rename .ep template to .ep1 to get rid of it. Nothing renders... When i remove .ep1, .tt starts rendering
20:40 sri so remove the spare template
20:41 sri i don't think we verify that a renderer exists when looking for a template
20:42 skaurus sri: if i correctly understand you yesterday ("skaurus: I thought Mojolicious take all available template renderers, sorted them in some way and then tries to find template in on of these formats      sri: thats exactly what it does") mojo should look to templates ep1, tt, think "oh, i don't know about ep1, but i have tt renderer..." and render ut
20:43 sri dunno
20:43 skaurus can it be feature request?
20:44 sri a feature request is specific and contains a failing test case
20:44 skaurus so i should add new test if i wish to change current behaviour?
20:44 sri thats the usual way
20:45 skaurus ok, thank you
20:45 skaurus i have a feeling i'm already annoy you)
20:48 sri not at all, useless bug reports annoy me, not reasonable questions :)
20:51 skaurus cool :)
20:51 xantus useless bug reports is my middle name!
20:53 xantus looks like I'm going to stick stay here for a little longer.  I'm going to st louis for a few weeks
20:53 sri i think we still have a line ending bug in the DATA template code that might affect template detection on windows
20:53 xantus s/stick//
20:53 sri no idea how to verify/test it though
20:54 sri xantus: what made you change your mind?
20:55 sri btw. Mojo::Server::Hypnotoad yay or nay?
20:56 skaurus everybody loves hypnotoad, but how it correlate with forking?
20:56 sri it doesn't at all
20:56 skaurus should be some mnemonic...
20:56 sri how does unicorn correlate to foriking?
20:56 sri *-i
20:57 skaurus in no way
20:57 sri or nginx :)
20:57 skaurus EngineX :)
20:57 sri or apache
20:57 skaurus what is unicorn?
20:58 sri ruby server
20:58 skaurus ))
20:58 sri http://tomayko.com/writings/unicorn-is-unix
20:58 sri the image is pretty cool
20:59 skaurus perl6 and butterfly...
20:59 skaurus or perl5 and onion
20:59 sri butterfly can look good
20:59 skaurus ok, i agree :)
21:00 sri or Tornado
21:00 xantus sri: had a little meeting, talked things out
21:00 janus hypnofly
21:00 skaurus how about south park?
21:01 skaurus Moses
21:01 janus now that's forking :)
21:01 sri lol
21:01 skaurus http://static.tvfanatic.com/imag​es/gallery/south-park-moses.gif
21:02 sri only simpsons and futurama
21:02 sekimura left #mojo
21:04 skaurus hypnotoda, devil robot, brain slugs, ...
21:22 vel left #mojo
21:41 stephen left #mojo
21:42 perlite left #mojo
21:42 perlite joined #mojo
21:42 stephen joined #mojo
21:51 xantus all glory to the hypnotoad!
21:59 Htbaa left #mojo
22:01 ltriant joined #mojo
22:04 dotan joined #mojo
22:06 sri actually i'm going to use an external loader for the new prefork daemon
22:07 sri hypnotoad myapp.pl
22:07 sri it will default to production mode
22:07 sri and be tuned for production setups
22:08 sri only options are --daemonize, --config and --help
22:08 sri everything else is in the config file
22:08 sri (or the app)
22:08 sri oh btw.
22:09 sri ALL GLORY TO THE HYPNOTOAD!
22:16 fod joined #mojo
22:16 marcus hypnotoad is a good name
22:16 dotan left #mojo
22:24 MojoGuest670 joined #mojo
22:24 MojoGuest670 From: http://www.google.com/search?q=websocket+demo&​amp;ie=utf-8&amp;oe=utf-8&amp;aq=t&amp;rls=org​.mozilla:en-US:official&amp;client=firefox-a (1 hits)
22:24 MojoGuest670 left #mojo
22:25 MojoGuest433 joined #mojo
22:25 MojoGuest433 From: http://www.google.com/search?q=websocket+demo&​amp;ie=utf-8&amp;oe=utf-8&amp;aq=t&amp;rls=org​.mozilla:en-US:official&amp;client=firefox-a (3 hits)
22:25 MojoGuest433 is now known as individual
22:26 individual So does this websocket thing really work?
22:26 xantus it
22:26 xantus oops, it uses an http upgrade
22:27 xantus the browser has an open connection to the web server
22:28 xantus so the server can just send data back at any time
22:29 individual And the HTTP server is proxying to the irc network
22:32 xantus yes
22:33 xantus so the client does the parsing of the irc protocol
22:33 xantus thats my demo btw
22:33 xantus if you need the code its on github
22:33 individual I had my suspicions. :P
22:33 xantus http://github.com/xantus/mojo-websocket-examples
22:34 xantus yeah, the url gives it away
22:45 individual left #mojo
23:17 sri damn, textmate with missingdrawer is cool
23:42 MojoGuest952 joined #mojo
23:42 MojoGuest952 From: http://www.google.com/search?hl=en&amp;client=safa​ri&amp;rls=en&amp;q=websockets+demo&amp;aq=0&amp;a​qi=g1&amp;aql=&amp;oq=websockets+de&amp;gs_rfai= (1 hits)
23:43 MojoGuest952 left #mojo

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