Camelia, the Perl 6 bug

IRC log for #mojo, 2011-02-19

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

All times shown according to UTC.

Time Nick Message
00:09 marty left #mojo
01:23 perlrocks Twitter: "Mojolicious:: Lite is worth attending just to gain the material! Thank you. # Hokkaidopm" (ja) --tyamaguc07 http://twitter.com/tyamaguc​07/status/38770576571113472
01:24 Akron left #mojo
01:34 tholen_ joined #mojo
01:37 perlrocks Twitter: "Current projects @ tyamaguc07 Mojolicious:: Lite flowed I was also told to use. In that article to please w" (ja) --ABE_TAKASHI http://twitter.com/ABE_TAKAS​HI/status/38774229163966464
01:38 tholen left #mojo
01:42 stephen left #mojo
02:06 s0va_ joined #mojo
02:07 s0va_ hello there
02:07 s0va_ left #mojo
02:07 s0va joined #mojo
02:07 s0va hello :)
02:08 s0va is --listen daemon option broken in 1.1.x?
02:08 s0va i get this
02:08 s0va Can't use string ("https://*:3000") as an ARRAY ref while "strict refs" in use at /export/home/bfg/git/mojo/​lib/Mojo/Server/Daemon.pm line 62
02:09 s0va if i change line 62 to from $self->listen ||    to [ $self->listen ] seems to work again
02:14 mobyu joined #mojo
02:24 bigbootay joined #mojo
02:39 mobyu left #mojo
02:49 toei left #mojo
03:01 toei joined #mojo
03:35 meraxes left #mojo
03:45 meraxes joined #mojo
04:03 tl joined #mojo
04:05 tholen42 left #mojo
04:30 tempire s0va: should be an arrayref: (["https://*:3000"])
04:55 perlrocks Twitter: "Mojolicious-Plugin-GroupedParams 0.02 by IVSOKOLOV - http://frepan.org/~ivsokolov/Mojol​icious-Plugin-GroupedParams-0.02/" --cpan_new http://twitter.com/cpan_ne​w/status/38823972615229440
05:03 tholen_ left #mojo
05:03 kimoto left #mojo
05:04 kimoto joined #mojo
05:18 kimoto left #mojo
05:42 kimoto joined #mojo
05:49 bigbootay left #mojo
05:55 kimoto left #mojo
07:22 perlrocks Twitter: "What a popular Mojolicious (# hokkaidopm live at http://ustre.am/lW0e)" (ja) --hisaichi5518 http://twitter.com/hisaichi5​518/status/38860875972485121
07:22 perlrocks Twitter: "Try using mojolicious test. # Hokkaidopm" (ja) --chao2suke http://twitter.com/chao2suk​e/status/38860946818596864
07:22 perlrocks Twitter: "Mojolicious done in the morning served in the case of LT # hokkaidopm" (ja) --aloelight http://twitter.com/aloeligh​t/status/38860979152494593
07:22 kaare joined #mojo
07:23 perlrocks Twitter: "Looks like a good Mojolicious" (ja) --janus_wel http://twitter.com/janus_we​l/status/38861305716674560
07:32 sri good morning sunshines
07:46 non joined #mojo
07:53 tholen joined #mojo
08:09 MojoGuest503 joined #mojo
08:09 MojoGuest503 From: http://news.ycombinator.com/item?id=1277067 (249 hits)
08:10 MojoGuest503 left #mojo
08:25 MojoGuest441 joined #mojo
08:25 MojoGuest441 From: http://news.ycombinator.com/item?id=1277067 (250 hits)
08:26 MojoGuest441 left #mojo
08:36 MojoGuest23 joined #mojo
08:36 MojoGuest23 From: http://twitter.com/ (5 hits)
08:38 MojoGuest23 left #mojo
08:45 memememomo_ is now known as memememomo
08:45 memememomo is now known as memememomo_
08:56 mobyu joined #mojo
09:03 giwi left #mojo
09:12 aiglenoir left #mojo
09:18 aiglenoir joined #mojo
09:21 mobyu left #mojo
09:36 sri hmm, limechat for iphone is a lot better than colloquy actually
09:42 gabb morning guys
09:42 * sri waves
09:42 gabb I see mojo is coming along nicely sri :)
09:42 sri indeed it is :)
09:49 mobyu joined #mojo
09:52 tholen left #mojo
10:02 mobyu left #mojo
10:07 perlrocks Twitter: "@sharifulin Starting to praying after updating #mojo from #github and understand that backward compatibility matters. #Perl #Mojolicious" --su_bzero http://twitter.com/su_bzer​o/status/38902516003381248
10:17 crab hi.
10:22 aiglenoir left #mojo
10:23 marcus sri: agree
10:23 marcus sri: I use limechat on the iphone/ipad and the mac
10:23 aiglenoir joined #mojo
10:27 tholen joined #mojo
10:28 perlrocks Twitter: "@su_bzero mojolicious 1.x hasn't backward compability with 0.9 or smaller" --sharifulin http://twitter.com/shariful​in/status/38907809869471744
10:28 aiglenoir left #mojo
10:29 aiglenoir joined #mojo
10:31 aiglenoir left #mojo
10:33 perlrocks Twitter: "I thought was strange behavior Mojolicious use Mojo:: Base-base not working like syntax. Installation from the PPM?" (ja) --choplin http://twitter.com/choplin​/status/38909118890643457
10:34 aiglenoir joined #mojo
10:39 sherr left #mojo
10:39 sherr joined #mojo
10:46 sri su-bzero: you do realize that github is bleeding edge right?
10:47 sri su-bzero: while the github repo has been rather stable there are no guarantees at all
10:48 sri su-bzero: also, if you're worried about backwards compatibility you should never use anything marked experimental, undocumented or untested!
11:25 GitHub4 joined #mojo
11:25 GitHub4 mojo: master Sebastian Riedel * 4465c59 (3 files in 2 dirs): fixed small documentation bug - http://bit.ly/hzsUVy
11:25 GitHub4 left #mojo
11:33 plu left #mojo
11:35 fhelmber_ joined #mojo
11:42 kimoto joined #mojo
11:43 su-bzero sri: Ok, no problems. I just realize that mojo is in a high development mode... So we should be very careful. Yes, thank you for your point.
11:45 sri su-bzero: if you update to the released version of 1.11 you'll notice that ->async is bug for bug backwards compatible, and just deprecated now
11:48 kimoto left #mojo
11:52 * crab listens to the purple sunbirds singing outside his window
12:19 tholen left #mojo
13:20 breaker313 joined #mojo
13:23 s0va sri:
13:23 s0va question.
13:23 purl question. is &bar a closure? perl -le '{ sub foo { my $foovar="th0th rulz!"; sub bar { print "bar sez: @_; also, $foovar"; } } } &foo; &bar ("howdy");'
13:23 s0va mojo 1.1, ./app daemon --listen "http://localhost:3000" gives me not arrayref error
13:24 s0va has command line format changed or what?
13:24 s0va otherwise datemon, line 62, $self->listen  ... => [ $self->listen ] and works again
13:24 sri s0va: i don't get that error
13:25 s0va one sec.
13:25 purl time's up!
13:25 sri ./script/mojo daemon --listen http://*:4000
13:25 sri simple test
13:25 purl simple test is to see if setuid works.
13:25 * gabb hugs purl
13:25 * purl nibbles gabb's elbow
13:25 sri or actually just "mojo daemon --listen http://*:4000"
13:26 sri or even "mojo daemon --listen http://*:4000 --listen http://*:5000" for multiple
13:27 s0va wtf?!
13:29 s0va ok, this is weird. i got this error last night
13:29 s0va but now is gone.
13:29 s0va i feel stupid.
13:31 s0va sri: https://github.com/bfg/anyevent-mojolicious  <= daemon works, client should work, but i didn't test it, your ioloop tests fail, but i'm goint to fix issues.
13:32 s0va i just want to use AnyEvent::FastPing in mojo apps :)
13:32 mobyu joined #mojo
13:33 s0va performance: your loop implementation powered by epoll has lower latency, anyevent implementation can push the same or a little more reqs/sec...
13:34 sri s0va: interesting
13:34 s0va as i said, it's not my motivation to replace anything in mojo. i just want to use all the mojo goodies in anyevent environment
13:35 sri big performance difference would have been a surprise
13:35 s0va sri: it's normal that latency is bigger: on_read for example: i get read chunk, i get the buffer and i need to forward it by calling another sub
13:36 s0va sri: i agree :) i'm suprised that everything works so fast using anyevent ... for each io event two subs are called.
13:37 s0va but ipv6 and tls works out of the box, which is really sweet :)
13:37 mobyu left #mojo
13:38 s0va sri: question: i'm planning implementing class which would start mojo daemon in non mojo started app.
13:38 sri tls works out of the box with ioloop too
13:38 s0va i guess that it is completely doable to run multiple daemon instances in single process, isn't it?
13:38 sri if you have the optional prereq installed
13:39 sri we have unit tests running multiple daemons and clients inside the same process
13:39 s0va sri: true, never said that it doesn't; you're just very fascistic about io::socket::ssl version :) but i don't blame you
13:39 s0va ok, great, i'll check them out.
13:39 s0va now, the real question...
13:39 purl i guess now, the real question is whether i'd damage an ibook by leaving it turned on inside a bag for a bit
13:39 sri we have to be, lots of bugs in older versions
13:40 sri we are prolly the first to do non-blocking tls with it
13:40 s0va is it possible to run multiple daemon instances in single perl process each one running different mojo app?
13:40 sri sure
13:40 sri nothing unique about mojo apps
13:41 s0va sri: i just saw few accourances of $ENV{MOJO_APP} in source...
13:41 perlrocks Twitter: "just released Mojo-Logite module on CPAN #mojolicious" --areggiori http://twitter.com/areggior​i/status/38956300020752384
13:41 sri right, detection hacks
13:41 mobyu joined #mojo
13:42 sri there are more to make detection stuff work with plack
13:42 s0va hm :)
13:42 sri i hate them
13:42 sri but not much of a choice
13:42 s0va :) i was also wondering about creating some glue between mojo app and twiggy...
13:43 s0va but twiggy is not as cool as mojo::server::daemon
13:49 sherr left #mojo
14:09 marcus hmm, I really like webos
14:09 marcus too bad there isn't any phones with it on in norway
14:09 marcus I think it would have been much smarter for nokia to partner with HP
14:11 sri nobody takes webos seriously
14:12 sri hp is just not the company to make full stack work
14:13 marcus webos is way nicer than android and windows phone 7
14:13 marcus been playing with the simulator for a while now
14:13 marcus I think their tablet device will kick ass
14:13 sri hp will still somehow screw it up :)
14:13 sri i give it one maybe two years until they switch to android or wp7
14:15 marcus new webos even bundles node.js ;)
14:15 marcus their IDE is really nice too.
14:17 sri that's all meaningless, hp is incapable of growing an ecosystem around webos
14:18 mobyu left #mojo
14:19 marcus sri: it could be a refugee platform for apple developers running away from draconian TOS :p
14:20 marcus specially if they let their libraries be used in other platforms as well.
14:20 sri at the end of the day ios developers still make the most money
14:21 marcus unless they make ebook readers or streaming music players :p
14:21 sri poor amazon
14:21 marcus poor spotify
14:22 gabb webos tablet looks rather promising tbh
14:22 * sri bonks garu on the noggin
14:22 sri dreamhost guide is still missing from the wiki!
14:23 * sri cracks the whip
14:23 gabb marcus: was it you who tweeted Waiting for Superman?
14:24 gabb movie was awesome
14:26 marcus gabb: I don't think so. It's in my to see queue
14:30 sri isn't that movie about the american education system?
14:31 sri sounds rather uninteresting for europeans
14:32 sherr joined #mojo
14:34 marcus damnit, my plex installation refuses to update it's plugins
14:34 marcus Cowardly, I might add.
14:35 perlrocks Twitter: "Mojolicious Web Framework - Join the Perl revolution! http://bit.ly/eHiB7K" --delicious50 http://twitter.com/delicious​50/status/38969913393549312
14:37 gabb sri: yes, but its mindboggling, which makes it interesting, at least to me - maybe because I am kinda in edu atm
14:45 marcus yay, got it!
14:52 breaker313 left #mojo
14:54 jwang joined #mojo
15:38 perlrocks Twitter: "Perl took the time to pick up the error. . JSON old version of the module would work, did not render_json Mojo's. ." (ja) --hid_tgc http://twitter.com/hid_tgc​/status/38985666331152384
15:50 perlrocks Twitter: "I am writing to Mojolicious, listening Mujuice, both super:) # mojo # perl" (uk) --mshershnev http://twitter.com/mshershn​ev/status/38988805918826496
15:58 perlrocks Twitter: "Mojolicious-Plugin-CSRFDefender 0.0.5 by SHIBAZAKI - http://frepan.org/~shibazaki/Mojol​icious-Plugin-CSRFDefender-0.0.5/" --cpan_new http://twitter.com/cpan_ne​w/status/38990938223288321
16:03 julien joined #mojo
16:37 yko Moby announces new album!
16:37 yko oops, it's not #moby room
17:30 julien hi there
17:30 julien is there a way to have a configuration file that is not parsed at each request ?
17:31 julien like a persistent stash or something similar....
17:32 sri which deployment environment?
17:34 julien hypnotoad
17:34 purl it has been said that hypnotoad is at http://www.r33b.net/ or ALL GLORY TO THE HYPNOTOAD!
17:34 yko julien: http://mojolicio.us/perldoc?Mojolicious#defaults
17:35 sri yko++
17:35 yko also see http://mojolicio.us/perldoc?M​ojolicious/Plugin/JsonConfig
17:36 yko and it's stash key => config is stored into stash with that key
17:36 sri basically everything in your application instance is persistent
17:36 s0va sri: can you explain in a sentence: how does daemon start the app?
17:36 sri s0va: nope
17:38 s0va dammit. i'll have to read the code then.
17:38 sri yea, it goes through multiple mayers
17:38 sri *layers
17:39 julien sri: so if I modify a stash value while processing a request (hit counter for exemple), it will be persistent ?
17:39 sri no
17:39 * yko spent holydays once to get how does hypnotoad works
17:39 sri application instance, not controller
17:39 sri controller is request scope
17:39 julien ok
17:40 sri such a hit counter wouldn't work with hypnotoad in any case though
17:40 sri since it maintains a worker pool
17:41 sri at that point you need a database
17:41 julien and they don't share memory ?
17:42 sri copy on write
17:42 purl copy on write is not copying values before it is absolutely necessary.
17:43 julien ok, so as soon as I updated the counter, the new value is worker specific and the other workers don't have knowledge of it
17:43 crab something like that
17:48 MojoGuest604 joined #mojo
17:48 MojoGuest604 is now known as korshak
17:51 julien left #mojo
17:52 su-bzero left #mojo
17:52 julien joined #mojo
17:52 yko left #mojo
17:59 korshak left #mojo
18:00 yko joined #mojo
18:29 su-bzero[off] joined #mojo
18:29 yko_ joined #mojo
18:31 punytan left #mojo
18:33 yko left #mojo
18:33 yko_ is now known as yko
18:34 su-bzero[off] is now known as su-bzero
18:44 stephan48 left #mojo
18:48 s0va julien: use cache::mmap
18:48 s0va julien: http://search.cpan.org/~pm​h/Cache-Mmap-0.11/Mmap.pm
18:49 s0va i would subclass cache::mmap and provide singleton method.
18:49 stephan48 joined #mojo
18:50 crab use memcached and don't mess around with locking.
18:50 crab or use a database.
18:55 julien left #mojo
19:31 s0va sri: last test in ioloop.t expects IO::Socket::INET object from handle method...
19:31 s0va my implementation returns glob.
19:32 sri yea, added that test because we've had people depend on it :(
19:32 s0va that's why test fails. hm, i don't have any idea gow to convert glob reference (plain socket) to IO::Socket::INET
19:32 s0va ideas?
19:32 purl no idea, s0va...
19:32 sri no ideas
19:33 s0va http://cpansearch.perl.org/src/​GBARR/IO-1.25/lib/IO/Socket.pm
19:33 sri you can prolly just bless it into the class or so
19:33 s0va my $sock = $class->SUPER::new();  ${*$sock}{'io_socket_timeout'} = delete $arg{Timeout}; fuck ...
19:33 sri oh right, there is some magic in IO::Socket
19:34 sri we've been bitten by that with IO::Socket::IP
19:34 s0va is there any illegal way to write something like: $obj{*$sock} = $my_real_sock ?
19:35 sri you can only try i guess, or google
19:35 s0va well... i could just change isa_ok requirement :)
19:36 sri quite sure the tls tests have a similar requirement
19:36 s0va oh, look there is IO::Handle new_from_fd method
19:36 s0va let's see if it works :>
19:36 sri :)
19:37 s0va new_from_fd ( FD, MODE ) ... fuck, what mode should socket be in? 'w+'?
19:40 s0va sockets should be read/write, huh?
19:43 s0va Files=4, Tests=12,  4 wallclock secs ( 0.02 usr  0.01 sys +  1.26 cusr  0.42 csys =  1.71 CPU)
19:43 s0va Result: PASS
19:43 s0va YAAAY!!! :)
19:44 s0va return IO::Socket::INET->new_from_fd($plain_socket, 'r+') does the job
19:47 punytan joined #mojo
19:59 crab what on earth are you doing?
20:03 espent left #mojo
20:03 espent joined #mojo
20:08 GitHub66 joined #mojo
20:08 GitHub66 mojo: master Sebastian Riedel * f0b4abd (5 files in 5 dirs): added drain callback support for WebSockets - http://bit.ly/hQ4ov5
20:08 GitHub66 left #mojo
20:19 sri oh, i would have expected new_from_fd(fileno $plain_socket, 'r+')
20:21 julien joined #mojo
20:54 fhelmber_ left #mojo
20:57 Akron joined #mojo
21:14 kaare left #mojo
21:41 MojoGuest732 joined #mojo
21:41 MojoGuest732 From: http://www.google.com/url?sa=​D&amp;q=http://dev.xantus.org (3 hits)
21:41 MojoGuest732 left #mojo
22:48 jwang_ joined #mojo
22:55 fhelmber_ joined #mojo
22:55 jwang_ left #mojo
22:56 fhelmber_ left #mojo
22:56 fhelmbe__ joined #mojo

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