Camelia, the Perl 6 bug

IRC log for #mojo, 2012-12-02

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

All times shown according to UTC.

Time Nick Message
01:00 b2b joined #mojo
01:16 Mike-PerlRecruiter_ joined #mojo
01:59 ObseLeTe joined #mojo
03:04 asarch joined #mojo
03:07 noganex_ joined #mojo
03:21 mattastrophe joined #mojo
03:50 mattastrophe joined #mojo
03:57 Miked joined #mojo
04:41 xaka joined #mojo
05:23 Foxcool joined #mojo
06:44 marcus sri: do you really need separate timeouts per connection/session inside a spdy connection? I'd think the timeout belongs on the physical connection even if it's multiplexed.
07:14 Vandal joined #mojo
07:30 jwang joined #mojo
07:33 ladnaV joined #mojo
07:38 Vandal joined #mojo
07:43 sri marcus: not sure, maybe
08:07 ladnaV joined #mojo
08:22 d4rkie joined #mojo
09:15 Vandal joined #mojo
09:25 ladnaV joined #mojo
09:38 d4rkie joined #mojo
09:46 dotan joined #mojo
09:47 d4rkie joined #mojo
09:52 marcus_ joined #mojo
09:54 marcus_ joined #mojo
09:56 marcus_ joined #mojo
09:57 sri marcus: it would of course make everything a lot easier
09:58 marcus sri: I think it makes sense too, not sure why you would need separate timeouts for logical channels.
09:58 sri marcus, tempire, crab: you think it would be ok to break client_* and server_* methods in the trabsaction classes before 4.0?
09:59 sri they are only useful for implementing user agents and servers
09:59 marcus sri: you want to change their behavior or remove them?
10:00 sri move them to Mojo::Channel classes
10:00 sri they are basically the protocol implementations
10:00 marcus sri: so can't you just deprecate them and leave a warning alias for some releases?
10:00 sri you mean as dead code?
10:01 sinkovsky joined #mojo
10:01 sri hard to say if they would still work afterwards
10:01 marcus warn(" … has been deprecated, please use channel->client_* instead"); $self->channel->client_* ...
10:01 sri they have no direct tests
10:02 sri no, it won't work that way
10:02 dpetrov_ joined #mojo
10:02 marcus I'm sure we'll hurt *someone*'s feeling, but the breakage seems rather minor.
10:02 sri behavior of the new methods will be different
10:02 marcus ah, right.
10:03 sri we could leave them as dead code i suppose, and hope for the best
10:03 sri channels will be server/client specific too
10:04 sri Mojo::Channel::HTTP::Client
10:04 marcus sri: seeing as the spdy/multiplexing bit is new, we don't have to wory about how it works for them.
10:04 marcus only for the http client and maybe web socket.
10:05 sri i think the HTTP2 channel will have multiple HTTP channel children or so, kind of a meta channel
10:06 marcus friendship is magic :)
10:07 marcus_ joined #mojo
10:10 marcus_ joined #mojo
10:13 marcus_ joined #mojo
10:15 marcus_ joined #mojo
10:16 marcus_ joined #mojo
10:18 marcus_ joined #mojo
10:20 marcus_ joined #mojo
10:22 marcus_ joined #mojo
10:24 marcus_ joined #mojo
10:27 marcus_ joined #mojo
10:27 sinkovsky joined #mojo
10:29 marcus_ joined #mojo
10:31 marcus_ joined #mojo
10:33 marcus_ joined #mojo
10:35 marcus I'm starting to like sourcetree quite a bit.
10:37 marcus_ joined #mojo
10:39 marcus_ joined #mojo
10:41 marcus_ joined #mojo
10:44 marcus_ joined #mojo
10:44 DaTa marcus: stop it
10:44 marcus DaTa: stop what?
10:45 marcus_ left #mojo
10:46 DaTa marcus: that marcus_ joining and quitting all the time
10:46 marcus DaTa: ah. I just ignore joins/quits
10:46 marcus DaTa: it's just because it reconnects each time morbo reloads the app on file change. Parted #mojo now.
10:58 sinkovsky joined #mojo
11:01 dotan joined #mojo
11:27 batman left #mojo
11:30 batman joined #mojo
11:44 suy_ joined #mojo
12:11 sri https://github.com/kraih/mojo/issues/423 # i've opened a new future ticket for HTTP/2
12:14 sri i have no immediate plans to start working on the refactoring, perhaps it's a good task for someone who wants to get to know mojolicious internals
12:15 crab hi sri.
12:15 sri o/
12:22 sri the channel idea is stolen from netty btw. their abstraction is crazy, but i like the general design for protocol abstraction
12:22 sri i'm open for better abstraction ideas btw.
12:23 sri anything that makes adding new protocols easier is fine
12:32 sri maybe a task for a really good summer of code student too
12:44 b2b joined #mojo
12:45 nic In a full app, trying to do authentication, have $r->bridge->to('auth#check')
12:45 nic but it renders nothing
12:45 nic not a sausage
12:46 nic My hunch is that it's cos my bridge is too general
12:46 nic wondering if I need a pattern on it to match 'anything except auth'
12:48 nic In all the examples I've seen, the secure area is a separate path component (eg /admin)
12:49 nic but I need to cherry pick routes at the top level
12:53 Vandal joined #mojo
13:04 nic yeah, maybe that was the problem; changed it to $r->bridge->to(cb => sub {... and it now works fine
13:15 Mike-PerlRecruiter_ joined #mojo
14:03 b2b joined #mojo
14:16 [R] joined #mojo
14:21 [R] joined #mojo
14:27 Britzel joined #mojo
14:36 Vandal joined #mojo
14:38 mire__ joined #mojo
14:43 [R] joined #mojo
15:09 l3l1p joined #mojo
15:15 kitt_vl joined #mojo
15:28 ladnaV joined #mojo
15:30 niczero joined #mojo
15:32 niczero for the record, https://github.com/byterock/Mojo-Aut​hentication-Authorization-examples/ proves I was wrong above
15:34 gryphon joined #mojo
15:37 Vandal joined #mojo
15:48 sri hmmmm
15:48 sri if anyone wants to work on Mojo::Channel, i think the best refactoring strategy would be to start with Mojo::Server::Daemon and move all transaction management first
15:49 sri and only once that works move the server_* methods
15:54 d4rkie joined #mojo
16:05 sri it's a little complicated for the user agent, i wonder if there's another refactoring step that should be made there first
16:05 sri perhaps factor out connection management completely
16:06 alexeiras joined #mojo
16:11 nicomen joined #mojo
16:22 sri oh, firefox seems to call it channels too :)
16:24 sri http://hg.mozilla.org/mozilla-central/file/0352a​32fde64/netwerk/protocol/http/nsHttpChannel.cpp # yea... that's about the size of the entire mojolicious distribution :D
16:43 Vandal joined #mojo
17:13 dabudabu joined #mojo
18:13 ladnaV joined #mojo
18:17 sh4 joined #mojo
18:34 dpetrov_ joined #mojo
19:10 batman left #mojo
19:11 batman joined #mojo
19:15 b2b joined #mojo
19:19 marty joined #mojo
19:46 sinkovsky joined #mojo
19:52 rem_lex| joined #mojo
19:58 xaka joined #mojo
20:10 sh4|2 joined #mojo
20:17 b2b left #mojo
20:18 b2b joined #mojo
20:21 lukep joined #mojo
21:11 mire__ joined #mojo
21:23 sinkovsky joined #mojo
21:52 good_news_everyone joined #mojo
21:52 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/5m98CQ
21:52 good_news_everyone mojo/master 2a9d917 Sebastian Riedel: added recipe for rendering static files manually
21:52 good_news_everyone left #mojo
21:55 perlite_ joined #mojo
21:58 alexeiras joined #mojo
22:03 gryphon joined #mojo
22:12 good_news_everyone joined #mojo
22:12 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/OVtFuA
22:12 good_news_everyone mojo/master c4627f5 Sebastian Riedel: added custom response recipe
22:12 good_news_everyone left #mojo
22:22 good_news_everyone joined #mojo
22:22 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/GP1HdA
22:23 good_news_everyone mojo/master 8dc2915 Sebastian Riedel: tweaked custom response recipe
22:23 good_news_everyone left #mojo
22:49 atz joined #mojo
22:56 sinkovsky joined #mojo
23:08 Annihitek joined #mojo
23:13 jzawodn joined #mojo
23:17 sinkovsky joined #mojo
23:34 kitt_vl joined #mojo
23:48 lukep_ joined #mojo

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