[13:39] <Keybuk> \o/ Colin committed my D-Bus patch
[15:23] <Keybuk> sadmac2: does Fedora have Python 2.5?
[15:24] <AlexExtreme> Keybuk, it does
[15:27] <Keybuk> cool
[16:43] <keesj> Keybuk: d-bus support in upstart?
[16:45] <keesj> btw. I have some 'theory' questions
[16:45] <keesj> I have two jobs who "start on startup". wich is started first
[16:46] <keesj> I have this behaviour where the second jobs only gets started once the first is "done"
[16:49] <keesj> hmm .let me first think , this sounds like two questions
[17:09] <Keybuk> keesj: yeah, I'm playing with D-Bus at the moment 
[17:10] <Keybuk> both jobs started by the same event will be started at the same time
[17:36] <keesj> so two forks one after the other?
[17:40] <Keybuk> right
[18:04] <keesj> hmm, my scripts are quite large and I don't have  dev/fd/  entries. spawing the first /bin/sh migh also take some time. 
[18:04] <keesj> the basic idea is "main -> event_poll -> job_handle_event -> job_change_goal("start") -> job_change_state
[18:04] <keesj> "
[18:08] <keesj> and job_change_state will call job_run_process . 
[18:09] <keesj> I have no clue how big the defailt buffers are 
[18:14] <Keybuk> err, me neither ;)
[18:14] <Keybuk> a page I think
[18:14] <Keybuk> I probably picked something sensible like that
[18:14] <Keybuk> it's all handled properly anyway
[18:14] <Keybuk> if you don't have /dev/fd, it'll be passed on the command-line
[18:14] <Keybuk> in which case you'll hit the maximum command line length before anything else
[18:14] <Keybuk> that's quite small
[18:18] <keesj> food time :p