=== robins is now known as robinsmidsrod [08:36] morning! [08:37] I'm still looking for a solution to start a script when tty1 is in state "starting" [08:37] _and_ if installed/active when gdm/kdm is in state starting [08:38] but I would like do add only a new upstart script, and not modify other scripts [08:39] I have a condition that depends on tty and gdm/kdm, but if kdm/gdm is not running this is not working [08:39] my script should run before a user can login [08:40] and we have systems with *dm and without [12:31] ralfgro: start on starting tty1 or gdm doesn't work? [12:31] ralfgro: oh you need to run *after* those? [18:27] I wrote my upstart script, it starts up fine. It shuts down my services fine but it hangs after it shuts down my program [18:27] http://pastebin.com/aUg5ssm5 [18:27] The script actually starts 3 threads. Please help [18:29] Do not use ‘expect’ if you’re not absolutely sure of the forking behavior of the main process. You’ll confuse Upstart. [18:30] ok [18:30] The service script that is called calls three different scripts with & at the end to fork them. Is that why upstart is confused ? [18:32] That definitely doesn’t match what ‘expect daemon’ means. [18:34] oh ok [22:33] keybuk: Do you know how Erlang implements trivially easy and robust runtime code upgrades? I’ve thought a bit of how something similar could be done in C. [22:39] In fact, i’ll try to get around to writing an email to the Upstart mailing list, trying to articulate my general idea. [22:47] I don't know much about Erlang === Keybuk changed the topic of #upstart to: Upstart 1.0 "This is a fertile land, and we will thrive" | http://upstart.at/ [23:01] Woot, that was quick. Is that with proc connector goodness? [23:01] Ah, the release message doesn’t seem to say so. [23:05] no, that's me declaring that the version that's been stable for years is 1.0 ;-)