| * xnox ponders if jodh is lost in testing 1.6.1 | 15:07 | |
| === xnox changed the topic of #upstart to: Upstart 1.6.1 | http://upstart.ubuntu.com/cookbook/ | Post to mailing list if no response here: http://bit.ly/ooeNVv | ||
| xnox | slangasek: https://bugs.launchpad.net/libnih/+bug/672643/comments/2 <--- is your bug. | 15:22 |
|---|---|---|
| xnox | with a patch. | 15:22 |
| slangasek | xnox: sweet, so it was xfs :) | 17:03 |
| xnox | slangasek: i'm not sure which branches we can merge that patch. | 17:04 |
| xnox | slangasek: well there are other "popular" filesystems that are affected. | 17:04 |
| slangasek | xnox: packaging branches only, AIUI Scott is still upstream for libnih | 17:04 |
| xnox | slangasek: ack. but it does mean that potentially upstart does not do the right thing on xfs/reiserfs and other filesystems. | 17:06 |
| slangasek | right | 17:08 |
| slangasek | so we should definitely get that fixed | 17:08 |
| slangasek | would you mind opening a bug against the Debian package for me? | 17:09 |
| xnox | slangasek: ok. | 17:31 |
| slangasek | stgraber, xnox, jodh: so regarding https://code.launchpad.net/~stgraber/upstart/upstart-prctl/+merge/136759 | 18:23 |
| slangasek | has anyone actually talked to the desktop team (and in particular, Ted) about this constraint that upstart jobs won't work on 14.04 until you reboot? | 18:24 |
| slangasek | this is something they should be involved in discussing, we shouldn't assume we know the right solution for them | 18:25 |
| * xnox didn't discuss it with -desktop. | 18:25 | |
| xnox | but yeah, I agree we should. | 18:25 |
| xnox | something like a mail to -devel with cc to seb&ted? | 18:26 |
| stgraber | I believe we covered it in the spec before we asked the desktop team for feedback, but we didn't get any really useful feedback from them | 18:26 |
| stgraber | so yeah, might be worth asking specifically about the prctl case | 18:26 |
| stgraber | and explain exactly what won't work (spawning, events, ... will be fine, it's just respawning and stopping that won't quite work) | 18:27 |
| stgraber | well, stopping could probably work as long as we don't expect to get SIGCHLD | 18:27 |
| slangasek | stgraber: well, please reach out to Ted specifically | 18:29 |
| slangasek | he and his team will be the most impacted, I think, so he needs to understand the implications here | 18:29 |
| stgraber | ok, will send an e-mail this afternoon and Cc upstart-devel | 18:29 |
| slangasek | stgraber: perfect, thanks | 18:30 |
| slangasek | xnox: what is the value of d_type on these filesystems? | 22:48 |
| slangasek | xnox: is it set to DT_UNKNOWN? | 22:49 |
| slangasek | xnox: commented on the merge with suggestions for improvement | 22:56 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!