| === sadmac_ is now known as sadmac | ||
| === cwillu_ is now known as cwillu | ||
| h\h | hi | 12:09 |
|---|---|---|
| Keybuk | hi | 12:15 |
| Keybuk | how goes it harald? | 12:15 |
| h\h | yes | 12:19 |
| h\h | s/yes/fine, thanks/ | 12:19 |
| h\h | :) | 12:19 |
| h\h | Hey Keybuk, what is upstart 1.0 doing? :) | 12:19 |
| Keybuk | "what is" or "how is" ? | 12:20 |
| h\h | "how is" | 12:20 |
| Keybuk | mostly pinned down what I want to do | 12:21 |
| h\h | sry for my bad english :) | 12:21 |
| Keybuk | been doing some ground work and playing around really so far | 12:21 |
| Keybuk | not going to start coding in earnest for a few weeks yet | 12:21 |
| Keybuk | we're still in pre-beta, and I've still got a pile of boot speed related patches to do, as well as new m-i-t etc. | 12:21 |
| h\h | " boot speed related patches" | 12:22 |
| h\h | what are you going to do? | 12:22 |
| Keybuk | that's a long list ;) | 12:24 |
| Keybuk | https://wiki.ubuntu.com/FoundationsTeam/BootPerformance | 12:24 |
| h\h | Investigate why udevadm is taking too long, and improve where possible | 12:25 |
| h\h | 12:25 | |
| h\h | Done | 12:25 |
| h\h | what was it? module loading? | 12:25 |
| Keybuk | partially | 12:25 |
| Keybuk | other fixes Alan did for udev to use pselect and so on have helped too | 12:25 |
| h\h | upstream patches? | 12:26 |
| h\h | or can you point me to a patch? | 12:28 |
| Keybuk | it's all in 139 | 12:31 |
| h\h | ah, fine | 12:32 |
| h\h | are you starting dbus, hal from upstart now? | 12:34 |
| Keybuk | not yet, no | 12:34 |
| keesj | http://fosdem.unixheads.org/2009/maintracks/upstart.xvid.avi | 13:06 |
| keesj | (upstart talk at fosdem) | 13:08 |
| * Keybuk doesn't remember them asking for permission to video me ;) | 13:14 | |
| sadmac | win 21 | 13:18 |
| sadmac | shit, its too early | 13:18 |
| Jc2k | :) i do that twice a day sadmac :( :( | 13:22 |
| MarcWeber | The event based model is fine. However does init also provide some notion of state? | 16:32 |
| MarcWeber | Example: there is one job setting up network. So many jobs will start on started network. | 16:33 |
| MarcWeber | Now the os wants to install a new daemon which should only be started if network has been started as well. | 16:34 |
| MarcWeber | So is the way to start this daemon only if initctl status network = running ? | 16:34 |
| MarcWeber | Or does upstart start this daemon itself when recognizing there is a new job file depending "on started network"? | 16:37 |
| sadmac2 | MarcWeber: the next version will have this stuff. | 17:31 |
| MarcWeber | Ah thank you for this information. | 17:32 |
| MarcWeber | Which is the correct way to shutdown a system? | 18:23 |
| MarcWeber | for job in `initctl list`; do stop $job; done then sync and halt -f -p ? | 18:24 |
| ion_ | Pull the power cord. | 18:24 |
| MarcWeber | ion_: Can't, its a vserver :) | 18:24 |
| ion_ | Pull the power cord of the physical box it’s running on. :-) | 18:24 |
| * ion_ watches upstart.xvid.avi | 19:11 | |
| keesj | is the sound ok? | 19:14 |
| ion_ | Yeah | 19:29 |
| sadmac | Keybuk: your plumbing layer looks like a blue hamburger | 22:46 |
| damjan | Where can I fing a detailed reference of what can a job file contain? | 22:50 |
| sadmac | Keybuk: also we had upstart in F9 ;) | 22:53 |
| damjan | Keybuk: wow, you had some bad luck with git :) | 23:02 |
| damjan | Dunno, I don't consider me as even a programmer, but I find git easier than you experience. And better than darcs (the first DVCS I used), and mercurial (the second) | 23:03 |
| damjan | better, because although mercurial and darcs were easy to understand, while git only a bit harder, with git I could more easily get myself out of stupid situations I've put myself in | 23:04 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!