[01:57] how do I commit a merge? [01:57] in bzr [02:03] nvm. [04:56] does upstart need to have pid 1? [11:56] I wonder how upstart could solve the "boot splash problem". If there is no goal, how can a "percentage" be calculated? [11:57] Some sort of "spinning" is the only UI solution that seems to make sense [11:59] you could also have some sort of "icon parade" in the lower part of the screen so that you can get a sense of what has already started [14:26] I can't build current upstart trunk :-( [14:27] I did the ../nih/nihify && autoreconf -i && ./configure && make dance [14:27] I don’t think it would be very difficult to estimate the percentage based on previous boots. [14:28] yes but what will define the "100%" in order to calculate this percentage? when X started? [14:28] For instance, yes. [14:29] but that can happen at different times depending on the situation (e.g., a usb drive is plugged or not) [14:29] Yes. [14:29] As i said, estimate. [14:30] and when this estimation underestimates? you return the bar back? [14:31] or do you take something like the 200% of the estimation just to be sure? [14:31] No, just let it stay at 100%. [14:31] hmm, if it stays at 100% for long the user might think the system is stuck [14:32] The probability of that happening is low. It’s more likely it stays at a much smaller percentage (e.g. waiting for a device or network), at which time the bar could be replaced with a spinner until there’s progress again. [14:32] and there is also the question how much do you increment the bar for each "started" events? [14:33] That’s based on previous boots. [14:33] but since things start in parallel it is unlikely you can estimate that correctly (at least without a lot of samples) [14:34] I said previous boots. That’s where we get a lot of samples from. [14:34] If there isn’t enough certainty, just show a spinner and collect data. [14:35] So it is a spinner in the beginning and at some arbitrary boot it is transformed to a bar? [14:35] Yes. [14:36] Probably already at the second boot. [14:36] And the accuracy will increase after that. [14:36] interesting [14:37] ion_: does current trunk build for you? [14:38] I haven’t built it for a while, dunno. [14:38] I see. It fails to build here sadly :-( [14:49] darius12: what was the result of "make" ? [14:50] INSTALL_NIH does not appear in AM_CONDITIONAL [14:51] I added it manually and now it breaks inside libnih [14:51] *** No rule to make target `../m4/codeset.m4', needed by `Makefile.in'. Stop. [14:51] is it possible that I have an incorrect version? I bzr branch'ed from launchpad [14:52] I have rev 960 [14:52] I 'm trying to build it in ubuntu hardy [14:54] so you checked out trunk [14:54] and you ran autoreconf -i ? [14:54] first I run ../libnih/nihify [14:54] you nihify'd libnih? [14:54] and then autoreconf -i yes [14:54] the _correct_ sequence is ... [14:55] bzr checkout .../libnih/trunk libnih [14:55] bzr checkout .../upstart/trunk upstart [14:55] cd libnih [14:55] autoreconf -i [14:55] cd ../upstart [14:55] ../libnih/nihify [14:55] autoreconf -i [14:55] ie. you nihify _upstart_ rather than libnih [14:55] if you do it inside libnih, it'll likely futz things up [14:55] so you'd need to checkout again [14:56] I did it within upstart's dir [14:56] did you do the above sequence? [14:56] but I didn't autoreconf libnih before I do that [14:56] ah [14:56] so I 'll try again [14:57] If it works could you please just copy & paste the above snippet in HACKING? [14:58] I missed the "autoreconf -i" part inside libnih in the current instructions [14:58] yeah, I probably should [14:58] http://upstart.ubuntu.com/wiki/CompilingUpstart [14:58] has it [14:59] hmm, you may have to configure libnih too [14:59] I can't remember ;) [14:59] how do you build it? [15:00] it would also be nice then if there was a standard script for all this dance assuming standard directory names ("upstart" and "libnih") [15:00] this way you wouldn't forget yoursef :-) [15:02] heh [15:02] It's just always checked out [15:02] so you just "make" right? [15:03] ok, I did exactly the above sequence. Now it says: /bin/sh: ../libtool: No such file or directory [15:03] yeah [15:03] try a ./configure in libnih as well [15:03] ok [15:04] hmm, no it seems to be building :-) [15:04] s/no/now/ [15:04] yep, it built fine. thanks :-) [15:05] would you like me to pastebin the exact sequence so that you can put it in HACKING? [15:08] well, it's here if you want it: http://pastebin.ca/1021024 [15:08] bbl