[06:36] <persia> heh.  There's a race condition on liveCD boot, so that some shells get the disclaimer, and some don't.
[09:15] <persia> cjwatson: Just to confirm, rev 3854 did fix the issue we'd been seeing with partman-uboot
[09:16] <cjwatson> persia: good, thanks
[09:35] <ev> michaelforrest: any word from Amanda?
[09:35] <michaelforrest> ev: I am struggling to get copy written and she seems to be to busy to talk face to face :(
[09:43] <ev> yikes, okay
[11:13] <michaelforrest> ev: ok I have a layout and I am now waiting for legal feedback.
[11:13] <ev> hooray
[11:15] <ogra> cjwatson, ev, what would you guys think about something that calls 'date -s "<date of image build>"' to casper (at least in cases where the system date is < builddate)
[11:16] <ogra> its one date we know for sure and will fix a lot of the 01.01.1970 issues i guess
[11:26] <cjwatson> ogra: I thought the intent was to put this somewhere central, like hwclock; I've discussed that with Keybuk before
[11:26] <cjwatson> casper is the wrong place, since this problem is not confined to live CD booting; you can get the same thing if the PRAM battery on a Mac goes, for instance
[11:27] <cjwatson> and, as discussed with Keybuk, using a date such as that of the image build is unhelpful.  What we should do is use a single known static date which is clearly wrong but at least valid; that way you can take one look at a bug report and say "aha, I know, your problem is *this*"
[11:27] <cjwatson> the equivalent of 0xDEADBEEF in C code
[11:43] <persia> cjwatson: The trick is that some stuff breaks if the system clock is older than the filesystem creation time, which means that 0xDEADBEEF either needs to be arbitrarily in the future, or that it doesn't matter if we do that.
[11:45] <cjwatson> there are different degrees of breakage
[11:46] <persia> True.
[11:56] <Omahn> Has anyone else come across bug #530632 when upgrading 8.04 to 10.04?
[11:58] <ev> michaelforrest: http://people.canonical.com/~evand/tmp/ubiquity-greeter-2.png - does this look okay?  I realize the language list is a bit narrow, that's next on my list to fix.
[11:59] <michaelforrest> ev: yeah that's cool, but definitely sort out the language list :)
[11:59] <ev> will do
[11:59] <persia> ev: Do feel free to use 800 horizontal pixels if you run into otherwise insoluable issues.
[11:59] <michaelforrest> has the window got narrower?
[12:00]  * persia is delighted with the shrinking ubiquity
[12:00] <cjwatson> Omahn: we mostly only do initial installation here
[12:00] <michaelforrest> ev: it would look a bit more balanced if the window was wider
[12:00] <Omahn> cjwatson: No problem. That works fine ;-)
[12:06] <ev> michaelforrest: okay, I'll see what I can do there
[12:17] <ogra> cjwatson, ah, i didnt know you were on it already
[13:21] <CIA-3> ubiquity: evand * r3857 ubiquity/ (3 files in 3 dirs):
[13:21] <CIA-3> ubiquity: Set the width of the language selection treeview to the minimum
[13:21] <CIA-3> ubiquity: required to fit the contents. Do not allow it to expand.
[14:20] <CIA-3> ubiquity: evand * r3858 ubiquity/ (10 files in 6 dirs):
[14:20] <CIA-3> ubiquity: Move the greeter functionality into ubiquity itself, still
[14:20] <CIA-3> ubiquity: accessible with the maybe-ubiquity boot option.
[14:24] <cjwatson> I'm working on the corresponding gfxboot stuff
[14:28] <ev> awesome
[14:29] <CIA-3> ubiquity: evand * r3859 ubiquity/debian/changelog: LP bug reference for feature freeze exception request.
[14:45] <cjwatson> everything done but drawing the access icon ...
[15:14] <ogra> yay !
[15:14]  * ogra sees oem-config-gtk finish for the first time in a rootstock image
[15:15] <ogra> there is one ugly error message though about language packs not being installable (which is no wonder since i dont have NM in the image, so no network)
[17:43] <cjwatson> michaelforrest: how does http://people.ubuntu.com/~cjwatson/tmp/initial-icon.png look to you, as a first cut of the boot screen?
[17:45] <michaelforrest> cjwatson: I don't think it meets the remit of telling the user they can do something here
[17:46] <michaelforrest> cjwatson: and the icon would definitely need to be smaller and monochrome
[17:46] <cjwatson> do you have a suggestion?  I thought this was what we agreed the other day
[17:46] <michaelforrest> yeah I wasn't 100% sure what you were envisioning but you seemed to have clarity so now I know what you were talking about doing
[17:47] <michaelforrest> I think we will probably need to put some english text next to the icon
[17:47] <cjwatson> boo
[17:47] <michaelforrest> yeah, or just not have anything really
[17:47] <persia> Is the icon parseable by a screen-reader?
[17:47] <cjwatson> no
[17:48] <cjwatson> no screen reader will be running; this is the CD boot loader
[17:48] <michaelforrest> so we were looking at doing [keyboard icon] = [accessibility icon]
[17:48]  * persia isn't sure who it helps
[17:48] <michaelforrest> but I got stuck trying to find a keyboard icon
[17:49] <cjwatson> http://people.canonical.com/~cjwatson/tmp/initial-icon.png reload, smaller and grayscale, modulo image artifacts
[17:49] <cjwatson> persia: https://wiki.ubuntu.com/Lucid/GfxbootUpdate
[17:49] <michaelforrest> cjwatson: how do I get that people.canonical.com thing working?
[17:50] <michaelforrest> is it chinstrap?
[17:50] <cjwatson> no, the machine is currently called lillypilly
[17:50] <cjwatson> https://wiki.canonical.com/MachineOverview, and ssh to people.canonical.com
[17:50] <cjwatson> mkdir public_html
[17:50] <cjwatson> and copy stuff there
[17:50] <persia> Oh, if F5 works, then I suppose it doesn't matter
[17:50]  * persia quiesces
[17:51] <cjwatson> persia: pressing any key at this drops through to the previous style of boot screen
[17:52] <cjwatson> michaelforrest: the one on Applications -> Preferences -> Keyboard maybe?  it's a little hard to see ...
[17:52] <cjwatson> but it is sort of recognisable
[17:53] <michaelforrest> yeah I think it'll have to be!
[17:53] <michaelforrest> I'm doing a mockup quickly
[17:53] <michaelforrest> if these damn adobe products will cooperate
[17:54] <cjwatson> http://people.canonical.com/~cjwatson/tmp/initial-icon-keyboard.png - not brilliant
[17:55] <cjwatson> http://people.canonical.com/~cjwatson/tmp/initial-icon-keyboard-bigger.png - maybe works better a little bigger?
[18:03] <davmor2> cjwatson, ev:  Todays ubiq seems to be stuck at "looking for other operating systems" this is a whole drive install so there is only lucid on it.
[18:03] <michaelforrest> cjwatson: I've emailed you
[18:07] <michaelforrest> (http://assets.grimaceworks.com/boot-screen.png )
[18:07] <cjwatson> michaelforrest: nifty; can I just extract that bar across the bottom and slam it into my image?
[18:08] <michaelforrest> cjwatson: thanks - yeah if that's easiest
[18:08] <cjwatson> no point in trying to reproduce it in code, I feel
[18:08] <michaelforrest> I realigned the ubuntu logo too
[18:08] <cjwatson> hmm, not sure about that
[18:08] <michaelforrest> ideally it will be vertically centered in the black
[18:08] <cjwatson> if you do that, the logo will have to be moved up when the full menu is displayed
[18:08] <cjwatson> seems a bit awkward?
[18:08] <michaelforrest> oh the damn full menu
[18:08] <michaelforrest> I really don't mind that
[18:09] <michaelforrest> because the full menu is not on the primary journey
[18:09] <cjwatson> I do, 'cos I have to write the horrible horrible code to move it around ;-) but OK ...
[18:09] <michaelforrest> obviously I want you to animate it into position and then fade in the menu l)
[18:09] <cjwatson> die ;-)
[18:09] <michaelforrest> mark says he wants it!
[18:09] <michaelforrest> what can I do! ;)
[18:09] <cjwatson> http://paste.ubuntu.com/387145/ code so far
[18:10] <michaelforrest> hope nobody comes out with 72bit and 128bit computing any time soon then...
[18:10] <cjwatson> that bit?  that's legacy from suse, don't think it does anything on Ubuntu, I just haven't got round to removing it
[18:11] <michaelforrest> anyway it looks like the sort of code I'd write ruby code to write for me rather than writing it directly :)
[18:11] <cjwatson> that sounds even harder :)
[18:11] <michaelforrest> but so much more maintainable
[18:11] <michaelforrest> but I digress
[18:11] <cjwatson> not really, this is a real special-purpose programming language
[18:12] <cjwatson> it's not pretty, but I'm not sure writing a second-layer compiler for it would cause me to get things done any quicker :-)
[18:12] <cjwatson> I suppose I can slice out a defined area of the background image and shift it down a bit
[18:12] <persia> I believe Ubuntu can run on some Crusoe chips (at least old versions could) which were 256-bit internally.
[18:12] <cjwatson> as a quick hack
[18:22] <davmor2> cjwatson: bug 530861  I don't know why this has just seeming stopped doing anything,  I'm hoping it's not a race condition again cause I know how much you like those :)
[18:25] <cjwatson> davmor2: not having looked at the bug yet, it sounds like that blkid spin that's been reported
[18:25] <cjwatson> davmor2: look at 'ps axf' - is blkid in there, somewhere under os-prober?
[18:26] <cjwatson> unfortunately I can't reproduce it on my own systems or I'd have fixed it already :(
[18:26] <davmor2> cjwatson: yeap
[18:27] <davmor2> blkid -o value -s TYPE /dev/sda2
[18:27]  * cjwatson sends it off to util-linux
[18:27] <cjwatson> I guess I'll have to try to zen-debug that one
[18:29] <davmor2> cjwatson: I'll restart it after tea and see if I can reproduce if I can, I can always open up ssh again so you have access to a non working box
[18:32] <cjwatson> michaelforrest: ok, reproduced your mockup
[18:33] <cjwatson> davmor2: if you can reproduce it by simply running 'sudo blkid -o value -s TYPE /dev/sda2', could you leave it in that state?
[18:42] <cjwatson> michaelforrest: committed; awaiting feature freeze exception
[18:42] <cjwatson> thanks!
[18:44] <cjwatson> ev: reassigned the closest associated work item from foundations-lucid-gfxboot-update to you
[18:48] <ev> cool, thanks
[19:03] <davmor2> cjwatson: will do
[19:06] <michaelforrest> cjwatson: amazing
[19:13] <davmor2> cjwatson: just ran the blkid command and it exited after a second.  Retrying the install.  However I was overwriting vista before so if it works this time I'll drop vista back on and see if that is a factor
[19:14] <cjwatson> hm
[19:24] <davmor2> meh worked
[19:25] <davmor2> cjwatson: should ureadahead be running on the CD?
[19:26] <davmor2> cjwatson: I'm assuming that ureadahead is similar to sreadahead that caused the nice race conditions in the installer for karmic
[20:15] <davmor2> cjwatson: I can reproduce the issue installing over vista but now need to sod off.  2 things.  1 could it be because sda2 is in use by vista as a really partition and ubuntu as a logical? 2 I'll do it again tomorrow and hook up ssh properly for you.
[21:07] <ev> yay, FFe granted
[21:07] <ev> preparing an upload now
[21:10] <CIA-3> ubiquity: evand * r3860 ubiquity/debian/real-po/ (81 files): debconf-updatepo
[21:10] <cjwatson> I'll do gfxboot-theme-ubuntu later this evening
[21:10] <ev> cool, thanks
[21:10] <cjwatson> (on the n900 now so it'd be tricky from here)
[21:10] <ev> haha
[21:11] <CIA-3> ubiquity: evand * r3861 ubiquity/ (d-i/manifest debian/changelog debian/control):
[21:11] <CIA-3> ubiquity: Automatic update of included source packages: netcfg 1.51ubuntu2,
[21:11] <CIA-3> ubiquity: partconf 1.32ubuntu1, partman-base 138ubuntu1.
[21:19] <superm1> cjwatson, does all of the artwork in the ubuntu-cdimage branch still need to be updated for that change too then for stuff like xubuntu/mythbuntu/kubuntu?
[21:22] <cjwatson> superm1: I don't think so, due to the way I did it
[21:23] <superm1> cjwatson, cool :)
[21:23] <superm1> although the kubuntu media probably won't want to be booting in maybe-ubiquity mode yet i'd suppose
[21:31] <ev> I'll work on fixing that tomorrow
[21:31] <CIA-3> ubiquity: evand * r3862 ubiquity/debian/changelog: releasing version 2.1.29
[23:23] <cjwatson> superm1: yeah, was planning on enabling it image by image - my priorities are the Ubuntu ones, and it may even be that some of the others don't want it because of more complex boot menu requirements
[23:23] <superm1> right, sounds sensible