[01:25] <CIA-3> wubi: Agostino Russo * r180 trunk/ (debian/changelog src/wubi/application.py): Fixed CD menu reboot (LP: #543032)
[02:15] <CIA-3> wubi: Agostino Russo * r181 trunk/ (data/grub.install.cfg debian/changelog): Safe graphic mode boot option was disabled (LP: #550539)
[05:08] <cjwatson> dmarkey: sure, it would need to be (hd0,1) or whatever in order to actually work, and it's probably worth fixing it up, but it really doesn't actually mattere
[05:08] <cjwatson> *matter
[06:11] <kusum> can anyone here expalin me the preseed file
[08:05] <dmarkey> cjwatson: when instaling an ordinary ubuntu, it's (hd0,1), when installing paravirtualised under xen, its /dev/xvda. which unfortunately crashes pygrub
[08:06] <dmarkey> im pretty sure it was (hd0,1) in 9.10 under xen
[08:15] <dmarkey> what ever script populates that is trying to make the mapping from /dev/xvda to the disk number out of the bios, the paravirtualised xen doesnt have a bios
[08:15] <dmarkey> so i assume it falls back to xvda
[10:32] <ogra> cjwatson, i'm a bit lost with the OAMP images and could need some advice, the design is supposed to be as follows: install runs from SD (OAMP uboot cant boot from USB) targeting a USB disk but using SD for /boot, since the SD is mounted i cant format it but only replace uImage/uInitrd during install, but i need UUID->/boot for it in fstab after first boot
[10:33] <ogra> *OMAP
[10:33] <ogra> is there (hopefully not to intrusive) way to achieve that in ubiquity/d-i ?
[10:35] <ogra> i could use flash-kernel-installer for mangling fstab and putting the files in place but that feels very very ugly
[10:43] <cjwatson> the usual way to do that kind of thing is to have a partman fstab.d component that spits out the right line as appropriate
[10:43] <ogra> hmm, i couls use partman-uboot for that i suppose
[10:43] <ogra> *could
[10:44] <ogra> that still leaves the prob that the Sd cant be mounted on /boot at install time though
[10:45] <ogra> oh, wait, i could bind mount/umount it while flash-kernel-installer runs
[10:50] <dmarkey> cjwatson: hey, do you have a quick sec to talk about this root=(/dev/xvda,0) issue i've come across?
[10:50] <cjwatson> dmarkey: it's a consequence of not having device.map any more
[10:50] <cjwatson> it's not xen-specific or anything - it also shouldn't be harmful in any way
[10:51] <dmarkey> i tried to install 10.04 on a normal machine, and root=(hd0,0)
[10:51] <cjwatson> I do want to clean it up for 10.04 somehow, but essentially so that people stop asking about it rather than anything else :)
[10:51] <cjwatson> set root= is almost entirely irrelevant nowadays.  the UUID search command immediately afterwards is what normally *actually* finds the root partition
[10:51] <cjwatson> set root= is basically just insurance fallback
[10:52] <dmarkey> hmm, you see, xen uses pygrub, which relies on root= to be the proper notation
[10:52] <cjwatson> (and yes, the current situation means that that insurance fallback won't work properly, but there are lots of situations where it would go wrong anyway)
[10:52] <cjwatson> pygrub is grub legacy only, is it not?
[10:52] <dmarkey> no, grub2 aswell
[10:52] <cjwatson> oh, that's news to me
[10:52] <dmarkey> very new patches
[10:52] <cjwatson> they haven't been mentioned on grub-devel@
[10:52] <dmarkey> but they are in xen 4.0, which was released yesturday
[10:53] <cjwatson> anyway, I'll see if I can fix this up
[10:53] <dmarkey> that would be *great*
[10:54] <cjwatson> have been trying to get rid of device.map in general, and this is one of the bits of fallout
[10:55] <dmarkey> i see
[10:59] <dmarkey> having said that pygrub grub2 support is very limited, just enough to boot the system.
[11:11] <dmarkey> also, i'm ready to test that PAE image/initrd when you get the chance to build it
[11:12] <cjwatson> I can't do anything
[11:12] <cjwatson> I asked #ubuntu-kernel to apply an appropriate patch to build the udebs
[11:12] <cjwatson> but until that's applied, there's no point asking me I'm afraid
[11:13] <dmarkey> oh i see
[11:13] <dmarkey> sorry bout that
[11:13] <cjwatson> that's ok
[11:50] <ev> cjwatson: http://paste.ubuntu.com/410986/
[11:53] <cjwatson> ev: out of sheer paranoia I'd put a DebconfError guard around the db.get as well, and maybe ret > 0 rather than != to match the code below (or change the code below?)
[12:01] <cjwatson> dmarkey: ok, (/dev/xvda,*) thing should be fixed shortly
[12:04] <dmarkey> cjwatson: good stuff, may i ask what logic you are using for that
[12:06] <cjwatson> probe all devices explicitly in 'grub-probe --target=drive' if /boot/grub/device.map is missing
[12:07] <cjwatson> http://bazaar.launchpad.net/~ubuntu-core-dev/ubuntu/lucid/grub2/lucid/revision/1963
[12:09] <dmarkey> cool. will this be in the daily snapshot so i can test it?
[12:11] <cjwatson> dunno when I'll upload it exactly
[12:11] <cjwatson> sometime after beta-2
[12:13] <dmarkey> ok i'll wait
[12:27] <CIA-3> ubiquity: evand * r4057 ubiquity/ (bin/ubiquity-dm debian/changelog debian/ubiquity.templates):
[12:27] <CIA-3> ubiquity: Let the user know why we're starting a desktop session or rebooting
[12:27] <CIA-3> ubiquity: when ubiquity in only mode crashes. If in automatic mode, the
[12:27] <CIA-3> ubiquity: default is to reboot. Preseed ubiquity/reboot_on_failure to false
[12:27] <CIA-3> ubiquity: to start a desktop session instead.
[12:40] <CIA-3> ubiquity: evand * r4058 ubiquity/debian/changelog: Add LP freeze exception request bug for previous commit.
[12:42] <dmarkey> Kevin`: welcome
[12:49] <Kevin`> :/
[12:49] <Kevin`> my cable died for a while, no sync at all
[12:49] <dmarkey> how much of this stuff is backported to the debian installer
[12:50] <Kevin`> I assume that wasn't a question to me
[12:51] <dmarkey> anyone really
[12:52] <Kevin`> if you mean xen, they have/had a dedicated installer for it. i'm not sure if now, with proper pvops support, they have integrated the two. i'm kinda stuck on lenny
[12:52] <dmarkey> if we have ubuntu xen/domu support, it should be easy to get debian/domu xen support
[13:43] <rgreening_> ev: ping
[13:45] <ev> rgreening_: pong
[13:46] <rgreening_> how are ya?
[13:46] <rgreening_> ev: usb-creator seems to mount the partitions from my usb stick automatically. I guess this is ok, however, and interesting side effect is that if I now close usb-creator-kde without making the startup disk, it leaves the partitions mounted. This is problematic for current KDE as KDE's Solid back-end is still using HAL, and Solid/HAL cannot umount any device mounted by UDisks. This leaves the device hanging for the user. Thoughts?
[13:49] <ev> HAL is dead upstream.
[13:49] <ev> completely
[13:50] <ev> I'd say this is a bug in Solid more than anything else
[13:51] <ev> usb-creator leaves partitions mounted purposefully.  I don't want to get into the corner cases of is this a partition we mounted or was it already, is the user still using the partition, and so on.
[13:52] <ev> oh, missed the start of that -- I'm good, thanks!  How are you?
[13:52] <rgreening_> ev: I agree, however, Solid does not have a proper way to deal with UDisks atm
[13:52] <rgreening_> Im good. Going to UDS :)
[13:52] <ev> woohoo!
[13:52] <rgreening_> hehe
[13:53]  * ev shrugs
[13:54] <ev> so we definitely couldn't do anything about it for lucid - we're in hard freeze here, and I would hope that the Solid folks get their act together in the next six months.
[13:54] <rgreening_> ev: yeah. true enuff. Im hoping they get the backend updated.
[13:56] <rgreening_> ev: got any specific plans for maverick for the stuff you are working on? Anything I'd be interested in :)
[14:00] <ev> rgreening_: so we *may* have a summer of code student working on some usb-creator stuff.
[14:01] <rgreening_> cool. for mac? win? something else?
[14:01] <ev> <https://wiki.ubuntu.com/GoogleSoC2010/Ideas#usb-creator Improvements>
[14:01]  * rgreening_ reads
[14:02] <ev> in ubiquity-land we'll be doing a redesign
[14:02] <ev> we may also have a summer of code student to do wubi migration
[14:03] <rgreening_> cool
[14:04] <rgreening_> ev: we need to make sure that if the gtk frontend changes or any backend changes which could affect the kde frontend are taken into consideration for the GSOC project.
[14:04] <rgreening_> sounds good otherwise :P
[14:05] <ev> yarp
[14:05]  * rgreening_ wishes I had time ot get the server piece done this cycle... stupid $work
[14:05] <ev> entirely understandable
[14:05] <rgreening_> work wouldn't be a problem if I had been hired for the foundations team  ;)
[14:05] <rgreening_> ha
[14:05] <ev> heh, indeed
[14:06]  * rgreening_ is still interested in a job with canonical though
[14:15] <CIA-3> ubiquity: evand * r4059 ubiquity/debian/ (changelog oem-config.templates):
[14:15] <CIA-3> ubiquity: Make oem-config/install-language-support false by default. OEMs
[14:15] <CIA-3> ubiquity: should only be turning this on if they've provided all the possible
[14:15] <CIA-3> ubiquity: langpacks in oem-config/repository or can guarantee an Internet
[14:15] <CIA-3> ubiquity: connection (LP: #539710).
[14:18] <ev> they're always hiring :)
[14:40] <rgreening_> yup
[14:41] <rgreening_> just not in the right area .. foundations would have been awesome... and Im not into ARM as such...
[14:43] <kusum> Could somebody explain me the lines in preseed.cfg file please
[14:59] <cjwatson> kusum: not in such general terms
[15:00] <cjwatson> kusum: you need to ask specific questions
[15:02] <dmarkey> kusum: http://www.google.co.uk/search?hl=en&q=site:ubuntu.com+ubuntu+preseed&meta=&aq=f&aqi=&aql=&oq=&gs_rfai=
[15:13] <kusum> cjwatson: dmarkey : How is the wubildr.mbr handled when we select ubuntu on the grub
[15:20] <cjwatson> wubildr.mbr is a grub image that's configured to load the appropriate disk image using grub's loopback module, and then chain through to the grub.cfg in that image
[15:21] <cjwatson> see data/wubildr.cfg in wubi
[15:57] <shtylman> ev: http://arstechnica.com/tech-policy/news/2010/04/digital-economy-bill-passes-in-the-uk.ars < --- looks like you guys have it worse than us
[15:58] <ev> shtylman: not at all
[15:58] <shtylman> heh
[15:58] <ev> there's a lot of shouting going on over something that's not all that bad
[15:58] <shtylman> sounds pretty bad from the article
[15:58] <ev> the big thing that everyone is complaining about is that if you get caught breaking copyright three times, they can disconnect you from the internet
[15:59] <shtylman> sounds like the matrix :)
[15:59] <shtylman> if you die online... do you die in real life?
[15:59] <ev> that's already the case in the states, only they can do it on the first occurrence
[15:59] <shtylman> :(
[16:00] <ev> Keybuk pointed out something interesting though
[16:00] <ev> from his reading of the bill, he gathered that you could have someone disconnected from the internet if they violated the GPL
[16:01] <shtylman> oh... man...
[16:01] <shtylman> that would be epic
[16:01] <ev> hahahaha
[17:32] <superm1> ev, re r4057, would it be possible to call the fail late command before rebooting if it were preseeded to provide consistency with what happens if the failure occurs later?
[17:32] <ev> absolutely
[17:32] <superm1> (of ubiquity that is)
[17:32] <ev> sorry for the oversight there
[17:33] <superm1> great thanks
[17:40] <CIA-3> ubiquity: evand * r4060 ubiquity/ (bin/ubiquity-dm debian/changelog): Call ubiquity/failure_command if we crash in only ubiquity mode.
[17:50] <CIA-3> ubiquity: cjwatson * r4061 ubiquity/ (debian/changelog ubiquity/frontend/gtk_ui.py):
[17:50] <CIA-3> ubiquity: * GTK frontend:
[17:50] <CIA-3> ubiquity:  - Re-fetch translations before calling plugin_translate (LP: #552673).
[19:16] <CIA-3> ubiquity: evand * r3922 ubiquity.testing/ (3 files in 2 dirs):
[19:16] <CIA-3> ubiquity: * Clean up tests/run, only recgenerate templates if modified.
[19:16] <CIA-3> ubiquity: * Provide testPluginsLoadProperly and testReturnToPartitioning tests.