[00:00] or /host/ubuntu/mythtv I suppose [00:00] yep, although you cannot be completely sure about /host/ubuntu but will do for 8.10 [00:00] ./autopartition-loop:285: mkdir -p "/host${path%/*}" [00:01] so yes [00:02] xivulon, why wouldn't /host/ubuntu do it? isn't that where the loop image gets stored? [00:03] finished copying now [00:03] so seems syncio is in because it seems a bit slower, not 10x worse though as predicted by cking [00:04] superm1 yes, but in the future it might be /host/mythbuntu [00:05] xivulon, well what happens for kubuntu,xubuntu etc? [00:05] do they get /host/xubuntu? [00:06] no all in /host/ubuntu in 8.10, I meant things might change in future versions, but we will have time to do it properly then [00:10] hm - in a partman recipe, if I set a priority that is bigger than the size of the disk is, the install fails (with an unable to create partition error) [00:11] is that expected ? although the minimum size fits the disk [00:12] http://paste.ubuntu.com/53318/ <- that's the partman expert recipe - however the drives are 1GB in size. [00:13] the last partition is not created. [00:15] mathiaz: I'm totally not going there after midnight, but you can ask me again tomorrow ... [00:16] cjwatson: sure [00:17] xivulon, ah well that would just be modfiying the preseed anyway if that behavior changed [00:18] xivulon, i'll do some experiments with a preseed for master backend frontend then too and see how things work out [00:18] cool [00:18] have a build ready [00:18] test went well [00:18] other than nm [00:20] I think that the main advantage of allowing a backend is sparing me the trouble to answer "What is a backend? Why can't I press forward?" :P [00:21] evand wubi rev 509 is up [00:31] superm1 http://wubi-installer.org/devel/minefield/Wubi-8.10-rev509.exe [00:34] evand wubi 509 is required if we have lupin 111+ [09:27] hi, didn't notice that the kde3.5 -kde4 dicothomy has been abolished [09:28] will need a new revision for wubi [09:29] xivulon: Yes kubuntu is just kde 4.1.2 now [09:36] Ok uploaded wubi rev 5.10 (evand all yours), Davmor, please test http://wubi-installer.org/devel/minefield/Wubi-8.10-rev510.exe [09:36] davmor2 ^ [09:36] that also contains the uninstaller progressbar fix by the way [09:36] have to go now [10:02] http://cdimage.ubuntu.com/netboot/ - some folks here might be interested in that set of landing pages [10:02] they're just links to the right bits of the archive [10:05] cjwatson: sweet [10:08] cjwatson: Should those be scheduled to test cases, etc. as with other images? [10:10] persia: netboot is [10:10] Ah. I suppose I was confused by two images vs. 7 images, but it all becomes clear now. [10:11] http://iso.qa.ubuntu.com/qatracker/build/netboot/all [10:11] Yeah :) [10:12] I'll update for older releases as well when I have a chance [10:42] I've just done all the vista updates and am resaving the partitions. 15.85 GB for vista and nearly 2 gig for system partition so nearly 20gig total [11:35] evand can we have a new iso with new wubi (510) and lupin (115) and sysvinit (2.86.ds1-59ubuntu9)? [11:36] xivulon: is this to remedy the loop issue [11:37] xivulon: let's wait for the cron jobs to be turned back on [11:37] ok [11:37] Steve hasn't done that so it wouldn't surprise me if he's intentionally letting cdimage cope with the beta load for a while [11:38] davmor2: what loop issue do you refer to exactly? [11:40] xivulon: https://bugs.edge.launchpad.net/wubi/+bug/204133 this one [11:40] ah yes === davmor2 is now known as davmor2_lunch [14:30] xivulon, okay i downloaded it. i'm going to do a build today with my own patched network manager on our hotfixes PPA and make sure that network manager bug is resolved === davmor2_lunch is now known as davmor2 [16:25] cjwatson, cking was mentioning that it might be better to use dm-loop as opposed to loop for performance reasons, do you think it would be feasible? [16:28] I don't know anything about it - it doesn't *sound* difficult [16:31] would not think so, I have asked cking to join us though [16:31] hi cking, was discussing with cjwatson about your suggestion to use dm-loop as opposed to losetup [16:32] Ok. its a case of modprobing dm-loop and then I using dmlosetup (this is a symlink to the dmsetup binary) [16:32] does that amount to some arguments to dmsetup? [16:32] (otherwise we could change dmsetup-udeb to add that symlink I suppose) [16:33] is dmlosetup equivalent to mount in that setup? [16:33] cjwatson: The rune I used was: [16:33] dmlosetup loop1 /mnt/ntfs/wubi_ext3.img [16:34] and then mount -o loop /dev/mapper/loop1 /some/mount/point [16:34] that's it [16:35] ..of course the path names need to be changed to match the Wubi config [16:35] if I want to list the loop devices, can I still use losetup? [16:35] xivulon: not sure. never tried that [16:36] would be good to know as some script that does not actually mount the devices, might be using losetup to check the setup [16:37] or detach them [16:40] dmlosetup -f is reporting that this it's not currently implemented .. and AFAIK losetup won't help us for this case [16:54] xivulon: mm.. I don't know how to list the dmloop devices. [16:57] does $(dmlosetup /dev/loop$loopn) spit out the same string as losetup? [17:02] xivulon: ..just a mo. [17:09] xivulon: /dev/mapper is populated with loopN when loopN is created. You can therefore tell which dm-loop devices are being used [17:11] /dev/mapper/control should always exist for 0-N dm-loop devices [17:12] xivulon, can you pop in #ubuntu-mythtv-dev again [17:12] tgm has some more questions [17:13] the files that need to be modifed should only be autopartition-loop and initramfs/scripts/local [17:13] init.d scripts should be fine, have to check grub/grub-installer but should be ok [17:13] superm1 sure [17:13] #ubuntu-mythtv-dev again [17:13] #ubuntu-mythtv-dev again [17:13] ops paste gone wrong [17:20] dmlosetup doesn't exist in Ubuntu [17:20] cjwatson: it's a symlink to dmsetup [17:20] not here it ain't [17:20] command-not-found reckons it's not in intrepid [17:21] cjwatson: one needs to symlink it by (hand at the moment) - which needs to be fixed [17:21] ah [17:21] somebody should file a bug on devmapper about that [17:22] cjwatson: yep... I seemed to overlook it back in July when I started faffing around with dm-loop [17:31] cjwatson: I will create a bug report in a short while for this === robbiew is now known as robbiew-lunch === robbiew-lunch is now known as robbiew