| Rydia | Any unetbootin devs around? | 06:09 |
|---|---|---|
| xivulon | cjwatson, was reviewing initramfs/scripts/local, in relation to bug #226622 and noticed a FIXME (I belive from you) for the mount line | 11:43 |
| xivulon | # FIXME This has no error checking | 11:43 |
| xivulon | Would it be possible to put that errorchecking in (at this stage) so that we also address #226622 | 11:44 |
| xivulon | (and possibly do something similar for casper) | 11:44 |
| cjwatson | I don't think that was from me | 11:44 |
| cjwatson | the error checking (such as it is) happens later when the initramfs system notices that the root device hasn't really appeared | 11:45 |
| cjwatson | but yes, it does produce something unclear; though I'm a little worried that there may be some expected failures at that point | 11:45 |
| xivulon | can we maybe save $? and have a proper panic message within the if [ "LOOP" ] so that it is specific to wubi? Ideally the message should appear also if a user is on usplash | 11:47 |
| cjwatson | err, I'm really not sure, I didn't write that originally - it sounds reasonable | 11:48 |
| xivulon | a clear error when ntfs mount fail because of unclean shutdown, would greatly reduce the number of tickets | 11:50 |
| xivulon | +/- http://paste.ubuntu.com/10492/ | 12:08 |
| xivulon | doing that in casper might be slightly trickier though, since the partition is found heuristically | 12:11 |
| xivulon | in lupin-casper we could have an error message if iso-scan/filename is set but no ISO is found and similarly in custom-installation | 13:13 |
| xivulon | we should also check whether the filesystem is mounted ro instead of rw. | 13:13 |
| xivulon | evand ^^ | 13:13 |
| evand | indeed, that looks roughly reasonable. | 14:04 |
| xivulon | great, I will do a proper diff later on then, for initramfs-tools and upload some new code into lupin/hardy.proposed as well | 14:10 |
| xivulon | ps any idea when we will have new dailys with the latest fixes? | 14:11 |
| evand | cjwatson: ^ I've been wondering this myself. Is that known yet, or is it still in the planning stage? | 14:12 |
| xivulon | error codes for ntfs are listed in man ntfs-3g.probe(8) | 14:15 |
| xivulon | do we have error-remount-ro type of behaviour to worry about? | 14:17 |
| cjwatson | evand: for hardy you mean? | 14:18 |
| evand | cjwatson: sorry for not being specific, yes | 14:18 |
| cjwatson | evand: poke slangasek about that - he has all the equipment now to do it | 14:18 |
| evand | will do | 14:18 |
| xivulon | evand, at the end of autopartition-loop we run `apt-install lupin-suport`, when is lupin-support actually installed? in particular is it ok to call update-initramfs in .postinst and is /etc/default/locale available at that stage? | 14:38 |
| xivulon | this is re #136682 | 14:39 |
| xivulon | I would assume it is fine, but would like confirmation | 14:41 |
| evand | checking | 14:46 |
| xivulon | thx | 14:52 |
| evand | at least in the case of ubiquity, that happens sufficiently late for that to work. And yes, that should be fine. | 14:52 |
| xivulon | good, I'll let Tormod know | 14:53 |
| cjwatson | xivulon,evand: fancy doing a demo of wubi at UDS? | 15:11 |
| cjwatson | xivulon: I gather you're only there part of the week? | 15:12 |
| evand | sure, I'd love to help with that | 15:13 |
| xivulon | cjwatson: absolutely | 15:13 |
| xivulon | I will be there the last two days, so we can do that thu or fri! | 15:13 |
| cjwatson | ok, cool | 15:14 |
| cjwatson | I've asked Scott et al to pencil that in | 15:14 |
| xivulon | not with my laptop though because it does have ACPI issues :( (#146692) | 15:14 |
| evand | cjwatson: speaking of which, did you get the emails on two wubi sessions? | 15:14 |
| evand | I can furnish a laptop for the task. | 15:15 |
| xivulon | great | 15:15 |
| cjwatson | oh, I did, but my agenda is actually oversubscribed right now - if I get it down to the point where they'd fit, that would be fine, otherwise maybe put them into a whiteboard-scheduled slot? | 15:15 |
| xivulon | we can squeeze in 1 hour I guess if required (as mentioned in my reply) | 15:16 |
| evand | works for me. If we run out of time, we can always find vacant space in a hallway. | 15:17 |
| xivulon | sound good to me too! | 15:17 |
| xivulon | only requirement is for cking or some other kernel dev to be around when we discusso kernel stuff | 15:18 |
| cjwatson | xivulon: I'll check with Ben in case he has vacant space on his track - doesn't all need to be on the platform track of course | 15:23 |
| xivulon | cjwatson: thanks | 15:23 |
| xivulon | as mentioned by evand, might help to szaka on voip when that happens | 15:25 |
| cjwatson | the other one's actually already on my track | 15:25 |
| cjwatson | I've replied by mail | 15:25 |
| xivulon | souds good, is the migration topic going to be on a separate session? | 15:29 |
| cjwatson | yes | 15:29 |
| xivulon | can we have that too thu-fri pls? | 15:30 |
| xivulon | by the way I was thinking of providing a preliminary script to migrate people over since LVPM has not been update | 15:30 |
| xivulon | d | 15:30 |
| xivulon | should be a matter of copying over files + edit menu.lst/fstab + grub-install | 15:31 |
| xivulon | that is assuming that a target partition is already available | 15:31 |
| xivulon | talking of which... | 15:34 |
| cjwatson | I've made a note about your attendance times | 15:36 |
| xivulon | it might also be possible to migrate to a real partition without installing grub... I.E. reusing C:\ubuntu\disks\boot and only changing kopt root | 15:36 |
| cjwatson | target partition> this is why it needs to be integrated with ubiquity so that it gets a proper partitioner, boot loader installation, etc. | 15:36 |
| cjwatson | otherwise you're going to find yourself gradually reinventing ubiquity, even if it seems like a small job at first | 15:37 |
| xivulon | as mentioned in boston I am in favor of ubiquity integration :) | 15:37 |
| cjwatson | I think it's better to reinstall grub - otherwise a user might quite reasonably think they could delete Windows and then bang goes Ubuntu too | 15:37 |
| xivulon | makes sense | 15:38 |
| xivulon | the script mentioned above is only to "fill the hole" while LVPM and or Ubiquity are ready | 15:38 |
| xivulon | I would assume that the Ubiquity migration functionality is supposed to be used via CD, correct? | 15:56 |
| xivulon | any chance of making that also available as an app (possibly using unpartitioned space and/or partitions with no fs/files)? | 15:57 |
| cjwatson | I don't think it's fair to make assumptions about something that hasn't really been designed yet ... :-) | 15:57 |
| cjwatson | there's no reason you couldn't run ubiquity from a regular system as long as it had something stable to copy from | 15:58 |
| cjwatson | while it needs to know about CDs, it's not intrinsically tied to them | 15:58 |
| xivulon | Ah I was thinking more in terms of feature freeze | 15:59 |
| xivulon | since that would require a new ubiquity to be installed in 8.04 | 15:59 |
| evand | xivulon: I'm confused. What does this have to do with 8.04? | 16:04 |
| xivulon | evand, say users of wubi in 8.04 want to migrate to a real partition, can we still help them out once the migration facility is in ubiquity? | 16:10 |
| xivulon | one way is to wait for ubiquity 8.10 on CD, the other is to let such users install a version of ubiquity to 8.04 with migration facility and use that | 16:10 |
| xivulon | sorry if I wasn't clear | 16:11 |
| xivulon | I guess that they can use LVPM in the meantime, no biggie | 16:14 |
| evand | well, we can discuss this at length at UDS, but regardless of the technical merits, I don't think a brand new feature like this is going to make it into hardy-proposed. | 16:28 |
| xivulon | yeah was expecting that. | 16:30 |
| cjwatson | I'm sure it would be possible to build something unofficial | 16:30 |
| xivulon | since the migration tool will be unofficial anyway for 8.04 users (whethere my own scripts or LVPM) we might also consider some early PPP ubiquity build maybe | 16:31 |
| xivulon | cjwatson you beat me :) | 16:31 |
| xivulon | s/PPP/PPA | 16:32 |
| xivulon | evand I think that Tormod branch is almost ready for hardy.proposed (#136682 | 16:43 |
| xivulon | I would like to do some changes to the panic messages in lupin-casper re #226622 but I guess it is better to keep things separately | 16:44 |
| evand | xivulon: ok, I'll have a look at it and merge it in. | 16:46 |
| xivulon | can we also have ubuntu-testing to play with http://launchpadlibrarian.net/14240142/lupin-support_0.16%2Blocale110_all.deb | 16:46 |
| xivulon | evand I already have a hardy.proposed branch https://code.launchpad.net/~ubuntu-installer/lupin/hardy.proposed (which is now same as final) | 16:49 |
| evand | ok | 16:54 |
| mario_limonciell | evand, i saw that there is a udeb for gptsync. is that actually used anywhere right now/ | 16:54 |
| xivulon | not sure if overriding LANG in /etc/default/locale is desirable | 16:54 |
| xivulon | otherwise code looks good to me :) | 16:56 |
| mario_limonciell | either rdepends doesn't appear to work on udebs, or i don't believe it's in use at all | 16:56 |
| evand | mario_limonciell: as far as I can tell it's done inside partman now: | 17:00 |
| evand | partman-efi (11ubuntu3) feisty; urgency=low | 17:01 |
| evand | * Remove gptsync code; parted does this itself now as of version | 17:01 |
| evand | 1.7.1-3ubuntu3. | 17:01 |
| evand | parted (1.7.1-3ubuntu3) feisty; urgency=low | 17:01 |
| evand | [ Matthew Garrett ] | 17:01 |
| evand | * gptsync.dpatch: Perform automatic GPT/MBR partition table | 17:01 |
| evand | synchronisation (LP: #46853). | 17:01 |
| evand | http://launchpadlibrarian.net/6603773/gpt.diff | 17:01 |
| mario_limonciell | ah hm. i suppose that doesn't handle with recovery partitions well then | 17:01 |
| mario_limonciell | per scott davilla's indication to the mailing list, i started to dig a little bit into it | 17:02 |
| mario_limonciell | i wonder if that patch is literally lifted from gptsync | 17:03 |
| mario_limonciell | or what not | 17:03 |
| cjwatson | I have seen a report of problems with recovery partitions, yes | 17:03 |
| evand | curious, those mails landed in my inbox, but I didn't get them in the ML folder. | 17:03 |
| cjwatson | I understand it was pretty much from gptsync, but there may be some divergence | 17:03 |
| cjwatson | I'd appreciate somebody who can conveniently test it looking into it | 17:04 |
| cjwatson | mario_limonciell: BTW, I've heard reports that Dell is shipping some systems with four primary partitions, which is obviously very inconvenient if you want to install Linux later. Do you know anything about this? | 17:04 |
| mario_limonciell | cjwatson, this i've not heard. | 17:05 |
| mario_limonciell | could you point me to some indications/ i can poke around | 17:05 |
| mario_limonciell | cjwatson, re gptsync, why was the code integrated into partman rather than just living in the udeb and calling that binary in the first place, do you know/ | 17:06 |
| mario_limonciell | [sorry, my shift is broken right now due to vmware messing a few things up] | 17:06 |
| cjwatson | mario_limonciell: unfortunately it was just on IRC | 17:07 |
| cjwatson | mario_limonciell: (in theory) much more robust to do it inline in parted, required fewer scary partman hacks, etc. | 17:07 |
| cjwatson | ensures that people can't screw up their systems by running parted and forgetting to run gptsync | 17:07 |
| mario_limonciell | but less maintainable in the sense of newer gptsync versions etc | 17:07 |
| cjwatson | it's a trade-off | 17:07 |
| mario_limonciell | ah yeah i see. | 17:08 |
| cjwatson | gptsync surely doesn't change *that* often | 17:08 |
| cjwatson | and libparted ought to know everything about how to handle partitions, really | 17:08 |
| mario_limonciell | well unfortunately the version that is in debian and ubuntu doesn't handle the recovery partitions either | 17:08 |
| cjwatson | how are they supposed to be handled? | 17:08 |
| mario_limonciell | i sent a ping to the debian maintainer about updating it | 17:08 |
| mario_limonciell | well it's fixed in the 0.11 release | 17:08 |
| mario_limonciell | he just hasn't touched the package since 2006 | 17:08 |
| mario_limonciell | i'm a bit new to the whole thing, and was looking at it out of curiosity after seeing that there has been these successes with getting ubuntu on the apple tv | 17:10 |
| cjwatson | I suspect the answer is simply to forward-port patches from gptsync one by one | 17:13 |
| mario_limonciell | that's very attainable then. i'll make some notes of this conversation, scott may be convincing me to get one of these atv's in the near future so i may be able to help with that. | 17:14 |
| mario_limonciell | cjwatson, re the 4 partitions, i'll send some e-mails around. if you get any more indications of models or locales this is happening in, could you let me know/ | 17:16 |
| cjwatson | ok, I'll keep my ears open, thanks | 17:16 |
| cjwatson | seems like it should be a general policy to use logical partitions wherever possible | 17:16 |
| mario_limonciell | i agree, but unfortunately some of the overseas factory teams sometimes aren't very cognoscente of the possibility of people needing more than 4 partitions for any reason | 17:18 |
| cjwatson | that's why it needs to be policy rather than left up to individual initiative ;-) | 17:18 |
| mario_limonciell | ;] | 17:18 |
| cjwatson | Ubuntu only creates a primary partition now if there isn't one already, so it should be as well-behaved as you can get in this regard | 17:19 |
| mario_limonciell | cjwatson, initial responses back would like to know which locale at least. do you know where the person(s) who indicated this were located? | 17:39 |
| cjwatson | mario_limonciell: I looked at my logs; it was astronut on #ubuntu-devel, who I believe is in the US | 17:41 |
| cjwatson | 17:40 [Freenode] -!- astronut [n=astronut@sfnc-162-39-87-194.sandhills.us] | 17:41 |
| mario_limonciell | cjwatson, okay. i'll pass that on. thanks | 17:41 |
| cjwatson | he wasn't very clear, but he did claim that it came with four primaries | 17:41 |
| cjwatson | 20:39 <astronut> cjwatson: i'm not sure... thisone has a a tiny fat16, two large ntfs (C: and D:, d has label backup and had some empty folders and some data) and then a hidden fat32 w/ restore data | 17:41 |
| mario_limonciell | both for the tiny fat16 utility partition, the recovery partition, the factory process doesn't allow for non primary (internal limitations), but i'm not sure why a backup partition would have been created, especially being primary | 17:43 |
| cjwatson | can Windows cope with being booted off a logical partition? | 17:49 |
| mario_limonciell | I'm not sure | 17:50 |
| mario_limonciell | even if it couldn't, that backup partition if really necessary should be able to be logical | 17:51 |
| cjwatson | yeah, you'd have thought | 17:52 |
| mario_limonciell | I don't suppose windows can be booted off of a logical partition, because a DOS MBR can't address a logical partition in an extended partition | 17:55 |
| CIA-1 | lowmem: cjwatson * r74 ubuntu/ (3 files in 2 dirs): merge from Debian 1.28 | 18:01 |
| CIA-1 | lowmem: cjwatson * r75 ubuntu/debian-installer-startup.d/S15lowmem: revert to Debian's version | 18:02 |
| CIA-1 | lowmem: cjwatson * r76 ubuntu/debian/changelog: releasing version 1.28ubuntu1 | 18:03 |
| evand | Is there any reason why we don't currently disallow /boot (or by extension /) on XFS in ubiquity or include lilo to use with it in such circumstances? | 20:18 |
| CIA-1 | partman-target: evand * r723 ubuntu/ (debian/changelog finish.d/clear_partitions): | 20:39 |
| CIA-1 | partman-target: * Ensure that if we clear the root partition, / is owned by root:root | 20:39 |
| CIA-1 | partman-target: (LP: #224446). | 20:39 |
| xivulon | Hi tormod, I am Ago | 22:09 |
| xivulon | thanks a lot for the locale patch :) | 22:09 |
| tormod | xivulon: hi! it's an honour - wubi is the greatest thing | 22:12 |
| xivulon | ...after sliced bread | 22:12 |
| xivulon | ...and d-i | 22:12 |
| xivulon | ...and ubiquity | 22:12 |
| xivulon | :) | 22:13 |
| tormod | for linux adoption... since the live cd | 22:13 |
| tormod | or vfat support :) | 22:13 |
| xivulon | you seem capable with c also, and have windows | 22:15 |
| xivulon | there is one bug I cannot understand and have looked at the 200 times | 22:15 |
| xivulon | maybe a pair of fresh eyes would help | 22:15 |
| xivulon | bug #207137 | 22:16 |
| xivulon | there is a stand-alone app (see hampus comments) | 22:16 |
| tormod | I only have a USB drive with an (unbootable) ntfs partition, but I can boot into NT with vfat. | 22:16 |
| xivulon | basically this is a small app to extract a CD to an ISO | 22:17 |
| xivulon | code is very simple: gets a handle for the device, keeps reading it in a loop and pushes the bytes into a file | 22:17 |
| xivulon | yet for some users it does not work (usually at the end) and you get 2 or 3 (different) errors | 22:19 |
| xivulon | code is here: http://launchpadlibrarian.net/14044790/CD2ISO.zip | 22:20 |
| xivulon | compiled with http://www.codeblocks.org/ | 22:21 |
| xivulon | of course the above is an open invitation for anyone else since myself and hampus are out of idea | 22:21 |
| xivulon | s | 22:22 |
| xivulon | evand, you too ^ | 22:22 |
| xivulon | this is the last bug I am tracking for point release (have a few open ones but those should be manageable) | 22:23 |
| xivulon | tormod also added a small comment for #136682 | 22:27 |
| xivulon | ehm c++ | 22:28 |
| cjwatson | evand: we should definitely make it possible to install lilo. GRUB+XFS has a race condition; it may work for some people but it can't currently be made to work across the board | 22:59 |
| Zelut | cjwatson: are you around? | 23:17 |
| cjwatson | Zelut: for a short while | 23:17 |
| cjwatson | it's late for me and I have an early meeting tomorrow | 23:18 |
| cjwatson | feel free to leave a message or send mail | 23:18 |
| Zelut | I was going to do some preseed work to automate an ubuntu-server install. | 23:18 |
| Zelut | anything you might be able to point me to? | 23:18 |
| cjwatson | start with https://help.ubuntu.com/8.04/installation-guide/, and you probably also want to make sure you incorporate each of the things in /preseed/ubuntu-server.seed from an Ubuntu server CD unless you explicitly choose otherwise | 23:21 |
| xivulon | evand panic text re 226622 http://paste.ubuntu.com/10631/ | 23:36 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!