[10:49] <xivulon> TheMuso thanks will look into that
[10:49] <xivulon> basically current initrd ignores ro/rw flag whenever root is on ntfs and always mount rw
[10:50] <xivulon> that was because 1. ntfs-3g could not be remounted (then it could be remounted, then again it could not) 2. there is not fsck
[10:51] <xivulon> for fat I assume it gets mounted r/o and mounthost fails (at the moment there is a bug in that I was using awk which might not available at that point)
[10:56] <xivulon> I have in a note that the relevant initramfs code is:
[10:56] <xivulon> if [ ${readonly} = y ] && ([ -z "$LOOP" ] || [ "${FSTYPE#ntfs}" = "$FSTYPE" ]); then roflag=-r
[11:03] <xivulon> here is a revised (untested) /etc/init.d/mounthost (lupin-support package): http://paste.ubuntu-nl.org/59231/
[11:05] <xivulon> if mounthost above does not fix it, then it might be necessary to remove "|| [ "${FSTYPE#ntfs}" = "$FSTYPE" ]" from initramfs-tools/scripts/local
[11:22] <xivulon> by the way selecting more than 4GB in fvat should simply split root into multiple <4GB files (/ubuntu/disks/root.disk /ubuntu/disks/usr.disk /ubuntu/disks/home.disk...)
[11:37] <xivulon> mounthost take-2http://paste.ubuntu-nl.org/59234/
[11:45] <TheMuso> xivulon: Yes I saw that, I eventually saw that the disk size was not the issue.
[11:52] <xivulon> TheMuso, I would assume that installation goes fine but you cannot reboot into the installed system
[11:52] <xivulon> in case try to replace /etc/init.d/mounthost with http://paste.ubuntu-nl.org/59234/
[11:53] <xivulon> if that fails, second thing to try is to edit initramfs-tools/scripts/local removing check on ntfs
[11:53] <xivulon> and regenerate initrd
[12:43] <xivulon> What is the difference between console-keymaps-at/keymap and console-setup/layoutcode?
[12:45] <xivulon> On the same subject, is it necessary to preseed console-setup/modelcode?
[13:42] <cjwatson> xivulon: console-keymaps-at/keymap is obsolete; don't preseed it. You are unlikely to need to preseed console-setup/modelcode.
[13:43] <xivulon> cjwatson can you pls expand on "unlikely"
[13:44] <cjwatson> "not unless you have extremely unusual requirements and know what you're doing"
[13:44] <cjwatson> just leave it at the default
[13:47] <xivulon> default meaning "do not preseed". Only thing I want avoid is an interactive question.
[13:50] <cjwatson> it won't be asked on any platform you support
[13:51] <cjwatson> it's asked on m68k Sun, powerpc CHRP, sparc, and anything unrecognised
[13:51] <xivulon> great
[13:51] <evand> cjwatson: You moved bug 200423 from casper to ubiquity, but I'm not sure I agree.  Wouldn't it be best to do any sufficient memory checks in casper, where we don't have to worry about having enough memory to load X and ubiquity initially?
[13:51] <evand> Or perhaps you had a different solution in mind?
[13:51] <xivulon> last q on preseed, a user long ago suggested to set clock-setup/utc to false as being more appropriate for windows setup
[13:52] <cjwatson> evand: feel free to move it back if you prefer that; I was just thinking that casper is not in a position to make ubiquity crash differently
[13:52] <cjwatson> clock-setup/utc will already be set to false if os-prober detects Windows
[13:52] <cjwatson> so assuming that it *does* detect Windows in the wubi environment (is that true?0
[13:52] <cjwatson> )
[13:53] <cjwatson> if that's not the case, yes, preseed it
[13:53] <xivulon> do you suggest me to skip preseeding then or preseed it myself to false?
[13:53] <xivulon> I guess I'll preseed then
[13:53] <ubotu> Launchpad bug 200423 in casper "The installer CD doesn't warn the user that they dont have enough RAM to continue." [Wishlist,Confirmed] https://launchpad.net/bugs/200423
[13:54] <evand> you're a few minutes too late ubotu.
[13:54] <xivulon> evand see ^^ for mounthost fix, I have not tested it, can do so tonight and merge it in
[13:54] <xivulon> fell free to beat me to it ;)
[13:55] <xivulon> feel
[13:55] <xivulon> that should address bug #198007
[13:55] <ubotu> Launchpad bug 198007 in lupin "Don't use awk in early mounthost" [Low,Confirmed] https://launchpad.net/bugs/198007
[13:56] <xivulon> well done ubotu...
[13:56] <evand> heh, great
[13:59] <xivulon> ah I see cjwatson already addressed 191122!
[13:59] <simosx> hi! am looking into translating the ubuntu installer (what's the package that runs as soon as the CD boots up).
[14:00] <xivulon> sorry for the set-x I discussed with evand and decided to leave them for the alphas
[14:00] <xivulon> umenu/wubi
[14:00] <xivulon> simonsx https://translations.launchpad.net/wubi/ and https://translations.launchpad.net/umenu
[14:00] <xivulon> note that you need to be a member of Ubuntu Translators group
[14:01] <simosx> xivulon: many thanks. am member of the group.
[14:01] <xivulon> your contributions are most welcome
[14:17] <simosx> xivulon: I translated umenu but did not find the message:
[14:17] <simosx> "Press F4 to select alternative start-up and installation modes"
[14:17] <simosx> and "Test memory"
[14:17] <simosx> wubi is the tool that starts in Windows, isn't it?
[14:19] <cjwatson> simosx: those are in the gfxboot-theme-ubuntu package in Ubuntu
[14:19] <cjwatson> https://translations.launchpad.net/ubuntu/hardy/+source/gfxboot-theme-ubuntu
[14:19] <cjwatson> simosx: you should also look at https://translations.launchpad.net/ubuntu/hardy/+source/debian-installer
[14:19] <cjwatson> xivulon: not everyone uses Windows ;-)
[14:21] <simosx> cjwatson: thanks. I always try to append the package name to the URL https://translations.edge.launchpad.net/gfxboot-theme-ubuntu and it comes up negative.
[14:22] <cjwatson> that's for projects, not for packages in Ubuntu
[14:22] <cjwatson> two different namespaces
[14:25] <simosx> oh, I'll have to check more stuff again.
[14:26] <simosx> the UI in Ubuntu is almost all translated with the exception of a few messages. I am trying to figure out what package they come from and it's a tough process.
[14:27] <simosx> is there an irc channel where translators are around? or some input for this?
[14:27] <xivulon> ah yes I misread simosx original msg, anyway wubi/umenu translations would be quite handy...
[14:28] <xivulon> #ubuntu-translators
[14:29] <simosx> xivulon: that's good.
[14:34] <cjwatson> simosx: if they're installer strings, you're welcome to ask here
[14:55] <jstrunk> How can I regenerate the /etc/fstab that d-i generates during installation?
[14:59] <xivulon> jstrunk: http://bazaar.launchpad.net/~ago/partman-auto-loop/lupin-support/annotate/ago%40nbago-20080114235447-9rmsbro57lb2zsru?file_id=autopartitionloop-20070730115247-m21lgboyh63uf6mv-2
[14:59] <xivulon> see line 121-133
[15:00] <cjwatson> huh? that is not helpful
[15:00] <cjwatson> NOT EVERYBODY USES WUBI :-)
[15:01] <cjwatson> (that code does not work outside d-i)
[15:01] <cjwatson> jstrunk: there isn't really a way to do so outside the context of the installer, I'm afraid
[15:02] <xivulon> not sure jstrunk is outside d-i, in case I guess it might be possible to replicate /lib/partaman/fstab.d and run those scripts as indicated above
[15:03] <jstrunk> I am thinking about using systemimager to clone a system to about 200 workstations.
[15:03] <cjwatson> xivulon: please find this sort of information out before confusing people
[15:04] <cjwatson> replicating /lib/partman/fstab.d will not help outside d-i; parted_server won't be running and the amount of infrastructure you need to do that is enormous
[15:04] <jstrunk> It looks like the only things that change on disk with different hardware are Xorg, /etc/blkid.tab, and /etc/fstab
[15:04] <cjwatson> jstrunk: I'd suggest just writing it by hand, with the obvious boilerplate as constructed by d-i
[15:04] <cjwatson> ("by hand" => in code obviously)
[15:05] <cjwatson> we don't have a canned way to regenerate it in your type of environment
[15:06] <jstrunk> Thank you.
[15:06] <cjwatson> sorry that the answer isn't better, but really systemimager is going to need to supply identifiers for the partitions it's installing onto
[15:07] <xivulon> well apologies for any confusion, was simply pointing out to a code section I am familiar with that actually generates fstab (in d-i)
[15:07] <jstrunk> it can handle that, but I don't think it will handle the cdrom and floppy drives correctly.
[15:08] <jstrunk> I guess I could probe those and append them to what systemimager gives me.
[15:09] <cjwatson> yeah, I can see how it would be nice to reuse d-i's code for that
[15:11] <cjwatson> http://bazaar.launchpad.net/~ubuntu-core-dev/partman-target/ubuntu/annotate/cjwatson@canonical.com-20080306221605-6cuu03xjs8zh4g3d?file_id=x_Arch_Librarian_<arch@canonical.com> is the file in question (partman-target/finish.d/fstab_removable_media_entries), but you'd have to extract list-devices from debian-installer-utils as well, and remove the use of mapdevfs
[15:11] <cjwatson> and generally munge it about
[15:14] <jstrunk> Thank you for finding this for me.
[15:51] <CIA-23> ubiquity: evand * r2560 ubiquity/ (debian/changelog ubiquity/zoommap.py):
[15:51] <CIA-23> ubiquity: * Use a black border around tzmap points to assist the visually impaired.
[15:51] <CIA-23> ubiquity: * Flash the currently selected location on the tzmap so that it's easier to
[15:51] <CIA-23> ubiquity:  spot.
[15:51] <CIA-23> ubiquity: * When clicking on the tzmap, if the closest match is already selected, take
[15:51] <CIA-23> ubiquity:  the next closest.
[16:22] <CIA-23> ubiquity: evand * r2561 ubiquity/ (debian/changelog gui/glade/stepLocation.glade):
[16:22] <CIA-23> ubiquity: * Place the tzmap in an AspectFrame so it doesn't look distorted on
[16:22] <CIA-23> ubiquity:  widescreen displays (LP: #200152).
[16:45] <CIA-23> partman-target: evand * r712 ubuntu/ (debian/changelog finish.d/clear_partitions): * Whitelist /usr/local in clear_partitions (LP: #200582).
[17:13] <cjwatson> evand: sorry to be a pain, but /var/local too?
[17:13] <cjwatson> (it's in the FHS I believe)
[17:13] <evand> ah, I did a quick search through the FHS docs and didn't come across that one.  Will do.
[17:14] <mario_limonciell> evand, did you get a chance to talk to cjwatson about adding a preseed option for not setting the active partition on boot?
[17:14] <evand> oh, thanks for reminding me
[17:14] <cjwatson> I have an interview to conduct now, but briefly, sounds like a good idea; would need to be done in both grub-installer and lilo-installer, I expect
[17:15] <evand> ok, I'll take care of putting it in then.
[17:15] <mario_limonciell> thanks
[17:17] <evand> ah, it is indeed in the FHS
[17:22] <CIA-23> partman-target: evand * r713 ubuntu/ (debian/changelog finish.d/clear_partitions): * Whitelist /var/local in clear_partitions.
[18:35] <mario_limonciell> evand, it looks like oem-config didn't get installed in the noninteractive install
[18:43] <evand> mario_limonciell: ok, on it
[18:44] <mario_limonciell> thanks
[19:22] <xivulon> evand, have asked bean123 (grub4dos author) to have a look at truncation + other bug that came up today http://ubuntuforums.org/showthread.php?t=720618
[19:22] <xivulon> hopefully new code will be available shortly
[19:22] <evand> ah, nice.  Thanks
[19:28] <xivulon> TheMuso did you have a chance to test if mounthost code solves your vfat issues?
[20:45] <CIA-23> grub-installer: evand * r729 ubuntu/ (3 files in 2 dirs):
[20:45] <CIA-23> grub-installer: * Added the grub-installer/make_active debconf question. When
[20:45] <CIA-23> grub-installer:  preseeded false, this will skip the code that marks a partition as
[20:45] <CIA-23> grub-installer:  bootable.
[20:58] <evand> lilo on the way
[20:59] <mario_limonciell> evand, will that be testable in tomorrow's daily?
[20:59] <mario_limonciell> er when were you planning another ubiquity upload?
[21:02] <evand> mario_limonciell: if I can I'll do a ubiquity upload tonight.  I might have to run some tzmap changes by slangasek though.
[21:02] <mario_limonciell> okay sounds sane
[21:02] <evand> but that should be straightforward
[21:02] <mario_limonciell> well most ideally would like to see noninteractive fixed at the same time
[21:02] <mario_limonciell> to test both sequentially
[21:02] <mario_limonciell> so if it takes another day or so to sort that out first, that's more preferential
[21:05] <TheMuso> xivulon: Not yet, I had to go to bed. I'll have a look today, unless its been resolved.
[21:05] <mario_limonciell> glancing through the differences in gtk_ui and noninteractive, there are a variety of things that are done only in gtk_ui in the self.oem_config block
[21:07] <alefteris> cjwatson, I was not able to translate the "Normal" menu option that opens from gfxboot-theme-ubuntu F4 menu. Is this string unstranslatable?
[21:24] <cr3> is there a way to strip -updates from /target/etc/apt/sources.list from the preseed file?
[21:24] <cr3> I know there's something like this for stripping restricted: d-i apt-setup/restricted boolean false
[21:24] <cr3> I just can't find anything about updates
[21:25] <cr3> worst case, I can always do it in the late-command
[21:46] <TheMuso> evand: Alright, with beta approaching, is there anything I can do to try and get a11y working in only-ubiquity?
[22:09] <cr3> also, might there be a way to strip deb-src lines in sources.list from the preseed file?
[22:29] <Baldone> hello
[22:30] <Baldone> i am having problem installing ubuntu on my laptop
[22:44] <cjwatson> alefteris: err, yeah, oops, please file a bug
[22:44] <alefteris> ok
[22:51] <ago> evand lupin rev 91 is up
[23:02] <ago> TheMuso, updated mounthost code: http://bazaar.launchpad.net/~ubuntu-installer/lupin/hardy/annotate/agostino.russo%40gmail.com-20080311225057-2kq30onjg2p4wsqe?file_id=lupinsupport.mountho-20071025014343-c5dddjlu6r8jume9-1
[23:03] <ago> umenu rev 21
[23:04]  * ago needs CIA...
[23:05] <TheMuso> ago: Pulling the branch now.
[23:08] <TheMuso> ago: When I get the desktop cd synced, I'
[23:08] <TheMuso> ago: When I get the desktop cd synced, I'
[23:08] <TheMuso> gah
[23:08] <TheMuso> I'll make a custom desktop image and test it to see if it helps.
[23:10] <ago> TheMuso,  if you have already a loopinstallation, you can simply replace /etc/init.d/mounthost
[23:10] <ago> and see if you can boot with that
[23:11] <ago> do you recommend preseeding netcfg/dhcp_timeout?
[23:11] <TheMuso> ago: Right, just got to get into the install first, but that shouldn't be too difficult.
[23:11]  * TheMuso boots the box.
[23:12] <TheMuso> I can't answer that one I'm sorry.
[23:12] <ago> TheMuso there is no need to do a custom image anyway, install with an old ISO then mount root.disk and replace mounthost
[23:12] <TheMuso> ago: I know, but I was thinking of some other stuff I wanted to do, but that can wait.
[23:13] <ago> was just about to write that...
[23:58] <ago> evand, what would you suggest re #201046?