=== roxirc is now known as matttp [00:46] hey guys [00:47] this the channel where we ask about installing ubuntu? [01:58] cjwatson, evand, the option creating issues in update-initramfs was "group_id=0" [11:03] Have filed bug #187282 anyway, the option triggering the failure was "group_id" by the way [11:03] Launchpad bug 187282 in wubi "update-initramfs -u fails with "WARNING: /boot is ro mounted." even if /boot is rw" [Medium,Confirmed] https://launchpad.net/bugs/187282 [11:15] evand, cjwatson, re #175772 unlike stated in the email, I think that the fix simply involves removing the line: [11:15] if [ -f "$loop_file" ]; then [11:15] from update-grub [11:16] * xivulon thinks a bug comment would be better [11:25] evand, cjwatson, ignore my yesterday emails and look at https://bugs.launchpad.net/wubi/+bug/175772/comments/6 [11:25] Launchpad bug 175772 in wubi "Update-grub does not set kopt correctly in loopinstallations" [Medium,In progress] [11:26] was far too late yesterday... === cjwatson_ is now known as cjwatson [13:26] I have noticed that update grub defaults to /dev/hda1 if it cannot find a root_device. [13:26] Shouldn't that be /dev/sda1 which seems to be far more common these days? [13:37] that case should be considered a bug if it ever occurs so I think it's fine to leave it as it is [13:46] cjwatson: do you think it's inappropriate to upload a patched version of sysvinit today, or should I wait until Friday? I've tested it and found no issues, the code also looks ok as far as I can tell. [13:49] ick, I could've worded that better. [13:50] evand: are you aiming to get the whole of wubi into alpha 4? [13:52] ideally [13:52] but I wholly acknowledge it may be too late for that. [13:53] that would be nice :P [13:54] patch to 175772 is trivial (see last comment) [13:54] patch to umountfs is a bit more invasive but should be ok [13:56] for ref last comment in 151579 [14:02] evand: if so, then it's appropriate; if not, it should wait [14:03] evand: if so, of course, then you have a matter of hours [14:04] can you clarify, if what is so? [14:04] cjwatson: ^ [14:06] if you are aiming to get wubi support in alpha 4 [14:06] ok, thanks for the advice. [15:00] partman-target: evand * r694 partman-target.ubuntu/ (commit.d/clear_partitions debian/changelog): [15:00] partman-target: * Rework clear_partitions to not try to remove active mountpoints (LP: [15:00] partman-target: #186711). [15:03] oh good, I was about to ask you about that [15:04] I finished that yesterday, but I woke up this morning to a VM missing /boot and I'm trying to determine now if clear_partitions is the cluprit. [17:36] ubiquity: evand * r2424 ubiquity/ (debian/changelog ubiquity/frontend/gtk_ui.py): [17:36] ubiquity: * Add new nautilus gconf keys to the list we temporarily disable to [17:36] ubiquity: prevent constant pop-ups. [17:45] partman-target: evand * r695 partman-target.ubuntu/debian/changelog: releasing version 52ubuntu3 [17:46] looks good as far as I can tell (I tested several different partition table configurations with it present) [17:47] so my missing /boot was probably me canceling an install after it ran but before the grub-install component had a chance to run [17:47] and forgetting about it === michael_e_brown_ is now known as miichael_e_brown === miichael_e_brown is now known as michael_e_brown [20:25] ubiquity: evand * r2425 ubiquity/ (d-i/manifest debian/changelog): [20:25] ubiquity: * Automatic update of included source packages: console-setup [20:25] ubiquity: 1.21ubuntu2, migration-assistant 0.6.0, partman-target 52ubuntu3, [20:25] ubiquity: user-setup 1.16ubuntu3. [20:29] ubiquity: evand * r2426 ubiquity/ (aclocal.m4 configure configure.ac): Bump to 1.7.5 [20:31] ubiquity: evand * r2427 ubiquity/debian/po/ (79 files): debconf-updatepo [20:36] ubiquity: evand * r2428 ubiquity/debian/changelog: releasing version 1.7.5 [20:46] how can I seed variables from the command line? [20:47] cr3: debconf-communicate [20:47] echo "SET passwd/username evan" | debconf-communicate [20:48] evand: thanks man! [20:48] cr3: anytime [20:48] evand: now lets see if my package works :) [20:48] heh [20:50] evand: I'm getting: 10 hwtest-verify/plugins/exchange_prompt/secure_id doesn't exist [20:50] is it in the templates file of your package? [20:51] evand: yes, but the package is not installed yet [20:52] evand: basically, I want to set the variable and then try dpkg installing it [20:53] to avoid the question being asked? [20:54] evand: yep, and also to test that the variable is being substituted properly. this is ultimately meant to go in the preseed of a network install, so I want to test on the command line because the install takes forever :) [20:55] ah, then I should've pointed you at debconf-set-selections instead [20:55] it takes the same format as a preseed file [20:56] thanks, I'll try that [20:56] so echo "d-i foo/bar string barish" | debconf-set-selections [20:56] you're welcome [20:56] evand: that's even more perfect, I get to test the exact line from my preseed file [21:06] evand: worked on the command line, now to figure out why it doesn't work during the installation [21:06] because the database is locked by the install process [21:07] you can only modify questions as part of a component of the install (sourcing confmodule and using db_set) [21:07] oh, perhaps I misread [21:07] evand: I'm not quite sure I understand [21:08] I think I misread, you are doing this as part of a udeb and you're saying that it's not working within that context, right? [21:08] I thought you were saying using debconf-set-selections wasn't working while the install was running. [21:09] evand: right, so I'm netinstalling and using a preseed which has worked well for me so far. now, I added this line to pressed a package: [21:09] d-i hwtest-verify/plugins/exchange_prompt/secure_id string $secure_id [21:09] (where $secure_id gets substituted at some point, don't worry about it :) [21:09] haha that scared me for a moment [21:09] then, I install the extra packages I need with this line: [21:10] d-i preseed/late_command string apt-install hwtest-server-cli [21:11] shouldn't that be pkgsel/include hwtest-server-ci? [21:11] errm wait no [21:12] evand: in the sample preseed file, I see two ways: pkgsel/include and preseed/late_command. the former is used to install openssh-server in the sample and the latter to install zsh. [21:13] is hwtest-server-cli a udeb? [21:15] pkgsel/include is preferred if you can [21:15] the latter is an example in case you need to do more complicated stuff [21:25] cjwatson: did you get my PM from earlier today? [22:25] evand: no, hwtest-server-cli is not a udeb, does it need to be? [22:26] no, I was curious as if it was a component of the install process the way you installed it would be different. [22:32] evand: does the install log get stored somewhere so that I can review after the installation whether the variable was indeed preseeded? [22:32] /var/log/installer [22:42] cjwatson: You around? [22:54] evand: I didn't find anything interesting relating to preseeding my variables in /var/log/installer, and I do have DEBCONF_DEBUG=5 in my kernel parameters [22:58] cr3: hrm, I suppose you could copy the database out as part of late_command. cjwatson might know of a more elegant solution than that. [22:58] evand: it's getting late, and even later for cjwatson. I'll try again tomorrow at a more reasonable time. thanks for all the help dude! [22:58] yeah, I just noticed the time :) [22:59] anytime, goodnight