[01:24] cjwatson, evand: is the trunk branch now for karmic? would anyone object to an attempt to consolidate the oem-config/ubiquity codebase? or wait after UDS? === TheMuso_ is now known as TheMuso [07:18] ubiquity: superm1 * r3243 mythbuntu-ubiquity/ (debian/changelog ubiquity/frontend/mythbuntu_ui.py): [07:18] ubiquity: * Mythbuntu Frontend: [07:18] ubiquity: - Explicitly allow the user to progress when the graphics driver is [07:18] ubiquity: selected. For Wubi installs that contain an NVIDIA or AMD graphics [07:18] ubiquity: card, the installer would be stuck on this page because of being in [07:18] ubiquity: automatic mode. This page is not shown on Non-NVIDIA or AMD systems. [07:44] ^ i'm not sure if evan was planning another upload, so if so i'd like to sneak that in if possible. [08:29] xivulon: morning [08:30] morning [09:00] arg virtualbox crashes in jaunty as soon as I load a kernel [09:10] shtylman: I'd advise doing any work in a separate branch; it's easy to merge [09:10] shtylman: there's no need to wait until after UDS though [09:12] shtylman: the tricky question is how to merge the frontend implementations, given that (a) there are common methods in there (b) there are methods that need to diverge (e.g. the implementation of the language question) [09:27] davmor2: http://people.ubuntu.com/~evand/wubi/jaunty/wubi-r127.exe - Can you give that a go? === evand1 is now known as evand [09:28] evand1: will do just following some instructions that xivulon gave me first [09:30] evand have sent you a few emails [09:30] have to go shortly [09:30] see r127, and probably we need to remove /S flag from compact. You might want to play with compact a bit from within windows [09:31] It seems that /S does not actually recurse on the last argument of the compact command, but on the current dir... [09:32] have sent a patch via email to remove /S. have to go now [09:37] wubi: evand * r128 trunk/ (debian/changelog src/wubi/backends/win32/backend.py): [09:37] wubi: Remove the /S flag from compact calls. It forces compact to run [09:37] wubi: against the current directory. [09:40] davmor2: http://people.ubuntu.com/~evand/wubi/jaunty/wubi-r128.exe [09:40] evand np's thanks [09:59] evand: the lag just vanished from the install so that is fixed :) [10:00] great [10:00] evand: it was taking like 40 minutes to install [10:00] just the windows side [10:04] right linux part now [10:05] go on, go on little wubi install [10:05] Yay coppying files [10:07] great [10:57] davmor2 so last patch was ok right? [10:57] perfic [10:58] cool, that weird dos command syntax got me [10:58] xivulon: by the looks of things slangsek has already had the fix pushed too [10:59] yeah noticed the comment, great [11:00] ps did you test 344151 (comment 4), basically there was an issue when the windows username contains non-ascii chars, hopefully should be ok [11:03] Pass [11:03] I'll have a look at it now [11:04] you have to create a windows user with accented chars in the name and install from there [11:05] you should see the username box pre-filled with a sanitized version of the username with the non-ascii chars omitted [11:17] xivulon: Right I'm doing phillip where the first i has two .. above it and the second has ^ above it that should do it correct? [11:23] yes [11:23] xivulon: right so it shows up in wubi as phllp is that correct behaviour? [11:23] that is as expected, otherwise it would have crashed [11:24] cool in that case it works :) [11:24] thanks, please comment on the bug [11:24] I don't know about "correct", but it's better than crashing :-) [11:24] :) [11:24] (ideal might be to attempt to transliterate to ASCII, or to fix the whole installer workflow so that non-ASCII usernames work - we have a bug for that already) [11:25] by the way does the corrent locale affect the range a-z in a regex? [11:25] bug 344151 [11:25] Launchpad bug 344151 in wubi "[jaunty] Wubi crashes on startup to UnicodeDecodeError when not using English" [High,Fix released] https://launchpad.net/bugs/344151 [11:25] xivulon: yes [11:26] xivulon: if you want "lower-case character" in a locale-sensitive way, you should use [[:lower:]] [11:26] i do transform the string to lower before substituting in regex [11:26] [a-z] can give you broken results; for example in some locales it can include A-Y as well [11:27] the important thing is that it doesn't include non-ascii chars, if it is upper case it will be caught later [11:27] it might include non-ASCII characters in some locales [11:27] of course so might [[:lower:]] [11:28] [a-z] is probably a bit more likely to do what you want [11:28] thta is what I use currently [11:28] xivulon: I've added to the bug [11:29] but really it would be better to be more explicit (not for jaunty) [11:29] cjwatson: maybe a good topic for karmic uds :) [11:29] hardly necessary to bring up at UDS [11:29] UDS discussions are for big complicated things not individual bugs [11:30] Right so it's only me that's lost in this discussion then ;) [11:31] I will fix it properly after release, that was a quick patch in an internet caffe [11:31] xivulon: well I'm just running the rest of the install to ensure there are no issues [11:32] davmor2 thanks a lot for everything, most useful [11:32] ubiquity: evand * r3244 ubiquity/ (debian/changelog ubiquity/frontend/gtk_ui.py): [11:32] ubiquity: Initially set the extra autopartition options to an insensitive [11:32] ubiquity: state (LP: #362914). [11:33] .decode('ascii', 'ignore') would be one straightforward way in Python to throw away all non-ASCII characters [11:33] I tried that but it failed [11:34] WFM [11:34] maybe you tried .encode? [11:34] >>> 'dédéd'.decode('ascii', 'ignore') [11:34] u'ddd' [11:35] ah yes I used encode! bugge [11:35] r [11:35] I always get confused [11:36] evand not sure if you want to patch that, probably it's not worth it at this stage [13:18] evand was there any progress with the signing certificate? #204834 [13:20] xivulon: checking into it now [13:23] it was "within 24 hours" on 15 Apr [13:23] lamont would be the guy to ask === dpm_ is now known as dpm [14:57] evand: where should bug 363335 be reassigned? [14:57] Launchpad bug 363335 in installation-guide "launching Emenu causes "invalid CD detected" message" [Undecided,New] https://launchpad.net/bugs/363335 [15:00] cjwatson: sorted [15:00] thanks [15:10] kirkland: I heard through the grapevine that you have a new version of kvm that was rejected for upload to ubuntu available somewhere. Could you point me at it? I'm having issues whereby KVM only continues working when it has focus. [16:11] Lovely. I'm running into an issue where the installer is showing the resize widget for use_biggest_free. It does not cause the installer to crash or make the option unusable, though. [16:11] I'm looking into a fix [16:23] ubiquity: evand * r3245 ubiquity/ (3 files in 3 dirs): [16:23] ubiquity: Properly reset the 'after' partition bar when switching between 'use [16:23] ubiquity: the largest continuous free space' and 'install them side by side.' [16:24] I'm just double checking that. [16:24] is this release-critical? images have already been rolled ... [16:25] I don't think it's release critical, but it will look odd [16:25] basically, you'll get an incorrect display in the after partition bar (the resize widget with the wrong data) instead of the correct display for use_biggest_free. [16:27] best mention it on #ubuntu-release. I think it's probably too late unless we need to reroll for something else ... [16:31] ok, noted [16:31] thanks [16:36] evand: huh? [16:36] evand: that's a weird one [16:36] I used to get it a while ago [16:36] evand: btw, i had 4 kvm's that i let run for a week [16:36] and it's back with the latest kvm version [16:36] evand: i killed them after 7 days of uptime [16:37] wow, glad to hear. Equally I can no longer reproduce that bug. [16:37] evand: that is in deed a strange one [16:37] evand: can you tell me a bit more about how you're launching them? [16:37] kvm -m 768 -boot dc -hda hda.img -hdb hdc.img -no-quit -k en-us drom jaunty-desktop-amd64-20090420.iso [16:39] evand: why the -k en-us ? [16:39] evand: that should not be necessary [16:40] I think that's leftover from working around an old bug [16:40] evand: yeah, i fixed that in intrepid ;-) [16:43] ubiquity: evand * r3246 ubiquity/debian/changelog: Add LP bug reference. [16:59] evand, so it's not likely that these last changes from today for ubiquity will make it in for gold media? (I'm of course hoping to get the mythbuntu one in, but didn't want to raise it and cause re-rolls for everything else if it came to it) [16:59] not at present ... [17:13] would it be possible to upload the new ubiquity then to the archive and just request mythbuntu to get re-rolled with it and the others opportunistically? or is the preference to not upload it to the archive at all then? [17:56] evand, cjwatson: Have you ever noticed ? [18:32] mpt: ages ago, but in general trying to trawl through Launchpad for blueprints that we didn't write is a pointless exercise and I rarely bother [18:32] yeah [18:33] can't mark 'em implemented [18:33] can't mark 'em as duplicates [18:33] "This project uses blueprints" is disabled for ubiquity [18:33] can't shoot 'em [18:33] superm1: we wouldn't want to have packages out of sync with the images for final release [18:34] cjwatson, ah i see. [18:45] evand they just filed 364166, have posted a possible patch, not sure if you want to push it [18:47] xivulon: I suggest discussing whether or not it's release critical with slangasek [18:49] will do, might be good to branch off and prepare a build though, and have davmor2 play with it, just in case [18:55] xivulon: busy now :( might have time tomorrow [19:01] I can only test (fairly late) tonight myself