[00:08] umountroot take 2: http://paste.ubuntu.com/6726/ === cr3_ is now known as cr3 [09:09] saispo: so *what* exactly went wrong with germinate 0.44? What were the errors? [09:34] ubiquity: cjwatson * r2640 ubiquity/ (debian/changelog ubiquity/frontend/kde_ui.py): * KDE frontend: Translate the partition create dialog too. [10:26] ubiquity: cjwatson * r2641 ubiquity/ (3 files in 2 dirs): * Fix crash if opening /cdrom/.disk/info fails (with Jonathan and Evan). [10:30] cjwatson, I would appreciate some help with bug #204133 [10:30] Launchpad bug 204133 in wubi "wubi install unusable - Buffer I/O error on device loop0" [High,In progress] https://launchpad.net/bugs/204133 [10:31] myself, evand and cking are on it, but the resolution might not be trivial (see cking comments) [10:41] and it would involve some delicate files [10:44] I think adding me to the mix is unlikely to help at this point, frankly [10:44] I do feel that adding complexity to the system (as you suggest in your last comment) will only make it worse and even harder to debug [10:45] but honestly, my head hurts just looking at that bug and I need to sort out recruiting catch-up today [10:46] I'll wait for cking and evand to discuss further, if you have any idea on how to handle it, please let us know [10:46] it sounds to me as if Colin K simply misunderstood the order of events, and that talking to *him* on IRC would help [10:46] he's on UK timezone [10:46] ah is up on #ubuntu-kernel... [10:46] right, he'll be *much* better at dealing with this than going through me [10:48] ok, wanted also to make sure you knew about the bug [10:49] I do ;-) [10:52] ..me neither. I think we may need some expertise from someone like Colin Watson. [11:54] cjwatson: excuse me i'm not @work today [11:54] i can tell you that on monday [11:55] saispo: it's a shame you didn't tell me at the time, rather than just saying "there were some errors" [11:55] saispo: this means that the chances of me being able to fix it for 8.04 are greatly reduced [11:55] yep :/ it's a python error [11:55] are you building mythbuntu? [11:55] no [11:56] ok [11:56] , i build my own custom cd [11:56] don't move, i try to connect to my remote server [11:56] you should have filed a bug with the error dump [11:56] no, don't take the time :/ i have a lot of work yesterday [11:56] i powerup my laptop [12:07] cjwatson: can i copy/paste it here ? [12:09] paste.ubuntu.com [12:12] http://paste.ubuntu.com/6750/ [12:12] no problem with germinate 0.43 [12:15] huh, that's odd [12:16] eek, uncommitted change [12:16] sorry about that, will fix [12:17] ok :) [12:31] saispo: 0.45 making its way into the archive now; sorry about that glitch [12:34] no problem :) [12:34] thanks cjwatson [14:11] evand ping [14:12] xivulon: pong [14:13] ah 2 things,1 you may want to review the logs of #ubuntu-kernel we discussed on umountroot [14:14] 2 I have a new report of segfault, that is with a 2.36 build [14:14] good news is that I have a potential idea [14:14] would need you to upload a build to have them test it [14:15] can you pls replace src/wubi/info_networking.nsh with http://paste.ubuntu.com/6759/ [14:15] and upload a build? [14:15] will do [14:15] thx [14:25] cjwatson: fwiw, given slangasek's comments in bug 215347, I think I'm going to have to revert my changes to partman-target and ubiquity and hold them and the fix for the point that he raises in a branch until Intrepid. [14:25] Launchpad bug 215347 in ubiquity "freeze exception: clear_partitions warning." [Undecided,New] https://launchpad.net/bugs/215347 [14:25] evand, the above is only for having the users test with more debugging info, not really a fix per se (in case I wasn't lear) [14:31] evand: I'll follow up in the bug [14:31] xivulon: http://people.ubuntu.com/~evand/tmp/Wubi-8.04-beta-rev482.exe [14:33] great thx: http://ubuntuforums.org/showthread.php?t=751366 === evand_ is now known as evand [15:07] evand pls follow discussion on #ubuntu-kernel [15:18] debian-installer: cjwatson * r906 ubuntu/ (9 files in 4 dirs): * Move to 2.6.24-16 kernels. [15:20] debian-installer: cjwatson * r907 ubuntu/debian/changelog: releasing version 20070308ubuntu38 [15:21] xivulon: indeed [15:37] well apparently cking was not able to reproduce fs corruption, which is good news [15:38] but we are not sure why, since sync operations are slow and do not seem to show much respect for sysctl settings... [16:34] is it intended that the language selection on bootup on live cd does not any more affect LANG? now the Install icon isn't translated, even though the installer apparently gets its language info from the kernel parameter. [16:34] this has changed during the last few days, I took screenshots on 5th where the Install icon was translated (together with the menu entries etc. which have translations included even without language packs) [16:35] there was a bug about that, I'd be investigating it if I weren't on the phone [16:36] oh, I bet casper needs to be changed, sigh [16:37] ok. [16:39] argh. hard. [16:49] Mirv: bug 213930, FWIW [16:49] Launchpad bug 213930 in gfxboot-theme-ubuntu "Language selection does not affect live system, only menu" [Undecided,New] https://launchpad.net/bugs/213930 [16:59] cjwatson: thanks, I tried to find that [17:07] Evand in the bug report I do not see the debugging message I added, there should be for instance one "DetectNetworkSettings" after profile name [17:13] 7hrm [17:22] xivulon: fixed and tested, http://people.ubuntu.com/~evand/tmp/Wubi-8.04-beta-rev482.exe . Sorry about that. [17:22] np [17:22] err wait [17:22] that is built with 2.36 correct? [17:22] waiting [17:22] yikes, I just put that in the wrong place [17:22] ok, done [17:22] same link? [17:22] yes [17:23] thx [17:23] and yes, it is [17:47] evand I see lupin 0.15 in the manifest but only 0.14 in LP [17:49] forgot to push, fixed [17:49] thanks [17:53] bdmurray: would you mind testing the lates daily ISO and see if you can boot without hacks? [17:54] xivulon: Do you mean install using wubi from the latest daily and then see if I can boot normally? [17:55] yes please [17:55] The Ubuntu daily-live? [17:55] yes, we have some sysctl changes would like to know if they are sufficient [17:56] much appreciated [18:04] xivulon: I forget does it need to be an i386 iso? [18:04] you should use the same arch that was causing problems originally [18:26] partman-target: evand * r716 ubuntu/check.d/partitions_formatted: Fix output of partitions_formatted. [18:34] xivulon: I've just noticed that if you press "ok" when presented with "not enough free space" the setup still comes up. [18:38] bdmurray: can you post the log (in %temp%)? [18:39] evand, in my segfault hunt I found that most reports segfault at either registry_reader.nsh or the tooltip plugin [18:40] that is where the program stops, not necessarily where the problem starts [18:40] but there are new versions for both and it might be worth to try and update them, as usual it's a shot in the dark [18:42] xivulon: If we hope to get them in the final version for Hardy, we'll need to demonstrate that they actually fix the bugs in question. Now that we're in FinalFreeze this is especially true. [18:42] Are these install attempts using the new NSIS? [18:43] last 2 reports yes (482 should be using 2.36 unless I made a mistake) [18:44] argh [18:44] hard evidence is going to be difficult, will talk to the debian mainteiner of nsis in case he has any tip [18:44] I still cannot reproduce [18:45] I guess though there is no harm in doing a temp build with 2.36 and new tooltips/registry_reader and pass it to users with problems [18:45] Getting it in otherwise is going to be equally difficult, unless slangasek is more willing to let such changes in than I think as they don't have any dependencies in the archive. [18:45] yeah, definitely [18:46] Even for our own sanity at this point, we shouldn't be throwing entirely new versions of things in to try to solve the problems of a minority. [18:47] Well in the case of registry_reader the delta should be small... I already spotted a couple of suspicious differences... [18:49] I didn't write that part of the code, it's from Ecology2007, and I he sourced from an nsis script that has since been updated [18:51] ok [18:54] also neither is actually that essential, one is for tooltip, one is to detect ip/gateway to preseed in case there is no dhcp [18:56] right, but the new versions could trigger segfaults as well, no? [18:57] err could cause [19:00] yes, well I will create a branch with new tooltips and new reg_reader code with some special builds, better have it ready [19:00] if it solves some real user cases without too much troubles we will resume the discussion [19:00] fantastic, thanks [19:01] any word from newz on getting a direct download link for wubi on the website? [19:01] no I never talked to him [19:01] oh? I thought you were in discussions with him and slangasek, or was that just for metalink? [19:02] I only taled to slangasek re metalinks [19:02] on irc [19:02] xivulon: I've having a bit of hard time recreating it as I started cleaning up the drive. The free message said "4999 MB < 5000 MB" so maybe it only continues at the specific amount of free space? [19:02] ah you need 1 MB!!! [19:03] you can do it! [19:03] (probably 2 for some headroom) [19:04] That's not a problem, I thought we were talking about the fact that clicking on "OK" lets you proceed when you have 4999 MB free [19:04] ah sure, I will need the log for that ^ [19:04] I'll see what I can do [19:04] the log is in your user temp folder [19:05] %temp% [19:08] bdmurray: did you have a partial installation when you run the installer? or was that after having uninstalled? [19:08] well I'll see from the log... [19:11] evand on 204133 I think that the fs corruption was a coincidence (see my last comment), coupled with cking comments on irc about his inability to reproduce fs corruption, that is encouraging [19:14] indeed, here's hoping [19:14] xivulon: does http://pastebin.com/f56adc04b [19:14] ...have what you are looking for? [19:14] I'll start to put wubi through a stress test as well. [19:15] xivulon: re download button> So you never got a response to your email on the 1st to Matthew about the matter? [19:17] bdmurray, thanks [19:17] xivulon: I didn't realize the log was appended to [19:18] evand no, unless I missed it [19:18] but do not think so (no = no email response) [19:18] ok, I'll talk to him and see what's up. [19:18] thx [19:20] bdmurray I think that occurred because you where just at 5000 free [19:20] and when the log was written you dropped below (I left a second check just in case, but without quitting) [19:22] will fix it! [19:23] if you start with 4999 for instance the application should quit [19:23] in fact that is what happened in the later runs of yours [19:25] have to go, bdmurray please add a comment to 204133 when you are done over sysctl, if it does not work you may want to fiddle with the sysctl settings in /etc/init.d/lupin-sysctl and rise the numbers in there [19:26] see cking comment for a link on the topic [19:26] xivulon: yes, I've seen them [19:26] thanks a lot! [19:26] xivulon: after installing I click "close" and it is not responsding do you want the same log file? [19:27] hmm after installing you should see the reboot dialog, and the only button should be "finish" [19:29] It said / says "Installation Complete" (w/o the green bar all the way full as we've discussed) and only button available was "Close" [19:29] It also said "Completed" above the green bar [19:30] hmm strange it should move automatically to the reboot page [19:30] yeah send the log [19:31] maybe a screenshot too [19:31] okay, should I use the same green bar bug or a new one? [19:32] that will do [19:43] xivulon: ok, he's committed to having it on the "alternate install options" page, and he's going to start a discussion on whether its appropriate to have on the main downloads page. [19:46] evand thanks a lot [22:10] ~>air [22:39] er