[00:09] usb-creator: evand * r283 usb-creator/ (6 files in 6 dirs): [00:09] usb-creator: * Pulse the progress bar while flushing changes to disk. [00:09] usb-creator: * Fix unmounting the target partition at the end of install. [00:09] usb-creator: * Stop listening to devicekit events just before we enter the core install [00:09] usb-creator: routine. === NCommand1r is now known as NCommander [07:50] ev, cjwatson: I do not even reach 541607, in VM I get stacked at the kernel before the initrd is loaded [08:36] bug 541607 [08:36] Launchpad bug 541607 in wubi "Lucid: Wubi drops immediately into grub shell on reboot" [Undecided,New] https://launchpad.net/bugs/541607 [09:03] cjwatson: can you commit that wubi patch from last night? I'll build and upload a new wubi so we have something to point users at in the release notes. [09:06] ev: I haven't tested it *at all* [09:06] but I can if you want ... [09:06] hm, perhaps davmor2 would be so kind as to give it a spin first, once he wakes up [09:08] it should be no worse than the current situation - I've gone ahead and committed [09:08] r174 [09:09] thanks [09:21] base-installer: cjwatson * r395 ubuntu/ (4 files in 4 dirs): Add support for preempt kernel flavour on amd64 (LP: #541625). [09:54] tasksel: cjwatson * r1430 ubuntu/ (debian/changelog tasksel.pl): [09:54] tasksel: On new installs, include preseeded tasks in the choices list, regardless [09:54] tasksel: of filtering (LP: #540167). [10:04] tasksel: cjwatson * r1431 ubuntu/ (9 files in 2 dirs): [10:04] tasksel: Update Ubuntu tasks from seeds, removing edubuntu-ship-addon, updating [10:04] tasksel: uec's Key to cloud-init, and hiding eucalyptus-* on new installs. [10:06] tasksel: cjwatson * r1432 ubuntu/debian/changelog: releasing version 2.73ubuntu26 [10:07] base-installer: cjwatson * r396 ubuntu/debian/changelog: releasing version 1.103ubuntu5 [10:12] fail. I forgot about the keymap when pushing the greeter into ubiqutiy: https://bugs.edge.launchpad.net/ubuntu/+source/ubiquity/+bug/539255 [10:12] Ubuntu bug 539255 in ubiquity "[Lucid Alpha3] kbt-002 testcase failed, selecting a language in ubiquity, keyboard layout is always set to US" [High,Confirmed] [10:13] I don't think it needs to be explicitly selected here, maybe just run through the default [10:29] As your request I can try wubi in the pm, in Italy now it's 11.25 o'clock, I can try wubi about at 16 o'clock [10:29] ev: there's a new wubi I believe? [10:29] RobertoSbragia, davmor2: (09:15:58) Evan: anyone have a copy of Windows handy and the free time to give http://people.canonical.com/~evand/wubi/lucid/wubi-r174.exe a spin? [10:29] Hi, I got an email [10:30] ev: I'll give it a go now dude :) [10:30] davmor2: you rock [10:31] ev: I try but my head hurtz when I mosh to close to the desk ;) [10:32] haha [10:33] :-) see you later! [10:33] ev: http://www.google.co.uk/products?q=thermos+mugs&oe=utf-8&rls=com.ubuntu:en-GB:official&client=firefox-a&um=1&ie=UTF-8&ei=31KjS62GGc614ga65KX0CQ&sa=X&oi=product_result_group&ct=title&resnum=3&ved=0CCsQrQQwAg that should fix you tea issues :) [10:34] davmor2: :) [10:35] I wonder if they make those no-spill sippy cups in a size and thermal integrity adequate for tea [10:35] Oh yes [10:40] you get insulating ones that are designed to keep milk cold, but presumably roughly the same principle applies [10:41] ev: it's dling the image now [10:41] as long as the thing doesn't melt ... [10:41] haha, that is my concern [10:41] * ev will try to hunt one down and experiment [10:42] ev: just stick to the thermal one it has a lid and keeps it hot, I'd of thought that Ubuntu Shop would of had one but apparently not [10:42] apt-cdrom on usb live environment is known broken due to the lack of symlinks in vfat, right? [10:44] ew [10:52] YAY! well yay in that it got further than yesterday [10:53] davmor2, :-) [10:53] oh oh oh Installing system [10:54] ev: the first screen got to 814% which I find a tad bit concerning but other than that it seems to be trundling along nicely :) [10:55] we should really nail this progress bar breakage, we've had plenty of reports of it [10:55] but actual can't-install usually takes priority ... [10:57] ah! found it [10:58] cjwatson: indeed [10:58] if I remember correctly, it's due to nested progress bars, no? [10:58] I think so [10:58] incorrect nesting or something [11:00] reboot into system now fingers crossed everyone [11:01] I got plymouth [11:01] I got gdm [11:01] I got desktop [11:01] You guys ROCK [11:03] davmor2, \o/ [11:06] ev, cjwatson: it doesn't like plymouth on exit but other than that the system is fine :) [11:07] excellent [11:07] bit worrying that xivulon couldn't get past the kernel mind you [11:07] but probably a different problem ... [11:08] console-setup: cjwatson * r140 ubuntu/ (5 files in 2 dirs): [11:08] console-setup: Don't gzip font or ACM files in /etc/console-setup/. The space saving [11:08] console-setup: is negligible, and it's more efficient not to need to uncompress them at [11:08] console-setup: boot time. [11:10] console-setup: cjwatson * r141 ubuntu/debian/changelog: releasing version 1.34ubuntu12 [11:14] casper: evand * r786 casper/ (debian/changelog scripts/casper-bottom/41apt_cdrom): [11:14] casper: Don't let apt try to auto-detect the CD-ROM device using udev. We [11:14] casper: already know what it is, and using udev to find it again wont work [11:14] casper: for USB disks. [11:17] ev: is there away to tell wubi to only get the 32bit version? [11:17] wubi.exe --32bit [11:17] cool :) [11:18] ev, cjwatson: I'm going through all the desktops in 64bit now then I'll hit it again with 32bit [11:19] okay, symlink updates, new CD builds will have the fixed wubi [11:20] I'll release-note this [11:21] thanks [11:26] Meh Installer Crashed on kubuntu I'll grab the info === ogra_ is now known as ogra [11:28] davmor2: if it's console-setup crashed with code 141, it's already fixed in trunk [11:35] ev: traceback reads File "/usr/lib/ubiquity/bin/ubiquity", line 468, in main(oem_config) \"/usr/lib/ubiquity/bin/ubiquity", line455, in main install(query=options.query) \"/usr/lib/ubiquity/bin/ubiquity", line 243, in install ret = wizard.run() \"/usr/lib/ubiquity/ubiquity/frontend/kde_ui.py", line 388, in run self.debconf_progress_window.set_title( [11:35] davmor2: can you pastebin it [11:35] ev: no browser [11:36] I'm ubuntu-bugging it though [11:36] davmor2: apt-get install pastebinit ;) [11:38] ev: How do I then get the info from the qt window screen to the tty? [11:38] ev: is there somewhere I can grab the traceback from? [11:39] or I can take a photo for you :) [11:40] traceback should be in /var/log/installer/debug [11:44] ev, cjwatson: bug 541829 has that file attached [11:44] Launchpad bug 541829 in ubiquity "Kde wubi crashes on r174" [Undecided,New] https://launchpad.net/bugs/541829 [11:57] does that help? [12:04] cjwatson: so michaelforrest2 had a quick chat about this updating the installer thing, and he felt (and I agree) that this a general problem applicable to any piece of critical software on the system. If we're going to solve it for one thing, why not do it across the board? He also sees this as a documentation issue, that if we find a critical bug in the installer, then we should push a new ubiquity to -updates, and provide instructions on the wiki [12:04] michaelforrest2: do step in and correct me if I'm getting this wrong [12:05] michaelforrest2: the one thing we didn't cover is the release notes link. Do you agree with Mark that it should go as well? [12:09] the problem has always been that there is no convenient way for a user to update the installer, and statistically, installation blockers have tended to be bugs in the installer rather than in other bits of the system; it's a common-case optimisation [12:09] (well, or kernel or X bugs, but we can't realistically help with those) [12:09] we often *do* push a new ubiquity to -updates, and have provided instructions on the wiki in the past - it didn't help much [12:10] ev: I think the same goes yes. [12:10] cjwatson: really? Did we do this since I joined and I've just forgotten? [12:10] we certainly did it for hardy [12:10] Mainly because in this context it looks like *Ubuntu* release notes instead of *Installer* release notes === michaelforrest2 is now known as michaelforrest [12:11] I'm just concerned that this is a failing-to-learn-from-experience thing [12:11] we've tried the things being suggested here in the past when we had critical installer bugs, and continued to get flooded with bugs anyway, which suggests to me that it didn't really help users [12:11] cjwatson: in my opinion, this relates to larger issues of incompletely tested code, and is an ad-hoc solution to a class of problems that needs to be addressed on a more general level [12:12] yes, it's an ad-hoc solution [12:12] I'm not sure that necessarily makes it the wrong thing to do at the moment [12:12] so 1. we should not be expecting to release faulty code, and 2. we should think about a way that applications in general can show that they require critical updates [12:13] I believe that as long as it is possible to update the installer via the Live Session, then we are not preventing people from solving any problems that might arise. [12:13] but we're explicitly guiding people away from the live session in lucid [12:13] we've actually made this problem *worse* by active design decisions [12:13] no we're not - we've put a big 'Try' button in. [12:14] realistically, somebody who fails to install Ubuntu by the primary method is not going to reboot, press the Try button, and try an upgrade [12:14] they're going to throw the CD in the bin instead [12:14] I disagree. [12:14] They are presented two options. If the second doesn't work, and they are motivated to try Ubuntu, then they are likely to give the first option a go. [12:15] normally I'd say lets put it to user testing, but we're way out of time for that [12:15] and go through this convoluted process? I just don't believe that [12:15] The key thing here is that we shouldn't be releasing broken software! [12:15] though it seems though that we've already had plenty, given what cjwatson is saying about the bug count [12:15] obviously, but what do you want of us [12:15] We should not be patching the UI in the assumption that we will fail [12:16] we're patching the ui if and only if we've failed (in the proposed design) [12:16] If we keep the options to a minimum, it is far more likely that somebody will try both, and they can then use Google etc… to solve their problem. [12:16] but, again, experience from the past says they don't [12:16] This needs user testing. [12:17] Anyway this has come from Mark so you can raise it with him, but I tend to agree with his decision. [12:17] the button itself came from Mark in the first place, of course ;-) [12:17] of course. [12:17] this is the crazy world in which we live ;) [12:18] I mean, obviously I agree that we shouldn't be releasing broken software, but even aside from the "stupid" class of bugs (i.e. logic just wrong), the installer is always going to be more vulnerable to only-reproducible-in-certain-specialised-environments bugs than most [12:18] and I think we should be institutionally recognising that [12:19] we have had too many of the "stupid" class of bugs in the past, and we've been working to rectify that with things like better static analysis [12:19] and we're certainly in a much better position than we were in e.g. dapper [12:19] but I'm concerned about taking an entirely idealistic position here [12:20] ev: where does wubi look for une or is it still looking for unr? It can't located the meta and therefore can't get the image for ubuntu-netbook [12:23] * cjwatson tries to work out how to detect at partitioning commit time when partitions have been deleted [12:24] I think maybe if I leave a 'deleted' marker in the state directory it would help [12:53] cjwatson: going back to the design of this update thing. If we stuck with it, we could put an upstart job that starts on networking and tries to fetch the URL and drop it somewhere local that ubiquity will look. Though perhaps the time between networking and ubiquity is quite small and this is just trying to be too clever. [12:54] I like the general idea although sometimes networking isn't really available until network-manager comes up ... although mind you in that case it probably won't work in the ubiquity-dm session anyway :) [12:54] indeed, that's what I was thinking [12:54] there's no chance for them to select a wireless network anyway [12:54] the time between networking and ubiquity *actually displaying its UI* should be sufficient to do a wget [12:55] not if we keep fixing the slowness in ubiquity and scott keeps fixing the boot performance [12:55] perhaps I'll ask him for some sleep statements ;) [12:55] well, there's only so fast it can go off a CD ... [12:55] or even a USB stick [12:55] indeed [12:55] ev, have you seen davmor2's question about UNE+wubi? [12:55] ara: I did, then something shiny passed by [12:55] * ev checks again [12:56] ah, still looking for UNR [12:56] * davmor2 files bug [12:57] ev, thanks [13:02] wubi: evand * r176 trunk/ (data/isolist.ini debian/changelog): Ubuntu Netbook Remix is now the Ubuntu Netbook Edition. [13:02] ara: sure thing [13:03] ev: bug 541881 [13:03] Launchpad bug 541881 in wubi "Wubi is still pointing at unr and not une" [Undecided,New] https://launchpad.net/bugs/541881 [13:04] wubi: evand * r177 trunk/debian/changelog: LP reference for previous commit. [13:04] thanks [13:04] and uploaded as well [13:04] noted :) [13:50] hey [13:50] I'm willing to help testing the new Wubi. [13:52] yo2boy_: ask on #ubuntu-testing [14:17] ev, cjwatson: http://i44.tinypic.com/29ok1mp.png has just been dropped onto -testing [14:18] davmor2: fixed in r3941 [14:19] ev: thanks [14:19] davmor2: if they use the most recent CDs, they wont see that [14:19] err nevermind [14:19] they still will [14:19] but it's definitely fixed === NCommand1r is now known as NCommander [14:59] usb-creator: evand * r284 usb-creator/ (4 files in 4 dirs): Pulse when installing the bootloader. [15:04] partman-base: cjwatson * r195 ubuntu/ (debian/changelog debian/partman-base.templates lib/commit.sh): [15:04] partman-base: Don't warn about data loss on formatted/removed partitions when there [15:04] partman-base: are no such partitions (LP: #151266). [15:11] michaelforrest: ^- that's something you were asking for, IIRC [15:12] cjwatson: yes [15:13] why don't I get a link to launchpad from this bot? [15:13] partman-base: cjwatson * r196 ubuntu/ (8 files in 5 dirs): merge from Debian 139 [15:13] Error: Debian bug 139 could not be found [15:13] it probably doesn't notice the "LP: #nnnnnn" format in changelogs [15:14] bug 151266 [15:14] Launchpad bug 151266 in partman-base "partman/confirm should not include warning if there's no data left to lose" [Low,Triaged] https://launchpad.net/bugs/151266 [15:18] partman-base: cjwatson * r197 ubuntu/debian/changelog: releasing version 139ubuntu1 [15:19] ev: How does one get Migration Assistant to work? I've got Xubuntu 64 on a test system, with some text files I created to be migrated over. Installing Ubuntu 64bit now (Lucid) but Migration Assistant never runs [15:19] it worked for davmor2 just now on his test, but for some reason, not on mine... [15:20] bladernr_: mine was against vista :) [15:20] ahhh... there's the difference [15:20] bladernr_: it doesn't search for documents on Linux [15:20] as davmor2 suggests [15:20] ah hah... [15:20] that explains it and I am not, in fact, crazy... [15:20] well, I may still be... but that makes sense then... [15:21] hrmmm.. Ok... to be sure, I'll pop my "real" hd back in my netbook and see if I can grab stuff from the XP install there [15:22] crap... no I wont. netbooks aren't 64bit :( [15:22] doh! [15:22] Um, some are. [15:23] well... atoms arent AFAIK [15:23] Well, except the new ones :) [15:24] so out of curiosity, why doesn't it pull info from other linux installs? [15:26] bladernr_: cause you can link to it easily [15:27] davmor2: true... meh... I was just getting exited about a chance to be lazy and have the installer duplicate /home for me ;-) [16:01] Heyy! I want to help test the new wubi. === robbiew_ is now known as robbiew [16:02] hello? [16:04] I don't see any anwsers. Maybe I did this wrong.... I hate IRC. I want to help test the new wubi! E-mail me at [cfhs_1@yahoo.com] thanks! [16:40] ev: it would be good to have some kind of comment from somebody on bug 368060 about where it stands at the moment, if anywhere [16:40] Launchpad bug 368060 in ubiquity "Map of Kashmir when selecting the timezone is incorrect" [High,Triaged] https://launchpad.net/bugs/368060 [16:42] cjwatson: indeed, done [16:44] thanks [16:45] glad you clarified :-) [16:48] sure, my apologies for not being more vocal as I work to resolve this [16:49] I spoke at length and in person with Ken the other day about it [16:49] so he should have everything he needs to complete it [16:49] I'll continue to check in with him to make sure it's on track for beta 2 [16:54] great, thank you [16:54] I thought it was on track, but just wanted to make sure [17:01] I'm going to upload a new ubiquity now that beta 1 is out the door [17:03] actually, I'll wait for partman-base to build [18:32] hello i am willy to test wubi if the need still is here [18:33] I think we're OK for the moment, thanks [18:34] not a problem [23:52] ubiquity: evand * r3961 ubiquity/ (d-i/manifest debian/changelog): [23:52] ubiquity: Automatic update of included source packages: base-installer [23:52] ubiquity: 1.103ubuntu5, console-setup 1.34ubuntu12, partman-base 139ubuntu1.