[06:31] cjwatson - Bug #717699 showed up on the iso tracker for 10.04.2, can you have a look? [06:31] Launchpad bug 717699 in casper (Ubuntu) "Lucid 20110211.1-desktop and 20110211.3-alternate amd64, gfxboot bug (affects: 1) (heat: 6)" [Undecided,Confirmed] https://launchpad.net/bugs/717699 [08:17] cjwatson, is that not the bug triggered when the usb device is built on a system which is too far away from the system on the image ? the gfxboot option thingy [09:31] might I kindly ask that someone NEW oem-config-slideshow-ubuntu [09:40] skaet: what apw said. I've duped it [09:40] not a bug in 10.04.2 as such, more a bug in later releases [09:44] cjwatson, is this somewhere that the dreaded hybrid thingy might hep [09:45] help [09:45] yes, but it's not quite that simple either - ev has been wrestling with it [09:46] we still (potentially) have to make some modifications to the image [10:07] ev: done [10:07] cjwatson: thanks bunches [10:55] skaet: lucid DVDs are in place now [10:55] jibel: ^- [10:55] (for i386 too, I mean) [11:01] cjwatson, thanks, syncing. [11:02] skaet, there's no upgrade test cases on the tracker ? [13:53] cjwatson, skaet , xubuntu i386 images failed to boot, with no bootable medium found. [13:54] odd error. can anyone else reproduce that? [13:54] charlie-tca reproduce that [13:55] it'll be three hours absolute minimum before I can look, as I'd already started rsyncing another image [13:55] perhaps somebody else can track down what's wrwong? [13:55] *wrong [14:00] bug 718749 [14:00] Launchpad bug 718749 in ubiquity (Ubuntu) "Xubuntu i386 Lucid 10.04.2 images will not boot (affects: 1) (heat: 6)" [Critical,New] https://launchpad.net/bugs/718749 [14:02] not a ubiquity bug [14:02] it's only a ubiquity bug if it actually gets as far as the installer [14:02] (for the record) [14:02] sorry, I didn't know which package it was for sure [14:04] install dumpet, see what it says for that image [14:05] (dumpet -i foo.iso) [14:11] desktop image - does not contain an El Torito bootable image. BootRecordIndicator: 2 [14:12] that would do it [14:12] alternate 386 is the same [14:25] the only obvious thing that might have gone wrong is that it *really* doesn't like -joliet-long [14:25] * cjwatson digs up the build logs [14:25] oh shit [14:25] + echo -n "-joliet-long " > $N.mkisofs_opts [14:26] > is NOT the same as >> [14:26] sorry folks, mea culpa [14:26] i386 DVD will be affected too [14:27] Thanks for looking [14:28] thanks for spotting it. will fix ASAP [14:29] rebuilds in progress [14:32] hmpf, someone edited the cdimage branch directly on antimony [14:33] the branch, or the working tree? [14:33] /srv/cdimage.ubuntu.com shows me a diff [14:34] in buildlive and publish-release [14:34] yes, those are long-standing, please leave them in place until I figure out what to do about them [14:34] hmm, and the buildlive change is from last year (!) [14:34] ah, k [14:35] if its on purpose i'm fine [14:35] I know about it, at least [14:35] k [14:48] cjwatson, which images are slated for rebuild at this time? (looks like some need to be marked rebuilding on the tracker, ubuntu DVD, Xubuntu, ???) [14:48] I'll mark them now [14:48] just Xubuntu plus Ubuntu DVD i386 [14:49] and DVDs weren't posted anyway [14:51] Xubuntu marked now for rebuild [15:15] jibel, upgrade test cases added for all except xubuntu, will add those with the new images [15:15] thanks for catching it. :) [15:19] skaet, seen that. Thanks. [15:22] skaet, we have verified all the images excepted server. 2 major issues have been identified so far: bug 718749 (rebuilding in progress) and bug 645818 (not a bug in lucid) [15:22] jibel: Bug 718749 on http://launchpad.net/bugs/718749 is private [15:22] jibel: Bug 645818 on http://launchpad.net/bugs/645818 is private [15:22] ubot4, you're wrong [15:22] Factoid "you're wrong" not found [15:23] skaet, I've run the upgrades for K/Ubuntu desktop last week without any problem. [15:23] argue with the 'bot? ;-) [15:25] jibel, thanks for the update. Can someone with an existing lucid system confirm they can make usbs/cds and install a 10.04.2 system that way? (ie. not impacted by 645818) [15:26] sure, we'll find a volunteer :-) [16:36] skaet, I'll change the notice on the tracker to mention the issue with USBs created on 10.10+, do you agree ? [16:38] jibel, good idea. :) [16:44] skaet, this lbm filename limit prob, is there a bug for that ? [16:45] cjwatson, ^^? [16:47] skaet, and can you remind me why this puppy is even on the CD ? [16:48] apw, its the image from bjf and sconklin we'll be using for 10.04.2 [16:49] skaet, LBM seems a strange package to end up on a CD ever ? [16:50] * skaet puzzles, and goes back to check some IRC logs... LBM? [16:50] linux-backport-modules-* [16:52] * skaet nods [16:56] apw, bjf - [01:01] genisoimage: Error: CD1/pool/main/l/linux-backports-modules-2.6.32/linux-backports-modules-compat-wireless-2.6.34-2.6.32-28-generic_2.6.32-28.27_i386.deb and CD1/pool/main/l/linux-backports-modules-2.6.32/linux-backports-modules-compat-wireless-2.6.34-2.6.32-28-generic-pae_2.6.32-28.27_i386.deb have the same Joliet name [16:56] [01:01] Joliet tree sort failed. [16:56] skaet, why are LBM packages being seeded on the CDROM ? [16:57] phone, minute [16:58] tgardner, will let the expert comment. :) [16:59] skaet, the problem is that the meta packages are also too long in some cases, [16:59] so changing their name without a transitional package seems a problem [17:00] but having the transitional package will trigger the same problem [17:00] apw, those are _new_ meta packages since Lucid was released, right? [17:00] tgardner, the too long ones are new yes [17:01] in all cases the names which are too long were not in the -release pocket [17:02] the DVD includes practically everything that's in supported [17:02] (technically, supported-common, which includes supported-kernel-common) [17:02] so linux-backports-modules has been there for a long time [17:04] skaet: anyway, the new DVDs haven't finished building yet [17:04] -joliet-long should have worked around all this [17:04] * skaet will keep fingers crossed [17:04] cjwatson, ok so if we rename shorter which isn't impossible, what can we do about transitional pacages [17:04] packages, to pull people to the new names, is there some way we can exclude those from the CD? [17:05] DVD [17:05] sure, but why don't we just ignore the problem for lucid [17:06] cjwatson, i am all for ignoring it :) [17:06] and then we shouldn't have to worry about it since we don't do transitionals for LBM for lucid->? anyway [17:06] cjwatson, correct [17:06] ok as there is no point release maverick we only have to worry about natty [17:06] and we can fix these names before we do lbm for there [17:06] we can just make sure that package names are unique in the first 64 characters in the future [17:07] I only applied the debian-cd change for lucid i386, so it'll blow up on us in the future again if we violate that [17:07] (and TBH, it may not be a massive problem as Joliet is just for Windows compat, but we should probably be a little careful in case we happen to change something that might affect Wubi - this shouldn't, though) [17:07] cjwatson, cool [17:08] cjwatson, we'll take steps to correct the meta package names for Natty and announce on ubuntu-devel, etc [17:08] cjwatson, is there an LP bug for this puppy? [17:08] doesn't look like it [17:08] I didn't mention one in the debian-cd commit message, anyway [17:08] cjwatson, ok don't think we need one just wanted to hoover it up if it was there [17:16] apw, ack [17:16] skaet, so i think we have a plan... do nothing and don't mess up natty [17:31] apw, heh, "do nothing" seems a bit ambiguous. I read it as no change for lucid, email to u-devel about 64 char restriction so we don't mess natty. ;) [17:32] yeah do nothing for existing releases, email out a warning for natty, and then get natty right [17:32] :) [17:33] since this is the first time it's bitten us in 6+ years, I'm not desperately worried about it being a pervasive problem [17:36] cjwatson, yeah, though it is always that annoying kernel team pushing the boundaries [17:43] apw, I pushed a patch to ubuntu-natty-meta that sets the format for future LBM compat-wireless package names [17:43] tgardner, you are a star [17:46] apw, which reminds me, we need to upload an LBM soon just as a placeholder or we'll end up having to write a MIR after the fact. [17:46] tgardner, ok ... what do we have which it would produce [17:46] apw, its always empty to begin with. [17:46] no compat wiress, no also, i guess smb's input might exist [17:47] I guess, if it hasn't gone upstream [17:47] if we didn't have to bump abi on all of these packages, it owuld be nice for them to be separate source [17:48] apw, don't go there :) way too much work. [17:48] yeah i konw, well unless we could make them dkms packages [17:49] we have 11.10 to think about that for [17:49] cjwatson: I removed a language pack from kubuntu lucid live if you want to respin it to stop oversizing [17:49] cjwatson, pitti, for the release change log, should it be since 10.04.1 or since 10.04 - what's the precedent? [17:50] since 10.04.1 [17:51] cjwatson, thanks [17:52] Xubuntu reposted [17:52] thanks [17:52] er, oops, wrong version [17:53] not December yet [17:53] ok, fixed, and Ubuntu DVDs posted [17:53] Riddell: ok, thanks - shall I just respin i386, since that's the only one affected? [17:54] cjwatson: yes [17:55] ok, respinning [18:10] * skaet -> lunch, biab [18:23] Xubuntu amd64 desktop image should be okay now, did not resize the ppc images [18:25] Can respin that image only, please. [18:26] to clarify: respin Xubuntu desktop amd64? [18:26] please. It should no longer be oversize [18:26] Riddell: Kubuntu still oversized [18:47] * skaet back [20:39] skaet: https://bugs.launchpad.net/ubuntu/+source/usb-creator/+bug/645818 [20:39] Launchpad bug 645818 in usb-creator (Ubuntu Natty) (and 3 other projects) "10.04.1 image created in Maverick does not boot in my Dell Mini9 (affects: 58) (dups: 4) (heat: 178)" [Critical,Triaged] [20:40] marjo, yes? jibel has documented workaround on isotracker. [20:40] skaet: ack [20:42] skaet: just FYI, per jibel's request, I created a 10.04.2 bootable USB with usb-creator-gtk on Lucid. I was able to boot to a live session with this USB image. [20:42] System that booted is an Acer Aspire One Model ZA3. [20:42] marjo, excellent thanks - good to know. :) [20:48] skaet: But the bug was for an image created on Maverick, not an image created on Lucid? [20:49] ScottK, yes, I plan on marking it invalid, after I put out the release notes. Just wanted it to show up so it didn't get overlooked. [20:49] I put a comment in the bug to that effect. [20:50] Ah. I see. [21:01] skaet: please don't mark the bug invalid for natty [21:01] we still want to figure something out for that - it's just hard [21:01] (and maverick too) [21:08] cjwatson, ack, was just planning on it for lucid === bjf is now known as bjf[afk]