[00:02] ev: I just realized I added the changelog line to the previous section marked as unreleased, but it appears to have actually been released! [00:04] or maybe I messed something up... [00:08] ev: oh nm... I see u fixed it, or merging did, however that works... [00:53] tzsetup: cjwatson * r519 ubuntu/ (debian/changelog tzsetup): Use '-T 15' rather than '--timeout=15', to avoid breaking busybox wget. [00:54] tzsetup: cjwatson * r520 ubuntu/debian/changelog: clarify [00:57] tzsetup: cjwatson * r521 ubuntu/debian/changelog: releasing version 1:0.26ubuntu8 [05:56] max? [09:12] amichair: indeed, I took care of it during the merge. No worries though, changelog mangling is often necessary when merging branches. [09:38] ubiquity: evand * r4070 ubiquity/ (debian/changelog ubiquity/frontend/kde_ui.py): [09:38] ubiquity: Translate the yes and no buttons on the quit dialog in the KDE [09:38] ubiquity: frontend (LP: #561876). [09:38] Launchpad bug 561876 in ubiquity "[Kubuntu Lucid] Quit dialog buttons not translated" [Undecided,New] https://launchpad.net/bugs/561876 [09:48] cjwatson, why does d-i always pick the right TZ for me, even if i use a plain US install and dont indicate at all that i'm in germany it picks europe/berlin ... thats awesome ! [09:48] geoip [09:48] thank IS and ev :-) [09:49] cool ! [09:49] and cjwatson for getting the ball rolling on that :) [09:55] choose-mirror: cjwatson * r624 ubuntu/ (choose-mirror.c debian/changelog): [09:55] choose-mirror: Expand CC.archive.ubuntu.com to archive.ubuntu.com if [09:55] choose-mirror: debian-installer/locale is set to "C" (LP: #550694). [09:55] Launchpad bug 550694 in choose-mirror "setting locale to "C" doesn't choose a generic mirror" [Undecided,New] https://launchpad.net/bugs/550694 [09:57] choose-mirror: cjwatson * r625 ubuntu/debian/changelog: releasing version 2.29ubuntu3 [10:08] cjwatson, so my partman issue seems to be actually a disk-detect issue ... i finally got set -x output in syslog for that, would you mind taking a look ? http://launchpadlibrarian.net/44009972/syslog [10:10] cjwatson, the "Unhandled fault" seems to happen actually before disk-detect even runs, is there anything else triggered by default when i run it ? [10:21] cjwatson, did you get my above ping ? [10:22] no [10:25] cjwatson, so my partman issue seems to be actually a disk-detect issue ... i finally got set -x output in syslog for that, would you mind taking a look ? http://launchpadlibrarian.net/44009972/syslog [10:25] cjwatson, the "Unhandled fault" seems to happen actually before disk-detect even runs, is there anything else triggered by default when i run it ? [10:26] i also have added set -x to check-missing-firmware now (see https://bugs.edge.launchpad.net/ubuntu/+source/linux-ti-omap/+bug/561426) that exposes the bus error but its not clear why exactly causes it [10:26] Launchpad bug 561426 in linux-ti-omap "partman dies when trying to detect disks due to kernel error" [High,In progress] [10:26] well, I mean it goes through main-menu and such, but you might also just be seeing log messages flushed out of order; it isn't conclusive [10:26] and the kernel error still shows up above it [10:26] parted_devices is consistently getting a bus error there [10:27] ah [10:27] * ogra looks for that [10:27] gah, no script [10:28] ~ # [10:28] ~ # parted_devices [10:28] [ 1034.439666] Unhandled fault: external abort on non-linefetch (0x1018) at 0x40200000 [10:28] Bus error [10:28] ~ # [10:28] ha ! [10:28] there you go [10:30] 'anna-install strace-udeb' and try stracing it? [10:35] ok [10:56] hi there [10:57] cjwatson, do you have any preferenc where to add omap in the HW list in libdebian-installer (alphabetical by name or arch or so ?) or should it just add it by gut feeling (seems dove and imx51 were just added in random places) [10:58] the existing arches seem to follow a schema sorted alphabetically by subarch [11:00] dunno really, stick it just after dove I guess [11:00] then at least the Ubuntu additions are together [11:00] ok [11:01] How do you debug something further than DEBCONF_DEVEL=developer ? [11:02] For example, I'm building a custom ISO based on lucid, and "anna" tells me that "no packages for kernel in archive" [11:02] I would pretty much like to know which package it is looking for [11:03] I tried poking around in the filesystem. I had some hope when I found /var/lib/dpkg , but not [11:04] anna-install is a minimalist dpkg replacement for .udeb as I understand, but doesn't it unwrap the package scripts somewhere before executing them ? [11:09] zimbatm: well, you should start by having the source code handy and tracing through it [11:10] but that error means that there are no entries in the Packages file on the CD with a Kernel-Version field matching the running kernel [11:10] rather than any specific package name [11:10] I'll clarify that message upstream [11:11] cjwatson, hmm [11:12] I'm using the latest packages from archive.ubuntu.com -> reprepro -> updated dist/pool on the CD [11:12] what's weird is that sometimes it work [11:12] that doesn't necessarily mean that the kernel you're booting the installer with matches, though [11:12] It's like they publish a new kernel, but *.udeb get updated later (that's what I suspect) [11:13] and you should also check the Packages files in dists/*/main/debian-installer/binary-*/ to make sure that they have Kernel-Version fields [11:13] your suspicion is incorrect [11:13] alright [11:13] however, the installer (dists/*/main/installer-*/current/images/) does need to be rebuilt against new kernel udebs, and *that* isn't immediate [11:13] so you should always make sure that they match [11:17] cjwatson, reprepro has fetcher the Kernel-Version tags correctly, now I will look for version correspondance [11:20] ogra: can you try applying http://paste.ubuntu.com/413578/, building, scping in the new libparted.so.0.0.1, and running parted_devices again? [11:57] cjwatson, will do [11:59] grrr ... cant wget from paste.u.c ... silly openid [12:00] Just use a different pastebin [12:09] libdebian-installer: ogra * r203 libdebian-installer/ (debian/changelog src/system/subarch-arm-linux.c): add beagle omap3 support [12:10] libdebian-installer: ogra * r204 libdebian-installer/debian/changelog: releasing version 0.68ubuntu2 [12:14] cjwatson, hmm, FTBFS [12:14] http://paste.ubuntu.com/413591/ [12:19] weird, is_apple is clearly defined in gpt.c [12:22] oh, its surrounded by the #ifdef USE_DMI [12:35] cjwatson, thanks a lot for your help. Updating the CD's vmlinuz and initrd made the trick. [12:48] ogra: whoops. Try http://paste.ubuntu.com/413606/ ? [12:50] cjwatson, i also have an issue with libdebian-installer somehow dh --builddirectory=build gets expanded to -O--builddirectory=build at build time, any idea why ? [12:50] its not that i would have changed anything but adding one line to the c source [12:51] what's wrong with that? [12:51] -O is documented in debhelper(7) [12:52] http://launchpadlibrarian.net/35017551/buildlog_ubuntu-lucid-armel.libdebian-installer_0.68ubuntu1_FULLYBUILT.txt.gz is the upload from Nov ... it doesnt have that ... http://launchpadlibrarian.net/44019882/buildlog_ubuntu-lucid-amd64.libdebian-installer_0.68ubuntu2_FAILEDTOBUILD.txt.gz has the -O [12:52] it looks broken that its directly attached to the -- [12:53] its the only difference i can see ... [12:54] looks unrelated to me [12:55] directly attached is correct - see debhelper 7.4.12 changelog for why it's there === rgreening_ is now known as rgreening [12:55] hmm, i would have read the manpage that it needs to be -O=--... [12:55] it works fine here the way it is. I think this is a red herring [12:55] but if thats ok, then i dont know whats wrong here ... [12:56] oh, wait, seems the autoreconf doesnt run [12:56] could you let me prepare a new source package? [12:57] I'll fix it [12:57] not all the stuff is in bzr for historical reasons, and it requires care. why didn't you use debdiff before uploading? [12:57] oh, sorry, i didnt know that [12:58] checking debdiff defends you against things you don't know :) [12:58] yeah, will do that the next time [12:58] thanks [13:00] libdebian-installer: cjwatson * r205 ubuntu/debian/changelog: Reupload with the right autotools bits. [13:01] libdebian-installer: cjwatson * r206 ubuntu/debian/changelog: releasing version 0.68ubuntu3 [13:05] that parted build looks better [13:06] * ogra waits for it to finish [13:37] cjwatson, successfully partitioned a disk on the beagle, go for an upload :) [13:41] cool, thanks [13:42] done [13:42] * ogra hugs cjwatson [13:43] now i only need the ugly flash-kernel-installer hackage [13:43] then beagle should work [13:48] silly trick for quickly testing grub changes [13:49] install new version and make very sure that grub.cfg is configured to show the menu by default, rather than using the default zero-delay hidden timeout [13:49] then: sudo kvm -hda /dev/sda [13:50] bit twisty, but a lot quicker than rebooting [13:51] as long as you make bloody sure not to actually boot an OS inside that kvm :-) [13:52] heh. [13:52] Would it be safe for dual-booters to boot an OS, as long as it wasn't the runing OS, or is it too much madness? [13:59] as long as no common partitions were mounted, I suppose [13:59] I wouldn't want to be the one to recommend this, though :-) [14:06] michaelforrest: https://wiki.ubuntu.com/InstallUpdatesWhenInstallingUbuntu [14:23] "No installable kernel was found in the defined APT sources." [14:23] whine [14:23] what magic am i missing ? === icarus902 is now known as icarus901 [14:27] dunno, can I see the log? [14:28] once i'm done, i want to see where it fails next when moving on without a kernel [14:28] oh, I can guess [14:28] i added it to the boot and installer seeds [14:28] base-installer: cjwatson * r399 ubuntu/ (debian/changelog kernel/armel.sh): Add armel/omap subarchitecture. [14:28] so i would have expected to be picked up from there [14:28] but i might miss something [14:28] that base-installer change should fix it [14:28] heh, thanks :) [14:29] base-installer: cjwatson * r400 ubuntu/debian/changelog: releasing version 1.103ubuntu7 [14:43] ubiquity: cjwatson * r4071 ubiquity/ (debian/changelog finish-install.d/07oem-config-user): [14:43] ubiquity: Update finish-install.d/07oem-config-user for new location of KDE's [14:43] ubiquity: oem-config-prepare .desktop file (LP: #557309). [14:43] Launchpad bug 557309 in ubiquity "no "prepare for shipping icon" in kubuntu alternate" [High,New] https://launchpad.net/bugs/557309 [15:45] how for the possibly stupid question but where can i get daily netboot images? [15:45] sorry* [15:47] http://cdimage.ubuntu.com/netboot/lucid/ has the links. the Xen build stuff is only on my hard disk right now though :) [15:54] even x86_64? [15:55] and the grub fix [16:09] cjwatson: have the install options been removed from the live cd out of curiosity. example install free software only? [16:14] should be on the f6 menu [16:14] dmarkey: maybe 64. grub fix should be in [16:20] ubiquity: evand * r3930 testing/tests/run-frontend: Add a testPageNavigation test for r4041 / LP #556180. [16:20] Launchpad bug 556180 in ubiquity "Buttons on ubiquity partition page doesn't work" [Undecided,Fix released] https://launchpad.net/bugs/556180 [16:22] modules are there anyway :) [16:36] ev: are you going to push usb-creator 0.2.22? There's an important fix for Format under the KDE front-end. [16:37] rgreening: will do [16:37] thanks ev [16:38] ev: new POT's need generation as well for the 'Erase Disk' change... so we get translations (unless you wish to revert that for now). [16:40] indeed, I'm going to send a mail to ubuntu-translations [16:40] and ubuntu-doc [16:41] cjwatson: grub bug is fixed, many thanks [16:41] cool [16:45] debian-installer: ogra * r1281 ubuntu/build/config/armel/dove/netboot.cfg: make sure dove also uses INITRD_FS = initramfs for netboot, else it fails to uncompress the initrd LP: #541399 [16:45] Launchpad bug 541399 in debian-installer "netboot image fails to boot." [Medium,In progress] https://launchpad.net/bugs/541399 [16:47] debian-installer: ogra * r1282 ubuntu/debian/changelog: update changelog for dove fix [16:55] usb-creator: evand * r302 usb-creator/debian/changelog: releasing version 0.2.22 [16:55] rgreening: ^ [16:56] ty [16:56] now to deal with the fallout regarding string changes in kubuntu-docs [17:24] michaelforrest: wget http://popcon.ubuntu.com/main/by_inst -O - 2>/dev/null | egrep -w "language-support-[A-Za-z].." | less [17:32] ev: why I like rubyL [17:32] : [17:32] file << line unless line =~ /^#/ [17:32] *LOVE ruby. [17:34] michaelforrest: re vowel removal> I had to take a series of AS/400 programming courses in university (don't ask why). We were told that the list of commands was designed by a "world renowned linguist" to be easily pronounceable and easily remembered. Not so much - http://whatis.techtarget.com/content/0,290959,sid9_gci827107,00.html === dpm is now known as dpm-afk [21:35] ev: I'm thinking of http://paste.ubuntu.com/413884/ as a general approach to this oem-config removal bug [21:35] * ev looks [21:41] cjwatson: that doesn't seem to handle the case whereby oem-config returns nonzero, but it otherwise looks good. [21:41] set -e [21:41] it could maybe do with a comment :-) [21:41] ahh [21:53] ubiquity: cjwatson * r4072 ubiquity/bin/ubiquity-dm: simplify set_locale [22:41] ubiquity: evand * r3931 testing/tests/run-frontend: Add testPluginTranslate test for bzr r4061. [22:46] ubiquity: evand * r3932 testing/tests/run-frontend: Clarify previous commit.