=== dpm_ is now known as dpm [09:52] Anyone want to hold up a ubiquity upload for anything they're working on? [10:04] evand: Isn't that question a little dangerous? :-) [10:04] haha, probably [10:06] ubiquity: evand * r3235 ubiquity/ (d-i/manifest debian/changelog): [10:06] ubiquity: Automatic update of included source packages: flash-kernel [10:06] ubiquity: 2.13ubuntu7, user-setup 1.23ubuntu20. [10:11] evand: have you suddenly got like the worlds flakiest connection [10:17] ubiquity: evand * r3236 ubiquity/debian/changelog: releasing version 1.12.10 [10:19] evand: have you had time to try out the new wubi? [10:20] not as of yet, but I'll to squeeze some in today [10:20] evand: No Probs is it a package or do I need to make it? [10:21] http://people.ubuntu.com/~evand/wubi/jaunty/wubi-r122.exe [10:23] evand: ago is on about test r124 so that will need to be built locally correct? [10:24] oh, I can build that quickly for you [10:24] one moment [10:24] evand: cool. xivulon showed me how to build it too not that hard once all the packages were in place :) [10:27] evand: it needed to include partman-auto too [10:27] argh, apologies [10:27] evand: https://bugs.launchpad.net/wubi/+bug/359716 this is the bug where the test request came from by the way :) [10:27] accepted a couple of hours ago so it may be taking a while to publish or something [10:27] Ubuntu bug 359716 in wubi "(jaunty beta)Wubi does not make sure that root.disk and swap.disk are not NTFS-compressed" [Medium,Fix committed] [10:30] http://people.ubuntu.com/~evand/wubi/jaunty/wubi-r124.exe [11:10] evand: ta [11:50] evand: Heya, would like to push http://paste.ubuntu.com/152711/ to usb-creator/trunk, can I do so directly? is this worth uploading to jaunty? [11:51] lool: please do (core-dev and installer-team have permission), and I'd say so on uploading it [11:51] (Yeah, I saw I had permission but wanted to double check as I never pushed there) [11:54] lool: ah, noted. Please feel free to commit anything to there. I have absolutely no problem with anyone from either of those teams making contributions without consulting me. [11:54] evand: I think the new partman-auto should be available now, if you'd like to respin ubiquity with it [11:55] cjwatson: will do [11:55] lool: does the translatable flag break something in that case? (I agree the fix is correct, just not sure of the consequences) [11:56] cjwatson: It will show broken icons in languages where it was translated, e.g. Finnish [11:56] (which is how I discovered it) [11:56] *no* I don't speak Finnish ;-) [11:56] right [11:57] makes sense [12:02] Quit in Finnish is insulting in French [12:03] oem-config: cjwatson * r660 trunk/ (d-i/manifest debian/changelog): Automatic update of included source packages: user-setup 1.23ubuntu20. [12:13] ubiquity: evand * r3237 ubiquity/ (debian/changelog ubiquity/components/migrationassistant.py): [12:13] ubiquity: Properly filter out partitions that are going to be formatted from [12:13] ubiquity: migration-assistant again after changes to partman broke the code [12:13] ubiquity: (LP: #349334). [12:14] hw-detect: cjwatson * r113 ubuntu/ (debian/changelog hw-detect.sh): [12:14] hw-detect: Don't use 'grep -w' in VFP-detection code, since busybox doesn't [12:14] hw-detect: implement it (LP: #359915). [12:14] ubiquity: evand * r3238 ubiquity/ (d-i/manifest debian/changelog): [12:14] ubiquity: Automatic update of included source packages: partman-auto [12:14] ubiquity: 84ubuntu4. [12:20] urgh, and now I find another thing that needs to be changed within ubiquity [12:20] oh well [12:20] actually it probably doesn't affect desktop installs so isn't absolutely vital [12:21] base-installer: cjwatson * r363 ubuntu/ (debian/changelog library.sh): [12:21] base-installer: Add linux-lpia to kernel_update_list metapackage ordering hack [12:21] base-installer: (LP: #359174). [12:23] I'm going to punt oem-config through a PPA before uploading it properly [12:24] I haven't uploaded yet [12:24] so there's still time if you want to squeeze it in [12:27] evand: I'm getting the creation of swap space in partition #1 of /host/ubuntu/disks/swap.disk failed [12:27] davmor2: does http://people.ubuntu.com/~evand/wubi/jaunty/wubi-r122.exe work? That is, is it a regression? [12:29] evand I'll check in a second I'm assuming it didn't work as xivulon said that 124 was the fix [12:29] evand: probably not worth the wait [12:29] ok [12:29] base-installer: cjwatson * r364 ubuntu/debian/changelog: releasing version 1.98ubuntu7 [12:33] ubiquity: evand * r3239 ubiquity/debian/changelog: releasing version 1.12.11 [12:48] cjwatson: bug 362874 I'm going to see if I can replicate it. If I can at what point would you want what logs if there are any formed at that point? [12:48] Launchpad bug 362874 in ubiquity "ubiquity 'back' goes back too far" [Undecided,New] https://launchpad.net/bugs/362874 [12:49] also would the logs be any use? [12:49] 'ubiquity -d' logs are best for this kind of problem [12:49] although I think that one is known [12:50] actually, TBH now that I look at it I don't think it needs more info [12:50] not for jaunty though [12:51] cjwatson: okay, I thought it was pretty self explanatory, I just wondered if logs would help as I'll be trying it in order to confirm [12:51] I marked it Triaged [12:52] I think we simply don't have the code yet to restart the partitioner in the middle like that [14:37] evand1: 122 didn't work on a compressed vista drive either [14:38] oh, apologies, I think I misunderstood what was happening [14:38] so creating swap only fails if the drive is compressed? [14:41] evand1: Will the error show up in wubi logs? And would it help if I could get partman and syslog of the wubi install too? [14:42] please include all the log files, yes. But is my question correct? Is creating the swap file only failing if the drive is compressed? [14:42] evand1: Sorry yes. If you run compress drive to save space in vista that is when the issue appears. [14:43] ok [14:43] Right I'll battle with the logs and see how far I get :) [14:51] evand: Right I've added what I can to bug 359716 [14:51] Launchpad bug 359716 in wubi "(jaunty beta)Wubi does not make sure that root.disk and swap.disk are not NTFS-compressed" [Medium,Fix committed] https://launchpad.net/bugs/359716 [14:51] thanks [15:00] evand: is that enough info for you? [15:03] davmor2: yes, though I suspect it will be xivulon who works around it in wubi [15:04] evand: No I just wanted to check that all the info was there before wiping vista off the drive to do some testing :) [15:04] ah, sure thing [15:20] Hey, with 8.10 x86 in kickstart, does the 'reboot' directive cause a problem with the preseed/late_command? it seems that if I have reboot in there that runs before late command [15:22] DogWater: no - 'reboot' just silences a note, that's all [15:22] ks_preseed d-i finish-install/reboot_in_progress note '' [15:22] IOW it causes the "Installation complete" message not to be displayed [15:23] hm, thats really strange then because if I #reboot in the ks it doesn't reboot obviously, but the late_command string is run, if I have reboot in the ks it reboots but the late_command is not run [15:23] does it maybe not run concurrently? what I mean is it possible that reboot could run during the execution of late_command? [15:24] no, that isn't possible [15:24] Kickstart commands don't really "run" anyway, they just preseed answers in d-i [15:24] (for the most part) [15:25] think of it as setting up a database in advance [15:25] ah, okay darn, i've reproduced this about a dozen times but i could be doing something wrong. [15:25] now, of course, if your late_command backgrounds itself, it won't necessarily complete before d-i gets round to rebooting [15:26] it just has a bunch of in-target commands, nothing backgrounded and then | true at the end to get past some bug you told me about awhile ago [15:27] do you mean "|| true"? [15:27] | true is what I've had [15:27] that might be your problem then ;-) [15:28] so it should be || true? [15:28] "| true" -> pipe output of this command to true (you never want this, to an extremely good first approximation); "|| true" -> ignore errors from this command [15:28] <_ruben> | true could have some nice (side)effects i'd say :) [15:29] hrm, im almost certain that you told me to put | true in there awhile ago but I dont really know ;-) i'll try it with || true [15:29] feel free to prove me wrong with logs, but I'd have to have been really outstandingly tired to recommend | true :-) [15:29] oh, its not really important [15:37] <_ruben> :) [15:41] oem-config: cjwatson * r661 trunk/gui/glade/step_timezone.glade: sync up timezone glade file with ubiquity, so that the map displays without unwanted stretching [15:42] whats important is figuring out why none of these jar files will run as applets, lol [15:49] I'm nearly certain now that bug 361112 is not our fault [15:49] Launchpad bug 361112 in ubiquity "standalone ubiquity won't let me alt-tab to status window" [High,Confirmed] https://launchpad.net/bugs/361112 [15:49] oem-config: cjwatson * r662 trunk/debian/changelog: releasing version 1.54.13 [15:53] debian-installer: cjwatson * r1087 ubuntu/debian/changelog: No-change rebuild to pick up new components. [15:58] debian-installer: cjwatson * r1088 ubuntu/debian/changelog: releasing version 20081029ubuntu34 === nxvl_ is now known as nxvl [16:29] cjwatson: preseed preseed/late_command string "in-target sed -i 's,http://192.168.25.200/ubuntu,http://mirrors.internaldomain.com/ubuntu,g' /etc/apt/sources.list; in-target sed -i 's,http://192.168.25.200/Ubuntu8.10,http://mirrors.internaldomain.com/ubuntu,g' /etc/apt/sources.list;in-target apt-get update;wget http://mirrors.internaldomain.com/in.php || true" [16:30] cjwatson: with | true or || true it doesn [16:30] doesnt seem to execute that if 'reboot' is in the kickstart [16:32] happy to look at a syslog [16:33] just the regular syslog from during the install? [16:33] and do you want it when it works or when it doesn't work? [16:34] installer syslog from when it doesn't work [16:34] hrm, it immediately reboots, is that syslog available somewhere in a normal environment? [16:35] /var/log/installer/syslog [16:36] hrm, i will check [16:36] i didnt think it kept the syslog from during the install [16:36] ubiquity: evand * r3240 ubiquity/ (debian/changelog ubiquity/frontend/gtk_ui.py): [16:36] ubiquity: Set the window manager type hint to dialog for [16:36] ubiquity: debconf_progress_window to prevent it from being hidden behind the [16:36] ubiquity: fullscreen ubiquity window when in only-ubiquity mode (LP: #361112). [16:37] DogWater: it does, for exactly this kind of debugging ;-) [16:39] Email it, or post it somewhere, or? [16:39] paste.ubuntu.com or e-mail to cjwatson@ubuntu.com, I don't mind [16:44] sent to email. [16:44] its kind of strange, it sees the early command and runs it but not the late command [16:44] nothing even about it in the syslog [16:53] Apr 17 10:51:18 kickseed: preseed command requires key, type, and optional value [16:53] looks like your problem [16:54] not sure why that's failing though [16:55] works for me in my test harness [16:55] could I have the full Kickstart file, please? [16:56] sure sure [16:56] (you can strip passwords obviously :-) ) [17:00] sent, keep in mind the late_command is a single line but it wraps in the email [17:03] DogWater: it's quite important that I have the actual file you're using, byte for byte [17:03] (aside from passwords) [17:03] please don't copy and paste === cr3_ is now known as cr3 [17:12] so you don't want me to edit out our public IPs, etc? [17:18] Okay, I resent it as an attachment [17:33] DogWater: oh, I don't mind about that [17:33] DogWater: but I need it to be transferred in a way that doesn't mangle newlines [17:37] hmm, this isn't failing for me with jaunty [17:40] ah, i'm using intrepid [17:42] yeah, trying that next [17:45] DogWater: ok, it's working for me with the intrepid mini.iso from http://people.ubuntu.com/~cjwatson/intrepid-busybox-fix/i386/netboot/mini.iso [17:45] DogWater: you are using that, aren't you? [17:45] (I haven't tried stock intrepid) [17:46] No, I use netboot [17:46] that *is* netboot [17:46] I'm just using the iso form factor [17:46] oh, right but I don't use the iso, right right. [17:46] but you're using the image from that directory? [17:47] let me make sure [17:47] I could've very well picked up the wrong one as I have like 70 different ones in my tftp root [17:48] Installer build: 20080522ubuntu23 ? [17:49] 7586372 Jan 21 14:08 ubuntu-installer/i386/initrd.gz [17:51] the build number's no indication, I don't think I changed it when rebuilding for the busybox fix [17:52] doesn't match the size I have here [17:52] okay, i'll get the version you specified and match the file size [17:53] http://people.ubuntu.com/~cjwatson/intrepid-busybox-fix/i386/netboot/ 404s [17:53] that matches the file size of intrepid as released, not my rebuild [17:53] oh, sorry this is the one specified in pxelinux [17:53] 7604962 Nov 4 07:43 initrd.gz [17:53] sorry, try http://people.ubuntu.com/~cjwatson/tmp/intrepid-busybox-fix/i386/netboot/mini.iso [17:53] that sounds more like the rebuild, yes [17:53] in that case I'm not sure what to say - I can't reproduce it here :-( [17:55] can i just put 'reboot' in the late_command? [17:56] or will that screw something else up [17:57] that will just break [17:58] I think the most practical way to get further information would be to boot the system with your usual boot parameters but also BOOT_DEBUG=3, which will give you shells at various points in the boot process - in the first of those, 'nano /lib/kickseed/kickseed.sh' and put 'set -x' on the second line, and then Ctrl-d until you get back to the installation as normal [17:59] then extract /var/log/installer/syslog at the end [18:07] would not having "" in the user add line cause a problem? [18:08] could you elaborate? [18:08] administrator --fullname --password stuff [18:09] the only thing i notice is that line is right above 'reboot' [18:09] the file you sent me had: [18:09] user administrator --fullname "" --password stuff [18:09] yeah it had --fullname "" [18:09] what about around the password? [18:09] that doesn't matter [18:09] okay [18:09] but --fullname --password stuff would be wrong [18:09] honestly my set -x log suggestion above is probably quicker than guessing [18:14] and that goes in the kernel line or the append line? [18:15] BOOT_DEBUG=3 goes in the append line; then follow my instructions above [18:23] Alrighty, i'm doing a CentOS 3 test atm then i'll go back to that one [19:38] cjwatson: about that, it appears that there is no usb support in that? [19:43] can't type anything [21:35] cjwatson: as broken as my stuff might be replacing reboot /w preseed finish-install/reboot_in_progress note fixed it [21:41] thankyewsir [23:24] cjwatson: any last minute kde installer things you want me to take a look at? [23:25] maybe check that we didn't break anything obvious with our last-minute hacking ;-) [23:25] I don't have anything OHMYGOD on my list [23:26] updated Kubuntu dailies are building right now [23:36] k [23:37] when those build I can run though some installs so test out various things