[12:42] hey all i've submitted a patch to a bug in partman-auto [12:42] https://bugs.launchpad.net/ubuntu/+source/partman-auto/+bug/83897 [12:42] should be a quick one somebody can bang out in a minute [12:43] I'm not convincd [12:43] convinced, even [12:44] $dev is the directory [12:44] e.g. =dev=hda [12:44] I know, but your patch is in the wrong place [12:44] I think it's the code calling get_auto_disks which is incorrect [12:45] get_auto_disks is called from more than one place and your patch would break the other callers, AFAICS [12:45] perhaps, although i ran through the install ok after applying that [12:45] specifically select_auto_disk [12:46] you'd need a >1-disk system without preseeding to see the breakage [12:46] ah i see [12:47] and just when i thought i was being clever. doh. [12:47] well then auto.d/initial_auto must need to do some massaging then? [12:47] jeckers`: try the patch I just posted to your bug? [12:48] will do [12:48] jeckers`: thanks for the report, though, this was the first I'd realised that was broken [12:48] jeckers`: and don't be put off, partman is a hairy piece of code to try to work on; good effort for getting as far as you did :-) [12:49] i spent the better part of 3 hours trying to figure out what the heck it was doing :) [12:50] my fault for not testing the fix in the first place - I was in Oslo and nowhere near any of my test rigs === jeckers` is now known as jeckers [01:01] yup your patch did the trick [01:02] thanks a lot [01:02] you'll probably be seeing more of me in the coming months [01:02] i'm going to be following along with feisty closely as it develops [01:03] hopefully i'll catch a few more good ones along the way :) [01:07] great, look forward to more installer manpower :) [01:08] will upload that patch now [01:09] cool thanks [01:09] and i'd be glad to give more installer manpower :) [01:10] feel free to ping me here for help === kro [n=user@pool-72-84-46-79.clrkwv.east.verizon.net] has joined #ubuntu-installer [01:11] and my partner in crime has arrived === cr3 [n=marc@pdpc/supporter/bronze/cr3] has joined #ubuntu-installer [11:22] hi all [11:23] ubiquity: cjwatson * r1848 ubiquity/ (83 files in 3 dirs): [11:23] ubiquity: * Display a warning message next to the password entry box if running in [11:23] ubiquity: debug mode, since the password will then end up in the log file via [11:23] ubiquity: debconf debugging. [11:33] ubiquity: cjwatson * r1849 ubiquity/ (debian/changelog ubiquity/components/console_setup.py): * Don't crash if /etc/X11/xorg.conf is unaccountably missing (LP: #83927). [11:35] ubiquity: cjwatson * r1850 ubiquity/ (configure configure.ac): bump to 1.3.19 [12:07] cjwatson: it's possible for me to backport this fix on edgy ? (https://bugs.launchpad.net/ubuntu/+source/partman-auto/+bug/83897) [12:17] saispo: yes, but only on top of the change in feisty that it was fixing ... [12:18] ok [12:18] it may fix the error with /dev/disc/disc0 things... [12:20] well, you can't use /dev/discs/disc0/disc any more either way, but yes, it could make that situation easier [12:24] because today i use a script but don't work on installer... [12:25] if [ cat /sys/block/hda/devices/media = "cdrom" ] ; then echo part-auto-sda.cfg; else echo part-auto-hda.cfg; fi [12:25] if it != cdrom he don't get the other file... [12:38] well, that's just bogus shell syntax [12:38] if that's literally what your script says [12:39] correct would be if [ "$(cat /sys/block/hda/devices/media)" = cdrom ] ; then echo part-auto-sda.cfg; else echo part-auto-hda.cfg; fi [12:39] ok, thanks [01:22] cjwatson: Are you shooting for an FF exception for the new partitioner work? [01:22] s/an/a/ [01:24] yeah [01:26] cjwatson: how comfortable would you be with me attempting one on say, Saturday morning, for Opera bookmarks importing and email account information importing? I've basically got the former done. [01:26] Or should I let it wait for Feisty+1? [01:26] well, can we land the ubiquity m-a work today? :) [01:26] if so extra little bits don't count for FF IMO [01:26] really? That's amazing. [01:26] But yeah, it's there for you to merge [01:27] I figured you were waiting for m-a main promotion [01:27] oh, I need to hassle about that, yes [01:27] brb, preparing meeting agenda for today and then lunch [01:27] lazy core devs :-) [01:27] ok [01:51] how can i debug why a package for a local repository i make on the cd is installing in the base installer without any apt-install command or other ? [01:51] s/for/from/ === cr3 [n=marc@pdpc/supporter/bronze/cr3] has joined #ubuntu-installer [04:39] cjwatson: where i can find docs about seeds structure files ? [04:40] http://wiki.ubuntu.com/SeedManagement [04:41] oh you mean the STRUCTURE file? [04:41] nono [04:41] description of a seeds [04:41] with = or == [04:41] you mean the text bits at the top of the seed files? [04:42] they're ignored - the only reason they have that format is that the seeds used to be stored in a wiki [04:42] ok [04:42] no difference between = or == ? [04:43] in the wiki two years ago, they were different heading levels [04:43] it makes no difference whatsoever now [04:44] germinate ignores all lines that don't start with " * " [04:46] k [04:47] i have a problem which i don't understand... [04:47] when i include a package in my seeds file wich is in my local repository on the cd, the base installer install it [04:47] and i don't want... [04:47] Dependency problem ? [04:47] Germinate Problem ? [04:52] i have no clue... [04:59] you probably gave it a Priority: of required, important, or standard [05:00] in the packaging ? [05:01] Priority: optional [05:01] or else something depends on it in the base system [05:01] maybe..., how can i see germinate output ? [05:01] or you put it in the minimal seed and cdimage inflated the priority for you [05:01] sudo apt-get install germinate [05:01] man germinate [05:02] k :) [05:02] thanks for your help, will search in this road [05:50] ubiquity: cjwatson * r1851 ubiquity/ (73 files in 4 dirs): [05:50] ubiquity: * Import translations for Cancel, Back, Forward etc. buttons from gtk+2.0 [05:50] ubiquity: 2.10.9-0ubuntu1 (LP: #43915). [06:12] blink, this is a fun merge === cjwatson attempts the m-a merge the other way round to see if that's any easier [06:18] Really? I thought I was fairly up to date. [06:19] The only issue should be glade, but that's as easy as just cutting the MA section out and pasting it into the most recent copy. [06:36] Did I do something wrong in bzr? === secureboot_ [n=lstclair@coriakin.cse.psu.edu] has joined #ubuntu-installer [07:09] is there any way to do encrypted partitions from the installer in feisty? any plans to enable this? [07:11] if not, is there any way to use etch's partman when building your own custom installer? [07:13] evand: I just made a bunch of changes ... [07:18] ow, merging .po files is so insanely painful [07:19] I'll just msgmerge it [07:33] evand: could you set your editor to never use tabs in Python code? space-indentation is usually considered better there [07:38] cjwatson: I thought I was, but sure thing. [07:40] looks like maybe you did part-way through the project [07:40] ah [07:40] anyway, not *right* now, mid-merge :-) [07:40] just setting up the test rig [07:40] heh [07:41] note that there was a diff in bin/ubiquity which I dropped (@VERSION@ -> 1.3.8, looked accidental) and an extra change in gtkui.py where it looked like you missed a new-partitioner bug fix (it was an "if 'UBIQUITY_NEW_PARTITIONER' not in os.environ:" around gparted_loop()) [07:41] hmm, dinnertime [07:42] will return afterwards and hopefully commit [07:42] ah, I probably did that recently, in the last merge [07:42] secureboot_: don't think the pieces are in place; if cryptsetup is ever moved to main it may be possible to assemble them from Debian [07:42] secureboot_: not unless you're a d-i god [07:42] (etch's partman) [07:43] cjwatson: thanks again, hopefully this will be much easier going forward. [07:43] yeah, I expect so [07:43] cjwatson: sad. [07:45] evand: I think you get to write the changelog entry, too. Any suggestions? [07:52] cjwatson: uh, "Merged in migration-assistant work, moved the user-setup page to after migration-assistant..." [07:53] here be dragons? [07:54] everything is pretty high level as this is going in all at once [07:54] those two things were obviously the main focus [07:56] oh, and made user-setup able to use users from m-a [08:02] if you want me to describe the feature set in the commit, let me know and I'll give you a run down [08:27] ubiquity: cjwatson * r1852 ubiquity/ (92 files in 8 dirs): merge migration-assistant branch from Evan [08:28] blam [08:28] [ Evan Dandrea ] [08:28] * Merged in migration-assistant work. Moved the user-setup page to after [08:28] migration-assistant (and thus also after the partitioner, since [08:28] migration-assistant needs to know which partitions are to be left [08:28] evand: let me know if that's inaccurate [08:28] intact). Made user-setup able to make use of users from [08:28] migration-assistant. [08:30] looks good [08:30] cjwatson: thanks === evand crosses fingers and hopes everything works as well as it did in his testing [08:36] ubiquity: cjwatson * r1853 ubiquity/ (4 files in 3 dirs): indentation, minor style nits [08:36] (easier to do it straight away, I think) [08:37] et in vim will do it, right? [08:38] yeah [08:38] augroup myauto [08:38] au! [08:38] [blah] [08:38] au FileType python setlocal tabstop=8 softtabstop=4 shiftwidth=4 expandtab [08:38] augroup END [08:38] is what I have [08:38] thanks [08:39] but my .vimrc has all sorts of stuff in it and some of it's global; I have autoindent on for everything [08:39] oh, and let python_highlight_all = 1 is handy too [08:40] I have the habit of just doing a :syn on [08:40] I'm not sure it necessarily highlights everything by default [08:40] hrm, I'll have to play around with it then [08:41] ooh, there's a changelog plugin === cjwatson makes the productivity error of reading vim's help files [08:41] ubiquity: cjwatson * r1854 ubiquity/ubiquity/frontend/gtkui.py: iter -> iterator (avoid shadowing builtins) [08:41] whoops [08:42] r1854 is one that python_highlight_all is handy for catching [08:42] it highlights builtins in cyan (here, anyway) [08:44] ah, indeed it does. Very nice. [08:46] evand: I think UBIQUITY_MIGRATION_ASSISTANT is implemented slightly dubiously - doesn't it need to fix up the page numbers too after removing a page? [08:47] as it is the page numbers skip if you're running without --migration-assistant [08:47] these are all nits though, it's looking pretty good [08:47] of course I've not run it on a system that actually has anything to migrate ;-) [08:48] evand: btw, might be fun for you to set up bzr-cia and have your commits logged here too [08:48] cjwatson: definitely to cia [08:48] cjwatson: It's skipping pages? [08:48] evand: mkdir -p ~/.bazaar/plugins; cd ~/.bazaar/plugins; bzr get http://bazaar.launchpad.net/~jelmer/bzr-cia/main/ cia [08:49] thanks [08:49] then 'bzr cia-project ubiquity' in your branch and make sure the bzr nick is right [08:49] ok [08:49] in ~/.bazaar/bazaar.conf, 'cia_user = evand' or whatever you're comfortable with and 'cia_send_revno = true' [08:50] at least that's what I prefer [08:50] ok [08:50] when you say page numbers, do you mean it's skipping actual pages or just the numbers on the labels? [08:50] and finally, 'bzr submit-cia' after each commit, or put 'post_commit = bzrlib.plugins.cia.post_commit' in the appropriate bit of ~/.bazaar/locations.conf [08:50] evand: just the numbers, the "Step N of M" bit [08:51] ahh, I actually kind of completely forgot about that [08:51] it goes up to "Step 8 of 8" now even if m-a is disabled [08:51] like I say, a nit [08:52] ok, so I'm assuming the workflow you'd prefer is what we've been doing? I'll pull trunk, make some changes and then when I fix a bug you'll merge again. Or would you prefer plain old patches or something else? [08:52] I'd imagine not the latter. [08:53] the former, definitely [08:53] I figured :) [08:53] ideally you should be in core-dev eventually and maintain it directly, of course [08:54] Yeah, I really need to get on the getting work into REVU so I can first get into MOTU [08:54] in retrospect maybe I should have made it writable by ubuntu-installer [08:54] but never mind now [08:55] eh, it's something I've been aiming for anyway, so this is just more incentive. [08:55] ubiquity: cjwatson * r1855 ubiquity/ (d-i/manifest debian/changelog): [08:55] ubiquity: * Automatic update of included source packages: apt-setup 1:0.16ubuntu3, [08:55] ubiquity: migration-assistant 0.3.1, partman-auto 62ubuntu4, [08:55] ubiquity: partman-basicfilesystems 51ubuntu1. [09:02] ok, I should be all set up for CIA [09:02] Should I do it on both m-a and my ubiquity branch or would you prefer just one of them? [09:03] why not both, and I'll tell the CIA bot to mention those commits here [09:03] ubiquity: cjwatson * r1856 ubiquity/debian/changelog: releasing version 1.3.19 [09:04] ok, great. Already done [09:04] what's the project name (cia-project)? [09:05] hrmm, where do I set that? [09:05] I have the post commit hooks [09:05] bzr cia-project in the branch [09:05] oh, right! [09:05] it knows about the project name and the branch name - the branch name comes from 'bzr nick' [09:06] the bot only needs to be told about project names though [09:06] there you go, ubiquity 1.3.19 uploaded, just under the feature freeze wire. :-) [09:06] ok I did a bzr cia-project migration-assistant in m-a [09:06] haha, awesome! Thanks again for all of your help. [09:06] all right, the bot'll know about that now [09:07] and bzr cia-project ubiquity.ma [09:07] creative, I know [09:07] no, that should be bzr cia-project ubiquity and bzr nick ma [09:07] oh crap [09:07] otherwise it won't link it up properly [09:07] it's ok, you can change it easily [09:07] ok, just did [09:08] do I need to do anything else, like add these to my projects list on cia.navi.cx? [09:08] I think it should do that magically when you commit [09:09] great [09:14] oh, I forgot to note in the changelog how to invoke it [09:14] maybe you could post to ubuntu-devel@ about that? [09:14] I left a note in the spec status whiteboard about it [09:15] will do [09:15] I was planning on doing a planet post as well [09:19] good plan === philwyett [n=philwyet@bb-87-81-146-45.ukonline.co.uk] has joined #ubuntu-installer [10:14] migration-assistant: evand * r33 migration-assistant/ (firefox-import.c firefox-import.h import.c): Added Opera support. [10:15] neat!