[00:50] <CIA-11> ubiquity: dmitrij.ledkov * r5637 trunk/ubiquity/ (frontend/gtk_ui.py segmented_bar.py): pep8 fixes
[10:31] <CIA-11> ubiquity: dmitrij.ledkov * r5638 trunk/ (bin/ubiquity-wrapper debian/changelog):
[10:31] <CIA-11> ubiquity: Prevent thunar from automounting, in ubiquity-wrapper (LP: #719338).
[10:31] <CIA-11> ubiquity: Automounting should probably be address in some better way on Xfce.
[10:43] <CIA-11> ubiquity: dmitrij.ledkov * r5639 trunk/ (3 files in 3 dirs):
[10:43] <CIA-11> ubiquity: Make `$ bzr bd -S' and `$ bzr bd' do something useful, i.e. generate
[10:43] <CIA-11> ubiquity: nice & clean source package or perform clean local builds.
[10:43]  * xnox wants to release ubiquity
[10:45] <cjwatson> Be my guest
[10:45] <xnox> =)
[10:47] <cjwatson> Actually
[10:47] <cjwatson> I was going to excise Python 2 support
[10:47] <cjwatson> Mind if I finish that?
[10:50] <xnox> go ahead.
[10:51] <cjwatson> xnox: ^-
[10:51] <xnox> I am working on a partman-auto-lvm bug, so if I get that done, it might even end up in ubiquity as well =)
[10:51] <xnox> ubiquity has all fixes which should unbreak xubuntu ISOs.
[10:52] <xnox> as well as ubuntu affecting bugs.
[10:55] <cjwatson> xnox: Sorry, don't know if you got this
[10:55] <cjwatson> 11:47 <cjwatson> Actually
[10:55] <cjwatson> 11:47 <cjwatson> I was going to excise Python 2 support
[10:55] <cjwatson> 11:47 <cjwatson> Mind if I finish that?
[10:55] <xnox> i wonder if my irc works.
[10:55] <xnox> yes, go ahead. I did reply....
 go ahead.
 xnox: ^-
 I am working on a partman-auto-lvm bug, so if I get that done, it might even end up in ubiquity as well =)
 ubiquity has all fixes which should unbreak xubuntu ISOs.
 as well as ubuntu affecting bugs.
[10:56] <cjwatson> Yeah, my connection dropped :-/
[10:57] <cjwatson> Thanks
[10:57] <xnox> ah =)
[11:15] <CIA-11> base-installer: cjwatson * r976 ubuntu/ (debian/changelog kernel/tests/runtests):
[11:15] <CIA-11> base-installer: * Cherry-pick from upstream (Ben Hutchings):
[11:15] <CIA-11> base-installer:  - base-installer/kernel/tests: Fix regex syntax error in runtests.
[11:30] <CIA-11> base-installer: cjwatson * r977 ubuntu/debian/changelog: releasing version 1.122ubuntu9
[11:32] <CIA-11> ubiquity: cjwatson * r5640 trunk/ (28 files in 8 dirs): Remove Python 2 support. We aren't going back now.
[11:32] <cjwatson> I suspect there are still a bunch of now-unnecessary str() calls in the KDE frontend, but life's too short
[11:36] <xnox> meh
[11:37] <xnox> I am enjoying old partman bugs
[11:37] <xnox> Ian Jackson: "I should have the option to destroy data."
[11:37] <xnox> made me giggle
[11:57] <xnox> cjwatson: given an empty partition table, parman creates new partition table and then it discovers there is left over lvm metadata, what should it do:
[11:58] <xnox> (i) thow a fit with a cryptic message and bail out
[11:58] <xnox> (ii) offer to remove & reuse, or go back to manual partitioning
[11:58] <xnox> (iii) auto-clear that metadata and continue as per original plan
[11:59] <xnox> --
[11:59] <xnox> this is bug 154086 or http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=542441
[11:59] <ubot2> Launchpad bug 154086 in partman-auto-lvm "Installing to HDD with previous ubuntu fails to create fresh LVM claiming group already in use" [High,Confirmed] https://launchpad.net/bugs/154086
[11:59] <ubot2> Debian bug 542441 in partman-lvm "old LVM data is not erased" [Normal,Open]
[12:00] <cjwatson> Not (i).  I would be slightly inclined to say (ii), but I can see the justification for (iii) because that's effectively what would happen if it were a real partition not LVM.  However, have a think about the effects of code to do (iii) on re-opening things like encrypted containers.
[12:00] <cjwatson> I believe that current code attempts to do (ii), possibly buggily, so I would tend to try to make that work rather than changing the intent.
[12:00] <xnox> cjwatson: (i) is the current behaviour =)
[12:01] <cjwatson> In practice, but I'm fairly sure the intent is (ii) and that it fails due to races or similar.
[12:01] <cjwatson> For instance there are already suitable debconf templates for (ii).
[12:01] <cjwatson> Somewhere.
[12:01] <xnox> cjwatson: it says "The vgname you chose is already in use. Lower priority to be offer to choose a different name." Upon restarting partman you notice the re-create partition table & old lvm group and lv volumes "recovered" out of the blue =)
[12:02] <xnox> yeah the second time you run it. (ii) happens.
[12:02] <cjwatson> Yes, I'm not saying it works right now.  But look at the code, not the behaviour.
[12:02] <xnox> ok. I will make it try to do (ii) instead of throwing a fit in corner cases.
[12:03] <cjwatson> I vaguely recall some bug where Phil Hands made it work by inserting sleeps.  That's obviously not a correct fix, but indicates that there's a race which should be fixed by inserting proper synchronisation.
[12:04] <xnox> not really a race.... it does a name check " does this vgname in use? " just before creating it. It does it via a call to lvs, which finds the zombie VG
[12:05] <xnox> his fix was to add an unconditional call to do ROW = RM-on-Write =)
[12:05]  * xnox a bad pun on COW
[12:06] <cjwatson> I thought part of the problem was that udev rules were still working on it.
[12:08] <xnox> regardless of udev, vgs rescans to find new vgs. Since the current one we have does not implement caching daemon.
[12:09] <xnox> and even the new one that does caching daemon can still invalidate cache & find the zombie vg.
[12:10] <cjwatson> OK
[12:54] <davmor2> cjwatson: the lvm encrypted password doesn't work if you use # in the password,  we are wondering if it is an issue with keyboard layout
[12:55] <cjwatson> Certainly possible
[12:55] <cjwatson> Try the US location
[12:58] <xnox> davmor2: the 94 printable ascii characters should be safe.
[12:58] <xnox> and # is one of them.
[12:58] <cjwatson> No, because it is in a different location on US and UK keyboards.
[12:59] <cjwatson> I mean, it *should* be safe, but it's certainly plausible that the keyboard layout hasn't been configured at the right time.
[12:59] <xnox> cjwatson: and there is this bug open about console-setup that keyboard layout options do not propagate into initramfs......
[13:00] <cjwatson> Quite so.  This is one of the reasons we need it in the initramfs.  It used to work. :-/
[13:00] <davmor2> xnox, cjwatson: it turns out it is he used @ and it is on shift 2
[13:00] <cjwatson> Indeed, so that's the US location.
[13:52] <Mez> cjwatson: https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1047384
[13:52] <ubot2> Launchpad bug 1047384 in ubiquity "System Encryption Password set before setting keyboard locale" [Undecided,New]
[13:56] <xnox> Mez: yes, it is.
[13:56] <xnox> Mez: and it must be....
[13:57] <Mez> xnox: Why must it be /
[13:57] <xnox> Mez: I will comment on the bug. One moment please
[13:57] <ogra_> xnox, well, you coudl do the kbd setup twice
[13:57] <ogra_> (in the backend)
[13:57] <xnox> ogra_: there is better way ;-)
[13:58] <Mez> xnox, then the locale needs to not be set before booting
[13:59] <xnox> Mez: you can if you press Esc and F6 at boot.
[14:01] <xnox> Mez: you can also change keyboard layout in the indicator in the top right? can you not?
[14:02] <Mez> xnox: both of those REQUIRE people to be aware of the problem.  They REQUIRE special knowledge.  It shouldn't mean that someone installing thinks they've broke their system because it won't accept their password
[14:02] <Mez> (Also the keyboard item doesn't work, as it's a password input box)
[14:03] <xnox> Mez: i know, i understand. I am writting up solutions in the bug report.
[14:03] <cjwatson> This is unresolvable by direct sequencing; my understanding was that the keyboard indicator was being included to break the deadlock
[14:03] <cjwatson> Shouting at us won't help
[14:03] <Mez> cjwatson: apologies, was just trying to add emphasis (as no voice).
[14:03] <xnox> Mez: This will need design changes, but I think I can add an extra feature here at the same time.
[14:08] <xnox> Mez: thank you for "discovering" this bug =)
[14:11] <xnox> cjwatson: commented on the bug report. What do you think about using $random keyfile to do cryptsetup and ask for real encryption passphrase after the keyboard layout has been chosen?
[14:12] <xnox> cjwatson: this should also allow keeping $random keyfile in the initramfs for OEM style full-disk encryption, where puppet/chef/etc. kick in on first boot to setup the correct luks passphrases (this is not used by OEMs but it is used by $bigcorps)
[14:13] <xnox> cjwatson: this should make it easier to setup multiple passphrases, which is currently not supported.
[14:13] <cjwatson> I'm not sure I have the brain for this at the moment :)
[14:14] <cjwatson> It's an interesting idea but I don't use cryptsetup myself as yet so I don't really have a feel for it.  Ask Steve maybe?
[14:30] <stgraber> xnox: going with a random key, then adding the real one and removing the random one sounds like a good idea
[14:30] <stgraber> xnox: and I never saw any issue with cryptsetup when running multi-key system, so it should be fine to implement
[14:45] <xnox> as I quick hack, I will unhide the passphrase, such that people atleast see what they are typing =/
[15:26] <zyga> hey
[15:26] <zyga> I've spotted an issue with the photo taking stage
[15:26] <zyga> the camera turns off/on when the user presses the picture button
[15:26] <zyga> for two of my cameras that de-stabilizes the picture quality
[15:27] <zyga> and the result is a very dark / greenish photo
[15:27] <zyga> while normal preview is perfectly fine after a second
[15:27] <zyga> is that reported?
[15:28] <xnox> zyga: search launchpad & report if it wasn't.
[15:28] <xnox> zyga: and if you didn't find it, we can always mark it as duplicate.
[15:28] <xnox> zyga: there are not that many people testing the webcam to be honest =/
[15:30] <ogra_> xnox, everyone who does a laptop install i would guess
[15:31] <xnox> ogra_: either it's perfect or not tested, first bug report about the webcam page =)
[15:31] <xnox> that I see/read
[15:31] <cjwatson> It's certainly not the first ever
[15:31] <ogra_> i could file some but i'm not sure its only ac100/tegra :)
[15:31] <ogra_> and i dont have a spare x86 laptop to verify
[15:32] <ogra_> for me the image list on the right is clearly empty
[15:32] <ogra_> and the preview pic isnt shown at all
[15:32] <ogra_> (if i click on "take picture" i get a proper pic though)
[15:37] <zyga> xnox, will do
[15:37] <zyga> xnox, sorry, I was prepping some dinner :)
[15:37] <zyga> xnox, against ubiquity?
[15:38] <zyga> ogra_, probably v4l
[15:38] <xnox> zyga: ubiquity will do, we will redirect it if it's not ubiquity problem.
[15:38] <xnox> but video problem.
[15:38] <zyga> awesome, thanks
[16:15] <zyga> xnox, reported as 1047493
[16:15] <zyga> xnox, reported as bug 1047493
[16:15] <ubot2> Launchpad bug 1047493 in ubiquity "Camera photo taken by ubuqity is dim despite good quality preview" [Undecided,New] https://launchpad.net/bugs/1047493
[16:15] <xnox> thanks
[16:16] <CIA-11> ubiquity: dmitrij.ledkov * r5640 trunk/ (debian/changelog ubiquity/frontend/base.py):
[16:16] <CIA-11> ubiquity: Fixed slideshow_get_available_locale missing new locales in
[16:16] <CIA-11> ubiquity: extra_slides_dir. (LP: #1046511)
[16:17] <CIA-11> ubiquity: dmitrij.ledkov * r5641 trunk/ (debian/changelog ubiquity/frontend/base.py):
[16:17] <CIA-11> ubiquity: Fixed slideshow_get_available_locale missing new locales in
[16:17] <CIA-11> ubiquity: extra_slides_dir. (LP: #1046511)
[16:54] <zyga> xnox, feel free to retweet: https://twitter.com/zygoon/status/244108311077715970
[16:59] <xnox> zyga: ;-)
[17:43] <bdmurray> bug 1047316 seems to have a pretty simple patch
[17:43] <ubot2> Launchpad bug 1047316 in casper "casper in initrd dies trying to find filesystem.squashfs on raid device" [Undecided,New] https://launchpad.net/bugs/1047316
[17:45] <CIA-11> ubiquity: dmitrij.ledkov * r5642 trunk/ (bin/ubiquity-wrapper debian/changelog):
[17:45] <CIA-11> ubiquity: Remove powerpc/ps3 low-memory profile, as it doesn't do anything any
[17:45] <CIA-11> ubiquity: more. gnome-session-remove command is no longer present, and the init
[17:45] <CIA-11> ubiquity: processes it tried to stop no longer exist under those names. We do
[17:45] <CIA-11> ubiquity: want a low memory profile, which ubiquity-only mode now.
[17:45] <xnox> gotta love spending time trying to understand what a piece of code is doing.....
[17:45] <xnox> .... only to find out it does nothing.
[17:48] <cjwatson> More relevantly, the ps3 port is dead :)
[17:51] <CIA-11> ubiquity: cjwatson * r5643 trunk/ (9 files in 5 dirs): Remove the rest of the long-dead PS3 port.
[17:51]  * cjwatson wields the axe today
[17:52]  * infinity sheds a tear for the PS3.
[19:18] <CIA-11> ubiquity: dmitrij.ledkov * r5644 trunk/ (bin/ubiquity-wrapper debian/changelog):
[19:18] <CIA-11> ubiquity: ubiquity-wrapper: Add support for udisks2 inhibit, drop devkit-disks
[19:18] <CIA-11> ubiquity: inhibit. (LP: #719338)
[19:42] <CIA-11> ubiquity: dmitrij.ledkov * r5645 trunk/ (d-i/manifest debian/changelog):
[19:42] <CIA-11> ubiquity: Automatic update of included source packages: base-installer
[19:42] <CIA-11> ubiquity: 1.122ubuntu9.
[23:46] <CIA-11> ubiquity: dmitrij.ledkov * r5646 trunk/ (debian/changelog gui/gtk/ubiquity.ui): Prevent progress label to expand & shrink the window (LP: #1046241)
[23:47] <CIA-11> ubiquity: dmitrij.ledkov * r5647 trunk/ (gui/gtk/stepPartAdvanced.ui ubiquity/plugins/ubi-partman.py): Use the joined up Add/Remove/Change... toolbar icons as per spec.
[23:48] <CIA-11> ubiquity: dmitrij.ledkov * r5648 trunk/debian/changelog: releasing version 2.11.31