[10:21] <ev> so funny story, apt-clone isn't exactly fast in repacking 500MB worth of debs
[10:21] <ev> I should really finish the wireless page in 11.10
[10:21] <ev> a
[10:21] <ev> as having an Internet connection completely avoids this problem
[10:22] <ev> plus everything else in the installer that's improved with a network connection
[10:22] <ev> geoname, langpacks, downloading updates, etc
[10:47] <ev> ugh, I failed to consider that the cache on the target system will have much newer versions of all the repacked debs, obviously
[10:52] <cjwatson> hiya
[10:52] <ev> welcome back
[10:55] <cjwatson> how's the week been so far?
[11:00] <ev> a bit too quiet
[11:01] <ev> no shortage of issues, but nothing that's causing a stir
[11:01] <ev> not as far as I'm aware, anyway
[11:03] <ev> good holiday?
[11:03] <cjwatson> not too bad thanks, good to have a break from work
[11:04] <ev> are going to take advantage of the double long weekends?
[11:04] <ev> are you*
[11:04] <cjwatson> plans for the day include bug 746313, bug 759545, maybe bug 764893, and bug 683904 if I get a chance; unless there's something more urgent I should be working on
[11:04] <ubot2> Launchpad bug 746313 in ubiquity "partman should reuse existing BIOS Boot Partition" [Undecided,Invalid] https://launchpad.net/bugs/746313
[11:04] <ubot2> Launchpad bug 759545 in grub2 "user prompted to update unmodified grub configuration during Ubuntu server upgrade" [Medium,Confirmed] https://launchpad.net/bugs/759545
[11:04] <ubot2> Launchpad bug 764893 in os-prober "os-prober: does not detect Ubuntu in btrfs subvolume" [High,New] https://launchpad.net/bugs/764893
[11:04] <ubot2> Launchpad bug 683904 in grub2 "natty: memtest86+ fails to run, reboots immediately" [High,Confirmed] https://launchpad.net/bugs/683904
[11:04] <ev> dear. lord.
[11:04] <cjwatson> I'm planning to actually get some sleep
[11:05] <ev> haha, a good idea, I'm sure
[11:06] <ev> I'm currently working on bug 766171, but I'm growing tempted to switch off the upgrade option for installs lacking an Internet connection
[11:06] <ubot2> Launchpad bug 766171 in apt-clone "apt-clone does not repack debs that it will not be able to download when there isn't an Internet connection" [High,Confirmed] https://launchpad.net/bugs/766171
[11:06] <cjwatson> might not be a bad idea ...
[11:07] <ev> yeah, I wanted to give it my best go, as it's a really nice feature to have, but it's growing increasingly complex as I discover new facets to the bug.
[11:44] <ev> okay, mpt and I talked it over and we're going to keep the option but change the text dependent on whether there's an active Internet connection (like how the prepare page works): https://docs.google.com/View?docid=0AU5sFuLRpCpBZGZra2pqY2pfMTAxZ25rcnBnNXY&hl=en#4_5_1_Automatic_partitioning_o_8475526086986065
[11:44] <ev> cjwatson: do you think this is a sufficiently critical problem to warrant the late string change?
[11:44] <ev> err addition
[11:45] <ev> personally, I think displaying something in English is better than hiding the option and thus offering no explanation at all
[11:46] <ev> (oh, and this is in the "Ubuntu {older version} on it..." box)
[11:53] <Kurisutian> genec: Hey there! Got a minute to go through the installe process with btrfs?
[11:57] <Kurisutian> cjwatson: Are you there?
[11:59] <Kurisutian> cjwatson: I tried to install natty server 64bit from beta2 on btrfs as you recommended but it still did not work at all... got the same errors and weird behavior of the installer
[12:20] <Kurisutian> genec: Are you there?
[12:22] <cjwatson> Kurisutian: I'm retrying it at the momenet
[12:22] <cjwatson> *moment
[12:23] <cjwatson> ev: I don't see the proposed change in the Google docs link - can you point me to it specificaly?
[12:23] <cjwatson> oh, I see it now
[12:23] <cjwatson> ev: we should keep the "System-wide settings will be cleared" bit
[12:24] <genec> Kurisutian: yes, had a connection issue but solved now
[12:24] <cjwatson> (IMO)
[12:24] <ev> cjwatson: agreed.  mpt, are you okay with that?
[12:24] <cjwatson> ev: I think at this point, I'm leaning towards preferring to disable the option for 11.04 than change the string, though
[12:25] <cjwatson> I mean, if there's no network connection
[12:25] <Kurisutian> genec cjwatson: I did the install again, but it still wanted to mount a non existing /home mountpoint (I'm only using /). Also when mounting the btrfs mounted partition manually first without manually re-formating it to btrfs, mount complains it wouldn't be an ntfs partition???
[12:25] <cjwatson> Kurisutian: do you have full logs?
[12:25] <ev> cjwatson: oh? Could you elaborate on why?
[12:25] <mpt> ev, done
[12:26] <Kurisutian> cjwatson: Where can I find them exactly? Machine is still running so I could get them rather easy out there....
[12:26] <cjwatson> ev: because it's a new feature, and obviously this situation isn't working well; it would be better to leave people in the same state they were in for 10.10, rather than giving them an option we already know has problems
[12:26] <ev> mpt: thanks
[12:27] <cjwatson> Kurisutian: /var/log/syslog and /var/log/partman if the installer is still running; /var/log/installer/syslog and /var/log/installer/partman if you managed to work around problems and have rebooted into the installed system
[12:27] <genec> Kurisutian: did you create a new partition table when using the installer's partitioner?
[12:27] <Kurisutian> genec: Yes, I did try that also... those were fresh harddrives..... didn't work either.....
[12:28] <ev> cjwatson: to clarify, do you mean disable it completely, even when there is an Internet connection?
[12:28] <cjwatson> ev: no, I mean disable it only if there is no network connection
[12:28] <ev> okay
[12:28] <cjwatson> we already have UI that says that things may not work ideally if you don't have a connection
[12:28] <Kurisutian> cjwatson genec: Also when setting the keymap to german (which would be the layout I need) it's going to be completely ignored by the installer.....
[12:29] <ev> my concern with doing that is two people will see different results without understanding why
[12:29] <ev> granted, we already have a fair bit of that on the partitioning page
[12:29] <cjwatson> ev: one of those people would have had an X on the second screen already
[12:31] <ev> I'm not convinced there's a clear connection between that and not having an upgrade option
[12:31] <cjwatson> it's certainly not perfect, but eight days before release I'd prefer to disable a known-problematic code path rather than introduce UI to warn about its problems
[12:31] <cjwatson> I'd feel differently if the upgrade option had been present in 10.10
[12:31] <mpt> cjwatson, if we disable it, we will need to at least explain why, in-place, so that someone doesn't wonder why they see different options from someone else (or from the screenshot in a tutorial or book)
[12:32] <mpt> and that itself would involve a new string
[12:32] <mpt> Other than that I don't mind much either way
[12:32] <cjwatson> mpt: as ev says, there are already several cases in automatic partitioning where people see different options
[12:32] <cjwatson> so one more, as a we-need-to-work-around-this-at-short-notice matter ... *shrug*
[12:33] <cjwatson> (BTW, this is my own position rather than a veto or anything)
[12:34] <mpt> cjwatson, sure, but they're at very different scales of knowing. Hopefully you are much more aware that you previously had Ubuntu 10.10 on this computer than that your router needs rebooting, for example.
[12:35] <Kurisutian> cjwatson: Can I send those files to you?
[12:35] <cjwatson> mpt: Uh, that's not what I'm talking about
[12:35] <cjwatson> mpt: most people have no idea that the reason they don't have a resize option is that their system already has four primary partitions
[12:35] <cjwatson> for example
[12:36] <cjwatson> the UI design offers different options in that case, but does not "explain why, in-place"
[12:36] <cjwatson> Kurisutian: yes
[12:37] <Kurisutian> cjwatson: On their way through dcc.... you just need to accept them
[12:37] <mpt> cjwatson, I would have if I could have. :-) The problem in that case was that it just got way too long and the explanation didn't help you make a decision anyway.
[12:37] <cjwatson> Kurisutian: I can't accept DCC
[12:37] <cjwatson> Kurisutian: meet NAT
[12:37] <Kurisutian> cjwatson: EMail?
[12:37] <cjwatson> Kurisutian: sure.  cjwatson@ubuntu.com
[12:38] <cjwatson> mpt: I'm not saying it would have been easy, but nevertheless, such situations are there already and create divergence from screenshots in tutorials or books; those tutorials or books already have to say that you may see different options (and the ones I've seen generally do)
[12:39] <Kurisutian> cjwatson: They should be in your Inbox right now.....
[12:40] <ev> for what it's worth, I do think this is slightly easier for the user to resolve
[12:40] <ev> if you have four primary partitions *shrug*
[12:40] <ev> but all this takes is connecting to the Internet, which isn't made clear to them
[12:40] <cjwatson> there are a number of cases where it isn't practical to connect during installation
[12:41] <cjwatson> sure, *sometimes* it's relatively easy ...
[12:42] <ev> sometimes? I would think generally.  We have a wireless indicator and it connects automatically for Ethernet via DHCP.
[12:42] <cjwatson> well, seeing as I was just entirely away from any internet connectivity at all for four days, I beg to differ
[12:42] <cjwatson> (I didn't have a laptop with me, but I might have done)
[12:43] <ev> I can't imagine someone taking their laptop on holiday with the intention of installing an operating system on it.
[12:43] <ev> it strikes me as the kind of thing you do with as many tools around you as you can find
[12:43] <ev> though, this is all rampant speculation
[12:43] <cjwatson> forgive me, but this is an awfully first-world attitude
[12:43] <cjwatson> internet connectivity is *not* universal
[12:43] <Kurisutian> cjwatson: Did the email arrive yet?
[12:43] <cjwatson> Kurisutian: please don't nag
[12:44] <cjwatson> yes it's arrived, I'll look as soon as I can
[12:45] <cjwatson> ev: anyway, I'm slightly losing track of the connection between this argument and the original point :-)
[12:45] <Kurisutian> cjwatson: I'm just asking if it went through.... Sometimes our server (even in the first world) tend to mess up once in a while..... alright, thank you very much for that.... :-)
[12:45] <genec> ev, as an example, there are large portions of the US that still only have dialup
[12:45] <ev> True, though I think we're diverging from the point somewhat, which is whether the extra explanation afforded to users who might be able to connect to the Internet and thus have a much better upgrade experience outweighs the confusion but minimal code delta of disabling it
[12:45] <cjwatson> ev: I would have thought that, in cases where it was easy for them to connect to the internet, they'd have already done that to make all the checkmarks go green
[12:46] <ev> cjwatson: my concern there is that the indicator doesn't really stand out, which is why I wanted to finish the wireless page
[12:46] <cjwatson> if they reach a later point without that, then either they're not reading what's in front of them (and there's only a limited extent to which we can help by adding more text) or it isn't straightforward for whatever reason
[12:46] <ev> my hope would be that the note about them connecting to the internet would prompt them to look around the screen
[12:47] <genec> Kurisutian: PM?
[12:47] <ev> again, I'm speculating
[12:47] <cjwatson> ev: but then we'd have to deal with redrawing those options when they connect, which is more code
[12:47] <cjwatson> (which I don't think is there yet?)
[12:47] <Kurisutian> genec: Sure
[12:47] <ev> I've got most of the patch done, actually
[12:47] <cjwatson> ev: I really don't think we should be adding such code eight days before release, if it's not already landed
[12:47] <cjwatson> for oneiric, by all means
[12:48] <cjwatson> Kurisutian: (FWIW, my test btrfs install seemed to work fine)
[12:49] <cjwatson> Kurisutian: ah; so you were installing onto an existing btrfs filesystem, correct?
[12:49] <ev> it's largely the same code that was already in ubi-parted moved to the frontend, but sure, I don't think this is ultimately worth the release team at-large having to take time out to mediate such a battle, when there are much more pressing matters to deal with :)
[12:49] <Kurisutian> cjwatson: I tried every possible combination but nothing worked
[12:49] <cjwatson> Kurisutian: it's a yes-or-no question :-)
[12:49] <ev> so I'll disable it when there's no Internet connection
[12:50] <ev> for O I'd much rather find some way to solve this for the no-Internet case
[12:50] <ev> rather than have to have the distinction
[12:50] <ev> oh and of course get that wireless page in
[12:50] <cjwatson> agreed, if it's feasible
[12:51] <Kurisutian> cjwatson: this time: Yes
[12:51] <cjwatson> Kurisutian: have you already filed a bug report about this?
[12:51] <Kurisutian> cjwatson: No (persuming this is also a YES/NO question.... ^^)
[12:52] <cjwatson> Kurisutian: can you please file one about this specific situation attaching those logs, on the partman-target package in Ubuntu?
[12:52] <cjwatson> Kurisutian: other failures probably aren't quite the same
[12:52] <genec> Kurisutian: perhaps it might be worthwhile to use fdisk and/or dd to prevent confusion in the installer
[12:52] <cjwatson> genec: I would rather fix the problems
[12:52] <cjwatson> in this case, it looks like the clear_partitions code isn't aware of the submodules scheme
[12:53] <genec> ahh, sorry about that.
[12:53] <cjwatson> I wish we could use 'btrfs subvolume set-default' without confusing other things ...
[12:53] <cjwatson> it would be much better than having to change everything that calls mount
[12:53] <Kurisutian> cjwatson genec: Also a clean install didn't work.... so is that related, do I have to mention that, too in my bugreport? Or is that something else?
[12:54] <cjwatson> Kurisutian: it's probably unrelated, and should be tracked separately
[12:54] <cjwatson> Kurisutian: as I said, a clean install worked perfectly for me ...
[12:54] <cjwatson> I would need logs from that case
[12:56] <Kurisutian> cjwatson: that's what I did before, but this won't work either..... that's when the installer complains about not being able to mount /home even though I don't have a seperate /home partition... I'll recreate that scenario and send you the logs right away
[12:57] <ev> mpt: there's a call about the installer for 11.10 at 1pm.  If you're interested, I'm in one of the phone rooms.
[13:00]  * cjwatson dials in
[13:03] <cjwatson> Kurisutian: I can see something wrong with keyboard layout, indeed ...
[13:04] <Kurisutian> cjwatson: that ain't to bad. I've been living in the states for a while and remember their layout fairly well.... ^^
[13:18] <Kurisutian> cjwatson: I just did send you the new logfiles for a regular installation that won't work for me.... I think we should start here since that is the more common thing that would happen....
[13:38] <Kurisutian> cjwatson: Thanks for looking into that. I have to take of right now. Maybe I can get a hold on you later if you ain't that busy! Again thank you, I really appreciate it!
[15:10] <CIA-7> ubiquity: evand * r4707 trunk/ (debian/changelog ubiquity/plugins/ubi-partman.py):
[15:10] <CIA-7> ubiquity: Disable the upgrade/reinstall option when no Internet connection is
[15:10] <CIA-7> ubiquity: available. See LP 766171 for details.
[15:10] <ubot2> Launchpad bug 766171 in apt-clone "apt-clone does not repack debs that it will not be able to download when there isn't an Internet connection" [High,Confirmed] https://launchpad.net/bugs/766171
[15:25] <CIA-7> ubiquity: evand * r4708 trunk/ (147 files in 3 dirs): Update translations from Launchpad.
[15:29] <CIA-7> console-setup: cjwatson * r407 ubuntu/ (debian/changelog setupcon):
[15:29] <CIA-7> console-setup: Weaken test for whether /usr is mounted; testing for /usr/share is
[15:29] <CIA-7> console-setup: sufficient, and fixes operation in d-i.
[15:31] <CIA-7> console-setup: cjwatson * r408 ubuntu/debian/changelog: releasing version 1.57ubuntu20
[15:32] <ev> oh
[15:32] <ev> should I wait for that?
[15:32] <ev> lets go with yes
[15:33] <cjwatson> it shouldn't affect ubiquity
[15:33] <cjwatson> though it would be nice to be consistent, if this might be the final upload
[15:33] <ev> one can always dream
[16:52] <CIA-7> ubiquity: evand * r4709 trunk/ (debian/changelog gui/gtk/ubiquity.ui):
[16:52] <CIA-7> ubiquity: Don't ever show scrollbars in the slideshow. The window growing
[16:52] <CIA-7> ubiquity: slightly is not ideal, but still better than this. See LP 529201
[16:52] <CIA-7> ubiquity: for details.
[16:52] <ubot2> Launchpad bug 529201 in ubiquity "ubiquity shows scrollbar in installation slideshow" [High,Confirmed] https://launchpad.net/bugs/529201
[17:20] <CIA-7> partman-auto: cjwatson * r610 ubuntu/ (4 files in 3 dirs):
[17:20] <CIA-7> partman-auto: Add $reusemethod internal specifier, which excludes the partition if a
[17:20] <CIA-7> partman-auto: partition with the same method already exists.
[17:25] <CIA-7> installation-guide: cjwatson * r491 ubuntu/ (build/entities/urls.ent debian/changelog): Update url-us-keymap (thanks, Manfred Hampl; LP: #741134).
[17:29] <CIA-7> installation-guide: cjwatson * r492 ubuntu/ (debian/changelog en/hardware/supported/i386.xml):
[17:29] <CIA-7> installation-guide: Document that i586 processors and i686 without cmov are no longer
[17:29] <CIA-7> installation-guide: supported (LP: #688195).
[17:31] <CIA-7> installation-guide: cjwatson * r493 ubuntu/debian/changelog: releasing version 20100518ubuntu4
[17:33] <CIA-7> ubiquity: evand * r4710 trunk/ (debian/changelog ubiquity/plugins/ubi-partman.py):
[17:33] <CIA-7> ubiquity: Defend against temporary mountpoints not existing before attempting
[17:33] <CIA-7> ubiquity: to remove them (LP: #759716).
[17:45] <CIA-7> partman-auto: cjwatson * r611 ubuntu/ (debian/changelog recipes/atomic recipes/home recipes/multi):
[17:45] <CIA-7> partman-auto: If there is an existing BIOS Boot Partition, reuse it rather than
[17:45] <CIA-7> partman-auto: creating a new one (LP: #746313).
[17:50] <CIA-7> partman-auto: cjwatson * r612 ubuntu/lib/recipes.sh: fix comment
[17:52] <CIA-7> partman-auto: cjwatson * r613 ubuntu/debian/changelog: releasing version 93ubuntu15
[21:02] <cjwatson> ev: still around?
[21:46] <CIA-7> grub-installer: cjwatson * r1201 ubuntu/ (debian/changelog grub-installer):
[21:46] <CIA-7> grub-installer: Remove grub-gfxpayload-lists in situations where we need to remove
[21:46] <CIA-7> grub-installer: grub-pc (LP: #765270).
[21:48] <CIA-7> grub-installer: cjwatson * r1202 ubuntu/debian/changelog: releasing version 1.60ubuntu3
[22:18] <superm1> hmm, something in 2.6.6 seems to have busted --automatic installs
[22:22] <cjwatson> it busted all installs
[22:22] <cjwatson> waiting to get that grub-installer change accepted before doing an upload with Evan's fix for that
[22:22] <cjwatson> assuming that r4705 matches the failure you're seeing
[22:25] <superm1> i'll try that fix and see
[22:27] <superm1> ah yeah that looks like it's it