[00:21] <CIA-3> wubi: Agostino Russo * r84 trunk/ (4 files in 4 dirs):
[00:21] <CIA-3> wubi: Reverted previous changes as suggested by Evan Dandrea and instead
[00:21] <CIA-3> wubi: will hide Wubi option in CD menu if the drive is larger than
[00:21] <CIA-3> wubi: max_iso_size unless a --force-wubi argument is used (LP: #338015)
[01:24] <CIA-3> wubi: Agostino Russo * r85 trunk/ (47 files in 3 dirs):
[01:24] <CIA-3> wubi: Modified English strings to make them less technical and added
[01:24] <CIA-3> wubi: missing ones
[01:39] <CIA-3> wubi: Agostino Russo * r86 trunk/ (47 files in 3 dirs): More text string edits
[04:28] <esp1> i have used the alternate installer and installed encrypted root system a few times, and it struck me that the data is written in the same order to the disk every time.....isn't this a weakness in the setup, and give an unnessesary attack vector at the crypto?
[08:18] <CIA-3> wubi: Agostino Russo * r87 trunk/ (9 files in 4 dirs):
[08:18] <CIA-3> wubi: * Ask for confirmation before quitting
[08:18] <CIA-3> wubi: * Some text edits
[08:18] <CIA-3> wubi: * No need to set the title of the message boxes
[08:21] <CIA-3> wubi: Agostino Russo * r88 trunk/ (45 files in 2 dirs): Updated gettext template and translations
[08:55] <CIA-3> ubiquity: evand * r3076 ubiquity/ (d-i/manifest debian/changelog):
[08:55] <CIA-3> ubiquity: Automatic update of included source packages: apt-setup
[08:55] <CIA-3> ubiquity: 1:0.37ubuntu10, console-setup 1.28ubuntu6, debian-installer-utils
[08:55] <CIA-3> ubiquity: 1.66ubuntu1, hw-detect 1.71ubuntu5, kboot-installer 0.0.1ubuntu9,
[08:55] <CIA-3> ubiquity: localechooser 2.09ubuntu2, partman-auto-loop 0ubuntu16, partman-base
[08:55] <CIA-3> ubiquity: 129ubuntu1, partman-basicfilesystems 62ubuntu2, partman-partitioning
[08:55] <CIA-3> ubiquity: 64ubuntu4, user-setup 1.23ubuntu14.
[08:56] <cjwatson> evand: looks like you omitted to push your clock-setup change from a while back, so I got a reject ... could you push it to lp:~evand/clock-setup/adjtime or something and I'll merge it?
[08:57] <evand> sure, sorry about that
[08:59]  * evand turns a fake wheel next to bzr
[09:00] <CIA-3> ubiquity: evand * r3077 ubiquity/debian/po/ (79 files): debconf-updatepo
[09:04] <evand> cjwatson: pushed to the location you specified
[09:06] <cjwatson> ta
[09:10] <CIA-3> clock-setup: cjwatson * r205 ubuntu/ (debian/changelog finish-install.d/10clock-setup): merge lp:~evand/clock-setup/adjtime
[09:23] <cjwatson> evand: thanks
[09:23] <CIA-3> clock-setup: cjwatson * r206 ubuntu/debian/changelog: releasing version 0.97ubuntu3
[09:23] <davmor2> xivulon: long time no see dude :)
[09:23] <evand> sure thing
[09:25] <xivulon> hi
[09:25] <xivulon> it's complicated for me to use irc during the day
[09:25] <davmor2> :)
[09:25] <CIA-3> ubiquity: evand * r3078 ubiquity/debian/changelog: releasing version 1.11.15
[09:25] <xivulon> you have to show up at night :)
[09:26] <xivulon> evand have to go shortly
[09:26] <xivulon>  bittorrent downloader does not
[09:26] <xivulon> work
[09:26] <xivulon> you can run with --no-bittorrent
[09:26] <xivulon> I will have a look in the coming days
[09:26] <evand> xivulon: ok, I sent you an email.  A new CD is being built with r88 as we speak.
[09:26] <xivulon> Steve enabled the torrent only yesterday
[09:27] <xivulon> cool
[09:34] <davmor2> evand: If you can give me a ping once it's done I'll have a play :)
[09:37] <evand> davmor2: I have to run over to LHR in a bit, but sure, I'll let you know once I've verified that it's not completely broken :)
[12:43] <CIA-3> installation-guide: cjwatson * r449 ubuntu/debian/changelog: releasing version 20081208ubuntu1
[13:11] <CIA-3> choose-mirror: cjwatson * r608 ubuntu/ (4 files in 2 dirs):
[13:11] <CIA-3> choose-mirror: Allow preseeding the default CC.archive.ubuntu.com setup by preseeding
[13:11] <CIA-3> choose-mirror: mirror/http/mirror to CC.archive.ubuntu.com, and likewise for
[13:11] <CIA-3> choose-mirror: mirror/ftp/mirror (LP: #18225).
[14:21] <CIA-3> choose-mirror: cjwatson * r609 ubuntu/debian/changelog: releasing version 2.27ubuntu4
[14:37] <CIA-3> installation-guide: cjwatson * r450 ubuntu/ (debian/changelog en/appendix/preseed.xml):
[14:37] <CIA-3> installation-guide: Document new scheme for preseeding the use of CC.archive.ubuntu.com
[14:37] <CIA-3> installation-guide: mirrors (LP: #18225).
[14:38] <CIA-3> installation-guide: cjwatson * r451 ubuntu/ (build/entities/common.ent debian/changelog): Bump kernelversion to 2.6.28.
[15:05] <CIA-3> installation-guide: cjwatson * r452 ubuntu/debian/changelog: releasing version 20081208ubuntu2
[15:29] <davmor2> evand: 20090305.1 dl'd no umenu though :(
[15:36] <davmor2> Also I stopped wubi from accessing the web to ensure that it was the version on the cd being used :)
[15:46] <cjwatson> I've tagged all the old ubiquity releases in bzr, at least those I could identify automatically; the only tags in place were 1.8.0 and everything since 1.9.19
[15:46] <cjwatson> helps a bit when looking at old bug reports
[15:47] <cjwatson> "bzr log --line | grep 'releasing version' | sed 's/:.* / /' | tac | while read rev ver; do bzr tag -r$rev $ver; done" is the rune I use for this kind of thing
[16:13] <evand> davmor2: how did you test for umenu?
[16:15] <davmor2> evand: Yes sorry wrote that because it didn't auto start like it used to.  It's actually part of wubi now correct?
[16:17] <evand> yes
[16:17] <evand> it should autostart though
[16:17] <evand> I'll look into it
[16:17] <evand> pulling down a CD now
[16:21] <davmor2> evand: permission denied removing C:\ubuntu\uninstall-wubi.exe  I'm guessing because it is being used at the time
[16:22] <evand> davmor2: I think the uninstall is broken.  I remember a note from xivulon, but I'm about to go on a call so I can't look it up
[16:23] <evand> might want to use an old wubi executable to remove it
[16:23] <davmor2> no probs
[16:27] <davmor2> evand: It deletes it if you use the uninstall option from a new instance of wubi
[16:29] <evand> davmor2: from an email from Ago:
[16:29] <evand> 4) Other outstanding bugs I am aware of:
[16:29] <evand> * Uninstaller does not work when the uninstaller itself is in the
[16:29] <evand> directory to be deleted (which is currently the case)
[16:32] <davmor2> evand: definitely no autostart though
[16:32] <evand> noted
[16:33] <davmor2> other than that seems okay :)
[16:35] <davmor2> doesn't spit the cd out either
[16:40] <evand> davmor2: does a wubi install work?
[16:40] <davmor2> Yes :)
[16:41] <evand> FANTASTIC
[16:41] <davmor2> evand: the install is fine removal autorun and cd eject are the only issues.  I've installed it twice now :)
[16:43] <evand> great
[19:15] <davmor2> evand: Wasn't there a bug about using grey to represent the new full drive in ubiquity?
[19:27] <davmor2> the bar in kubuntu is now grey again so it looks like it is empty
[20:46] <evand> https://bugs.edge.launchpad.net/ubuntu/+source/ubiquity/+bug/336755 - can you just add a note in there to remember to address it in the kde frontend as well?  Thanks!
[21:57] <Wellark> btw, does ubiquity support setting up full encryption (root + swap) by now?
[22:01] <cjwatson> no, unfortunately not
[22:01] <cjwatson> it's becoming an increasing priority though ...
[22:04] <Wellark> true.. laptop users would love that...
[22:06] <Wellark> maybe I get interested hacking on d-i and ubiquity ^^
[22:07] <Wellark> ubiquity encryption support would be nice and shouldn't be that hard either
[22:16] <cjwatson> would involve figuring out how to integrate partman-crypto
[22:19] <Wellark> partman has crypto already, no?
[22:19] <cjwatson> in the partman-crypto component, yes :-)
[22:19] <Wellark> ah, OK :)
[22:19] <cjwatson> but there's no UI frontend code for it in ubiquity yet
[22:22] <Wellark> well, if I have some spare time in the near future I might look into this first: https://bugs.edge.launchpad.net/ubuntu/+source/debian-installer/+bug/337957
[22:23] <Wellark> btw, could someone set priority for this bug.. I'm thinking like "High" or something..
[22:29] <cjwatson> which installer - alternate or desktop?
[22:29] <cjwatson> for 337957
[22:30] <cjwatson> BTW, people should relax about the priority field. Mostly I don't care about it for bugs I'm actually working on.
[22:30] <cjwatson> ok, alternate
[22:31] <Wellark> but someone else might care ;)
[22:31] <cjwatson> who else?
[22:33] <Wellark> don't tell me you are the only one working on d-i? O.o
[22:34] <cjwatson> I'm not, but the priority field just isn't actually that important to developers except in a few cases :)
[22:34] <cjwatson> anyway, the problem is that os-prober doesn't do anything with the encrypted filesystem, not even "I have no idea what to do with this"
[22:35] <Wellark> OK. I kinda guessed something like that
[22:36] <cjwatson> any idea how it could go about recognising an encrypted Windows XP partition?
[22:38] <Wellark> well, it wouldn't have to recognise the partition, but at least double check where to install grub
[22:38] <cjwatson> http://www.truecrypt.org/ says "No TrueCrypt volume can be identified (volumes cannot be distinguished from random data)"
[22:38] <cjwatson> um, yeah, the problem is that conflicts with a general objective I have to ask as few questions as possible
[22:38] <cjwatson> on systems where we believe we're the only installed operating system, therefore, we just go ahead and install grub
[22:39] <cjwatson> (0,1,0) (12748,254,62)  1       32256-104864094719      104864062464    primary
[22:39] <cjwatson> unknown /dev/sda1
[22:39] <cjwatson> parted doesn't recognise it, of course
[22:39] <cjwatson> Wellark: what does 'fdisk -l' say for that partition?
[22:39] <cjwatson> specifically I'm interested in the partition type code
[22:45] <Wellark> cjwatson: I attached the output to the bugreport
[22:45] <Wellark> /dev/sda1   *           1       12749   102406311    7  HPFS/NTFS
[22:45] <cjwatson> ok, at least it's 0x7
[22:45] <Wellark> yes, but it might be anything
[22:46] <cjwatson> sure, but a 0x7 partition is some indication that we aren't alone on the disk
[22:47] <Wellark> I could argue that if user decides to install /boot on anything else than the first partition then it would be nice to ask where to install grub
[22:49] <btm> it appears that passwd/root-password-crypted may not be working with jaunty. with the SHA512 switch in LP:51551, is there a different debconf variable to be used now or maybe it's broken?
[22:49] <cjwatson> Wellark: I think a better test would be "are there any partitions on this drive that weren't created by Ubuntu" or similar
[22:50] <cjwatson> btm: you'll need to use SHA512-crypted passwords, of course ...
[22:50] <Wellark> cjwatson: yes, and the point is that it's doable :)
[22:50] <cjwatson> Wellark: never said it wasn't, just said it wasn't trivial
[22:50] <Wellark> neither did I :D
[22:51] <cjwatson> btm: (hmm, maybe we should check the crypt format in use and supply appropriate parameters to chpasswd)
[22:51] <cjwatson> btm: anyway, please file a bug on https://launchpad.net/ubuntu/+source/user-setup/+filebug if it isn't working
[22:52] <btm> cjwatson: I'll try with a SHA512 password. It's currently an MD5 password (auto install from intrepid)
[22:53] <cjwatson> btm: right, at present it needs to be SHA512
[22:53] <cjwatson> btm: I'd appreciate a bug on that, actually - we could be a little better about backward-compat
[22:55] <btm> cjwatson: np, will test to make sure sha512 works and file.
[23:00] <Wellark> ok. g2g. cu!
[23:04] <btm> cjwatson: openssl and friends appear to not support sha512 password generation, is there an included tool?
[23:10] <cjwatson> err, I'd have to dig, I didn't make this change myself
[23:11] <cjwatson> chpasswd -S maybe?
[23:11] <cjwatson> $ echo cjwatson:foo | chpasswd -S
[23:11] <cjwatson> cjwatson:$6$K./rc/OhIRi$ylKWgewTkGP3TyXfwj8nnKyIhph66WucLseLjGKKzRM0oRcuRzng2szcC/JZpY13dLxmlILx7eSfdfMHTruH40
[23:12] <cjwatson> maybe chpasswd -S -c SHA512 actually
[23:12] <cjwatson> that salt seems too short for SHA512
[23:13] <btm> it should respect ENCRYPT_METHOD in /etc/login.defs
[23:14] <cjwatson> well, it does seem to be respecting it, except that it's using an MD5-length salt
[23:14] <cjwatson> so I think there's a bug lurking
[23:14] <cjwatson> in the meantime, 'chpasswd -S -c SHA512' seems to work
[23:20]  * cjwatson debugs a bit
[23:23] <cjwatson> actually, I think I'm just on crack. Of course SHA512 has a variable salt length. 'chpasswd -S' alone should work
[23:36] <btm> i take it 'd-i passwd/root-password-crypted' wouldn't be stored such that I could verify it with 'debconf-get-selections --installer' for security reasons? I see the other settings (like passwd/root-login boolean true) but password still isn't working.
[23:53] <CIA-3> pkgsel: cjwatson * r132 ubuntu/debian/ (changelog pkgsel.templates postinst):
[23:53] <CIA-3> pkgsel: Run updatedb at the end of pkgsel, unless pkgsel/updatedb is preseeded
[23:53] <CIA-3> pkgsel: to false (LP: #8195).
[23:53] <cjwatson> btm: indeed, it's cleared out of the debconf database towards the end of installation
[23:55] <CIA-3> pkgsel: cjwatson * r133 ubuntu/debian/changelog: releasing version 0.20ubuntu13
[23:57] <CIA-3> installation-guide: cjwatson * r453 ubuntu/ (debian/changelog en/appendix/preseed.xml): Document pkgsel/updatedb (LP: #8195).