[09:15] <cjwatson> xnox: ubiquity/crypto-done - doit.  partman-crypto/finish.d - could I have a version of that MP against lp:~ubuntu-core-dev/partman-crypto/ubuntu, rather than confusingly against upstream?
[09:15] <xnox> cjwatson: ok. yes, will do.
[09:16] <xnox> sorry about confusing MP
[09:38] <CIA-7> ubiquity: dmitrij.ledkov * r5581 trunk/ (12 files in 8 dirs): (log message trimmed)
[09:38] <CIA-7> ubiquity: Merge lp:~dmitrij.ledkov/ubiquity/crypto-done:
[09:38] <CIA-7> ubiquity: * Make the UI layout test print the failed page/step name.
[09:38] <CIA-7> ubiquity: * Add crypt security key setup page and move crypt/lvm checkboxes to the
[09:38] <CIA-7> ubiquity:  ask page.
[09:38] <CIA-7> ubiquity: * Refactor password validation from ubi-usersetup into a reusable
[09:38] <CIA-7> ubiquity:  component (for ubi-partman).
[09:49] <xnox> cjwatson: updated MP for partman-crypto https://code.launchpad.net/~dmitrij.ledkov/partman-crypto/finish.d/+merge/118699
[09:50] <cjwatson> LGTM
[09:52] <xnox> cool, thanks.
[10:07] <xnox> cjwatson: partman-crypto uploaded.
[10:07] <xnox> cjwatson: ubiquity release?
[10:09] <cjwatson> can you or do you need me to?
[10:24] <xnox> cjwatson: yeah, I can do it. Just asking if it's ok.
[10:25] <cjwatson> Yes, please do
[11:36] <CIA-7> ubiquity: dmitrij.ledkov * r5582 trunk/ (d-i/manifest debian/changelog):
[11:36] <CIA-7> ubiquity: Automatic update of included source packages: flash-kernel
[11:36] <CIA-7> ubiquity: 3.0~rc.4ubuntu17, yaboot-installer 1.1.21ubuntu1.
[11:38] <CIA-7> ubiquity: dmitrij.ledkov * r5583 trunk/debian/ (77 files in 2 dirs): Update translations.
[11:40] <CIA-7> ubiquity: dmitrij.ledkov * r5584 trunk/debian/changelog: releasing version 2.11.19
[12:32]  * ogra_ curses ... 
[12:32] <ogra_> seems my flash-kernel hook lands in the right place now (if i only wouldnt always forget to make it executable !!)
[12:38] <ogra_> xnox, potential testers for you ... https://lists.ubuntu.com/archives/ubuntu-users/2012-August/262852.html
[12:39] <xnox> ogra_: nah =) won't have raid in it yet
[12:39] <ogra_> ah
[12:39] <xnox> ogra_: plus the build is currently failing on a unit test =(
[12:39] <ogra_> hmpf
[12:39] <xnox> so not uploading just yet, although i already tagged it =(
[13:07]  * ogra_ durses about bug 1028905
[13:07] <ubot2> Launchpad bug 1028905 in cdrom-detect "cdrom-detect in quantal omap4 hangs trying to look for install media on an extended partition" [High,Confirmed] https://launchpad.net/bugs/1028905
[13:07] <ogra_> *curses too
[13:24] <ogra_> Aug  8 13:22:49 main-menu[302]: INFO: Menu item 'pkgsel' selected
[13:24] <ogra_> Aug  8 13:22:52 pkgsel: dpkg-divert: error: `diversion of /usr/sbin/update-initramfs to /usr/sbin/update-initramfs.distrib by pkgsel' clashes with `local diversion of /usr/sbin/update-initramfs to /usr/sbin/update-initramfs.flash-kernel-diverted'
[13:24] <ogra_> Aug  8 13:22:52 main-menu[302]: WARNING **: Configuring 'pkgsel' failed with error code 2
[13:24] <ogra_> *sniff*
[13:24] <ogra_> cjwatson, ^^^
[13:25] <cjwatson> ahaha.  um.
[13:25] <ogra_> i could probably remove the diversion in a 99-flash-kernel script or so ... but ...
[13:26] <ogra_> ... it really feels like we should have that diversion logic in live-installer itself somehow
[13:26] <cjwatson> That wouldn't help this!
[13:26] <cjwatson> It's probably simplest to make pkgsel check whether the diversion's already there.
[13:27] <cjwatson> Ugly, but fairly unavoidable I think
[13:27] <ogra_> i guess it would make sense to name it the same then in f-k
[13:27] <cjwatson> Doesn't make a difference
[13:28] <cjwatson> Trying to avoid the clash by naming everything the same would be dangerous, because - even in the best likely case - pkgsel would then remove the diversion before it's time
[13:28] <ogra_> well, i just meant from a practical POV you already have a name in pkgsel you can check for
[13:28] <cjwatson> pkgsel doesn't need to check for a particular name
[13:29] <cjwatson> All it needs to do is check whether *any* diversion is already there
[13:29] <ogra_> ok
[13:30] <cjwatson> This is basically because diversions aren't stackable
[13:30] <cjwatson> Do you need me to do the pkgsel fix, or can you take care of it?
[13:30] <ogra_>  if $(dpkg-divert --list|grep update-initramfs);then foo;fi
[13:30] <ogra_> does that look ok ?
[13:30] <ogra_> (for the check
[13:30] <ogra_> )
[13:30] <cjwatson> You don't need $()
[13:30] <cjwatson> if dpkg-divert --list | grep -q update-initramfs; then ...
[13:31] <ogra_> ah, i thought because of the pipe ...
[13:31] <ogra_> yep
[13:31] <cjwatson> However
[13:31] <cjwatson> Doesn't --listpackage work
[13:31]  * ogra_ will take care ... just wanted to make sure i look for the right bit
[13:32] <ogra_> does that work for local diversions ?
[13:32] <ogra_> else i need to change the way f-k does the diversion ... currently trhats local
[13:32] <cjwatson> if chroot /target dpkg-divert --listpackage /usr/sbin/update-initramfs | grep -q .; then ...
[13:32] <ogra_> without package name
[13:32] <cjwatson> slightly more precise I think
[13:32] <ogra_> k
[13:33] <cjwatson> the man page documents it to work for local diversions
[13:33] <cjwatson> unfortunately dpkg-divert --listpackage exits zero even if there's no diversion, otherwise we could lose the grep
[13:54] <CIA-7> ubiquity: dmitrij.ledkov * r5585 trunk/ (debian/changelog tests/test_gtkui.py): Mark test_pages_fit_on_a_netbook as expected failure. See comment.
[18:29] <soren> Is there a simple way to build an install CD based on Ubuntu release, -updates + an extra repo (like a PPA or whatnot)?
[19:45] <soren> I guess if I can figure out myself how to stitch together the dists/ and pool/, the rest is a matter of copying the stuff from an existing ISO..
[19:46] <soren> and what else?
[19:46] <soren> An extra keyring.
[19:47] <stgraber> soren: live or alternate image?
[19:47] <soren> alternate.
[19:47] <soren> Well, server really, but ykwim.
[19:47] <infinity> People still install servers from CDs?
[19:48] <stgraber> ok, so you might be able to find some things in the debian-cd and cdimage branch of ~ubuntu-cdimage, though quite a few of the distro scripts are private
[19:48] <soren> infinity: The first one, yes.
[19:48] <stgraber> but you're at least looking at a d-i rebuild for the new keyring
[19:48] <soren> Apparently.
[19:49] <infinity> soren: Unless this is a product you're pressing (for customers, trade shows?), it's probably way less hassle to just install with a 12.04.1 CD, and then add the PPA after the fact.
[19:49] <infinity> (Or, if you were pre-seeding, do it in late-commands)
[19:49] <soren> infinity: I'm having to deal with an environment with no Internet connection at all.
[19:49] <soren> At. all.
[19:50] <soren> Apparently, they still exist.
[19:50] <infinity> Ah.
[19:50] <infinity> Banks, for one.
[19:50] <soren> Yeah?
[19:50] <infinity> Though, I'm pretty sure my bank is all OS/{360,390,400} in the non-connected back rooms.
[19:51] <soren> Sounds like they know how to party.
[19:51] <infinity> And how.
[19:51] <soren> stgraber: Ah, the key is embedded in d-i? Yeah, I guess that makes sense.
[19:52] <infinity> So, yeah.  As stgraber says, there's a bit of rebuilding and jiggery-poking, if you actually want your PPA bits to supersede the original CD in the initial install.
[19:52] <infinity> But there's the second option of just having a second package pool that has your junk in it.
[19:52] <soren> I was just about to suggest that.
[19:52] <infinity> And then just installing from that, post-install.
[19:52] <soren> Yeah.
[19:52] <stgraber> second pool + preseed early/late command using apt-key might be much easier indeed
[19:53] <stgraber> just ship the key + second pool the media, load the key, add the entry, update, install your stuff, success
[19:53] <soren> Hm...
[19:53] <stgraber> without actually touching anything from the original media except for the preseed files
[19:53] <soren> I guess I could provide two CD's.
[19:53] <soren> Not even the server CD has space left these days.
[19:54] <infinity> Oh, true, it needs remastering regardless, if you need to add to it.
[19:54] <infinity> Of course, "CD" doesn't have to be a CD.  Just because they have no internets doesn't mean they can't boot from USB. :P
[19:54] <soren> Yeah.
[19:55] <soren> I think the requirement was "an ISO". If they're shoving it onto a USB stick anyway (which I hope to $DEITY they are), it can be as big as I want it to. Almost.
[19:56] <soren> Yeah, that totally works.
[19:56] <soren> Fiddling a bit with the install options to ninja my archive key in should even be rather simple.
[19:57] <soren> stgraber, infinity: Thanks guys.