[10:01] <mpt> hi evand, I'm watching someone installing Ubuntu for the first time on a Pentium M
[10:01] <evand> hello
[10:01] <evand> painful experience?
[10:01] <mpt> it's revealing some deficiencies in progress feedback
[10:02] <mpt> e.g. we've been looking at a disabled version of the keyboard layout page with a busy cursor for the past two minutes or so
[10:04]  * mpt realizes "Require a password to log in" is bad wording
[10:05] <evand> do you have any suggestions, assuming that it would be difficult to give finer grained progress status?
[10:16] <mpt> bbiab
[10:25] <evand> sure thing
[11:22] <evand> wow, lots done on ubiquity while I was away on vacation.  Thanks for the hard work, guys!
[13:06] <StevenK> evand: I was testing Wubi trying to install UNR, and the Install button is a no-op. It seems that Wubi is expecting the CD to identify itself as Ubuntu Netbook Remix, whereas in .disk/info it's Ubuntu-Netbook-Remix. Any suggestions?
[13:06]  * evand digs
[13:08] <StevenK> evand: Trying to fix get_iso_info (or some such, I was digging a few days ago) to return the CD name without dashes results in an exception, but I don't remember which. :-/
[13:09] <xivulon> Hi StevenK, I have been on holiday on and off, and have done little work but will resume these days
[13:10] <StevenK> xivulon: Hey!
[13:10] <xivulon> if you are around after 10pm london time, it's best for me as I can actually code
[13:11] <StevenK> That's my tomorrow morning, which should be fine
[13:13] <xivulon> What is the content of .disk/info for UNR?
[13:14] <StevenK> Ubuntu-Netbook-Remix 9.10 "Karmic Koala" - Alpha i386 (20090817)
[13:14] <StevenK> For example
[13:15] <evand> StevenK, xivulon: how about this: http://pastebin.ubuntu.com/254494/
[13:16] <StevenK> evand: I did something similar to that, and it blew up
[13:16] <evand> lovely
[13:16] <evand> gotta run, lunchtime appointment
[13:17] <xivulon> don't remember if the match is against the name or the group name, in the latter case you have to replace '-' with ''
[13:21] <xivulon> it's a name check, so yes evand your patch should be ok, have to check whether the name is used to compose the ini file group name though
[13:22] <StevenK> xivulon: I'm happy to test it if you give me a .exe to try.
[13:22] <StevenK> When I tried to build Wubi, it didn't look the same at all
[13:23] <xivulon> but don't think so
[13:24] <xivulon> StevenK, what happened when you tried to build it? Any error message?
[13:25] <StevenK> xivulon: Nothing like that, the .exe I ran only wanted to provide information
[13:26] <xivulon> ?
[13:27] <StevenK> xivulon: I'm not sure if I still have a copy of the .exe
[13:27] <xivulon> cannot test it now anyway
[13:29] <StevenK> xivulon: If you look at it when you can, that would be fantastic.
[14:21] <kirkland> evand: hiya
[14:21] <kirkland> evand: today's my first day back after holiday
[14:22] <evand> kirkland: no worries, hope you had a good time :)
[14:22] <kirkland> evand: absolutely!  i went back to the highlands
[14:23] <kirkland> evand: take your dame up there sometime, it's awesome :-)
[14:23] <evand> oooh, will do
[14:23] <evand> looks very relaxing
[14:23] <evand> just got back from Paris, which is nice, but certainly not the trip to take for a rest
[14:24] <kirkland> evand: i'm downloading the daily iso's now
[14:24] <kirkland> evand: did you commit http://pastebin.ubuntu.com/250846/ yet?
[14:24] <evand> kirkland: nope, I'm hoping you can give it a look over first
[14:24] <kirkland> evand: okay, i'm looking at it now
[14:25] <evand> I've tested it and it works for me, but your mileage may vary
[14:25] <evand> thanks
[14:28] <kirkland> evand: okay, one thing ...  (and you're just repeating what i did) ...l
[14:28] <kirkland> evand: +	error "There were no usable swap devices.  Exiting."
[14:29] <kirkland> evand: i'm wondering if that should be a warning, exit 0, instead of an error, exit 1
[14:29] <kirkland> evand: b/c most systems i install have no swap space
[14:29] <evand> good call
[14:29] <kirkland> evand: which is valid/secure
[14:29] <evand> indeed
[14:29] <kirkland> evand: so this could be handled 2 ways
[14:29] <kirkland> evand: i think we should probably prevent ecryptfs-setup-swap from even being called, if there are no swaps
[14:30] <kirkland> evand: in which case, we can leave this an an error()
[14:30] <evand> actually, I'd prefer to call it in user-setup, if the encrypted option is selected
[14:30] <kirkland> evand: such that if a person (or process) is calling ecryptfs-setup-swap, and there is no swap, then error makes sense to me
[14:30] <kirkland> evand: "call it" = what?
[14:31] <kirkland> evand: call ecryptfs-setup-swap?
[14:31] <evand> sorry, to clarify, I'd like to execute ecryptfs-setup-swap in the user-setup-apply script if the encrypted home directory option is set, rather than executing it somewhere in partman's apply routines.
[14:32] <kirkland> evand: if so, then just put something like [ $(wc -l /proc/swaps) -gt 1 ] && ecryptfs-setup-swap
[14:32] <kirkland> evand: oh, yes, absolutely!
[14:32] <evand> sure
[14:32] <kirkland> evand: basically, in user-setup, then, only call ecryptfs-setup-swap if there are swaps to encrypt
[14:33] <evand> right, I'm just being pedantic on what the best place is to handle the "no swap" condition, but I think handling it in user-setup if you're keen on keeping "no swap" as an error condition in ecryptfs-setup-swap
[14:33] <kirkland> evand: do you have a system with encrypted-lvm?
[14:33] <evand> nope, but I can create one with kvm
[14:33] <kirkland> evand: i could be persuaded either way
[14:34] <kirkland> evand: okay, i just want to make sure that ecryptfs-setup-swap doesn't error out non-zero in cases where you're actually "okay"
[14:34] <evand> indeed
[14:34] <kirkland> evand: ie, 1) no swap, 2) lvm-encrypted swap
[14:34] <kirkland> evand: throwing a warning is fine, i think
[14:34] <kirkland> evand: but exiting 0 would be proper, i believe
[14:35] <evand> indeed
[14:35] <kirkland> evand: also, do we need to talk to mpt or someone about notifying the user at this point in user-setup that hibernation will be disabled?
[14:35] <evand> so are you happy to commit this upstream and release a new version, or should we carry a patch in ubuntu?
[14:35] <kirkland> evand: oh, no, i'll commit this upstream
[14:35] <evand> okay great
[14:35] <kirkland> evand: i'll do that this morning
[14:36] <evand> awesome, thanks
[14:36] <mpt> evand, oh, sorry, I forgot to get back to you about the progress feedback
[14:36] <kirkland> evand: you bet, thanks for the enhancement
[14:36] <evand> I'll upload a version of user-setup as soon as we have that in the archive then
[14:36] <kirkland> evand: cool
[14:36] <evand> mpt: no worries
[14:37] <kirkland> evand: once this is there, let's ping kees, but he told me last week that he was cool displaying the option to encrypt home again by default
[14:37] <evand> okay
[14:37] <kirkland> evand: he wanted the default value to be "require a password to login"
[14:37] <evand> yeah, best to have him look it over early on
[14:37] <kirkland> evand: right
[14:37] <kirkland> evand: he wanted to get mpt
[14:37] <mpt> evand, I think a good thing to do would be to switch panes instantly when you click Back/Forward, showing a spinner in the middle of the destination pane until its contents are ready
[14:38] <kirkland> evand: mpt's feedback on somehow *discouraging* users from the "no password" option :-)
[14:39] <mpt> evand, completely unrelatedly, I realized while watching this morning that "Require a password" should probably be "Require my password"
[14:40] <evand> mpt: so just so I'm clear, you're on the keyboard page and the partition page is up next.  You press next, the partition page appears but all you see is the "Prepare disk space" heading, which, along with the buttons at the bottom, is disabled.  When the interface is ready, everything that was disabled is enabled, and the rest of the interface is shown.  Is that correct?
[14:40] <evand> mpt: okay, I'll make that change now
[14:40] <evand> kirkland: noted
[14:41] <mpt> evand, yes, plus the spinner
[14:42] <mpt> actually, I might have reported this before somewhere
[14:42] <evand> mpt: indeed, just left that out for brevity
[14:42] <evand> mpt: do let me know if you have a bug number for this already, otherwise I'll file one to track it
[14:43]  * mpt waits for Launchpad to respond
[14:44] <evand> haha
[14:44] <evand> Ubuntu, now booting faster than Launchpad page requests.
[14:45] <kirkland> evand: :-)
[14:45] <kirkland> evand: *that's* funny
[14:46] <mpt> evand, found it: I suggested the same solution for bug 336751
[14:46] <evand> kirkland: in fairness, I stole that line from Keybuk
[14:46] <kirkland> evand: heh, i'll slap him on the back, then
[14:47] <mpt> evand, if you have a screenshot of the partitioning step handy, I can answer cjwatson's question in that report
[14:47] <evand> http://people.canonical.com/~evand/screenshots/ubiquity/1.12.3/4-automatic-partitioning.png
[14:47] <evand> a little old, but probably still useful
[14:48] <kirkland> evand: okay, on second thought, I am going to make the no-swaps-available situation exit 0 with a warning in ecryptfs-setup-swap
[14:48] <kirkland> evand: your filter logic contains a bunch of smarts to determine if there's anything to be encrypted
[14:48] <kirkland> evand: and you'd nearly have to duplicate that in user-setup otherwise
[14:49] <evand> well, it's your filter logic, but indeed
[14:49] <evand> good call
[14:49] <evand> (or someone's from upstream)
[14:49] <kirkland> if [ -z "$swaps" ]; then
[14:49] <kirkland>         warn "There were no usable swap devices to be encrypted.  Exiting."
[14:49] <kirkland>         exit 0
[14:49] <kirkland> fi
[14:50] <kirkland> evand: i think that makes sense
[14:50] <kirkland> evand: i'll document the behavior accordingly in the manpage
[14:50] <kirkland> (which doesn't yet exist, I see :-)
[14:50] <evand> haha
[14:55] <CIA-33> ubiquity: evand * r3376 ubiquity/ (3 files in 2 dirs):
[14:55] <CIA-33> ubiquity: Use "Require my password" rather than "Require a password" for the
[14:55] <CIA-33> ubiquity: sake of clarity. Thanks Matthew Paul Thomas.
[14:56] <mpt> \o/
[14:57] <mpt> evand, so the advanced one still looks like this? http://www.oreillynet.com/onlamp/blog/images/gparted.png
[14:57] <CIA-33> ubiquity: evand * r3377 ubiquity/gui/qt/app.ui: Whoops. Forgot the necessary change for KDE for the previous commit.
[14:57] <evand> mpt: nope, not for a long time
[14:57] <evand> mpt: http://people.canonical.com/~evand/screenshots/ubiquity/1.12.3/4-manual-partitioning.png
[14:57] <evand> (the shot you posted was from when we xembeded gparted)
[14:58] <mpt> ah, thanks
[14:58] <mpt> I should have guessed that from the previous URL :-)
[15:00] <mpt> bbiab, meeting...
[16:27] <mpt> evand, I commented in bug 336751
[16:28] <mpt> evand, but that's only a subset of the general suggestion to use a spinner whenever loading the next page is taking too long (more than about 2 seconds)
[16:28] <mpt> evand, do you want a separate bug report for that?
[16:29] <evand> please, if you have time
[16:29] <mpt> ok
[16:40] <mpt> evand, implementing something like this might be really interesting for Ubuntu: http://blog.mozilla.com/metrics/2009/07/30/an-improved-experience-for-2000000-non-firefox-users/
[16:42] <evand> mpt: Colin mentioned a while ago that they had previously floated the idea of a "Problems installing?  Let us know here: http://url"-type message on the CD sleeve.
[16:42] <evand> I imagine following what Mozilla did wouldn't be too difficult
[16:42] <evand> just need to get buy in
[16:43] <evand> I'm all for it, I'll start by asking Colin what he thinks
[17:32] <mpt> bug 414912
[18:28] <eeejay> hey evand
[18:29] <evand> hi
[18:30] <eeejay> evand: i have a merge request you will absolutely love
[18:30] <eeejay> evand: https://code.launchpad.net/~eeejay/ubiquity/bug_386982
[18:32] <evand> ah cool
[18:32] <evand> I'll merge that now
[18:32] <eeejay> evand: sweet, thanks
[18:32] <eeejay> evand: one of the cases you gave me wouldn't work because of that bug :P
[18:39] <CIA-33> ubiquity: evand * r3378 ubiquity/ (debian/changelog ubiquity/segmented_bar.py):
[18:39] <CIA-33> ubiquity: Reset the internal pointer to the segment to be resized when
[18:39] <CIA-33> ubiquity: clearing the contents of the segmented bar in the GTK frontend
[18:39] <CIA-33> ubiquity: (LP: #386982). Thanks Eitan Isaacson!
[18:40] <evand> eeejay: ^ Thanks!
[18:40] <eeejay> evand: thank you ack
[18:40] <eeejay> back
[19:59] <Kamilion> Where does debian-installer load storage controller modules from when Detect Disks asks for modules and 'none of the above' is selected? (Which directory on a fat32 stick or floppy?)
[19:59] <Kamilion> I've got a newer system with a LSI SAS2008 adapter supported by the mpt2sas.ko module. It exists in linux-image-2.6.31-5-server, but not in linux-image-2.6.31-5-generic which the Alpha 4 Server cd boots with. I've copied the module onto a USB stick, and insert it when the installer tells me to, but it never seems to find the .ko file. What directory am I supposed to place it in?
[20:30] <Kamilion> Hm, getting an error: "Failed to determine the codename for the release." at "Install the base System". any ideas?
[20:33] <Kamilion> found the cause, drive had become unmounted