[00:47] <CIA-4> usb-creator: rgreening * r112 usb-creator/ (3 files in 2 dirs): (log message trimmed)
[00:47] <CIA-4> usb-creator: Update KDE ui (replace some Qt elements with KDE ones and connect some signals to the ui)
[00:47] <CIA-4> usb-creator: Reorganize some code (move some UI initialization to __initUI)
[00:47] <CIA-4> usb-creator: Make reference to backend really private
[00:47] <CIA-4> usb-creator: Try and fix format missing volume crash with a try/except block
[00:47] <CIA-4> usb-creator: Add source_selection_changed and connect appropriate signal to it
[00:48] <CIA-4> usb-creator: Add isEnabled check for the persist ui elements to get_persistence
[08:23] <mattn_> hi
[08:23] <mattn_> i'm trying to set up an modified live-cd at our company
[08:23] <mattn_> it works already great
[08:23] <mattn_> but now i'm trying to add a casper script to modify some gconf settings and desktop stuff
[08:23] <mattn_> adding a new script to /usr/share/initramfs-tools to the casper-bottom folder doesn't help
[08:23] <mattn_> seems the new script is not executed at all
[08:24] <mattn_> is there any documentation available on how to customize casper?
[08:24] <mattn_> even update-initramfs -u isn't helping
[08:40] <waters> Problem with install .. I just downloaded the iso .. however .. when i try to install .. i have to install in safe graphic mode ... or i only get the blue screen ... Any ideas on this ... Main problem is the only CD i had after i get to 60% tells me it's an input/output error ... so .. i can use the thing from fedora to make a live os on my flash drive ... boot that way .. but .. I cant get it to run in safe mode ... dont know how
[09:10] <waters> Help .. i need some help with installing ...
[09:11] <waters> all i get is the blue screen of death ... unless i run in safe graphics mode ...
[09:13] <davmor2> waters: try asking on #ubuntu  this channel is really meant for development use
[09:14] <davmor2> #ubuntu+1 for karmic
[09:22] <evand> waters: are you installing using Wubi?
[09:22] <waters> nope
[09:22] <waters> i live cd
[09:22] <waters> a live 9.04 cd
[09:23] <waters> i can get it to kinda work if i use the safe graphics mode option ... but doesnt go all the way
[09:23] <evand> I'm confused then, what is this blue screen you're talking about?  Ubuntu doesn't have that kind of kernel panic mechanism yet
[09:24] <waters> i think it's an Xwindows problem ..
[09:24] <waters> like if i do ctrl+alt+f1 and get into a console
[09:24] <waters> i can do dmesg
[09:25] <mattn_> davmor2: isn't this the correct channel for my question, too? also tried #ubuntu-devel
[09:25] <waters> and i see an error drm:i915-GET_VBLANK_COUNTER} *ERROR* tying to get vblank counfor disabled pipe 1
[09:26] <evand> lets start with the safe graphics mode problem
[09:27] <evand> can you get to a console and type ubuntu-bug xorg
[09:27] <waters> yep ...
[09:27] <evand> please describe the graphics problem in as much detail as you can (you're running the live CD, you had to run safe graphics mode, etc)
[09:28] <davmor2> mattn_: that would probably be best asked here.
[09:28] <mattn_> davmor2: thanks - then i'll idle around here and wait for help ;)
[09:29] <waters> safe graphics is the only way i can see anything when i start the install ... it get's to 60% .. then tells me that the CD or CD-ROM has a problem .. INPUT/OUTPUT ... ok .. so i check the desk .. from the start up ..and it doesn't find a problem ..
[09:29] <waters> so .. i try to install normal ... no screen ...
[09:29] <waters> screen is blue ..
[09:30] <evand> waters: right, so the fact that you're getting into safe graphics at all is a bug that I'm kindly asking you to file using ubuntu-bug xorg
[09:30] <waters> yeah
[09:30] <evand> waters: now, regarding the input/output issue, if you can get to the try or install option in safe graphics mode
[09:31] <waters> i did .. and it get's to 60%
[09:31] <evand> select install, run it until it crashes, and then put your /var/log/syslog on http://pastebin.ubuntu.com, I'd be happy to look at it further
[09:31] <waters> ok... thanks
[09:31] <evand> sure thing
[09:33] <waters> installing ...
[09:37] <evand> mattn_: the easiest way is going to be unpacking an existing CD, modifying from there, and then rebuilding the initramfs: https://help.ubuntu.com/community/LiveCDCustomization#Boot init
[09:37] <mattn_> evand: that's what i'm doing
[09:38] <mattn_> just added a new script to the casper-bottom dir and ran update-initramfs -u
[09:38] <mattn_> inside the chroot of course
[09:41] <evand> mattn_: when you boot the CD, hit F6, then escape, then add break=top to the kernel command line parameters.  Is your script in /scripts/casper-bottom, is it +x?
[09:44] <mattn_> yes - it is executable
[09:44] <mattn_> will try that break=top once i've burned and created the latest version
[11:39] <mattn_> evand: no, my customized script is not there - thanks for pointing this switch for me. so i'm doing something wrong while creating the initrd
[11:39] <mattn_> that's something i can work with - thanks
[11:40] <evand> anytime
[12:13] <CIA-4> user-setup: cjwatson * r183 ubuntu/ (debian/changelog user-setup-ask):
[12:13] <CIA-4> user-setup: Allow underscores in usernames, matching change in adduser 3.110ubuntu3
[12:13] <CIA-4> user-setup: (LP: #388822).
[12:15] <CIA-4> user-setup: cjwatson * r184 ubuntu/debian/changelog: releasing version 1.26ubuntu2
[12:26] <CIA-4> iso-scan: cjwatson * r265 ubuntu/debian/ (69 files in 2 dirs): merge lp:~al-maisan/iso-scan/ubuntu
[13:54] <rgreening> good day evand
[13:54] <evand> hello
[13:54] <rgreening> did you have a chance to look at my changes evand
[13:55] <rgreening> I'd like to add the wrappers to gtk_frontend.py and merge kbackend.py changes to kbackend.py
[13:55] <rgreening> I tested the kde side, and it works!!!!!! :)
[13:56] <evand> no, I haven't yet
[13:56] <rgreening> yay
[13:56] <evand> but fantastic
[13:56] <rgreening> I still have the io/watch/child functions to wrap.. but the timeout_add is completed.
[13:57] <rgreening> evand: http://roderick-greening.blogspot.com/ - have a quick look at the second post. describes what I did.
[14:02] <evand> good deal
[14:03] <rgreening> :)
[14:05] <rgreening> so, I think it's merge worthy into main backup.py and moving the required stuff to gtk_frontend.py using a quick wrapper to the same call. It's simpler on the gtk side, as we are only using the wrapper to move the function to frontend.
[14:12] <evand> okay, I'll try to take a closer look once I've cleared some other work off my plate.
[14:13] <rgreening> k\
[14:13] <rgreening> I'll see about the remaining gobject stuff.. see how to port them...
[14:14] <rgreening> can't wait to finish this part... then it's just cleanup tasks (translations, packaging, etc).
[15:34] <cr3> evand: reping, might you have a moment to help me understand some preseeding of ubiquity?
[15:35] <cr3> cjwatson: you might know this on top of your head: when installing from the desktop image, the content of .disk/info gets copied to one of the files under /var/log/installer. does the same happen when netinstalling from the alternate image? I'm guessing not because the netinstall could just as well be done directly from the archive
[15:38] <cjwatson> cr3: bug 364649
[15:38] <ubottu> Launchpad bug 364649 in ubiquity "Please include installation media build number in installation logs" [Wishlist,Triaged] https://launchpad.net/bugs/364649
[15:39] <cjwatson> cr3: i.e. I don't think it does yet, and indeed it isn't guaranteed to exist in general
[15:39] <evand> cr3: sure
[15:42] <cr3> evand: so, to follow up on the thread for preseeding the clock-setup, I'm not sure I understand how to determine the package for which to define the preseed. for example, is this right: ubiquity clock-setup/ntp-server string foo.com
[15:43] <cr3> evand: I tried grepping around ubiquity for ntp-server but I couldn't find anything, so I'd like to have a better understanding of what happens
[15:45] <evand> cr3: it's in clock-setup, which ubiquity uses.  apt-get source ubiquity; cd ubiquity/d-i/source/clock-setup
[15:45] <evand> ah, I misread
[15:45] <evand> d-i or ubiquity is fine, they'll both work
[15:48] <cr3> evand: hm, how does that work? re. d-i or ubiquity? does that mean anything supported by d-i could be passed along by ubiquity?
[15:49] <cjwatson> err, the owner is not really very connected to that question
[15:49] <cjwatson> the first field of a preseed line specifies the owner of the question from debconf's point of view
[15:49] <cjwatson> this is used for garbage-collection when packages are removed
[15:50] <cjwatson> "d-i" is a special owner that says "don't copy this question to the installed system", for d-i
[15:50] <cjwatson> since we don't copy most questions to the installed system at all for ubiquity yet, it doesn't make a whole lot of difference, but when we implement that I guess we'll allow both "d-i" and "ubiquity" to have the same meaning
[15:51] <cjwatson> in general when preseeding things that are part of the installer you should use "d-i" (or "ubiquity" is OK too in the case of the desktop CD). When you expect preseeding to be processed by a package that's installed on the target system, use the appropriate package name
[15:53] <cr3> cjwatson: when you say the "the owner is not really very connected to that question", do you mean clock-setup/ntp-server for example which is not really connected to any owner and could potentially be given foobar as an owner, with the only difference that it won't be garbage collected and may affect the package foobar inadvertently?
[15:57] <cr3> cjwatson: potentially, I could reuse the preseed I use for the alternate image and just augment it with a few ubiquity-specific seeds and I should be good to go
[16:00] <evand> cr3: I believe that's exactly what he's saying, but you should still use d-i so it will be garbage collected :)
[16:01] <cr3> evand: right, I was just checksumming that my understanding was correct with some convoluted example :)
[16:01] <evand> heh, sure
[16:03] <shtylman> evand: I have a few ideas I have been throughing around for the kubuntu ubiquity installer (http://shtylman.com/stuff/kubuntu_installer/ubiquity_idea1.png) .. the biggest question I have is about the sidebar on the left...those are one work descriptions of the task and would probly require new translation entries...would that be acceptable if we went with something like this for kubuntu?
[16:04] <rgreening> blingy shtylman :P
[16:04] <shtylman> rgreening: :) one does try
[16:04] <cr3> evand: another question for you: what's the difference between the ubiquity and user-setup owners? for example, I see both being used for passwd/username
[16:05] <evand> creating strings for translation isn't a big deal for something like this, but my concern is screen space.  Those strings could get quite large when translated and we're already quite tight on space
[16:05] <evand> unfortunately I don't have a better suggestion :-/
[16:06] <evand> cr3: probably the difference of grabbing them via debconf-get-selections
[16:07] <cr3> evand: in case I want to add more preseed variables, how can I determine whether a particular variable might need to be grabbed via debconf-get-selections?
[16:07] <cjwatson> cr3: if you give it "foobar" as an owner, it will get copied to the installed system and never ever be removed unless you install and then purge a package called "foobar"
[16:07] <cjwatson> cr3: but otherwise yes
[16:07] <shtylman> evand: thanks...I will think about that...didn't really thing about horizontal space :/ (damn my large monitors)
[16:07] <evand> shtylman: my fault for not pointing this out earlier, but today is FeatureDefinitionFreeze.  I'm not entirely sure if that means community specs have to be done as well, though I would assume so.
[16:07] <cjwatson> cr3: by "the owner is not really very connected to that question" I meant that what owner you happen to set has nothing to do with whether anything supported by d-i can be passed along by ubiquity
[16:08] <evand> https://blueprints.edge.launchpad.net/ubuntu/+spec/kubuntu-karmic-ubiquity is looking a bit bare at the moment :)
[16:08] <cjwatson> cr3: you should never be using "user-setup" as an owner; that's a bug
[16:08] <cjwatson> cr3: as evand says, it might come from debconf-get-selections. Don't use debconf-get-selections.
[16:08] <cr3> evand: one of my concerns is that if I attempted to reuse the preseed I use for the alternate image for ubiquity, then passwd/username wouldn't've worked because I would've had to add an extra line for user-setup too
[16:08] <evand> apologies for not being more clear on that
[16:08] <cjwatson> cr3: use the installation guide instead; we add things that are actually useful there
[16:09] <cjwatson> right now, debconf-get-selections just gives you *everything*, including loads of stuff that's internal and shouldn't be preseedd
[16:09] <shtylman> evand: oh my...didn't know that...been all tied up with OO work...
[16:09] <cjwatson> +e
[16:09] <shtylman> ok...I will get on the spec fill out...with this mockup and see what I can produce
[16:09] <cr3> cjwatson: I've been using the preseed.txt.gz from the installation-guide as the base for my preseed template, which has been very useful. should I be expected to be able to use the exact same file for ubiquity as well?
[16:11] <cr3> by the way, I grab that preseed.txt.gz for each release and just add a few macros here and there to keep the diff between my file and the default one in the guide as short and readable as possible
[16:11] <cjwatson> cr3: largely
[16:12] <cjwatson> cr3: https://wiki.ubuntu.com/UbiquityAutomation documents the major differences (subject to errors and omissions)
[16:12] <cr3> cjwatson: excellent! I'll give that a try and follow up if there are any such errors and/or omissions. thanks for the reference!
[16:13] <cr3> using the same file for both alternate and desktop will make things much easier to maintain
[16:21] <cjwatson> I see you were the last person to edit that wiki page ;-)
[16:25] <cr3> cjwatson: I have to do so many things so far and wide, I don't even remember what I might've edited and it was only a month ago :)
[17:04] <evand> kirkland: Is there anything else that needs to be covered, or you would like to add to https://wiki.ubuntu.com/ServerKarmicEncryptedSwap ?
[17:05] <xivulon> evand, we can ship r136
[17:05] <xivulon> I have already starting moving forward to 9.10
[17:05] <xivulon> should be able to play with grub2 soon
[17:06] <evand> xivulon: indeed, I ran it by slangasek this morning and I just have to find time today to push it to releases.ubuntu.com
[17:06] <xivulon> cool
[17:06] <evand> xivulon: as today is Feature Definition Freeze, can you ensure that the Wubi specification for Karmic is drafted and ready to be reviewed?
[17:07] <xivulon> sure
[17:08] <evand> great, thank you
[17:08] <xivulon> will have to do it tonight do not have the gobby log with me (is it available online?)
[17:09] <xivulon> on colinux integration, did you have a chat with the kernel team?
[17:09] <evand> did we take notes in gobby?
[17:09] <xivulon> yes
[17:09] <evand> xivulon: not yet, please make that an action item of the spec
[17:09] <xivulon> I did
[17:09] <evand> one second, I'll pull them down
[17:11] <evand> I've emailed them to your gmail account
[17:11] <xivulon> I think that daemonccc is also working on that (see #376807)
[17:16] <evand> I've also dumped the notes here: https://wiki.ubuntu.com/FoundationsTeam/Specs/KarmicWubi and linked everything up.
[17:16] <xivulon> ot: was looking at genie programming language today (vala with python syntax), I love that stuff!
[17:17] <evand> I still think that the kernel patches are a *huge* blocker to official support and distribution from Ubuntu, but I'll follow up with the kernel team per the specification
[17:17] <evand> neat
[17:17] <xivulon> thanks
[17:19] <xivulon> as to genie it would be really cool if it cool spit out c code without gobject when objects are not used, and maybe integrated with tinypy with gobject introspection for embedding...
[17:19] <xivulon> </ot>
[17:22] <kirkland> evand: i've been seeing your updates/changes
[17:22] <kirkland> evand: let me read in full
[17:23] <kirkland> evand: looks great; simple, to the point; major points covered
[17:24] <evand> fantastic
[17:24] <xivulon> evand, shall I put everything in that spec? it's really 2 or 3 things: colinux, wubi-migration, wubi-enhancements, macbook (unlikely)
[17:26] <evand> I'd put them all in, but do note that individual bits can still be dropped and the spec can be considered complete for karmic
[17:27] <evand> case in point, colinux
[17:27] <xivulon> and macbook...
[17:27] <evand> right
[17:27] <xivulon> we already have a page for the migration tool by the way
[17:28] <evand> indeed, I'd just reference that
[17:28] <evand> as I believe we marked it as approved for 9.04
[17:28] <evand> or 8.10 :)
[17:29] <xivulon> 9.04, I guess we changed our views a bit
[17:29] <evand> oh, good call
[17:29] <xivulon> just to confirm, we will have migration inside of standard ubiquity, so we will not need a stage 1
[17:30] <xivulon> previously we agreed to have a stage 1 from within ubuntu, in karmic we decided to actually autodetect wubi installations
[17:32] <evand> right, I think this is best done within the live environment.  I see no reason for any work outside of that.
[17:32] <xivulon> yes that is in line with what discussed at last UDS, I will change the previous wiki
[17:37] <waters> Hi .. when i start ubuntu (a new install) all i get is a blank screen .. i can ctrl+alt+f1 and get to a console ... when i go dmsg i get this error: [drm:i915_get_vblank_counter} *ERROR* trying to get vblank count for disabled pipe 1.. Anyone got time to help?
[17:41] <xivulon> mpt, evand I had mentioned this change to the partitioning page UI in california, created a new page: https://wiki.ubuntu.com/UbiquityPartitioningPageProposal
[17:41] <xivulon> still think it's a good approach
[17:43] <evand> xivulon: I started a conversation with Ivanka (head of the Design and User Experience team) at UDS on ubiquity usability.  They seem to be a bit too busy for us at the moment, but she said the installer was on her radar, and I'll be sure to bring that design up when our turn comes for their resources.
[17:44] <cjwatson> waters: I think you'll need to ask either the kernel or X folks; it's a bit beyond me at least
[17:45] <waters> ok .. thanks
[17:46] <cody-somerville> cjwatson, Someone from Debian asked me to advocate LP #387944
[17:46] <ubottu> Launchpad bug 387944 in busybox "busybox-initramfs does not include wc" [Undecided,New] https://launchpad.net/bugs/387944
[17:51] <cjwatson> cody-somerville: yes, I'm aware of it and am merging busybox at the moment
[17:51] <cjwatson> I'm already planning to include that change in my upload
[17:51] <cody-somerville> cjwatson, Thanks.
[17:51] <cody-somerville> :)
[17:51] <mpt> xivulon, I like the first part
[17:53] <evand> mpt: any idea on when and if you guys will be looking into the installer?
[17:53] <mpt> evand, no, sorry
[17:53] <mpt> I'd dearly love to
[17:53] <evand> no worries at all
[17:53] <evand> we've already got plenty to do this cycle :)
[18:02] <xivulon> do you think that it would be possible to upgrade the packages concurrently with a Wubi migration? Say migrating 9.04 wubi installation via a 9.10 CD. Certainly it would be impossible to upgrade some of the packages.
[18:12] <CIA-4> usb-creator: rgreening * r113 usb-creator/ (TODO usbcreator/kde_frontend.py):
[18:12] <CIA-4> usb-creator: Begin working on translation support for kde_frontend.py (needs lots of love)
[18:12] <CIA-4> usb-creator: Add warning dialog for when user cancels progress dialog and update connects
[18:12] <CIA-4> usb-creator: Remove unnecessary update_persist_label() (update directly in connect instead)
[18:12] <CIA-4> usb-creator: Ensure Make Startup Disk button is disabled from start
[18:12] <CIA-4> usb-creator: Update TODO
[18:13] <evand> xivulon: one step at a time :)
[18:13] <evand> I'm happy to just have the migration working for 9.10
[18:14] <kirkland> cjwatson: got a minute?  i need to chat with you about /var/lib/ecryptfs/$USER
[18:15] <kirkland> cjwatson: 3 users in 3 days have either re-installed jaunty, or upgraded from jaunty -> karmic
[18:15] <kirkland> cjwatson: they backed up or preserved $HOME, but did not do the same for /var/lib/ecryptfs
[18:16] <kirkland> cjwatson: i'm going to need to fix this in a Jaunty SRU for the jaunty->karmic upgrade scenario
[18:16] <kirkland> cjwatson: i wanted to bounce the idea off of you before i implement it
[18:17] <kirkland> cjwatson: i'd create an /etc/init.d/ecryptfs, that does a mkdir -p /home/.ecryptfs && mount -o bind /home /home/.ecryptfs
[18:17] <kirkland> cjwatson: /home/.ecryptfs will be a bind mountpoint only;  but will be established at boot, before any user logs in and mounts their homedir
[18:18] <kirkland> cjwatson: the config data that lives in each user's /var/lib/ecryptfs/$USER dir would be moved to /home/.ecryptfs/$USER/.ecryptfs
[18:18] <kirkland> cjwatson: which would move it out of /var, and into their unmounted home dir
[18:19] <kirkland> cjwatson: users are good with backing up $HOME, so they'd keep a copy of their config data in that case
[18:33] <cjwatson> kirkland: not today, sorry; can you send me a mail instead?
[18:34] <kirkland> cjwatson: sure
[18:34] <kirkland> cjwatson: i may well be able to attach a patch by that point
[18:34] <kirkland> cjwatson: after my spec definitions are done
[18:34] <kirkland> cjwatson: just one quick question ....
[18:34] <kirkland> cjwatson: are you diabolically opposed to a /home/.ecryptfs bind mount?
[18:34] <kirkland> cjwatson: would you prefer /.home instead?
[18:34] <cjwatson> sorry, I *really* can't do this right now
[18:35] <cjwatson> I'm trying to fill out financial declaration forms for school, and am only at the computer looking for information
[18:36] <kirkland> cjwatson: good luck ;-)  ttyl
[18:41] <xivulon> evand, https://wiki.ubuntu.com/FoundationsTeam/Specs/KarmicWubi
[18:42] <evand> xivulon:  awesome!  I'll take a closer look when I get back from walking the dog.  Thanks a bunch for doing that so quickly!
[18:49] <xivulon> evand, np, feel free to edit
[19:16] <xivulon> mpt, evand fyi, I made a couple of enhancements to https://wiki.ubuntu.com/UbiquityPartitioningPageProposal
=== nxvl_ is now known as nxvl
[22:06] <cr3> might it be possible that late_command runs very early when attempting to preseed ubiquity? I've been seeing in-target errors popping by immediately after the preseed is being retrieved
[22:25] <cjwatson> I'm not aware that preseed/late_command is processed by casper/ubiquity at all