[12:20] <cjwatson> xivulon: you have mail about a crash while ejecting the CD in wubi
[12:22] <xivulon> cjwatson, thanks could not test the physical CD yet because of size issues
[12:23] <cjwatson> the code looks pretty hosed when compared to documentation I googled up on ejecting CDs
[12:23] <cjwatson> might be better to just comment it out for now
[12:23] <xivulon> sure
[12:23] <xivulon> I submitted quite a few changes rev 281
[12:23] <xivulon> and some more coming
[12:24] <xivulon> Unfortunately 7z and makeimage make nsis crash. I am downloading windows stack tracing software
[12:26] <cjwatson> rather you than me
[12:26] <xivulon> cjwatson, if you found any info on ejecting the CD please send it over
[12:28] <cjwatson> xivulon: http://www.thescripts.com/forum/thread124082.html though I don't know how to adapt it to nsis
[12:29] <cjwatson> http://vbnet.mvps.org/index.html?code/faq/eject.htm
[12:33] <xivulon> thanks will have a look
[12:34] <xivulon> cjwatson that's the guide I used
[12:35] <cjwatson_> 23:28 <cjwatson> xivulon: http://www.thescripts.com/forum/thread124082.html though I don't know how to adapt it to nsis
[12:35] <cjwatson_> 23:29 <cjwatson> http://vbnet.mvps.org/index.html?code/faq/eject.htm
[12:35] <cjwatson_> 23:29 <cjwatson> I googled for "winmm mcisendstring eject CD" and found several relevant hits
[12:35] <cjwatson_> not sure how much you got, my router threw a wobbly
[12:36] <xivulon> cjwatson I tried to implement first guide in nsis, but as mentioned did not test at all. In theory the function calls should be equivalent
[12:36] <xivulon> obviously theory and practice are not the same thing
[12:37] <xivulon> maybe I missed the :\\?
[12:38] <xivulon> System::Call "winmm.dll::mciSendString(t 'open $cddrive\ type cdaudio alias cdrom', n, i 0, *i 0)i.r2"
[12:47] <cjwatson> but AIUI that just opens a handle
[12:47] <cjwatson> it's the "set cdrom door open" that actually ejects
[12:47] <cjwatson> and "close cdrom" to close the handle again
[12:56] <CIA-18> ubiquity: cjwatson * r2244 ubiquity/ (debian/changelog scripts/install.py):
[12:56] <CIA-18> ubiquity: * Purge ucf's state for /etc/papersize so that it will be recreated
[12:56] <CIA-18> ubiquity:  properly (LP: #128258).
[01:00] <xivulon> cjwatson you are right
[03:37] <xivulon> evand, cjwatson I might have fixed most if not all of the wubi crashes
[03:37] <xivulon> will upload soon
[03:38] <cjwatson> neat
[03:38] <cjwatson> pkl has unionfs stuff in git now that's looking good
[03:38] <cjwatson> I'm testing it
[03:38] <xivulon> it's too late for me to test that, 230am here
[03:38] <xivulon> tomorrow
[03:39] <xivulon> /xivulon moves to the bright side
[03:51] <evand> awesome all around
[04:06] <xivulon> cjwatson, evand, new wubi version is up, also updated the binary
[04:07] <xivulon> Did not test lupin stuff at all nor physical CD
[04:08] <xivulon> Eject, at least with emulated CD, does not seem too crash
[04:09] <xivulon> rev282
[04:09] <cjwatson> ok, you're just in time, I'll test that along with unionfs
[04:10] <xivulon> To test uninstallation, please start with a clean sheet, uninstallation works for me
[04:10] <cjwatson> right, assuming I can get to a clean sheet :-)
[04:10] <cjwatson> I uninstalled but I still seem to have a grub4dos hanging around
[04:10] <cjwatson> is there a brute-force way to remove that and get back to the Windows bootloader?
[04:11] <xivulon> remove folder, registry key, use easybcd for vista boot menu, and any *wubildr* files you can find
[04:11] <cjwatson> ok
[04:12] <xivulon> Have to sleep, long day tomorrow. Pls send me an email to let me know how it goes
[04:12] <evand> goodnight xivulon
[04:13] <xivulon> night everyone
[04:33] <cjwatson> urgh, now wubi is set up to download the ISO
[05:17] <evand> how's it going?
[05:17] <evand> otherwise, that is
[05:58] <cjwatson> waiting for wubi ...
[05:58] <cjwatson> otherwise unionfs seems happy
[06:00] <evand> wonderful
[06:00] <cjwatson> it's at 84%
[06:00] <cjwatson> on the upside I am catching up on an awful lot of mailing lists
[06:02] <cjwatson> wow
[06:02] <evand> hahaha
[06:02] <cjwatson> I just got to the end of my 'mailboxes' list in mutt
[06:02] <cjwatson> it's been about two years since that last happened
[06:02] <evand> wow
[06:02] <evand> mine is just over a page
[06:57] <cjwatson> ah, now this is more like it
[06:57] <cjwatson> took a bit of persuasion to get wubi to boot
[07:01] <cjwatson> still no automatic ubiquity launching though
[07:01] <cjwatson> oh, whoops, that's a bug I saw a while back and forgot to fix
[07:04] <evand> whoops, I forgot to mention that from my limited testing.
[10:19] <CIA-18> migration-assistant: evand * r59 migration-assistant/ (11 files in 3 dirs):
[10:19] <CIA-18> migration-assistant: * Use stat instead of the DT_ macros to avoid issues with fuse.
[10:19] <CIA-18> migration-assistant: * Add a debug log.
[10:19] <evand> hooray for not committing things
[10:22] <evand> cjwatson: Is the ubuntu-installer branch of ubiquity broken?
[10:22] <evand> bzr: ERROR: Can't rename /srv/sm-ng/push-branches/00/00/10/68/.bzr/repository/lock/deenek5py8.tmp to /srv/sm-ng/push-branches/00/00/10/68/.bzr/repository/lock/held: /srv/sm-ng/push-branches/00/00/10/68/.bzr/repository/lock/held already exists
[10:22] <evand> debcommit: commit failed
[10:33] <CIA-18> migration-assistant: evand * r60 migration-assistant/debian/changelog: releasing version 0.5.0
[12:14] <CIA-18> ubiquity: cjwatson * r2245 ubiquity/debian/ (changelog init): * Fix automatic-ubiquity option handling in init script.
[12:14] <cjwatson> evand: try it now but make sure you're using sftp: rather than bzr+ssh: for the time being
[12:14] <cjwatson> I fixed it by hand following advice from Robert Collins
[12:48] <CIA-18> ubiquity: cjwatson * r2246 ubiquity/ubiquity/frontend/noninteractive.py: typo
[01:06] <xivulon> cjwatson, can you detail your problems about using ISOs?
[01:07] <xivulon> probably the current restriction is that I also check the iso name
[01:07] <xivulon> for ubuntu it must be gutsy-desktop-i386.iso (as specified in isolist.ini)
[01:07] <xivulon> ISO works fine for me
[01:07] <cjwatson> xivulon: *install* from ISO can't work as I explained at length the other day
[01:08] <cjwatson> and in particular it doesn't work because partman-auto-loop fails to mount /host
[01:08] <cjwatson> (which is a consequence of a lupin bug but ...)
[01:08] <xivulon> cjwatson, by install from ISO, I mean loopinstallation
[01:09] <xivulon> as lupin bugs I still did not have a chance to boot that, so debugging was not very easy
[01:09] <xivulon> in fact lately I just concentrated on the user interface waiting for segfaults/unionfs to be resolved
[01:12] <cjwatson> xivulon: I'm sorry, I'm confused
[01:12] <cjwatson> I was up until 7am and then got four hours of sleep, bear with me
[01:12] <cjwatson> xivulon: ok, the problem is that lupin mounts the Windows filesystem as /isodevice
[01:12] <xivulon> ah I thought you were in US
[01:12] <cjwatson> due to find_iso
[01:13] <cjwatson> no, just on a US timezone :)
[01:13] <xivulon> same here, went to sleep at 3 up at 7...
[01:13] <cjwatson> and then partman-auto-loop tries to mount the same filesystem as /host
[01:13] <cjwatson> which fails because it's already mounted
[01:13] <cjwatson> so OK, I guess it's not an intrinsic problem, we should just fix it in lupin/partman-auto-loop
[01:13] <xivulon> cjwatson, I am at work now, I can have a look only tonight
[01:14] <xivulon> As for avoiding installation when the ISO is run in "read-only" mode
[01:14] <xivulon> I guess that should be a check within Ubiquity rather than wubi
[01:15] <xivulon> Moreover it would be needless to disable ubiquity completely, since if the user does have more than 1 partition, he may still be able to install from a read-only ISO installation
[01:16] <cjwatson> yeah, I think I'm just wrong due to sleep deprivation
[01:16] <xivulon> Ideally when giving the option to choose a drive/partition, the ones that are already mounted should not make the list
[01:17] <cjwatson> oh, I meant to copy up my hacked ISO
[01:17] <xivulon> and if no drive/partition is available ubiquity should simply display a message and exit
[01:20] <xivulon> re branding: tonight I can either work on the dynamic rebranding or testing lupin, I'd go with the second
[01:21] <xivulon> to do static graphical rebranding is trivial, just change the icon and header.bmp within wubi/image with another image of the same size/format
[01:21] <xivulon> I still have not received anything from the artwork team
[01:22] <cjwatson> http://people.ubuntu.com/~cjwatson/tmp/gutsy-desktop-i386-wubi.iso
[01:23] <cjwatson> I love rsync
[01:23] <cjwatson> that has the always-tries-to-download ISO bug you mentioned, along with the lack of a leading backslash in the boot menu entry that I mentioned by mail and fixed
[01:23] <cjwatson> I got it to boot using editbcd
[01:24] <cjwatson> it has Phillip's unionfs fixes, which are working well for me
[01:24] <xivulon> thanks for that
[01:24] <xivulon> I'll test it tonight
[01:25] <cjwatson> and it has a nasty little partman-auto-loop bug which prevents installation
[01:25] <cjwatson> apply http://codebrowse.launchpad.net/~ubuntu-core-dev/partman-auto-loop/ubuntu/revision/cjwatson%40canonical.com-20070920103812-mgae9pjcor3xh5hy?start_revid=cjwatson%40canonical.com-20070920104108-tfnyb57itaid5v9y to /bin/autopartition-loop to fix that
[01:25] <xivulon> Was about to ask just that
[01:25] <cjwatson> that's as far as I got
[01:25] <cjwatson> xivulon: do you have an executable somewhere with your CD2ISO fixes?
[01:26] <xivulon> yeah in my usual place: http://wubi-installer.org/devel/minefields/Ubuntu-7.10-alpha.exe
[01:27] <xivulon> that does not include the backslash fix though
[01:27] <xivulon> also now it points to the latest daily build if it has to download anything AND to the metalinks on wubi-installer.org, we may want to change isolist.ini and the metalink files
[01:27] <cjwatson> ok, but that's workaroundable if you know how
[01:28] <xivulon> lates daily buld = latest ISO daily build
[01:28] <cjwatson> getting lupin and friends working is massively more important than branding
[01:28] <xivulon> I agree
[01:28] <cjwatson> I suspect that exe is too late for our daily build
[01:29] <cjwatson> looks like it from the timestamps
[01:29] <xivulon> But as mentioned I can only play with that tonight
[01:29] <cjwatson> I'm going to let a few things settle and do another CD build later on
[01:29] <xivulon> the exe is not an issue, people can simply download it separately for testing puroposes
[01:30] <cjwatson> does it behave in the same way when downloaded separately and when run from the CD?
[01:30] <xivulon> it should
[01:30] <cjwatson> ok, that's good
[01:30] <xivulon> the only difference is that if a CD is detected the distro options list is limited to the CD
[01:31] <cjwatson> right, but that's a property of the CD being inserted rather than a property of where wubi is run from?
[01:31] <xivulon> and rebranding should kick in (disabled at the moment)
[01:31] <xivulon> yes
[01:32] <cjwatson> ok, I'll try this out this afternoon
[01:32] <xivulon> when you run wubi, first thing it checks for a CD, and changes the distrolist and branding accordingly
[01:32] <cjwatson> this hsa been a fun ride so far :-)
[01:32] <xivulon> basically if you have a CD you should see only "Ubuntu" in the distro selector and see "Use CD" in the size selector
[01:33] <xivulon> if you have no CD you should see "Ubuntu|Kubuntu" in the distro selector and "Read Only" as the only "extra" option in the size selector
[01:34] <xivulon> If you select Use CD, the CD content is NOT extracted to disk. If you use Read Only the ISO content is extracted to disk.
[01:34] <xivulon> In both cases the kernel/initrd are extracted to disk
[01:35] <xivulon> glad you are enjoying it so far, I moved from hacking linux/d-i (which I like) to maintain nsis (which I hate)
[01:42] <cjwatson> well, observe my trick of turning as much of it as possible into d-i hacking ;-)
[01:42] <cjwatson> hmm, partman-auto-loop removes the .disk file and recreates it itself using dd
[01:42] <cjwatson> should it just accept one that already exists?
[01:42] <xivulon> ubi maior minor cessat
[01:43] <cjwatson> and yet your classicist credentials are so much better than mine. :)
[01:44] <xivulon> cjwatson, I though that the idea was that we test for the existance of *.disk and if any is found we abort the installation
[01:44] <xivulon> This was the behaviour in old version of wubi
[01:44] <cjwatson> but doesn't wubi preallocate the .disk files?
[01:44] <xivulon> Not anymore
[01:44] <cjwatson> so by the time partman-auto-loop runs they'll exist
[01:45] <xivulon> nope, there should only be a disks folder with an empty boot folder in it
[01:46] <xivulon> I can generate the files from within wubi if you want, the only advantage is that I can use a windows api to preallocate the space without zeroing the full thing
[01:46] <cjwatson> ah
[01:46] <cjwatson> ok, no problem then
[01:46] <cjwatson> "This shows that the ancient Romans were quite good at text compression. Probably their Empire fell when they accidentally deleted all the copies of pkzip from their abaci."
[01:47] <cjwatson> no, I'm happy with it done in Linux
[01:47] <xivulon> good one :)
[01:47] <cjwatson> makes the sanity check easier
[01:48] <xivulon> I agree, that's why I disabled it
[01:48] <cjwatson> although yes it's slower, should do something about that
[01:48] <xivulon> so if you see any .disk file you should refuse to overwrite them. The user has to uninstall explicitly if they want to overwrite stuff.
[01:49] <cjwatson> yep, I do that
[01:49] <xivulon> great
[01:49] <cjwatson> you get a "Loop-mounted file systems already present" error
[01:49] <cjwatson> or at least you should :)
[01:51] <xivulon> Can we add a non geek line as well? something like: "If you want to reinstall Ubuntu, please first uninstall the existing installation and then run the installer again"
[01:51] <cjwatson> post-beta
[01:51] <cjwatson> yeah, I was just trying to avoid branding; could you file a bug on partman-auto-loop and I'll come up with something?
[01:52] <xivulon> you mean for the non-geek message?
[01:52] <cjwatson> yes
[01:53] <cjwatson> I try to avoid saying "Ubuntu" in d-i strings because it requires branding for derivatives
[01:53] <xivulon> partman-auto-loop does not use Launchpad as its bug tracker.
[01:53] <cjwatson> https://bugs.launchpad.net/ubuntu/+source/partman-auto-loop/+filebug
[01:53] <cjwatson> i.e. the source package in Ubuntu, not the product
[01:53] <xivulon> of course
[01:55] <xivulon> https://bugs.launchpad.net/ubuntu/+source/partman-auto-loop/+bug/141217
[01:59] <cjwatson> thanks
[02:12] <xivulon> By the way, I should be in Boston wed/thu/fri I hope I'll be able to attend some of the installer meetings.
[02:22] <cjwatson> excellent
[02:22] <cjwatson> you're based in London aren't you?
[02:24] <xivulon> yeah
[02:24] <xivulon> want a beer?
[02:26] <cjwatson> I was thinking more of a hack session in the London office :)
[02:26] <cjwatson> it's much more worth it now than it would have been early on when so few of the pieces were present
[02:26] <cjwatson> (but sure, perhaps a beer afterwards ...)
[02:29] <xivulon> I'd love to, but the only issue with that is that I work during the week and I'll be away the next 2 weekends
[02:30] <xivulon> I can try to pop in "early" on next monday/tuesday. Early being 5pm
[02:31] <cjwatson> hmm, not ideal
[02:31] <cjwatson> I'm not sure I can get down then
[02:31] <cjwatson> I'm due in London on 9 Oct
[02:44] <xivulon> I can get a few hours off on the 9th then, hopefully there should not be much to do by then
[03:30] <xivulon> shall I add a --safe-mode switch that sets up menu.lst for the installer with noacpi and safe screen resolution?
[03:32] <cjwatson> I think it should be consistent with what we make available on our normal CDs
[03:32] <cjwatson> which means "safe graphics mode" adding the xforcevesa command-line option
[03:33] <xivulon> sure
[03:34] <xivulon> to make it simple for users, I was thinking that we may add one or two shortcuts, like wubi-safe-mode.exe, or have a special link on the CD webpage
[03:39] <cjwatson> how about getting Henrik to make it an option in winfoss
[03:39] <cjwatson> or make it a checkbox on the wubi UI
[03:39] <cjwatson> the latter might be better actually
[03:39] <cjwatson> and I repeat, please "safe graphics mode", it's not analogous to Windows safe mode and I don't want it confused
[03:42] <xivulon> what about noapic stuff? I know you mentioned it should be fixed properly, but for instance for my laptop I need acpi=off
[03:44] <cjwatson> right but that doesn't make it a safe mode
[03:44] <cjwatson> it makes it a workaround for your laptop
[03:44] <cjwatson> it's not correct to call it safe across the board
[03:45] <cjwatson> if it were, after all, we'd probably just make it the default
[03:45] <xivulon> What about "unsupported hardware fix" or something?
[03:45] <cjwatson> no
[03:46] <cjwatson> I don't want this sort of thing to get out of sync between wubi and our regular installer
[03:46] <cjwatson> this is hugely important - we can't support it if it's out of sync
[03:47] <cjwatson> and I really do think this should be handled by filing a bug on the kernel and getting them to add a quirk
[03:47] <cjwatson> it's usually straightforward if you attach dmidecode output
[03:48] <xivulon> As command line I can support as many options as you want (and will support the same parameters), but on the user interface I am a bit space constrained
[03:56] <cjwatson> xivulon: how about a button that pops up a dialog into which you can type in extra boot options if you need them?
[03:56] <cjwatson> there are lots of different workarounds that one might potentially need on different hardware, and I'm not sure enumerating them in the UI is worthwhile
[03:59] <xivulon> My idea is was to make the geeks happy via command-line parameters and non-geeks happy with a single "catch-them-all" workaround
[03:59] <cjwatson> but it is NOT a single catch-them-all workaround
[03:59] <xivulon> We used to have an "advanced" page, we might resurrect that, but I like it much better with a single page.
[03:59] <cjwatson> it fixes your laptop
[03:59] <cjwatson> it fixes a number of people's systems
[03:59] <cjwatson> it is far from complete
[04:01] <cjwatson> I agree that a single page is nicer, but on the other hand advanced boot options might not belong on the main interface
[04:01] <xivulon> I understand that but if we can cover most common cases in one go the ones that are left out may well read a guide and submit the parameters via command line
[04:01] <cjwatson> maybe thinking of it as a dialog rather than a page would be better
[04:01] <cjwatson> but people will check the "make my hardware work" box and then complain because they're one of the people whose hardware it breaks
[04:02] <xivulon> I see but people with complex setup will still need to read a lot of info to guess what is the right parameters for them
[04:02] <cjwatson> or some people will probably need some combination of the usual workarounds but not all
[04:02] <xivulon> Why should we force all the others to read the guide when we can fix most of them in one go?
[04:02] <cjwatson> because we can't fix most of them in one go
[04:02] <cjwatson> we will cause confusion by providing this "make it work" box that doesn't
[04:03] <cjwatson> I honestly think you're extrapolating from the specific to the general
[04:03] <cjwatson> and it is very non-Ubuntu-like to provide a workaround checkbox like that when it should really be fixed properly in the kernel, imo
[04:03] <cjwatson> if we provide a workaround checkbox, we'll stop hearing from people about it
[04:04] <cjwatson> and it will become harder to make it Just Work across the board
[04:05] <cjwatson> ultimately Ubuntu should not require this sort of magic workaround boot option at all on common hardware. Making it too easy to find the workaround does make it easier for affected users in the short term, but it unfortunately also makes it more likely that there will be more affected users in the long-term, because the affected users will be less likely to file bugs
[04:05] <cjwatson> I feel quite strongly about this :)
[04:07] <xivulon> Well it's clearly a trade off: bug reports (= number of happy new users tomorrow) vs number of happy new users today
[04:08] <cjwatson> we're still in the early adopter phase
[04:08] <cjwatson> well, maybe late early adopter :)
[04:08] <cjwatson> it's really important to get good bugs from early adopters - there are orders of magnitude fewer people using Ubuntu now than we hope will use it in the future
[04:09] <xivulon> All I am saying, is that if you go for bug reports, I don't see much point in putting up a page with cryptic things like "acpi", "vesa" & co. I'd rather just support command line options.
[04:09] <xivulon> Anything on the user interface has to be grandma-proof
[04:09] <cjwatson> it needs to correspond to the things people can google for under "why won't my <system name here> boot with Ubuntu?"
[04:10] <cjwatson> so inventing cutesy names for it like "safe mode" is just going to confuse, I think
[04:10] <cjwatson> it is very important to be consistent with other Ubuntu installation methods
[04:11] <cjwatson> I think a command line option that goes --boot-args="acpi=off noapic" would be useful, since then we could provide instructions to afflicted users
[04:11] <xivulon> Yeah but if they bother googling for it they can just add a command line parameter, no need to sugar that with a user interface that will scare the crap of the 90% of the other users
[04:11] <cjwatson> and the instructions could take the opportunity to say "please file a bug about this so that we can make sure people aren't affected in future"
[04:11] <cjwatson> sounds reasonable
[04:13] <xivulon> Ok so no change to user interface and I will support --boot-args="*" and maybe some other convenience flag
[04:13] <cjwatson> paradoxically I think that might actually get more users on broken hardware. the reason is that if you have a "safe mode" item that isn't absolutely 100%, people in the 100% minus (whatever) are likely to give up when safe mode doesn't work for them - obviously Ubuntu is just fucked on their hardware
[04:13] <cjwatson> whereas if the interface is boot options, they have to google for it and they might well find that there's something that works
[04:14] <cjwatson> basically if we're into safe mode territory we have already lost the most timid users
[04:14] <cjwatson> and a sizeable proportion of the rest are likely to persist for a little while if it's "but I just need to find the right rune"
[04:16] <Knuta> is a d-i package with the https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.17/+bug/73647 fix already uploaded to some secret place, or do people still need to rebuild things to netboot with bnx2? The "One-liner fix." comment in the bug is not very informing :-|
[04:16] <xivulon> I'd add a page to web interface of the CD: if wubi does not work for you see a list of possible reasons and solutions, wich will explain common options and have a link to launch wubi with appropriate parameters
[04:16] <cjwatson> Knuta: edgy isn't likely to get a point release, but bnx2 should be included in 6.06.2 when it's released (fairly soon)
[04:17] <Knuta> cjwatson: I'm using dapper
[04:17] <cjwatson> ah, your URL said 2.6.17 so I assumed you meant edgy
[04:17] <Knuta> cjwatson: I think the bug covers both
[04:17] <cjwatson> "One-liner fix" was Martin's comment justifying targeting it for 6.06.2
[04:18] <cjwatson> the kernels for that haven't been uploaded yet though (and in fact this is the major outstanding blocker for 6.06.2)
[04:19] <Knuta> people are kicking me around because they can't install ubuntu on the new servers :-|
[04:19] <Knuta> what does "fairly soon" mean? a week? a month?
[04:20] <cjwatson> I believe current schedule is beginning of Oct
[04:27] <xivulon> cjwatson did you see my comments on winfoss wubi troubleshooting page?
[04:29] <cjwatson> no
[04:29] <cjwatson> oh, maybe
[04:29] <cjwatson> 15:16 <xivulon> I'd add a page to web interface of the CD: if wubi does not work for you see a list of possible reasons and solutions, wich will explain common options and have a link to launch wubi with
[04:29] <cjwatson>                 appropriate parameters
[04:29] <cjwatson> that?
[04:29] <cjwatson> seems reasonable
[04:30] <xivulon> there was another piece
[04:30] <xivulon> Basically have a Winfoss page "Wubi Troubleshooting" and a sister page "LiveCD Troubleshooting"
[04:31] <xivulon> Then have a block for each common problem with a checkbox close to it. Something like: my screen resolution is *##~* Y/N
[04:31] <xivulon> Then a button with "Try Wubi With Common Workarounds"
[04:32] <xivulon> That button will launch wubi with appropriate boot parameter and show another winfoss page
[04:32] <xivulon> Submit bug report so that we can improve support for your hardware in the future
[05:24] <xivulon> cjwatson you mentioned you might have built another ISO with the patch. Please let me know if that is the case.
[05:25] <cjwatson> with which patch?
[05:25] <xivulon> http://codebrowse.launchpad.net/~ubuntu-core-dev/partman-auto-loop/ubuntu/revision/cjwatson%40canonical.com-20070920103812-mgae9pjcor3xh5hy?start_revid=cjwatson%40canonical.com-20070920104108-tfnyb57itaid5v9y
[05:26] <cjwatson> oh, no, not yet
[05:26] <cjwatson> needs a ubiquity upload first, for which I was sort of planning to wait for translations
[05:26] <cjwatson> you can just edit it on the fly
[05:26] <xivulon> and anything else that you might have fixed in the meantime
[05:28] <xivulon> np
[05:38] <xivulon> Had another idea re safe mode stuff, now the grub menu that launches the installer/liveiso is hidden away...
[05:39] <xivulon> I could simply add other items to menu.lst, and have a 5 sec black screen that says "Hit esc if you have problems booting with wubi..."
[05:41] <xivulon> The grub titles might be things like: "Standard|Workaround for graphics problems|Workarounds for XYZ problem..."
[05:42] <cjwatson> that would be a bit better
[05:43] <cjwatson> I'm still concerned that it wouldn't match Ubuntu, and that we don't have room for more menu entries on Ubuntu CDs (it would scroll and the user experience gets messy)
[05:43] <cjwatson> but I see I am not going to persuade you with that argument
[05:43] <cjwatson> have you at least filed a bug about your own laptop?
[05:44] <xivulon> It's already there
[05:45] <xivulon> No need to persuade me, I'll do what you say, just throwing out ideas
[05:45] <cjwatson> well, I'm assuming that this is coming from wubi users as well as just you
[05:47] <xivulon> Well graphics was less than an issue for us because we used the alternate iso, I'd expect to see more tickets there once we work with live cd
[05:47] <xivulon> We had a few users that required acpi fixes
[11:42] <xivulon> cjwatson,evand I have uploaded a new build including latest revisions (no change on my part)
[11:59] <CIA-18> ubiquity: cjwatson * r2248 ubiquity/ (d-i/manifest debian/changelog):
[11:59] <CIA-18> ubiquity: * Automatic update of included source packages: hw-detect 1.53ubuntu2,
[11:59] <CIA-18> ubiquity:  kboot-installer 0.0.1ubuntu4, localechooser 1.38ubuntu2,
[11:59] <CIA-18> ubiquity:  migration-assistant 0.5.0, partman-auto-loop 0ubuntu7, partman-base
[11:59] <CIA-18> ubiquity:  107ubuntu3.
[12:03] <CIA-18> ubiquity: cjwatson * r2249 ubiquity/debian/changelog: releasing version 1.5.16
[12:10] <xivulon> cjwatson,evand will have to upload again
[12:13] <xivulon> done rev291