[09:28] <davmor2> evand: I'm just running a kubuntu install test on fat 32 against 136 of wubi.  xivulon ask me if I could remind you about building it proper though :)
[09:28] <evand> indeed, I plan on it
[09:41] <davmor2> evand: Cool :)
[10:02] <CIA-9> anna: cjwatson * r417 ubuntu/debian/ (17 files in 2 dirs): merge from Debian 1.34
[10:03] <CIA-9> anna: cjwatson * r418 ubuntu/debian/po/ (ast.po et.po hy.po kk.po): apply Ubuntu branding to new translations
[10:11] <CIA-9> anna: cjwatson * r419 ubuntu/debian/changelog: releasing version 1.34ubuntu1
[10:14] <CIA-9> netcfg: cjwatson * r637 ubuntu/debian/ (15 files in 2 dirs): merge lp:~al-maisan/netcfg/ubuntu
[10:25] <CIA-9> cdrom-checker: cjwatson * r248 ubuntu/debian/ (14 files in 2 dirs): merge from Debian 1.16
[10:27] <CIA-9> cdrom-checker: cjwatson * r249 ubuntu/debian/po/ (ast.po et.po kk.po): apply Ubuntu branding to new translations
[10:31] <CIA-9> cdrom-checker: cjwatson * r250 ubuntu/debian/changelog: releasing version 1.16ubuntu1
[10:42] <CIA-9> clock-setup: cjwatson * r207 ubuntu/debian/ (16 files in 2 dirs): merge from Debian 0.98
[10:43] <CIA-9> clock-setup: cjwatson * r208 ubuntu/debian/changelog: releasing version 0.98ubuntu1
[11:28] <CIA-9> finish-install: cjwatson * r828 ubuntu/ (18 files in 3 dirs): merge from Debian 2.23
[11:29] <CIA-9> finish-install: cjwatson * r829 ubuntu/debian/changelog: releasing version 2.23ubuntu1
[12:27] <CIA-9> lilo-installer: cjwatson * r435 ubuntu/debian/ (16 files in 2 dirs): merge from Debian 1.30
[12:28] <CIA-9> lilo-installer: cjwatson * r436 ubuntu/debian/po/ (ast.po et.po kk.po): apply Ubuntu branding to new translations
[12:29] <CIA-9> lilo-installer: cjwatson * r437 ubuntu/debian/changelog: releasing version 1.30ubuntu1
[13:16] <evand> cjwatson: hrm, given that the DVD images are already oversized, am I correct in assuming that https://blueprints.edge.launchpad.net/ubuntu/+spec/foundations-karmic-unr-with-dvd is a non-start without a move to dual layer images?
[13:18] <cjwatson> evand: it wouldn't surprise me in the least if there were a good amount of stuff we could trim off the DVD images
[13:18] <cjwatson> evand: we've never really put extensive effort into it
[13:18] <cjwatson> evand: we *might* need to look at nested squashfses though to make it practical - not sure
[13:19] <evand> won't we need to do that anyway?
[13:19] <evand> I cannot forsee us finding enough space to put multiple entire squashfs images on a single DVD
[13:19] <evand> but perhaps I'm missing the obvious
[13:20] <evand> re trimming > I'll make that a requirement of the specification
[13:20] <cjwatson> yeah, we probably would
[13:20] <cjwatson> the livefs build organisation gets a bit interesting
[13:21] <cjwatson> you either have to do multiple builds at once (slow if we're in a hurry) or you have to cache base system build results for later builds the same day (how do you force it to update when you need to?) or something more complicated
[13:22] <evand> hrm, indeed
[13:34] <CIA-9> lowmem: cjwatson * r80 ubuntu/debian/ (changelog control lowmem.postinst): merge from Debian 1.31
[13:35] <CIA-9> lowmem: cjwatson * r81 ubuntu/debian/changelog: releasing version 1.31ubuntu1
[13:48] <CIA-9> main-menu: cjwatson * r134 ubuntu/ (15 files in 3 dirs): merge from Debian 1.29
[13:48] <CIA-9> main-menu: cjwatson * r135 ubuntu/debian/po/ (ast.po et.po kk.po): apply Ubuntu branding to new translations
[13:56] <CIA-9> main-menu: cjwatson * r136 ubuntu/debian/changelog: releasing version 1.29ubuntu1
[14:09] <CIA-9> net-retriever: cjwatson * r359 ubuntu/debian/ (16 files in 2 dirs): merge from Debian 1.24
[14:11] <CIA-9> net-retriever: cjwatson * r360 ubuntu/debian/changelog: releasing version 1.24ubuntu1
[14:14] <CIA-9> pkgsel: cjwatson * r145 ubuntu/debian/ (69 files in 2 dirs): merge lp:~al-maisan/pkgsel/ubuntu
[14:15] <CIA-9> pkgsel: cjwatson * r146 ubuntu/debian/po/ (ast.po et.po kk.po): debconf-updatepo for new translations
[14:19] <CIA-9> pkgsel: cjwatson * r147 ubuntu/debian/changelog: RFC2822 date format
[15:02] <CIA-9> partman-auto-lvm: cjwatson * r221 ubuntu/ (18 files in 3 dirs): merge from Debian 33
[15:03] <CIA-9> partman-auto-lvm: cjwatson * r222 ubuntu/debian/po/ (ast.po et.po kk.po): debconf-updatepo for new translations
[15:11] <CIA-9> partman-auto-lvm: cjwatson * r223 ubuntu/debian/ (64 files in 2 dirs):
[15:11] <CIA-9> partman-auto-lvm: Update translations of Ubuntu-specific strings from my last conveniently
[15:11] <CIA-9> partman-auto-lvm: available Launchpad export (2009-04-16); we'll update again after Karmic
[15:11] <CIA-9> partman-auto-lvm: translations open.
[15:13] <CIA-9> partman-auto-lvm: cjwatson * r224 ubuntu/debian/changelog: releasing version 33ubuntu1
[15:15] <CIA-9> partman-basicfilesystems: cjwatson * r576 ubuntu/debian/ (68 files in 2 dirs): merge from Debian 63
[15:18] <CIA-9> partman-basicfilesystems: cjwatson * r577 ubuntu/ (mountoptions/ext2_defaults debian/changelog): Drop ext2 relatime default, as it's now the default in the kernel.
[15:18] <CIA-9> partman-basicfilesystems: cjwatson * r578 ubuntu/debian/changelog: releasing version 63ubuntu1
[15:22] <rgreening> good day evand
[15:22] <evand> hello
[15:22] <rgreening> evand: the YES, NO, MAYBE range(3), etc can be stripped right?
[15:22]  * evand checks
[15:23] <rgreening> I believ ethe functionality is in the back-end now
[15:23] <rgreening> Backend.IMG, etc...
[15:23] <evand> yes
[15:24] <rgreening> do you want to clean that up or will I just go ahead and make the change
[15:24] <evand> by all means (I'm a bit busy this afternoon trying to finish off drafting specifications)
[15:25] <evand> davmor2: http://people.ubuntu.com/~evand/wubi/karmic/wubi-r136.exe - it will be on tomorrow's daily-live, assuming they build properly
[15:25] <rgreening> ok, I'll clean it up...
[15:27] <CIA-9> partman-lvm: cjwatson * r1214 ubuntu/debian/ (23 files in 2 dirs): merge from Debian 66
[15:30] <CIA-9> partman-lvm: cjwatson * r1215 ubuntu/debian/changelog: releasing version 66ubuntu1
[15:32] <evand> thanks
[15:34] <rudiger> My Scenario: primery part with WinXP, extended part with ubuntu 9.04, grub loader.
[15:35] <rudiger> my task: want to install Vista at pimary part. (sorry)
[15:36] <CIA-9> partman-partitioning: cjwatson * r707 ubuntu/debian/ (7 files in 2 dirs): merge from Debian 66
[15:37] <rudiger> question: How to avoid damaging the mbr and grup booting while installing vista?
[15:37] <CIA-9> partman-partitioning: cjwatson * r708 ubuntu/debian/changelog: releasing version 66ubuntu1
[15:37] <evand> rudiger: you cannot, as far as I know.  You'll just have to put GRUB back in the MBR after Vista installs.
[15:38] <CIA-9> usb-creator: rgreening * r105 usb-creator/usbcreator/kde_frontend.py:
[15:38] <CIA-9> usb-creator: Implement Solid support to replace gnomevfs support from PyGtk front-end. WIP.
[15:38] <CIA-9> usb-creator: Begin to port setup_sources_treeview and setup_targets_treeview
[15:38] <CIA-9> usb-creator: - start by adding Solid support - WIP.
[15:39] <rudiger> evand: thanx for info. How to reinstall GRUB back in the MBR?
[15:42] <cjwatson> 'sudo grub-install hd0', assuming that it's your first (or only) hard disk
[15:44] <cr3> evand: thanks for the answer regarding passwd/allow-password-weak. by the way, I've been meaning to ask you about the difference between the ubiquity and the user-setup lines when preseeding a desktop installation
[15:45] <rudiger> cjwatson: Hm, I guess, I'll have to do this after booting ubuntu from cdrom? (As I will not be able to boot ubuntu from harddisk any more after vista install.)
[15:45] <cr3> evand: one of the reasons I ask is that I grepped the ubiquity code for "passwd" before asking you about the weak password and I couldn't find anything because it turned out this was a user-setup thing, which is fuzzy in my mind
[15:48] <CIA-9> usb-creator: rgreening * r106 usb-creator/usbcreator/gtk_frontend.py:
[15:48] <CIA-9> usb-creator: Remove some code no longer used by gtk_frontend.py (now implemented in back-end)
[15:48] <CIA-9> usb-creator: - YES, NO, MAYBE range(3)
[15:49] <CIA-9> tzsetup: cjwatson * r502 ubuntu/debian/ (16 files in 2 dirs): merge from Debian 1:0.25
[15:49] <cjwatson> rudiger: yes, you can do it from a live CD
[15:50] <cjwatson> cr3: regarding your pxelinux.0 question last night, you can either take it from the package or from the archive - they're bit-for-bit identical so it probably won't matter
[15:51] <rudiger> cjwatson: Fine. :-)
[15:51] <cr3> cjwatson: thanks for the follow up, I decided to go for the archive approach because it was much simpler to implement
[15:51] <rudiger> cjwatson, evand: thanx for your support.
[15:52] <CIA-9> tzsetup: cjwatson * r503 ubuntu/debian/po/ (ast.po et.po kk.po): debconf-updatepo for new translations
[15:53] <CIA-9> tzsetup: cjwatson * r504 ubuntu/debian/changelog: releasing version 1:0.25ubuntu1
[15:53] <cr3> when I preseeded the desktop installation with my own package information, checkbox foo/bar boolean true, it seems like those values didn't get preseeded when calling "in-target apt-get install checkbox" in ubiquity/success_command
[15:55] <cjwatson> probably doesn't get copied over to /target
[15:56] <cjwatson> workaround: debconf-copydb configdb targetdb --owner-pattern checkbox --config=Name:targetdb --config=Driver:File --config=Filename:/target/var/cache/debconf/config.dat
[15:56] <cjwatson> we should probably do something a bit more sensible with preseeds that don't have "d-i" or "ubiquity" as owners, much as d-i does
[15:57] <cr3> cjwatson: or, since I only expect to use those preseed variables under /target rather than during the ubiquity install, should I debconf-communicate them as part of my ubiquity/success_command script?
[15:57] <cjwatson> you could do that too, which is easier depends on your preseeding setup
[15:58] <cr3> cjwatson: that would enable me to use the same ubiquity/success_command as the one I use in the d-i late_command
[15:58] <cr3> cjwatson: although, adding that debconf-copydb line to the d-i late_command script wouldn't hurt either
[15:58] <icarus901> cjwatson, is there any preseed magic that can suppress "Could not get identity of device /dev/xxx - Invalid argument"
[15:59] <cjwatson> cr3: it probably would, cdebconf's debconf-copydb doesn't support --owner-pattern
[15:59] <cjwatson> icarus901: I'll need more context for that error message, preferably full logs
[15:59] <cr3> cjwatson: if I were to set preseed variable in the success_command/late_command scripts, is there a simpler way than this: echo set foo/bar boolean true | in-target debconf-communicate
[15:59] <cr3> cjwatson: ouch, thanks for the warning!
[16:00] <icarus901> cjwatson, alright; i'll get back to you on that. short story: xen paravirtual kernel + standard netboot installer initrd. i'm not overly surprised by the message, as it is harmless in this case
[16:01] <cr3> cjwatson: by the way, would you like me to report a bug against ubiquity regarding: we should probably do something a bit more sensible with preseeds that don't have "d-i" or "ubiquity" as owners, much as d-i does
[16:08] <cjwatson> cr3: debconf-communicate> that seems pretty simple to me
[16:08] <cjwatson> icarus901: ok, thanks. I doubt there's a way to suppress that with preseeding
[16:08] <cjwatson> icarus901: (well, I suppose there might be, it depends where it is and what the device is)
[16:09] <cjwatson> cr3: yes please
[16:09] <icarus901> cjwatson, looks like it's related to probing by or prior to partman
[16:17] <cjwatson> parted_devices would be what I'd expect, but I can't speculate further without logs
[16:19] <CIA-9> yaboot-installer: cjwatson * r266 ubuntu/debian/ (21 files in 2 dirs): merge from Debian 1.1.15
[16:21] <CIA-9> yaboot-installer: cjwatson * r267 ubuntu/debian/changelog: releasing version 1.1.15ubuntu1
[16:29] <icarus901> cjwatson, is this sufficient? http://137.99.12.21:9055/
=== dpm_ is now known as dpm
[16:31] <cjwatson> icarus901: yes, it is. I'm afraid right now there is no way to exclude devices at that level
[16:31] <icarus901> cjwatson, that's a shame. i'd be comfortable tweaking code though. pointers to where i should begin?
[16:33] <cjwatson> /lib/partman/init.d/30parted, or init.d/parted in the partman-base source package
[16:34] <cjwatson> look in the loop that iterates over the output of parted_devices
[16:36] <CIA-9> installation-report: cjwatson * r67 ubuntu/ (15 files in 3 dirs): merge from Debian 2.39
[16:36] <icarus901> cjwatson, thanks much
[16:37] <cjwatson> if you figure out how we can detect the problem and skip it automatically without breaking non-Xen systems, let us know :-)
[16:37] <CIA-9> installation-report: cjwatson * r68 ubuntu/debian/changelog: releasing version 2.39ubuntu1
[16:38] <icarus901> cjwatson, will do
=== cjwatson_ is now known as cjwatson
[17:25] <CIA-9> base-installer: cjwatson * r371 ubuntu/ (11 files in 5 dirs): merge from Debian 1.101
[17:27] <CIA-9> base-installer: cjwatson * r372 ubuntu/debian/changelog: releasing version 1.101ubuntu1
[17:30] <CIA-9> debian-installer-utils: cjwatson * r677 ubuntu/ (block-attr debian/changelog): merge from Debian 1.69
[17:31] <CIA-9> debian-installer-utils: cjwatson * r678 ubuntu/debian/changelog: releasing version 1.69ubuntu1
[17:33] <CIA-9> partman-ext3: cjwatson * r757 ubuntu/debian/ (po/et.po changelog po/bn.po po/fi.po po/sk.po): merge from Debian 57
[17:34] <CIA-9> partman-ext3: cjwatson * r758 ubuntu/ (3 files in 2 dirs): Drop ext3/ext4 relatime default, as it's now the default in the kernel.
[17:35] <CIA-9> partman-ext3: cjwatson * r759 ubuntu/debian/changelog: releasing version 57ubuntu1
[17:37] <CIA-9> partman-auto: cjwatson * r294 ubuntu/debian/ (po/et.po changelog po/bn.po po/sk.po): merge from Debian 86
[17:38] <CIA-9> partman-auto: cjwatson * r295 ubuntu/debian/po/et.po: debconf-updatepo for new translations
[17:39] <CIA-9> partman-auto: cjwatson * r296 ubuntu/debian/changelog: releasing version 86ubuntu1
[17:48] <CIA-9> rootskel: cjwatson * r342 ubuntu/ (6 files in 2 dirs): merge lp:~al-maisan/rootskel/ubuntu
[17:51] <CIA-9> rootskel-gtk: cjwatson * r114 ubuntu/ (debian/changelog src/usr/share/graphics/Makefile): merge lp:~al-maisan/rootskel-gtk/ubuntu
[17:52] <CIA-9> rootskel-gtk: cjwatson * r115 ubuntu/debian/changelog: coalesce previous UNRELEASED changelog entry
[18:01] <rgreening> evand1: ping
[18:01] <evand1> rgreening: pong
[18:01] <rgreening> hey
[18:01] <rgreening> evand1: does usb-creator use the scripts/install.py?
[18:03] <evand1> yes
[18:05] <rgreening> hmm... where does it get imported/called? I must be blind...
[18:05] <rgreening> nm... I am blind... found it
[18:09] <cr3> cjwatson: reported bug #387400 regarding preseeding packages in ubiquity where the owner is not "d-i" nor "ubiquity", enjoy :)
[18:09] <ubottu> Launchpad bug 387400 in ubiquity "Ubiquity should support preseeding of other packages" [Undecided,New] https://launchpad.net/bugs/387400
[18:10] <cjwatson> thanks
[18:12] <cr3> cjwatson: no rush in fixing, you've given me plenty of workarounds :)
[18:19] <CIA-9> usb-creator: evand * r107 usb-creator/debian/usb-creator-kde.install: Fix a typo in usb-creator-kde.install
[18:59] <cr3> cjwatson: will in-target work when using debconf-communicate or do I need to chroot /target? it seems that my success_command called "echo 'set foo/bar string blah | in-target debconf-communicate" but none of the variables seem to appear in the debconf database
[19:02] <cjwatson> in-target probably won't preserve stdin
[19:02] <cjwatson> just use chroot /target, there's no great need for in-target here
[19:02] <cjwatson> also your code as pasted into IRC has a missing closing single quote
[19:37] <CIA-9> casper: cjwatson * r646 trunk/ (debian/changelog scripts/casper):
[19:37] <CIA-9> casper: If LIVE_MEDIA_PATH is set on the command line, record it in
[19:37] <CIA-9> casper: /etc/casper.conf for the benefit of ubiquity.
[19:39] <CIA-9> ubiquity: cjwatson * r3282 ubiquity/ (ubiquity/casper.py bin/ubiquity-dm debian/changelog):
[19:39] <CIA-9> ubiquity: Factor out /etc/casper.conf parsing from ubiquity-dm to a separate
[19:39] <CIA-9> ubiquity: Python module, with a few corner-case bugs fixed along the way.
[19:43] <CIA-9> ubiquity: cjwatson * r3283 ubiquity/ (debian/changelog scripts/install.py):
[19:43] <CIA-9> ubiquity: Honour LIVE_MEDIA_PATH in /etc/casper.conf when looking for files
[19:43] <CIA-9> ubiquity: normally found under /cdrom/casper.
[19:45] <CIA-9> ubiquity: cjwatson * r3284 ubiquity/ubiquity/components/summary.py: missed a bit for LIVE_MEDIA_PATH handling
[19:45] <icarus901> cjwatson, the exception being thrown is of type Warning and has priority high. I can't see any particularly clean place to toss a conditional to ignore the error though
[19:45] <DGMurdockIII> is the source code for the ubuntu oem installer avable?
[19:45] <icarus901> setting priority to critical certainly skips it
[19:46] <icarus901> though that strikes me as beating it into submission rather than something more appropriate
[19:46] <cjwatson> DGMurdockIII: yes, 'apt-get source oem-config'
[19:46] <cjwatson> icarus901: I'd have thought it would be easier just to skip opening the device in the first place
[19:47] <DGMurdockIII> is there a svn or git site
[19:47] <cjwatson> bzr
[19:47] <cjwatson> http://wiki.ubuntu.com/InstallerDevelopment
[19:48] <DGMurdockIII> witch one s the oem verson
[19:50] <cr3> cjwatson: preseeding other packages still doesn't seem to work, I tried both in-target and chroot /target: http://people.ubuntu.com/~cr3/ubiquity.seed
[19:50] <cjwatson> DGMurdockIII: I believe I already said. oem-config is the package name
[19:50] <cr3> I also looked for logs from the installed system under /var/log/syslog, but zgrep'ing around for certification didn't show anything
[19:51] <cr3> so, I guess I need to > /target/success_command.log manually from the success_command script
[19:51] <cjwatson> cr3: debconf-communicate might need an owner; i.e. 'chroot /target debconf-communicate checkbox'. Furthermore your syntax is wrong for debconf-communicate. If you just want to give something preseed file input then debconf-set-selections might be more appropriate
[19:52] <cjwatson> (and debconf-set-selections will need an owner as the first field of its input)
[19:53] <cr3> cjwatson: I looked around for the syntax to set a variable with debconf-communicate but the manpage only has an example for get, where should I look for a set example?
[19:53] <cjwatson> it takes debconf protocol input; debconf-devel(7) documents it
[19:54] <cjwatson> like I say I suspect debconf-set-selections is actually closer to what you want
[19:55] <Torgoton> How many choices do I have when it comes to running the installer on a 486 box? I don't have a CD drive, but I do have a 1GB CF card.
[19:57] <rgreening> evand1: qui gui.. haha .. I must have been thinking in french, non? :)
[19:57] <cjwatson> Torgoton: there's a netboot/386 installer image - I think you'll have to use that
[19:57] <cjwatson> http://archive.ubuntu.com/ubuntu/dists/jaunty/main/installer-i386/current/images/netboot/386/
[19:58] <rgreening> packaging needs lots more love evand1. i.e. we need to move usb-creator to usb-creator.gtk and put a replaces in... etc. If it's ok with you, I'll leave that bit for you to play with :)
[19:58] <cjwatson> I don't believe we build any other installer images suitable for <i586
[19:58] <Torgoton> cjwatson: I was afraid of that. My 36MB RAM seems to not work with the netboot installer. Not enough room to decompress files, or so it seems. Installation gets partway through and stops. Tried many times, but I'm not an expert with Ubuntu.
[19:59] <cjwatson> it might not be *too* hard to build one; 'apt-get source debian-installer' and poke around in build/config/i386/
[19:59] <Torgoton> cjwatson: But I'm happy to try again. Would you recommend 8.04.2 or 9.04?
[19:59] <cjwatson> it probably doesn't make a lot of difference
[19:59] <Torgoton> I'd need another Ubuntu machine for that, yes? Easy to set up on another machine. ... building an installer. Sounds kinda fun.
[20:00] <cjwatson> yeah, make sure it's of the same release as the installer you're trying to build
[20:00] <cjwatson> 'sudo apt-get build-dep debian-installer' first
[20:01] <Torgoton> If you don't mind, I'll start the 8.04.2 and keep y'all posted on how the installation goes. Might it work on a low mem machine? I do have a swap partition on the drive, but don't know how to enable that during the install. Here goes. (I'll capture your notes for building an installer for later too. Thank you!)
[20:03] <cjwatson> if the disk is pre-partitioned, you could use 'swapon /device/name' from the shell on alt-f2 to enable swap
[20:03] <cjwatson> I think that trick is likely to work better on 9.04; I rather suspect that 8.04.2 sometimes unnecessarily disabled swap during partitioning even if the disk was already partitioned exactly correctly
[20:03] <Torgoton> Great! I can do that as soon as I can get to that shell, right?
[20:04] <Torgoton> heheheh OK. Restarting with 9.04!
[20:05] <cjwatson> yeah, you'll definitely run into bug 287660 otherwise
[20:05] <ubottu> Launchpad bug 287660 in partman-base "Partman reports changes to a disk when there are none." [High,Fix released] https://launchpad.net/bugs/287660
[20:06] <Torgoton> Highly inconvenient bug.
[20:17] <Torgoton> ok cjwatson, I got the Low Memory note, and am now at the "Ubuntu installer main menu". I can switch to tty2, but I don't see any /dev/sdx or /dev/hdx devices for swapon. Are they enabled yet, or do I have to get to "Download installer components" and select an IDE driver?
[20:40] <cr3> I've preseeded two ubiquity installs of karmic, one on a system with broadcom ethernet and the other with intel ethernet. in both cases, the wired network seems to be down, is this a problem in karmic or might there be a problem with my preseeding?
[21:07] <rgreening> evand1: Is it possible to move the gobject watch/timer/timout code into the front-end? If so, I think I can do similar with Qt code. then we can simply make calls to self.frontend.PROPERTY or self.frontend.FUNCTION which will do the gobject or Qt appropriate stuff..
[21:36] <CIA-9> usb-creator: rgreening * r108 usb-creator/usbcreator/kde_frontend.py:
[21:36] <CIA-9> usb-creator: Fixup QProgressDialog to show window title and a label
[21:36] <CIA-9> usb-creator: Fix a typo for new Solid changes
[21:36] <CIA-9> usb-creator: Remove unnecessary import of gobject
[21:54] <cjwatson> Torgoton: you have to get past "download installer components" before you're guaranteed to be able to get at disks
[21:55] <cjwatson> cr3: is the Broadcom one bnx2? if so, bug 384861
[21:55] <ubottu> Launchpad bug 384861 in linux "Broadcom NetXtreme II (BCM5708) not detected by installer [karmic]" [High,In progress] https://launchpad.net/bugs/384861
[21:55] <cjwatson> cr3: what driver does the Intel one use?
[21:56] <Torgoton> thanks, cjwatson. My machine is taking its sweet time. Not quite there yet. When I do get to download components, which should I select?
[21:56] <cjwatson> I can't really say without being in front of the machine myself, I'm afraid
[21:56] <cjwatson> firstly, it's ages since I did a lowmem install; secondly, it's somewhat hardware-dependent ...
[21:59] <cr3> cjwatson: e1000e
[22:01] <cjwatson> cr3: I'm not quite sure what might be going on there. Can I see the syslog?
[22:02] <cr3> cjwatson: the machine's been flushed, I really need to keep the syslog before doing that. I'll reinstall...
[22:02] <cjwatson> it doesn't seem to be one of those with separate firmware
[22:06] <Torgoton> cjwatson: Thanks. I don't really know what each module does, exactly, and I'm trying to keep the number low because of memory. <shrug> We'll see if the menu comes back and go from there.
[22:19] <cr3> cjwatson: since the machine is remote and I don't have network access, only kvm access, I have captured a screenshot of what might be the relevant part from syslog regarding eth0: http://people.ubuntu.com/~cr3/screenshot.jpg
[22:28] <Torgoton> cjwatson (or anyone, really): I have my component menu (finally!). I'm thinking IDE-modules for the hard drive, right? But do I also need nic-pcmcia-modules if the PCMCIA NIC I have is already working? Could I install from an ISO on a CF card if I add iso-scan (I'd put the 9.04 alternate ISO on there)? Is block-modules needed? Is kernel-image needed? Are any partman modules needed? PATA? storage-core-modules?
[22:32] <cjwatson> Torgoton: 1) ide-modules is possible but depending on the controller it might be pata-modules instead; 2) you don't need nic-pcmcia-modules if your PCMCIA NIC already works; 3) iso-scan yes; 4) you probably don't need block-modules; 5) you don't need kernel-image; 6) you're going to need *some* partman modules, at least -base -partitioning -target -basicmethods -basicfilesystems, -auto for guided partitioning, -ext3 ...
[22:33] <cjwatson> ... if you want ext3 support; 7) PATA see 1; 8) you need storage-core-modules if you want to be able to mount ISO9660 filesystems or work with USB disks
[22:55] <CIA-9> grub-installer: cjwatson * r785 ubuntu/ (8 files in 2 dirs): merge from Debian 1.38
[22:55] <Torgoton> cjwatson: Thank you so much. Digging in...
[22:57] <CIA-9> grub-installer: cjwatson * r786 ubuntu/ (debian/changelog grub-installer):
[22:57] <CIA-9> grub-installer: Go back to using update-grub -y for GRUB Legacy for now; our grub
[22:57] <CIA-9> grub-installer: package is a bit old and still requires this.
[22:58] <CIA-9> grub-installer: cjwatson * r787 ubuntu/ (debian/changelog grub-installer): Default to grub2 for GPT systems.
[22:59] <CIA-9> grub-installer: cjwatson * r788 ubuntu/ (debian/changelog grub-installer): Allow grub/grub2 choice for ext4, though still default to grub2.
[23:00] <Torgoton> (FWIW, I went with ide-modules and partman-auto. partman-base, -partitioning, -target, -basicmethods, -basicfilesystems were not offered, and I hope are automatically selected.)
[23:02] <Torgoton> partman-base is getting installed.
[23:03] <Torgoton> loaded, rather.