=== superm1|away is now known as superm1 [08:25] davmor2, hi, did you have a chance to test wubi in 8.10? [08:26] I am stacked with mount problems ("Device or resource busy" when mounting the ntfs device) [08:26] xivulon: I did yesterday it dropped into initramfs [08:26] seems consistent with the above [08:27] can you pls boot in verbose mode to find out more? [08:27] check /proc/mounts to see if the windows device is mounted and if not try to mount it manually with [08:27] xivulon: https://bugs.edge.launchpad.net/wubi/+bug/268123 [08:28] ah missed that [08:28] can you try to mount manually? [08:28] verbose mode showed nothing up it went through all the hw setup and stopped dead [08:29] xivulon: I can try yes take me a second or two to setup just checking everything off at the minute [08:31] sure [08:36] I asked cking [08:41] Right okay back with you [08:43] and added a couple of comments to your bug report, this is a show stopper :( [08:44] Mounting /dev/sda1 on isodevice failed: invalid argument [08:51] mount -t ntfs /dev/sda1 /isodevice [08:51] or [08:51] mount.ntfs /dev/sda1 /isodevice [08:51] first should give "no such device" [08:51] second should give "device or resource busy" [08:53] xivulon: confirmed [08:54] cking is on it, hopefully will get a fix soon [08:54] evand, cjwatson FYI ^^ === superm1 is now known as superm1|away [10:55] cjwatson: did you have any chance to look into 243105? [10:55] (posted on # [10:55] ubuntu-kernel as well by mistake) [11:03] xivulon: no, sorry [11:04] np, I will stick with ISO extraction then [11:07] I have recoded almost all of wubi into python, good news is that I can probably keep it well below 2M, but it is not completed yet, and it is probably to late for 8.10 [11:08] it might be still a valid option for a stand alone release though, also because a bittorrent download manager is included [15:17] evand, I'm half-way through writing up feedback on usb-creator [15:17] mpt: thanks! Very much appreciated. [15:18] base-installer: cjwatson * r332 ubuntu/ (debian/changelog library.sh): releasing version 1.86ubuntu5 [15:24] ubiquity: cjwatson * r2806 ubiquity/ (debian/changelog scripts/install.py): Never select /dev/ramzswap* as a hibernation target. [15:28] cjwatson: hey there... [15:29] cjwatson: i'd like some help/review with http://pastebin.ubuntu.com/45354/ when you get a second [15:30] kirkland: I'm not entirely convinced that copying the mdadm template into partman-target (a) is a good idea to start with (where do translations go?) and (b) will actually work (will the value get transferred over to the debconf db in /target?) [15:30] kirkland: have you tested this? [15:30] cjwatson: no, i haven't [15:30] cjwatson: well, i have, but it hasn't worked for me yet :-) [15:30] kirkland: (also, minor point, please use [ foo ] || [ bar ] rather than [ foo -o bar ] - the rules for the former are labyrinthine and so it's best avoided) [15:31] cjwatson: but i spent most of yesterday with my wheels spinning at this point [15:31] the rules for the latter, I mean [15:31] cjwatson: fair enough, i'll fix the -o || bit [15:31] kirkland: I have to step out for a moment for the school run, but let me think about it [15:31] * kirkland takes a note to use the word "labyrinthine" in the future :-) [15:31] cjwatson: thanks [15:32] cjwatson: i wanted to put at least some code in front of you, to get some help [15:32] cjwatson: and tis why its in a pastebin and not in a bug :-) not ready to be picked up [15:32] kirkland: also you should break out of the loop after you've asked the question once [15:33] cjwatson: i've already made that change locally :-) [15:33] just to nitpick the code that's there ... [15:33] cjwatson: no worries, i fixed that immediately upon posting the paste [15:36] cjwatson: those couple of nits are fixed in http://pastebin.ubuntu.com/45357/ [15:36] cjwatson: the templates text is still in there, until I get advice on how to solve that === superm1|away is now known as superm1 [16:06] cjwatson: fwiw, I opened a bug on the issue, https://bugs.launchpad.net/ubuntu/+source/mdadm/+bug/268580 [16:06] cjwatson: in case you need to communicate about it asynchronously === superm1 is now known as superm1|away [16:58] I would like to determine when the installer fails to complete and I'm contemplating two approaches for which I'd appreciate feedback from this channel... [16:59] 1. report syslog to an external server, for which support was introduced in hardy, and monitor failures from that server [17:00] 2. have a script repeatedly run on the machine itself and check the last timestamp in the syslog, report if it has been inactive for more than x seconds [17:00] remote syslog sounds easier [17:00] the problem with approach #1 is that it requires a more elaborate setup with an external server, whereas the problem with #2 is that it might not be as reliable [17:01] though I suppose that you could do #2 with a preseed/early_command that forks off a daemon process [17:01] you could fetch an architecture-specific C program [17:02] cjwatson: could I have a udeb installed before early_command kicks in? === superm1|away is now known as superm1 [17:05] you could boot with anna/choose_modules=udebname [17:05] if it is of a certain type. what udeb? [17:35] cjwatson: sorry to nag, i understand it's getting toward the end of your day... i could use that advice on partman-target at your earliest convenience [18:42] evand, sent (finally!) [18:42] sorry it's a bit long, I went into a fair bit of how and why as well as what [18:42] yay [19:05] The how and why part was quite helpful, thanks. I'm going to re-read this a few times, digest it, and then give a shot at implementing the suggestions. [21:53] ubiquity: cjwatson * r2807 ubiquity/ (debian/changelog scripts/install.py): [21:53] ubiquity: Fix fully-qualified path generation while walking the source filesystem; [21:53] ubiquity: dirpath is already prefixed with self.source. (This made no difference [21:53] ubiquity: in practice because os.path.join throws away previous components upon [21:53] ubiquity: encountering an absolute path, but was inelegant.) [21:55] ubiquity: cjwatson * r2808 ubiquity/ (debian/changelog scripts/install.py): Make the exclusion of /etc/fstab from bulk copying more precise. [22:05] Ugh, sometimes I want to throw GTK into a furnace. [22:08] what's it doing to you? [22:10] * cjwatson fails to reproduce bug 251640. I wonder if I fixed that by accident [22:10] A few things. I'm having a bit of difficulty figuring out why labels in a vbox who's parent is the window itself are centering despite being left aligned. I also ran into some trouble in using the container option on a hscale (same problem as last cycle), but I avoided it by not using it. [22:10] err container option on a option button [22:10] as well as the alignment option on the label itself, I think there's an "X align" property; set that to 0.0 [22:11] I avoided it by instead packing the hscale below the option, rather than in it (inside a vbox with a label) [22:11] indeed, already set [22:11] I suspect that what you have is a label whose text is centred within the label (e.g. if you had multiple lines they'd be centred with respect to each other) but not within its container [22:11] oh, hmm [22:11] it's really quite odd. [22:12] perhaps put other dummy elements in and try to figure out which component is centred [22:12] are you sure it's the label rather than (say) the vbox? [22:13] as a child of the parent window the vbox has no control over the spacing to the left and right of a child, I thought [22:13] indeed there are no options available under packing in glade [22:13] and spacing only affects to the top and bottom of the child widgets [22:14] I mean, is the vbox definitely extending all the way across or is it perhaps that the vbox only occupies a horizontal region in the centre? [22:14] ah, the vbox itself definitely extends to the edges of the window [22:14] ok [22:14] at least that's my impression from clicking on it in glade [22:14] stick the glade file somewhere and I'll see if I can make it out [22:14] much appreciated, will do [22:15] (unless you want to battle with it on your own of course) [22:16] http://evalicious.com/tmp/usbcreator.glade [22:16] if you can't figure it out quickly no worries, I'll take a look at it later and work at it until I get something usable [22:16] * evand departs for a bit [22:20] if I install in some language, will that be reflected in syslog? [22:21] I mean, will the month like "Sep" be formatted differently? [22:21] evand: I think the problem is that both the image and the frame are set to expand within the hbox, and GTK has to put the space *somewhere* [22:22] evand: if you set both the image and the frame not to expand, it seems to fix the alignment [22:22] evand: if that puts the image too far to the left, you can stick an alignment round it [22:22] cr3: busybox syslog isn't localised, so no [22:23] evand: I have a nagging feeling I've missed some other reason, but at any rate that seems to fix the immediate problem and seems closer to the right model [22:24] cjwatson: thanks [22:25] kirkland: OK, sorry I didn't get round to you before now [22:25] kirkland: firstly, I think this should be in partman-md, not partman-target; you can add a new check.d script for it there [22:26] kirkland: *bing* sorry, a lightbulb just went off [22:26] kirkland: you want the translated text in mdadm too, right? great. put it there, and *also* put it in mdadm-udeb, along with the check.d script! [22:26] no more translation syncing problem [22:27] that will make it much more comfortable and less namespace-violating to call the template mdadm/boot_degraded [22:28] kirkland: you'll need a slightly awkward trick to get the answer to the question into /target; this is necessary because (presumably) you don't want to ask the question when mdadm is actually installed, but instead during partman [22:29] kirkland: this trick is: db_get mdadm/boot_degraded; echo mdadm mdadm/boot_degraded boolean "$RET" | debconf-set-selections [22:29] with a comment explaining that you're doing this so that the value gets set in /target's debconf database before mdadm is installed [22:32] kirkland: in the template, you should avoid "yes"; we try to write templates so that they're agnostic of changes in the debconf frontend. We usually prefer language such as "you might want to boot your system even with a degraded RAID if ..." [22:32] kirkland: I think perhaps the short description could be clearer. Maybe "Do you want to boot normally even if your RAID becomes degraded?" [22:33] that's all from me [22:33] cjwatson: thanks, i'll review this in a bit [22:33] cjwatson: chin deep in another problem :-) [22:34] cjwatson: smartpm related :-P thanks for that btw... [22:34] * kirkland is kidding, btw [23:19] cjwatson: sorry, I'm not sure I understand your terminology. There aren't any images in the file. Also, what frame and hbox? Perhaps this would be more quickly solved (assuming you still have the modified file and saved it) of sending it back and I'll diff against my local copy. [23:21] I tried going as far as setting the expand property on everything to False and that didn't make the "To try or install Ubuntu..." string align properly. [23:21] * evand continues to dig [23:22] (or for that matter "Source disk image" and USB disk to use") [23:22] evand: I mean the objects named "image3" and "frame5" [23:23] which are elements of the hbox indirectly containing "warning_dialog_label" [23:23] oh, you're not talking about warning_dialog at all, are you ...? [23:24] drat [23:29] ah, sorry [23:29] I should've been specific [23:29] the window creatively named window [23:30] the warning dialog is actually lifted straight from ubiquity (with modification on the warning as Abort makes the usability folk cringe). [23:30] evand: I think this is essentially http://bugzilla.gnome.org/show_bug.cgi?id=101968 [23:31] evand: GtkLabels try to avoid the label getting "too wide", and as a result refuse to use all the available horizontal space [23:32] evand: there's an easy workaround in this case, though; put the label inside a one-element hbox, and ensure that the label doesn't expand [23:32] evand: it still won't take up the whole width, but at least it will be at the LHS of the window [23:33] I like the new "Add Parent" submenu in glade-3 [23:33] ah! I thought I had tried that but apparently not [23:33] much appreciated! [23:34] and agreed on that submenu. [23:34] for some reason I still have glade-2 installed. I think it's there to remind me of how bad things used to be. [23:34] (alongside glade-3) === superm1 is now known as superm1|away