[01:05] jtholmes: please read http://help.ubuntu.com/community/ReportingBugs; at present your installer bugs are coming in in a state where we can't really do much with them, but if you followed the advice on that page then we would be able to start work on them without having to go back to you to ask for more logs [01:14] ok will reread reporting bugs [01:15] thanks [01:37] cjwatson, #360587 am at the point of this misinformation in step 7 of 7 in ubuntu installer [01:38] do you want the logs at this point or when the installer completes, which it should complete successfully in about 20 minutes [01:41] jtholmes: either is fine [01:42] ok [02:34] cjwatson: so...I think I have a problem... [02:34] not to be vague or anything... :) [02:34] if you have a moment...take a look at line 54 and subsiquently like 58 of oem-config.py in the main directory of oem-config [02:36] http://bazaar.launchpad.net/~ubuntu-core-dev/oem-config/trunk/annotate/head%3A/oem-config [02:36] does that pairing of if and else make sense to you? or am I missing something...I only ask cause for some reason I am having trouble running oem-config with -f gtk_ui [02:37] could be unrelated...but I just found that and the else seems ambiguous... [03:04] cjwatson: well, that aside...I have pushed again and have also tested this version [03:04] added the cairo_extensions file as we talked about [08:32] morning evand nice break? [08:34] indeed it was, thanks. How was yours? [08:35] evand: Worked most of it :) [09:08] shtylman: the else there is correct at the indentation level it has - for/else is a special (and useful!) construction in python. http://docs.python.org/reference/compound_stmts.html#the-for-statement [09:09] shtylman: what trouble were you having? [09:09] shtylman: thanks, I'll have a look at your branch again [09:43] http://people.ubuntu.com/~evand/tmp/new_timezone_map_london.png - waiting to hear back from kwwii on a missing highlight image, but this is what they're proposing. [09:43] what's the diff like - just replacement images and a cairo tweak for the box corners? [09:44] I think we've missed the boat for RC, but may be able to squeeze it in for final if the diff is minimal [09:50] replacement images, and no changes to the code, though kwwii has some slight changes he wants there (box color, roundness, 80% opacity). [09:50] the full details are in an email I just CCed you on [09:50] ok === dpm_ is now known as dpm [09:54] evand: is Mark in the office? [09:55] Not sure, I didn't go in today (though coordinating going in with Ken might be a good idea) [09:55] evand: also, have you checked whether there's (translated?) documentation featuring the current installer layout that'd need to be updated? [09:55] checking now [09:58] evand, cjwatson: m-a still appears on dvd when doing whole drive install [09:58] evand: he's in the office according to clan - I've asked her to poke him about this [09:58] cjwatson: great, thanks [09:59] davmor2: what version of ubiquity? [10:00] 1.12.8 [10:01] argh, I'll look into it [10:14] perhaps I'm not searching in the right place, but I've checked out the ubuntu-doc bzr branch, I've looked through doc.ubuntu.com, and did a Google Images search on *.ubuntu.com, and the only place I can find pictures of the time zone map used is on the community documentation pages: https://help.ubuntu.com/community/GraphicalInstall [10:15] yeah, those are about the places I'd have looked [10:15] so in that case, you should still notify ubuntu-doc@ and ubuntu-translators@ just in case, but we're probably in the clear for UI freeze purposes [10:16] okay [10:16] will do [10:22] well, I'll save this as a draft until I can generate a screenshot using kwwii's diff (he's testing it some more). [10:27] Mark has signed off, I see [10:28] I should be done in a bit, been having hardware problems this morning...seems my x60s chose today to crap out on me === mpt_ is now known as mpt [10:38] evand,kwwii: how far are you from an upload? [10:38] slangasek was just about to start rolling at least some of the images, but could take this if it's really close [10:39] as soon as I get the patch from kwwii and the missing UTC-9.5 image I can upload [10:40] evand: let me get that image to you first then I can try to finish the py [10:40] ok [10:40] sorry, this system messing up was not how I expected to start my day [10:40] inkscape kept crashing every time I would export a b [10:40] pic [10:42] yikes [10:59] evand: just to confirm both 32bit and 64bit dvd installs have run m-a [11:11] well, I have succesfuly got ubiquity to crash :-) [11:11] Oh? How so? [11:12] changed stuff in the python :) [11:12] ah :) [11:12] I looked into how you do the roundings on the boxes...funky stuff that [11:12] I assumed it would be much simpler :p [11:16] it's taken from the partition bar code, which is taken from banshee [11:20] evand: the last thing I am trying to do is to make the dot appear in the middle of the box...do you know if that is in segmented_bar.py or somewhere else? [11:20] can you elaborate on what you mean by making the dot appear in the middle of the box? [11:21] do you mean have the time zone selection dot vertically align with the time zone selection box? [11:21] the dot that shows the city should be vertically aligned to the box [11:21] right :) [11:24] ahhh, this is coded to move to dot depending on where it appears on the screen, right? [11:24] right, that's in timezone_map, around line 248 [11:24] eh? [11:25] newy is the y coordinate of the top left corner of the time text, and the box is built around it [11:25] if I remember correctly [11:25] * evand pokes [11:27] right, try changing line 248 to this: [11:27] newy = pointy - (ybearing / 2) [11:31] Is there anything else I can help with? [11:32] nope...it is (ybearing + 2) btw [11:32] I will send you a dif in a second [11:35] cjwatson: 42mins for dvd install which is better than intrepid. It's all the stuff it removes that takes forever [11:42] probably the maintainer scripts; need to convert those to triggers [11:56] kwwii: if the goal is for the dot to be vertically aligned with the text in the box, then surely it should be ybearing / 2 and not ybearing + 3 (which would break anyway if we ever changed the size of the text) [11:58] evand: ahhh, right that is probably better [11:58] note that I reduced the font size to 12 [11:59] indeed, those changes look fine [11:59] excellent :) [11:59] does mark need to see another screenshot before this gets uploaded? [11:59] as the alignment of the box to the point has changed [12:00] no, he has given the thumbs up, so it should be final...and the centered dot was in the pic I sent him already :) [12:00] great! [12:00] looks like we are done! [12:00] thanks for the help, sorry for the trouble [12:00] ubiquity: evand * r3214 ubiquity/ (41 files in 3 dirs): [12:00] ubiquity: Update timezone map images. Slight changes to the color and size of the [12:00] ubiquity: time zone time box. [12:00] no worries at all. Thanks a lot for all of your help with this. The final artwork looks fantastic. [12:01] it really does look much better, eh? [12:01] I am happy with it now :) [12:01] absolutely [12:05] ubiquity: evand * r3215 ubiquity/debian/changelog: releasing version 1.12.9 [12:06] evand: Oooh! I guess it won't hit RC? [12:06] StevenK: I believe slangasek has been kind enough to wait for it [12:06] evand: Although I'm impressed with the changes I saw in the 0413.3 daily. Sydney is in the right spot! [12:07] And it is usable on a MID sized screen :-) [12:07] StevenK: thanks. We have shtylman to thank for much of that, as it was his algorithm that we went with in the end. [12:07] re MID> Very good! [12:07] quite glad to hear that [12:18] evand: did you apply the map changes to oem this time? [12:19] davmor2: as I understand it, the changes were not requested for the oem-config map [12:19] evand: How come? Thats the one endusers will see..... [12:20] and there's slightly more involved there than just copying the module containing the map code (it requires a class from the segmented_bar module, which doesn't exist in oem-config, though could be ported) [12:20] davmor2: because these changes are happening so late, we want to make as few of them as possible [12:20] evand: Fair enough [12:26] shtylman produced a branch for oem-config that merged some earlier changes, but I'm still looking over it [12:26] (including copying CairoExtensions across) [12:27] ah, good deal [13:00] evand: erm, you did update the color coded pic as well, right? [13:01] evand: it was in the ones I sent to you...and it had an improvement [13:01] yes [13:01] the coast of Greenland, right? [13:02] yeah, and north of portugal [13:06] indeed [13:12] evand, cjwatson: bug 360844 is this actually a bug? I'm not sure as no partitions are listed as changed [13:12] Launchpad bug 360844 in ubiquity "ubiquity says partition table on sdb changed and it wasnt" [Undecided,New] https://launchpad.net/bugs/360844 [13:15] yes [13:16] jtholmes: nice catch, I only have single drives so I've never seen it :) [13:16] the process of automatically using swap on /dev/sdb is causing parted_server to think that the partition table on /dev/sdb needs to be reread [13:20] yes i thought that may be the case [13:24] there is also another bug or two associated with the side by side process but it is esoteric and when i run into it again i will report it [13:26] that repartioning process is particularly hard to code for [13:59] cjwatson: is there any reason you can think of that you can't use something like http://192.168.0.1/ks.php?type=ubuntu for the kickstart path in Ubuntu? [14:45] cjwatson: is there a historical reason why desktop/ubiquity.png is in the ubiquity package rather than the ubiquity-ubuntu-artwork package? (Bug 322174) My guess would be that the ubiquity logo doesn't have the same trademark policy, even if the ubuntu logo is contained within it, though I don't have a strong understanding of trademark law. [14:45] Launchpad bug 322174 in ubiquity "Ubiquity uses hard-coded references to Ubuntu rather than looking to /etc/lsb-release" [Undecided,New] https://launchpad.net/bugs/322174 [14:53] DogWater: that should work fine - what problems are you seeing? [14:54] evand: it'd probably be ok to put it in the artwork package for karmic [14:54] ok [14:54] thanks [15:34] https://bugs.edge.launchpad.net/ubuntu/+source/ubiquity/+bug/361112 looks potentially serious, though admittedly doesn't prevent installation. I'm looking into why it's occurring, given that debconf_progress_window is modal. [15:34] Launchpad bug 361112 in ubiquity "standalone ubiquity won't let me alt-tab to status window" [Undecided,New] [15:38] cjwatson: it downloads a file called http in the 'fetch' directory [15:40] * cjwatson raises an eyebrow. It should give you a *directory* called http there [15:40] DogWater: could I see the installer's syslog, please? [15:58] cjwatson: duh, i apologize you're right it does create a directory [16:03] cjwatson: ever figure out yet why the installer cannot destroy LVM? [16:07] DogWater: I thought I'd fixed that ...? [16:08] I'll try it again but i believe it still stops and asks if you really want to remove the LVM partition [16:08] even if its in your kickstart/whatnot [16:36] cjwatson: does %pre work in kickstart in Ubuntu or do you use preseed preseed/early_command ? [16:39] %pre should work fine [16:40] argh, %pre runs before dhcp [16:40] (well, before the 2nd dhcp) [16:46] is there any way to run a command after the kickstart file is downloaded and after dhcp comes up? it looks like for whatever reason %pre runs after kickstart but before network [16:47] you could use preseed/early_command for that [16:49] preseed preseed/early_command or just preseed/early_command? [16:49] 'preseed preseed/early_command string blah' [16:50] ah, for whatever reason the same string in there doesn't work that works in %pre [17:30] DogWater: what string is that? [17:48] evand: I don't suppose that bug 289663 might be fixable for final? I'm going to have to carry over some release notes text otherwise [17:49] Error: Could not parse data returned by Launchpad: The read operation timed out (https://launchpad.net/bugs/289663/+text) [18:20] cjwatson: i keep getting the following: http://people.ubuntu.com/~kirkland/Screenshot-QEMU.png [18:20] cjwatson: trying to install to a RAID1 [18:21] cjwatson: this is in kvm, with qcow2 disk images [18:23] kirkland: can I get syslog? [18:23] cjwatson: i'll need to reproduce it again, but when i do so, i'll post [18:24] cjwatson: i'm retesting with dd /dev/zero pre-allocated raw disk images now [18:42] cjwatson: okay, same thing with zero'd, pre-allocated disk images too [18:42] cjwatson: http://people.ubuntu.com/~kirkland/syslog [18:42] cjwatson: http://people.ubuntu.com/~kirkland/partman [18:43] cjwatson: perms, one second ... [18:43] cjwatson: okay, those URLs are ready [18:44] cjwatson: hmm, looks like some OOM errors [18:44] cjwatson: this install only had 256MB of memory, and no swap [19:09] yeah, that seems to be the cause [19:09] base installation died the first time round, and then it retried, at which point /target was indeed unclean [19:19] cjwatson: okay, yeah, i've confirmed [19:19] cjwatson: we used to be able to install the server with just 128MB of ram [19:20] cjwatson: now it requires 128MB of mem + 128MB of swap, or 256MB of mem [19:23] cjwatson: the warning on the page about not having swap should be sufficient [19:23] cjwatson: i'm just too cultured to actually read the warning about the installation failing if there's not enough memory :-) === davmor21 is now known as davmor2 [22:32] evand: hey [22:32] evand: you mentioned once upon a time about kvm's that mysteriously die overnight [22:33] evand: are you still experiencing this?