[09:25] ubiquity: evand * r3919 ubiquity/ (4 files in 2 dirs): Removed the alpha warning. Welcome to Ubuntu 10.04 Beta 1. [10:27] ev hallo [10:27] hiya [10:27] mark is happy with this version of the slideshow: http://people.ubuntu.com/~dylanmccall/ubiquity-slideshow-ubuntu/redesign-lucid/ubuntu-transitions/slides/index.html#controls [10:27] so let's get that in :) [10:28] I still need to get a proper ubuntu logo though [10:29] console-setup: cjwatson * r133 ubuntu/debian/ (changelog console-setup.initramfs-top): Fix incorrect FONT and ACM tests in initramfs script. [10:30] that's good, considering it ended up on omgubuntu last night :-/ [10:30] indeed [10:33] michaelforrest: did you get a hold of Marcus for that? [10:33] I just emailed him [10:33] and will now start hassling [10:33] they have been going into super-minute detail on it this week [10:33] so I think I may have to give him a chinese burn [10:34] lol [10:45] ev: what format do I need here? Do you think I can just send a png at the right size? [10:46] svg would be preferable (so that we have the original source in the tree), but png will do if that's all you have [10:53] for those playing along at home, bug 537986 is the freeze exception request for the labelled gtk entries on the user setup page [10:53] Launchpad bug 537986 in ubiquity "Freeze exception request: inactive labels on the user setup page." [Undecided,New] https://launchpad.net/bugs/537986 [11:06] ev: what's happening with foundations-lucid-oem-config? [11:07] cjwatson: I'm going to have to defer the last bit. My time has been focused on dealing with the list of design requested changes to ubiquity. [11:07] which is the last bit? [11:08] (I forget whether the non-ascii passwords stuff happened) [11:08] automatically remove language packs. I'm convinced that there are not problems with non-ascii passwords in GNOME these days. [11:08] didn't Dell request that bit? [11:08] or am I misremembering [11:08] I'm not sure who requested it, actually [11:08] otherwise I'd be inquiring as to why it was raised in the first place [11:09] superm1: were you looking for automatically installing language packs, or automatically removing them? [11:09] (oem-config) [11:09] oh, who requested the auto remove? I think that was Dell [11:09] I know that there *are* problems with non-ascii passwords in GNOME. [11:09] really? [11:09] Specifically for asian languages, where you need an IME. [11:09] can you give me a test case? [11:09] The issues are twofold: [11:09] IME doesn't work in ubiquity [11:10] 1) In GDM, it doesn't capture keystrokes fed to the IME to block timeout for timed logins. [11:10] right, so you can't enter IME-requiring passwords in the first place [11:10] 2) gnome-screensaver doesn't permit unlock. [11:10] I couldn't get ibus to talk to it [11:10] Oh, then you've already solved it differently :) [11:10] the comment in the specification is "non-ascii passwords are allowed in ubiquity, but disallowed in later password prompts like gksudo" [11:11] anyway, I agree that part can be dropped without major issue [11:11] for what it's worth, I did try gksudo and it worked fine with the set of characters I tried [11:11] That doesn't match my experience with it, but if ubiquity doesn't support an IME, it doesn't matter. Non-IME non-ascii seems to work OK. [11:11] for removing language packs, I'm inclined to say defer to beta-2 and I'll offer to do it? [11:11] if we defer it to beta 2, I'm happy to do it [11:11] Note that this is a regression from jaunty, where ubiquity *did* support the IME. [11:11] I don't want to dump things on you [11:12] what goes around comes around :) [11:12] hahaha [11:12] I probably deserve it, but whatever [11:12] I don't recall you ever dumping anything on me [11:13] persia: I'm not very familiar with IME. Did ibus exist in jaunty and did it communicate over dbus? [11:14] used to be scim, which probably wasn't dbus [11:14] ev: jaunty was scim, which used the GTK IM extensions. [11:14] ah, so I suspect that's why then [11:14] Indeed. [11:15] someday, when someone is crazy enough to rewrite ubiquity to talk to its frontend over dbus, all of this will go away [11:15] I'll give it a fiddle at some point, and maybe point some of the IME devs at it, but not likely in time for lucid. [11:15] and we'll all ride unicorns off into the sunset [11:15] cool [11:15] That will indeed be a great day :) [11:18] Oh, I should ask, rather than flailing more, is there a good way to get per-locale reconfiguration done after selecting language during install? [11:18] I would have thought to do it in casper, but with the new interface, I'm no longer sure this is the right answer. === _ruben_ is now known as _ruben [11:20] is what you're trying to do specific to a particular project, or is it general enough to be carried in Ubuntu proper? If the latter, you could modify the end of bin/ubiquity-dm (or I suppose we could add a hook for post-greeter steps) [11:20] It's bug #531155 [11:20] Launchpad bug 531155 in casper "gedit schemas is not registered during live session" [Undecided,New] https://launchpad.net/bugs/531155 [11:21] This ends up propagating into new installs too, forcing dpkg-reconfigure or use of the alternate installer. [11:21] I suspect there's a few other packages that have locale-specific logic in postinst. [11:21] (but gedit is the one that people complained about to me) [11:22] Does that belong in ubiquity-dm ? [11:25] definitely not ubiquity-dm then if it's affecting installs [11:25] I would imagine casper still [11:25] oh [11:26] no, that's obviously wrong [11:26] ignore me [11:26] Do we know the locale in casper still? I thought we didn't know it anymore because of the new gfxboot [11:26] indeed, we don't [11:26] I just didn't think that through enough before I started talking [11:26] hrm [11:26] I asked the reporter to file against casper because it would have been casper in karmic (and I hadn't looked at the lucid stuff enough at the time) [11:27] But I'm now convinced it's the wrong place. [11:27] I'm also convinced it's not a bug in gedit, because it's hard to detect encodings, so we want a narrow set for each locale. [11:28] (this will annoy folks who use one locale with documents from a different locale, but that's not a well supported use case right now anyway) [11:29] I'd say it's a bug in gedit for depending on the system locale rather than the user locale? [11:30] (if I'm understanding correctly - I'm just going on the conversation above, haven't checked the bug) [11:30] cjwatson: Essentially, gedit sets it's import filter lists at install time. I can see that argument, and can certainly go reopen the bug seb closed previously :) [11:30] (which was actually about cross-locale usage, and so somewhat different) [11:31] But I like the interpretation of it being a gedit bug. [11:33] michaelforrest: do you recall what we agreed to call the "Zone:" label now that the "Region:" label is gone? - http://people.canonical.com/~evand/screenshots/ubiquity/9.10-beta-candidate/2-timezone.png [11:44] Yah "Time Zone" [11:57] ah, duh [11:57] thanks [12:32] console-setup: cjwatson * r134 ubuntu/ (5 files in 2 dirs): [12:32] console-setup: Move boot tasks to a combination of two udev rules and a single Upstart [12:32] console-setup: job, ensuring that they're run at points when we are able to satisfy the [12:32] console-setup: constraints on the relevant ioctls. This may not work optimally with [12:32] console-setup: vgacon, but you have to go to special effort to use that with Ubuntu [12:32] console-setup: now, and it at least shouldn't break (LP: #524439). [12:51] bah, it's not possible to do something like strftime %X, but without seconds, is it? [12:58] ev: difficult. no, there doesn't seem to be [12:58] the separator isn't always : either [12:58] indeed [12:59] so looks like we're stuck with it unless we start providing a translation string for time format [12:59] but I think that's a bit much [13:00] for instance, as far as I can tell, in Chinese, you have something like 12时59分58秒 [13:00] how does the panel clock applet do it? [13:00] it has a "Show seconds" checkbox [13:00] maybe it's just locale-incorrect ... [13:00] does it? I have radio buttons for 12 hour and 24 hour time, despite being in en_GB.utf8 [13:01] hrm, I'll dig through its source though [13:01] maybe there's a trick to this [13:01] or maybe gtk already does the gettext thing [13:02] I have it set to not show seconds to avoid it waking up every second and sucking power [13:03] heh [13:06] #. Translators: This is a strftime format string. [13:06] #. * It is used to display the time in 12-hours format [13:06] #. * (eg, like in the US: 8:10 am). The %p expands to [13:06] #. * am/pm. [13:06] #. Translators: This is a strftime format string. [13:06] #. * It is used to display the time in 12-hours format (eg, like [13:06] #. * in the US: 8:10 am). The %p expands to am/pm. [13:06] #. [13:06] #: ../applets/clock/calendar-window.c:303 ../applets/clock/clock.c:445 [13:06] #: ../applets/clock/clock-location-tile.c:516 [13:06] msgid "%l:%M %p" [13:06] erm actually, not quite what we want [13:06] we want appropriate-hours rather than 12/24-hours [13:06] I think [13:07] indeed [13:07] you might be able to get that from nl_langinfo though? [13:34] I don't think so, at least not that I can see. My thought is to provide ubiquity/timezone_strftime_format, translated to %l:%M %p for en_US, and %k:%M for en_GB. [13:40] granted, that means we can't simply import translations from gnome-panel [13:40] hrm [13:42] though we could use this with the aforementioned strings from gnome-panel: http://paste.ubuntu.com/394018/ [13:54] * persia idly notes that there are locales that use neither 12-hour nor 24-hour time by convention (well outside the scope of this issue) [13:56] persia: who uses the binary clock by default? ;) [13:57] davmor2: Not that, but that here it is conventional to use numbers > 24 to denote times of day. [13:57] For instance, the trains run from about 4:00 to about 25:30 [13:58] (well, rather 4寺00分 to 25寺30分 === robbiew_ is now known as robbew === robbew is now known as robbiew [15:26] cjwatson, its actually both, installing them if they are available and selected, and a key to remove extras that were already installed and not needed [15:28] ev: ^- in that case I think we need to keep this on the list but move to beta-2 [15:28] cjwatson: okay [15:28] which indeed is what I did [16:07] ev: is your additional fix for bug 529366 likely to land for beta-1? and do you suspect a parted bug in and around all this? [16:07] Launchpad bug 529366 in usb-creator "Regression: usb-creator-gtk doesn't work as of 0.2.16" [High,Triaged] https://launchpad.net/bugs/529366 [16:07] probably not beta 1, moving the goalpost to beta 2... [16:08] and no, I don't think it's really a parted bug beyond the fact that the behavior changed in 2.1 [16:08] unless it's expected that you can tell the kernel to re-read the partition table with mounted partitions [16:09] err s/tell/get/ [16:09] you used to be able to do so partially. it was a bad hack at best. [16:09] maybe this is why Dell has been having trouble, since I know they were relying on that bad hack for recovery installs [16:09] anyway, separate bug about that ... [16:10] it's certainly best for any software we control to avoid relying on this [16:12] indeed [16:13] I never thought removing seconds from a clock would require a 2,900 line patch [16:13] (in fairness that's 99% translation data) [16:25] cjwatson: is there a place other than the list of languages in localechooser/languagelist I should be looking at when checking for sufficient support before adding new imported translations from GTK+? There seems to be quite a few this time around. [16:26] languagelist is canonical [16:26] if we don't have base installer translations, I haven't been importing the corresponding GTK+ translations [16:27] okay, thanks [18:06] installation-guide: cjwatson * r463 ubuntu/ (build/entities/common.ent debian/changelog): Bump release version and names for Lucid (LP: #501572). [18:08] installation-guide: cjwatson * r464 ubuntu/ (build/entities/common.ent debian/changelog): Bump kernelversion to 2.6.32. [18:09] installation-guide: cjwatson * r465 ubuntu/ (debian/changelog en/welcome/what-is-ubuntu.xml): Bump GNOME version to 2.30. [18:10] installation-guide: cjwatson * r466 ubuntu/ (build/entities/common.ent debian/changelog): Bump x11ver to 7.5. [18:12] installation-guide: cjwatson * r467 ubuntu/ (debian/changelog en/install-methods/automatic-install.xml): [18:12] installation-guide: Remove comments about automatic installation methods that have been [18:12] installation-guide: removed from Ubuntu (LP: #496554). [18:13] installation-guide: cjwatson * r468 ubuntu/ (debian/changelog en/appendix/preseed.xml): [18:13] installation-guide: Remove documentation of [18:13] installation-guide: base-installer/kernel/linux/initramfs-generators; setting this to [18:13] installation-guide: anything other than the default will cause the installer to fail [18:13] installation-guide: (LP: #415469). [18:17] installation-guide: cjwatson * r469 ubuntu/ (4 files in 2 dirs): Stop building for lpia. [18:22] installation-guide: cjwatson * r470 ubuntu/debian/changelog: sync up previous changelog [18:27] installation-guide: cjwatson * r471 ubuntu/debian/changelog: releasing version 20081208ubuntu6 [18:37] is this the place to ask about turnkey linux? [18:44] no, I think you'd want a turnkey channel, if there is one (I don't know) [18:45] well, googling freenode and turnkey brought me here, sort of [18:45] per this google hit: http://osdir.com/ml/linux.ubuntu.server/2008-09/msg00079.html [18:46] and a scan of the first page of hits didn't indicate any #turnkey-linux :-( [18:49] sorry, I don't think we can help you [18:49] perhaps there is a mailing list [18:49] perhaps [18:49] let me take a quick stab, though [18:50] turnkey linux does not have emacs on it by default...its ubuntu 8.0.4. I don't want to foul up my TKL, but I gotta have my emacs. What's the recommended apt repository? [18:51] I'm sorry, no idea [18:52] I wouldn't want to try to recommend something based on Ubuntu in case I screw it up for you. Please find somebody who knows about Turnkey ... [18:52] alright. I was hoping it was a generic enough question. maybe a different debian room? [18:53] no, you'll need something specifically turnkey. There may simply be no appropriate IRC channel, and you'll have to try e-mail [18:53] apt repositories are often project-specific [18:53] not necessarily always, but often, and I don't know whether that's the case here [18:54] feh. ok. yeah, there is a archive.turnkeylinux.org according to my sources.list (I think that's it..it's from memory yesterday) [18:55] but ther's also archive.ubuntu.com/unbuntu hardy [18:55] I wonder why it's not finding it there [19:16] i figured it out [22:29] console-setup: cjwatson * r135 ubuntu/debian/ (changelog control): [22:29] console-setup: Depend on kbd (>= 1.15-1ubuntu3) for a valuable loadkeys improvement. [22:29] console-setup: This was previously a recommends, but the udev rule won't work without [22:29] console-setup: kbd or console-tools anyway, I'm not really interested in supporting a [22:29] console-setup: system without either installed, and console-tools is thoroughly [22:29] console-setup: obsolete now. [22:30] console-setup: cjwatson * r136 ubuntu/debian/console-setup.console-setup.upstart: clarification of loadkeys behaviour [23:51] ubiquity: cjwatson * r3920 ubiquity/ (debian/changelog ubiquity/components/ubi-language.py): Fix label retranslation when clicking on update link (LP: #531864).