/srv/irclogs.ubuntu.com/2010/01/21/#ubuntu-installer.txt

=== robbiew is now known as robbiew-afk
=== robbiew-afk is now known as robbiew_
evargh, forgot that xkb-data-i18n hadn't been promoted yet00:21
cjwatsonev: *clicketyclick* fixed for the next publisher run00:28
CIA-41ubiquity: superm1 * r3696 ubiquity/ (scripts/install.py ubiquity/frontend/gtk_ui.py): whitespace cleanup05:52
CIA-41ubiquity: superm1 * r3697 ubiquity/ (debian/changelog scripts/install.py): Remove another reference to a long dead xserver-xorg debconf question.07:38
CIA-41ubiquity: superm1 * r3698 ubiquity/ (debian/changelog scripts/install.py): Don't try reconfiguring LRM, it's been superceded by DKMS based packages.07:40
CIA-41ubiquity: evand * r3699 ubiquity/ (10 files in 6 dirs): Provide an option to determine the keymap from a decision tree.08:49
CIA-41ubiquity: evand * r3700 ubiquity/ (5 files in 3 dirs):09:03
CIA-41ubiquity: * Add missing parameter to exception in bin/ubiquity.09:03
CIA-41ubiquity: * slideshow_get_available_locale is an instance method.09:03
evnote to self, run pychecker before upload09:11
evugh09:11
CIA-41ubiquity: evand * r3701 ubiquity/scripts/install.py: [pychecker] Remove unused exception parameter.09:28
evcjwatson: have you run into issues with bogl-bterm and kvm?  It throws a GPF for me in karmic and lucid.14:17
cjwatsonhaven't noticed it being *that* bad14:17
cjwatsonterminfo seems a bit misconfigured in d-i rescue mode, it's doing something wrong with backspace, but that could be the shell's fault14:18
evweird14:19
evby the way, there's a bug in debconf_ui, introduced by r3665.  self.db will always exist.  I think the solution is to get rid of the check in debconf_ui.debconf_communicator, but you might want to double check that, just to make sure I'm not railroading over a case that you had considered and I had not.14:21
cjwatsonhmm, I'd missed that14:24
cjwatsonI'm not sure removing the check is right though14:24
cjwatsonignore that.  you're quite right, ditch the check14:25
cjwatsonyou mean the hasattr bit?14:25
evyes14:25
cjwatsonhmm14:26
cjwatsonI think maybe we need to save the persistent d-c somewhere else14:26
evwhat about hasattr(self, 'db') and self.db is not None?14:27
cjwatsonthe intended behaviour is:14:28
cjwatson * if DEBIAN_HAS_FRONTEND, only ever instantiate self.db once, and on subsequent debconf_communicator calls just return the previous one14:28
cjwatson * otherwise, as normal14:28
cjwatsonWDYT about http://paste.ubuntu.com/360085/ ?14:29
cjwatsonor indeed the hasattr check might just be irrelevant since debconf_communicator should only ever be called if self.db is None, but I suppose it's a safety check14:31
evit will never hit, as BaseFrontend gets instantiated first14:32
evwhich sets self.db = None14:32
cjwatsonev: ah; then my patch combined with yours14:35
cjwatsonor indeed just 'if self.db is not None'14:35
cjwatsonhttp://paste.ubuntu.com/360089/14:36
evah, good call14:36
CIA-41ubiquity: evand * r3702 ubiquity/ (debian/changelog ubiquity/frontend/debconf_ui.py):14:39
CIA-41ubiquity: The frontend always has a db attribute, per r3665, so revise the check in14:39
CIA-41ubiquity: debconf_communicator. Thanks Colin Watson.14:39
CIA-41ubiquity: evand * r3703 ubiquity/ (debian/changelog ubiquity/components/install.py): Add missing function argument to Install's prepare.16:23
superm1ev, would you mind glancing over https://edge.launchpad.net/~superm1/ubiquity/mythbuntu-plugin-enhance r 3687, 3689 and 3690 to make sure those aren't doing anything to crackful before I continue wandering down that road?  I'm trying to convert as much of the deviations in mythbuntu_ui that are overriding functions (both in gtk_ui and install.py) directly into plugins and coming across some deficiencies16:26
evsuperm1: absolutely.  I'm knee deep in debconf_ui at the moment, but I'll try to find time for it later tonight.  If not tonight, definitely tomorrow.16:27
superm1cool thanks, i won't be touching again until after work anyhow16:28
dmarkeyis there a way to get the LVM partitioner to put the boot partition on partition 1?16:28
superm1note; if things look good, don't merge though - it's in an iffy state, i just want to make sure those particulars are good before i make the other bigger changes to mythbuntu_install.py16:29
evsurely16:29
CIA-41ubiquity: evand * r3704 ubiquity/debian/changelog: LP bug reference.16:42
evcjwatson: Just a heads up, debconf_ui is broken, with the bug surfacing in scripts/install.py.  It seems debconf doesn't like REGION in this context.  I'm looking into it (as I try to gather what I can for packages with debconf support in gtk oem-config), but I'm a bit out of my depth, so it might be a while.19:45
cjwatsonev: if PROGRESS REGION is leaking through to debconf, that's the bug; it's supposed to be handled by the filter.  We probably shouldn't be running install.py unfiltered19:51
CIA-41usb-creator: superm1 * r258 usb-creator/ (debian/changelog usbcreator/frontends/gtk/frontend.py):22:23
CIA-41usb-creator: If hiding the persistence and iso selection in the UI, make sure to do22:23
CIA-41usb-creator: so before window.show() to prevent weird sizing of the window22:23
CIA-41usb-creator: superm1 * r259 usb-creator/usbcreator/frontends/gtk/frontend.py: Leave backend.detect_devices where it was, in case it depended on those previous signal handlers22:27
StevenKcjwatson: Shall I update d-i for -11, or you'll take care of it?22:38
cjwatsonStevenK: if you have time, please go ahead22:39
StevenKcjwatson: Uploaded, thanks.22:56
=== robbiew is now known as robbiew_

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!