CIA-97 | console-setup: cjwatson * r148 ubuntu/ (788 files in 31 dirs): merge from Debian 1.55 | 08:23 |
---|---|---|
cjwatson | superm1: can't help that | 09:17 |
cjwatson | superm1: that's the only syntax supported by current syslinux | 09:17 |
=== persia` is now known as persia | ||
ev | yay! | 11:24 |
ev | progress is slightly broken (starts over between install and plugininstall (could use a rename)), but otherwise the install and plugininstall split works! | 11:25 |
ev | thanks to pychecker and pyflakes for catching lots of bugs that would've otherwise taken up a significant amount of time hunting down. | 11:26 |
CIA-97 | ubiquity: evand * r4161 maverick-redesign/ (7 files in 5 dirs): Split install.py into install.py (file copy) and plugininstall.py (everything else). The former is run with the in-memory debconf database, the latter is run with the normal debconf database. | 12:01 |
CIA-97 | console-setup: cjwatson * r149 ubuntu/debian/ (changelog keyboard-configuration.config): Go back to always asking layout before variant. | 13:17 |
CIA-97 | console-setup: cjwatson * r150 ubuntu/debian/ (changelog keyboard-configuration.config): | 13:21 |
CIA-97 | console-setup: Migrate from old console-setup/* to new keyboard-configuration/* | 13:21 |
CIA-97 | console-setup: question names. | 13:21 |
rgreening | hey ev, you around | 13:51 |
ev | hiya | 13:52 |
rgreening | how goes ev | 13:54 |
ev | pretty well. Ubuntu sprint this week, so covering off as much of the installer redesign as I can | 13:55 |
ev | how are you? | 13:55 |
rgreening | got an interesting bug in usb-creator (backend releated). If you use the --iso to add an image and specify the path in any form other than ./ it fails to add_image | 13:56 |
rgreening | I'm ok. Been busy of late. | 13:56 |
ev | ah, interesting | 13:56 |
ev | can you make sure a bug is filed for it? (I don't have free time to look at usb-creator bugs until at least feature freeze) | 13:57 |
ev | good deal | 13:57 |
rgreening | yeah, line 24 of base/backend.py is where it fails the 'exists' test | 13:57 |
rgreening | Im hoping to figure out why the exists test is failing. could be a python bug | 13:58 |
ev | doubtful | 14:01 |
rgreening | yeah, I think the exists in os.path behaves different than it's programmed. | 14:02 |
rgreening | ev: "On some platforms, this function may return False if permission is not granted to execute os.stat() on the requested file, even if the path physically exists." | 14:03 |
rgreening | from the docs. | 14:03 |
rgreening | ev: yep. sure enuff. setting the +x on the .iso and os.path.exists now detects the file | 14:04 |
rgreening | so, need to make a change or two to the code to deal with this. | 14:04 |
ev | +x sets the executable bit, that's not the same as being able to stat a file | 14:20 |
ev | but yeah, if you can't read a file as the user you're running usb-creator as, then usb-creator quite simply cannot do anything with the file. | 14:21 |
CIA-97 | ubiquity: evand * r4162 maverick-redesign/ (16 files in 6 dirs): Add an installer session panel. | 14:29 |
CIA-97 | usb-creator: rgreening * r314 trunk/ (debian/changelog usbcreator/backends/base/backend.py): Fix bug #608741. Now adding iso via commandline works as expected. | 14:30 |
ubot2 | Launchpad bug 608741 in usb-creator (Ubuntu) "adding image via commandline does not work correctly (affects: 1) (heat: 6)" [Medium,Confirmed] https://launchpad.net/bugs/608741 | 14:30 |
rgreening | ev ^ | 14:30 |
rgreening | ev, the docs for os.path.exists says the file can exist and be readable, but on some systems it will return false if +x not set. Seems to be the exact case. I switched to isfile, which is more accurate and fixes the issue. | 14:32 |
ev | okay, thanks | 14:33 |
rgreening | ev, whne will we release a new version? I'd like ot get this fix into lucid as soon as possible. | 14:34 |
ev | +x doesn't define whether a stat call works, but I see nothing wrong with that code | 14:34 |
ev | you're welcome to do an SRU for it | 14:34 |
rgreening | ok. well, shouldn't we put it in maverick first and then I can do the backport? Or ... | 14:39 |
rgreening | dunno if it's easier that way or not... | 14:40 |
rgreening | but it should def be released in maverick asap.. | 14:40 |
CIA-97 | usb-creator: rgreening * r315 trunk/ (6 files in 5 dirs): Bump version info to 0.2.23 | 14:41 |
rgreening | ev ^ | 14:41 |
ev | rgreening: you can create a ~usb-creator-hackers/usb-creator/lucid branch | 14:41 |
ev | and release from that | 14:41 |
rgreening | I have to run for a couple of hours (meetings). I'll touch back later and followup. | 14:42 |
rgreening | ty | 14:42 |
CIA-97 | ubiquity: evand * r4163 maverick-redesign/ (3 files in 2 dirs): Fix some apt deprecation warnings. | 15:07 |
CIA-97 | ubiquity: evand * r4164 maverick-redesign/ (bin/ubiquity-dm src/panel/panel.c): Actually run the panel in the installer session. | 16:08 |
ev | rgreening: sure thing | 16:12 |
=== cjwatson_ is now known as cjwatson | ||
CIA-97 | ubiquity: evand * r4165 maverick-redesign/scripts/ (install.py plugininstall.py): Wrap plugininstall's run method in a cleanup decorator. | 21:37 |
CIA-97 | ubiquity: evand * r4166 maverick-redesign/ubiquity/frontend/gtk_ui.py: Don't start installing when the user presses back from the partitioning page (whoops). | 21:39 |
CIA-97 | debian-installer: cjwatson * r1334 ubuntu/ (7 files in 2 dirs): Move to 2.6.35-10 kernels. | 22:31 |
CIA-97 | debian-installer: cjwatson * r1335 ubuntu/debian/changelog: releasing version 20100211ubuntu15 | 22:37 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!