[08:17] ScottK: the KDE frontend uses the same strings, so it will pick up the changes automatically. [08:19] mind you, the changes weren't raised because of legal concerns. mpt brought it up because "tick this box" is very poor wording, and we wanted to check with legal to be clear on what of the existing text we could remove. [09:44] ubiquity: evand * r4332 trunk/ (3 files in 3 dirs): [09:44] ubiquity: Run the Jockey backend with an in-memory debconf database, and run [09:44] ubiquity: the frontend asynchronously (LP: #627822). [09:44] that was way too easy in the end. Fail. [10:18] Hi all [10:18] I'm trying to make a remix. [10:19] I've been using UCK, and that has been going well. [10:24] I'm planing on carrying around my remix on a 1GB usb memory stick, so I have I bit of extra space to add some additional packages. [10:24] So I've added some localisations, and some extra foss packages such as thunderbird and gimp. [10:24] I want to add ubuntu-restricted-extras, but I want ubiquity to ask the user to agree to this with some notification of the licensing issues. If they don't agree to this, then it should not install it. [10:25] What would be the easiest/best way to archive this? [10:26] When I add packages with UCK, they get installed on the live cd, and then "magically" on the install system. [10:26] magically = I don't know how it happens :-p [10:29] GaryvdM: ubiquity in Ubuntu 10.10 does this already. [10:29] ev: Oh [10:29] I'll have to download the iso and try that... [10:29] ev: thanks. [10:30] mind you, it was broken until this morning [10:30] ubiquity 2.3.18 will have the fix === dmarkey_ is now known as dmarkey [11:55] superm1, cjwatson: can you look this over? http://paste.ubuntu.com/493568/ - I've discussed it with mpt and thoroughly tested it. It greatly improves the transition into the installer session, which sabdfl was hammering us about last cycle with the oem-config issues. [11:55] It also solves the problem of there being no indication to the user that a page is loading unless the mouse cursor is directly over the window. [12:13] looks OK - can we make sure that the root cursor gets set back in the event of a ubiquity crash? [12:14] indeed, it's done in the excepthook [12:14] as well as atexit [12:14] or do you mean even more paranoid than that? [12:14] like having ubiquity-dm do it [12:15] excepthook/atexit should be fine [12:16] okay, cool [12:21] ubiquity: evand * r4333 trunk/ (debian/changelog ubiquity/frontend/gtk_ui.py): [12:21] ubiquity: * Don't show the GTK frontend until it's done constructing the UI. [12:21] ubiquity: * Show a spinning cursor on the X root window in addition to the [12:21] ubiquity: ubiquity window, so it's clear that the installer is loading the [12:21] ubiquity: next page even when the cursor is off the main window. [13:44] ubiquity: cjwatson * r4334 ubiquity/ (debian/changelog gui/gtk/stepUserInfo.ui): [13:44] ubiquity: GTK frontend: Make user-setup entry boxes activate the Forward button [13:44] ubiquity: when Enter is pressed. [13:49] ev: Thanks. [13:49] (both for the answer and for the fix for 627822) [13:53] ubiquity: cjwatson * r4335 ubiquity/ (debian/changelog scripts/plugininstall.py): [13:53] ubiquity: Revert Debug::identcdrom=true for writable USB media; with changes in [13:53] ubiquity: apt 0.8.3ubuntu3, this is sufficient for stable apt-cdrom identification [13:53] ubiquity: of writable USB media (LP: #628290). [13:53] hooray [14:12] Hi - tried to debuild the latest ubiquity from lp:ubiquity, but I got this error: http://paste.ubuntu.com/493635/ [14:13] If I run make manully, it works fine [14:13] *manually [14:13] GaryvdM: run ./debian/rules update first [14:16] Ok - I'm trying that. It's busy downloading a bunch of tar/dsc files. [14:21] heads up - return_to_partitioning is broken with the move to parallel debconfs, on it. [14:22] ev: Thanks - debuild now working. [14:23] sure thing [16:57] ubiquity: superm1 * r4336 ubiquity/ (debian/changelog scripts/install.py scripts/plugininstall.py): [16:58] ubiquity: Fix progressbar reporting messages reporting missing from plugininstall.py [16:58] ubiquity: in ubiquity or oem-config. [16:58] ev, ^. that's an improvement over the current situation, but the plugininstall progress is still starting at 0 rather than 74 for standard install [16:58] ah thanks a bunch! [16:58] this at least gets the messages back, and axes the nastiness from counting plugins in install.py (which was needless with the split) [16:58] I've fixed return_to_partitioning, just testing it on the kde_ui [16:58] ah, noted [17:09] ubiquity: evand * r4337 trunk/ (3 files in 2 dirs): [17:09] ubiquity: Make return_to_partitioning work with parallel debconf instances [17:09] ubiquity: (LP: #637706). [17:12] 2.3.18 is getting quite large [17:13] yeah, it would be nice to see that ffe/uie bug approved to get it uploaded and make sure there is not any regressions coming in [17:14] indeed [17:14] * ev prods -release === cjwatson_ is now known as cjwatson [17:54] libdebian-installer: cjwatson * r211 ubuntu/ (12 files in 5 dirs): merge from Debian 0.76 [18:08] libdebian-installer: cjwatson * r212 ubuntu/debian/changelog: releasing version 0.76ubuntu1 [18:21] ubiquity: evand * r4338 trunk/ (3 files in 3 dirs): Added a Windows icon for the resize widget. Thanks Otto Greenslade! [18:37] ubiquity: evand * r4339 trunk/debian/ubiquity.templates: Fix non-ascii character in templates. [18:46] cjwatson: I saw that you closed the work item on improved subarch, does it mean my patch cut the mustard so to speak? :-) [18:46] I adjusted the hw-detect part of it, but yes [18:46] otavio said it was OK for squeeze [18:46] hw-detect: cjwatson * r144 ubuntu/ (archdetect.c debian/changelog debian/control): [18:46] hw-detect: * Backport from trunk: [18:46] hw-detect: - Add -g/--guess option to archdetect to ask for heuristics that make it [18:46] hw-detect: easier to bring up new boards, based on a patch from Michael [18:46] hw-detect: Casadevall. See [18:46] hw-detect: http://lists.debian.org/debian-boot/2010/08/msg00641.html for [18:46] hw-detect: rationale. Requires libdebian-installer4-dev (>= 0.76). [18:46] I've merged libdebian-installer into maverick, and am just backporting the hw-detect change now [18:48] cjwatson: gah, I was going to rework that patch for you to do the processing later [18:49] hw-detect: cjwatson * r145 ubuntu/debian/changelog: releasing version 1.73ubuntu5 [18:49] well, does what I just backported look ok? [18:49] I just moved the flag processing out of the loop [18:49] cjwatson: if you backported the last patch I sent, it was fine. [18:49] * NCommander looks [18:50] cjwatson: that looks sane. I can't properly test it until later today/tomorrow however. [18:50] and took out a now-unused header [18:50] cjwatson: thanks again. I owe you a beer or equivelent beverage at the next UDS/Rally [18:51] that's ok [18:52] is that everything that needs to go into maverick d-i for this? [18:52] cjwatson: I still need to work out a base-installer patch to use the new archdetect -g option. [18:53] sooner the better please, I want to avoid having to get final freeze exceptions wherever possible [18:53] cjwatson: I'll try and get it to you at the end of this week if possible. [19:10] ubiquity: evand * r4340 trunk/ (4 files in 3 dirs): Disable the alpha warning, albeit late. [19:10] ^ seems like a lot of overhead, but I couldn't think of a clean, faster way of doing that [19:45] ubiquity: evand * r4341 trunk/ (4 files in 4 dirs): [19:45] ubiquity: Add a command line option to not install a bootloader (-b, --no- [19:45] ubiquity: bootloader). [19:46] ubiquity: evand * r4342 trunk/debian/changelog: tidy changelog entry. [22:13] ubiquity: evand * r4343 trunk/ubiquity/frontend/gtk_ui.py: Make sure the set_cursor event gets processed. [23:53] ev, fastest way i could have thought of was bzr mv ubi-warning.py ubi-warning.py.disabled, but that seems functional enough what you did, and hopefully it is just a one line change in the future to flip it on and off