[00:09] evand: hmm, I tried the new zoommap changes out [00:10] evand: the problem I have now is that, when you try to move the mouse towards your home region, it tends to zoom to the edge of the map where you entered, typically some polar or ocean region [00:11] evand: I don't recognise polar regions very well, and the effect of this is that I suddenly have no idea where I am [00:11] evand: so I'm not sure this has helped, unfortunately :-( [00:11] evand: I wonder if a delay before zooming would help? Or maybe just go back to requiring a click before zooming? [00:12] huh, actually it seems to zoom to the top-left edge, or maybe the last place you zoomed to, or something; very oddd [00:12] odd [01:06] another problem is that the edges are small enough that it's easy to exit the zoom area by mistake, which I can then understand would require the zoom area to return to the last place you zoomed to [01:22] OK, I think this bit is uncontroversial [01:22] ubiquity: cjwatson * r2654 ubiquity/ (debian/changelog ubiquity/zoommap.py): * Allow diagonal scrolling in timezone map. [01:23] evand: the rest of what I've been playing with is http://paste.ubuntu.com/6948/ [01:24] evand: it stops it from instantly zooming in, and instead waits half a second, which seems to make it a bit easier to get to about the right area of the map before it zooms [01:25] evand: it also arranges for the zoomed position to be where you were on entry (or wherever you got the mouse to before the timeout fired) [01:25] evand: I widened the hover-to-move areas to 50, which felt a bit less fiddly to me, especially given that it now loses your zoom position when you move the mouse out of the map (the main downside of this change) [01:26] evand: but I left the step size at 20, since 50 was far too jumpy [01:26] evand: this feels better to me, but I haven't compared it with either the old map or the old-old map; please try it out? [01:28] (obviously this is on top of r2654 above) [01:45] evand: actually, the other thing that would make it a lot better would be if it only zoomed in by a factor of two; as it is, it zooms in such that the displayed image is exactly the size of the source image, which comes out as around 3x zoom, which is too far [01:46] cjwatson: man, you're there at a late hour. might you be up for a quick installer question :) [01:47] no, sorry [01:47] I need to focus on this so that I can get to bed [01:47] feel free to e-mail me though [01:48] cjwatson: sure thing [02:35] evand: http://paste.ubuntu.com/6953/ is what I have now. I adjusted the zoom (4.5x seems OK, though probably not quite perfect; to me, zooming out by much makes the points too close together, while zooming in too much makes it hard to see context) [02:37] evand: and I changed the hotspot selection to find the closest hotspot if there are several nearby, which allows vastly increasing the hotspot size, which makes it much easier to land on points if your coordination isn't perfect (this fixes a regression relative to the old map implementation, IMO) [02:37] evand: (actually, hotspot sizes are pretty much unnecessary with that, but never mind) === foka__ is now known as foka_ === cjwatson_ is now known as cjwatson [11:39] ubiquity: cjwatson * r2655 ubiquity/debian/ (79 files in 2 dirs): * Update translations from Launchpad. [11:43] ubiquity: cjwatson * r2656 ubiquity/debian/ (changelog po/de.po): [11:43] ubiquity: * Fix typo in German welcome message (LP: #195075). This needs to be fixed [11:43] ubiquity: in Launchpad or it *will* regress, but this will at least paper over it [11:43] ubiquity: for 8.04. [11:45] console-setup: cjwatson * r65 ubuntu/debian/ (changelog config.proto): [11:45] console-setup: * Honour OVERRIDE_ALLOW_PRESEEDING environment variable when deciding [11:45] console-setup: whether to allow preseeding, which is ordinarily only allowed in d-i [11:45] console-setup: (LP: #188492). [11:47] ubiquity: cjwatson * r2657 ubiquity/ (debian/changelog ubiquity/components/console_setup.py): [11:47] ubiquity: * Set OVERRIDE_ALLOW_PRESEEDING=1 while calling console-setup to force it [11:47] ubiquity: to allow preseeding (LP: #188492). [11:50] console-setup: cjwatson * r66 ubuntu/debian/changelog: releasing version 1.21ubuntu7 [13:25] evand: let me know when you're up - would like a phone call for plan-of-attack [13:50] debian-installer: cjwatson * r908 ubuntu/ (26 files in 3 dirs): [13:50] debian-installer: * Remove documentation of "Check CD for defects"; even in gfxboot we don't [13:50] debian-installer: have enough screen space for it. [13:53] debian-installer: cjwatson * r909 ubuntu/ (11 files in 3 dirs): * Resynchronise build/boot/x86/po4a.cfg with our real set of translations. [13:55] cjwatson: hi, reading scrollback [14:17] hi evand [14:17] hi xivulon, reading through your email [14:17] I see cjwatson already took care of 4, thanks!!! [14:19] 3 is the only point that might be worth discussing [14:20] basically I removed stack use for tooltips/network detection [14:20] also some calls were wrong [14:20] eg Advapi32::RegQueryValueExA(i, t, i, i, i, i) i [14:20] in registry_reader.nsh [14:20] 4th and 6th argumenst should be pointers [14:21] ok, do these still require the new nsis? [14:21] I used the new nsis in wubi-installer.org build [14:21] ok [14:22] as usual I cannot guarantee the fixes are enough, to avoid segfaults, but should be an improvement [14:22] how many reports of it (new nsis and your code changes) fixing the problem did you get? Just the one on Ubuntuforums, or were there others? [14:22] +1 [14:22] but no reports of segfaults so far [14:22] on 2000+ downloads [14:23] ok [14:23] test it first on XP since I only tested on vista [14:24] I'd like some feedback from bdmurray too [14:24] when he is up [14:24] I'll test it myself once I'm done with the tzmap stuff, but provided that goes OK, I'm happy seeing it merged and a new version uploaded to rookery. I imagine we should let slangasek know whats up though. [14:24] But he's on PST, so it will be a few more hours. [14:24] ok [14:25] rev 486 has a fix on rev 483 so whatever happens that should be in [14:25] feel free to merge the 2 branches [14:26] on the website I have 487 but did not push the code, that basically uses early_command to add "set -x" to autopartition-loop [14:26] to help debugging point #1 [14:26] that is why did not upload the code [14:49] cjwatson: if you'd like to have that call, I believe I am prepared for it now. Please call my VOIP extension though, as it wont be until wednesday until my replacement phone is working. [14:51] debian-installer: cjwatson * r910 ubuntu/ (33 files in 2 dirs): * Update help text translations from Launchpad. [14:51] added my thoughts to slangasek's proposal on bug 215347 [14:51] Launchpad bug 215347 in ubiquity "freeze exception: clear_partitions warning." [Undecided,New] https://launchpad.net/bugs/215347 [14:53] cjwatson: your changes to the tzmap widget had not occurred to me when I was fixing it but are nevertheless fantastic. [15:09] evand when you test wubi, can you please uninstall selecting "backup iso" and check that the iso ends up in ubuntu-backup (217126)? [15:09] sure [15:18] evand: I tried changing ubiquity code to run all bits as root, i.e the window manager, gnome-settings daemon, jat-spi-registryd, orca etc, and orca still zombified. Hell I even tried putting in code to kill pulseaudio, and while that worked, orca still acted the same. [15:18] evand: I'm wondering whether moving the accessibility code to be earlier than say the window manager will work? I'll try that tomorrow in any case. [15:21] TheMuso: I'll try to work through it tonight. I believe today is my last day to do my taxes, so I have to take care of that after my core hours, but hopefully it wont take too long and I can give the code a look. [15:21] (assuming of course that I don't find the time this afternoon) [15:26] evand: No worries, as I said I'll have another dig. [16:49] debian-installer: cjwatson * r911 ubuntu/debian/changelog: releasing version 20070308ubuntu39 [17:53] ubiquity: evand * r2658 ubiquity/ (debian/changelog ubiquity/zoommap.py): (log message trimmed) [17:53] ubiquity: * Various fixes to the new tzmap. Thanks Colin Watson. [17:53] ubiquity: - Wait half a second before zooming in. [17:53] ubiquity: - Zoom in from the position of the mouse, rather than the top left [17:53] ubiquity: corner. [17:53] ubiquity: - Widen the hover-to-move areas. [17:53] ubiquity: - Find the closest hotspot if there are several nearby. [17:54] evand, please don't forget to port those tzmap fixes to oem-config when you're finished up with touching them in ubiquity [17:55] mario_limonciell: thank you! I forgot to write that down in my TODO list. [17:56] thanks :) [17:57] ugh, FTTP cannot come soon enough. Comcast is terribly slow today. [17:58] i've been considering switching to ATT uverse myself in my area, but it appears a majority of the BW is restricted to TV/phone rather than normal intarweb [17:59] lame. They were outside running lines for FiOS yesterday, so we're anxiously waiting for that as I've heard nothing but good things about the QoS. [18:00] cjwatson is there anything on my side to do for 8497?. [18:01] shall I use an additional menu item with edd=on? [18:01] xivulon: if you like [18:02] but TBH I'd leave everything alone now unless you absolutely have to change it [18:02] this is not really the time for opportunistic changes ... [18:02] I have some instructions on the Wubi guide I guess I'll stick to that [18:03] I might add that after final for the stand-alone then [18:15] Have added bug #217348 as a reminder for post-release [18:15] Launchpad bug 217348 in wubi "Add edd=on boot option to address bug 8497" [Wishlist,Confirmed] https://launchpad.net/bugs/217348 [18:17] definitely shouldn't be added by default of course [18:18] if we thought that was a good idea we'd make it the default in the kernel :) [18:18] absolutely that is a menu that is shown when the user presses "esc" at boot [18:19] will make it clear [18:21] partman-target: evand * r717 ubuntu/ (5 files in 2 dirs): [18:21] partman-target: * Backed out check.d/12partitions_formatted. Moved the ubiquity version [18:21] partman-target: into partman-target. [18:31] ubiquity: evand * r2659 ubiquity/ (3 files in 2 dirs): * Removed 12system_partitions_formatted. It now lives in partman-target. [18:35] evand, cjwatson's revno 2657 should take care of keyboard preseeding should it not? [18:36] that was the idea, yes [18:37] awesome. that's the last bug that I've been tracking, so if nothing else shows up, all good for us :) [18:37] excellent [18:38] oem-config: evand * r447 oem-config/ (debian/changelog lib/zoommap.py): * Port tzmap usability fixes from ubiquity. === mebrown is now known as notmyrealname === notmyrealname is now known as different_anonym === different_anonym is now known as mebrown [21:17] grub-installer: cjwatson * r731 ubuntu/ (debian/changelog grub-installer): [21:17] grub-installer: * Backport from trunk: [21:17] grub-installer: - Run grub in the chroot for password encryption. [22:08] evand have uploaded new wubi / umenu with branch merge and po sync [22:10] xivulon: thanks! [22:11] will do some more tests later [22:12] bdmurray if you are around, would you mind testing http://wubi-installer.org/devel/minefield/Wubi-8.04-beta-rev487.exe? [22:12] (without ? in url) [22:13] xivulon: testing how? [22:14] mostly the windows side [22:14] did you experience any segfault in the past? [22:15] can you also check if the bar jam is addressed (didn't do anything aimed, but maybe other changes did help) [22:43] evand, based upon the number of fixes going into ubiquity trunk, are you planning one more ubiquity release pre-rc? If so, were you planning it against trunk, or a temporary RC fork? [23:25] evand, for your info we will need a fix to 216161 [23:26] the issue is that wubi is not creating virtual disks in vfat