[06:58] Good morning [07:26] i hope this is the right place to discuss about the changes to notify osd in karmic ie dx-karmic-notify-osd [07:28] in that there is a Consistent feedback for multimedia keys item [07:29] does this mean i will get a notification for every time i click play, forward, next on my keyboard. if so then is it not a abuse of the notification system. it really does not notify me of anything as i am the one who clicked and hence this would be useless [07:30] the correct way would be that banshee upon clicking play puts a notification of the song being played <--- and this is a bug on banshee bugzilla which is being/has been fixed and now if we give notification for play key. then each time i hit play there will be two notifications one for play key and one from banshee [07:31] am i wrong in interpreting this blueprint? [08:04] robert_ancell: why are bug 195737 and bug 392368 two separate bugs with two contradicting proposals? [08:04] Launchpad bug 195737 in compiz-fusion-plugins-main "Set "open" animation to glide 2, not glide 1" [Wishlist,In progress] https://launchpad.net/bugs/195737 [08:04] Launchpad bug 392368 in compiz-fusion-plugins-main "Set default window open/close animation to fade" [Wishlist,Triaged] https://launchpad.net/bugs/392368 [08:04] robert_ancell: I'm about to sponsor the first one, but the second one confuses me [08:06] good morning pitti [08:07] hey didrocks [08:08] pitti, the second one is not for sponsorship. There was a lot of comments in the former bug about using fade over glide. I opened the latter so the former can be closed without complaints [08:09] robert_ancell: so 195737 is "invalid"? (that's also an identified paper cut bug) [08:09] but why didn't you just update the bug title? [08:11] No, 195737 is valid. There was a request to change both animations to glide to be consistent which was agreed by everyone that this was better than the previous case (mixed animations). The papercuts project agreed. [08:12] As part of the discussion though people were suggesting that we shouldn't use glide at all, but instead use fade. This has had less discussion (there are almost certainly performance effects of using fade) [08:13] I took the position we should change from glide1+glide2 to glide2 for both and we may in the future change to fade if that is better (needs more discussion though) [08:13] robert_ancell: ok, could you please make 392368 a duplicate of 195737 and update the bug status/title? [08:14] I'm happy to sponsor this, but right now I have no idea what to change [08:14] so far I prepared a glide1 -> glide2 update, but then I noticed the other bug [08:14] pitti, they are not duplicates. 195737 is essentially "be consistent" and 392368 is "fade is smoother than glide" [08:15] but one says "set open animation to glide-2", the other "set open animation to fade" [08:15] they really shouldn't both be open [08:15] one says "open animation" the other says "open/close animations" [08:16] still, same contradiction for open [08:16] "Set "open" animation to glide 2, not glide 1" = "Set open animation to be the same as close animation (i.e. glide 2)" [08:17] "Set default window open/close animation to fade" = "Change window animations from glide to fade" [08:17] These are probably better description, I will update them if that makes more sense [08:18] thanks [08:19] pitti, updated [08:20] yop [08:20] asac: bug 280214 has a simple patch; I can't commit to ~network-manager, though, can you please apply it? [08:20] Launchpad bug 280214 in network-manager "Typo in /usr/share/doc/network-manager/README.Debian" [Low,Confirmed] https://launchpad.net/bugs/280214 [08:29] robert_ancell: xscreensaver> nice, does that mean that the new version is actually smaller than 5.07? [08:30] pitti, I didn't check size but it dropped a number of screensavers so I expect so! [08:30] \o/ [08:31] pitti, :) Is there an easy way to query the installed size from the .deb? [08:31] robert_ancell: dpkg -I .deb [08:31] robert_ancell: but .deb size is more interesting, though [08:31] 2160 now [08:31] (compressed) [08:31] 718k [08:31] pitti, also xscreensaver has been targeted by the papercuts proj so it will almost certainly be split (with only ~5 left in the default install) [08:32] nice [08:32] robert_ancell: 718k is -data ? [08:32] 448k data [08:33] 2M for -gl [08:33] ok, so it shrunk a little [08:52] good morning everyone [08:52] seb128 - i figured out the gnome-screensaver upgrade issue last night [08:53] hey chrisccoulson, hey seb128 [08:53] chrisccoulson: wow! [08:54] hi pitti [08:55] the screensaver issue is quite an easy fix. basically what is happening, is the old jaunty gnome-screensaver process tries to spawn the new karmic gnome-screensaver-dialog when you try to unlock the screen, and it passes it a deprecated option, causing it to fail to start [08:55] what's supposed to launch devicekit-power? [08:56] it doesn't seem to start up automatically =\ [08:56] the fix is basically http://bazaar.launchpad.net/~chrisccoulson/gnome-screensaver/upgrade-fix/revision/32#debian/patches/06_unbreak_upgrade_from_jaunty.patch [08:58] hyperair: d-bus activated [08:58] hyperair: if you have -1ubuntu1, you need to dist-upgrade (that was brokeN) [08:58] aha! [08:58] okay then =p [08:59] chrisccoulson: great work! [09:04] pitti - if you think the patch looks correct, i'll push it to the ubuntu-desktop branch:) [09:05] chrisccoulson: it makes sense, anyway [09:05] chrisccoulson: I guess you tested it :) [09:05] i did, and it seems to work ok [09:05] thanks [09:50] re [09:50] hello chrisccoulson pitti [09:50] nice for gnome-screensave [09:50] hi seb128 === baptiste_ is now known as baptiste [09:56] pitti: committed. thanks [09:56] asac: danke [09:58] pitti: http://pastebin.com/f1f5027de [09:58] pitti: withotut that i cannot link against libgudev [09:59] seems the .la files dont get the /usr/lib/ dir [09:59] asac: ah, indeed [10:00] asac: that doesn't break the .so -> so.0 symlink? [10:00] NM builds properly with it [10:00] let me check [10:00] -rw-r--r-- 1 root root 1022 2009-06-28 23:55 /usr/lib/libgudev-1.0.la [10:00] lrwxrwxrwx 1 root root 21 2009-06-28 23:59 /usr/lib/libgudev-1.0.so -> libgudev-1.0.so.0.0.0 [10:00] lrwxrwxrwx 1 root root 21 2009-06-28 23:59 /usr/lib/libgudev-1.0.so.0 -> libgudev-1.0.so.0.0.0 [10:00] -rw-r--r-- 1 root root 25800 2009-06-28 23:55 /usr/lib/libgudev-1.0.so.0.0.0 [10:01] asac: I built devkit-disks, g-d-u etc. against it just fine [10:01] asac: ah right, I thought libgudev was in /lib [10:01] no its not [10:01] you installed them manually in usr/lib, even though make install puts them in lib/ [10:02] right, then libdir makes sense [10:02] pitti: so what i could do is build NM ... however, building applet against NM would fail [10:02] $ grep libdir /usr/lib/libgudev-1.0.la [10:02] libdir='/lib' [10:02] seems the pkg-config files of NM would get busted this way [10:02] ah-ha [10:03] asac: ok, makes sense; want to upload yourself, since you have it ready? [10:03] pitti: yes. [10:03] pitti: i guess i shall commit it [10:03] please do [10:04] asac: I check udev in the meantime (when Keybuk uploads this, gudev will be built by udev source, and u-extra goes away) [10:04] pitti: ok. [10:04] asac: right, fixed in udev bzr head [10:05] pitti: 20090615+1-4 is still UNRELEASED. forgot to push? [10:06] asac: no, forgot debcommit -r; fixed and pushed, sorry [10:06] thx [10:06] asac: thanks for fixing this [10:11] pitti: is that a native branch? [10:11] asac: no orig.tar.gz, yes [10:12] pitti: the version confuses bzr builddeb then [10:12] i think [10:12] at least [10:12] it has a -4 (pkg revision) [10:12] it's not bd'ed [10:12] just debuild -S [10:13] yeah. bzr in theory bzr builddeb should work with all legal branches ;) [10:16] pitti: even my hand import of udev is failing at the moment [10:17] pitti: i might be dense or the previous upload was dirty; just debuild -S in the branch gives me this debdiff: http://pastebin.com/f62b82257 [10:18] asac: ah, you don't use -i? [10:18] asac: anyway, I can do the upload for you, might be quicker [10:18] * pitti does [10:19] pitti: yeah. but even then the Makefile.in stuff et all .... also consider to use a real native version (e.g. without -1) ... that would not confuse all the tools ;) [10:19] pitti: thanks [10:19] asac: uploaded [10:19] pitti: its all committed [10:19] cool [10:19] asac: debian/README.source has some bits, but it's pretty mad for this [10:19] since you manually have to generate the autobreak bits :/ [10:20] pitti: hmm. put it in debian/rules ;) [10:20] anyway. thanks. i will try to keep out of that ;) [10:20] well, the entire package will disappear in a matter of days :) [10:20] good ;) [10:30] pitti: am guessing that kay has rebased this tree in some way :-/ [10:31] yay rebase [10:32] ah, no [10:32] just a non-trunk merge [10:32] bzr fast-import is not very good [10:32] if someone does their own development, merges from trunk, then pushes [10:33] sometimes it can't work it out [10:38] of course, this now means I have to push --overwrite and upset everyone ;) [10:38] I'll remember to pull --overwrite :) [10:40] * didrocks prepares to be upset :) [10:54] pitti: uploaded GIT head [10:54] let's see if it builds ;) [10:54] Keybuk: thanks! [11:19] mpt, did any more action happen on this? http://live.gnome.org/Nautilus/ProgressWindow [11:19] andreasn, not that I know of [11:20] I wonder if it's easier to implement in baby steps. Right now the thing that hurts my eyes the most with that dialog is the lack of a word in the cancel button, so I'm looking into fixing that by hand [11:40] ok, I've cleaned my emails backlog from the weekend, took me some hours [11:46] seb128: \o/ [11:47] Keybuk: seems udev didn't actually get accepted? [11:48] yeah, dunno why ;) [11:49] think I forgot the tarball [11:49] * Keybuk uploads again [11:58] Keybuk: hm, some *.install file is broken, it seems (scsi -> ata renaming?) [11:58] pitti: yeah was just reading that [11:58] I think Kay dropped scsi_id.config ;) [12:02] * Keybuk tries again [12:10] oh arse, missed the udeb === MacSlow is now known as MacSlow|lunch === mvo__ is now known as mvo [12:56] hey mvo [12:57] hey seb128 [12:57] mvo, did you have a chance to review my update-notifier changes? ;-) [12:58] seb128: not yet, let me do it now [12:59] mvo, no hurry I was just curious don't bother if you are busy with other things === WelshDragon is now known as YDdraigGoch [13:13] seb128: merged [13:13] seb128: I upload in a bit [13:14] re [13:14] mvo, I was saying that I pushed a minor change [13:15] seb128: merged [13:15] dunno if you read that or wrote something [13:15] seb128: I upload in a bit [13:15] uploaded [13:15] yes, I saw the change [13:15] ok cool [13:15] * seb128_ hugs mvo === MacSlow|lunch is now known as MacSlow [13:39] Keybuk: I updated ubuntu-meta to not pull in udev-extras any more [14:06] udev-extras killed from archive, too [14:07] cool === SiDi_ is now known as SiDi === asac__ is now known as asac [15:29] pitti: hi. so can I try automagic python build on Jaunty? [15:31] rickspencer3: https://edge.launchpad.net/~pitti/+archive/apport-retracer has backports for hardy, intrepid, and jaunty [15:32] pitti: so should we go ahead and implement $quickly package? [15:32] rickspencer3: I did a ton of fixes and improvements over the weekend, so the upstream build/install part works wonderfully now [15:32] apport and jockey are fully converted [15:32] ! [15:32] rickspencer3: well, it doesn't build a debian/ dir for you yet [15:32] I can try it on bug-zapper later today [15:32] what does it do? [15:33] rickspencer3: "./setup.py build" and "./setup.py install" now DTRT [15:33] they know how to generate POTFILES.in, extract translations, where *.ui and D-BUS files need to go, all that [15:33] pitti: can we just write a few lines of code in package.py to kick of building the deb? [15:34] wow [15:36] rickspencer3: for a general template, yes; figuring out build/binary dependencies is the tricky part [15:45] mvo, glatzor_: How many top-level application categories are there for a .desktop file? Ubuntu has ~6 out of the box (Accessories, Games, Graphics, Internet, Office, Sound & Video), but how many others are there? [15:46] mpt: http://standards.freedesktop.org/menu-spec/latest/apa.html should have them [15:47] mpt: well, all of them :) [15:47] ah, great, thank you [15:49] mvo, is there an easy way at the moment of browsing the Ubuntu applications that are in those "Additional Categories"? [15:55] mpt: not from the top of my head, maybe seb128 knows? the information is available in /usr/share/app-install/data [16:39] kenvandine: for quite a while, most of the time the u1 file sync is disconnected for me; do you have this as well? is it intended? [16:46] rickspencer3, didrocks: https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-karmic-quickly is still in drafting, can this get ready this week? [16:46] pitti: yes ... sorry, thought it was done [16:48] rickspencer3: FYI, we are down to 4 specs in drafting === WelshDragon is now known as YDraigGoch === YDraigGoch is now known as YDdraigGoch [17:58] pitti, rickspencer3: I modified and integrated pitti's remark a week ago. [17:59] didrocks: oh, can you please set it back to 'review' then, when you are done with editing? [17:59] pitti: I just first wanted that rickspencer3 agrees to my changes. That's why I didn't set it to "review" yet :) [18:00] pitti: can you please send me or pastebin an example of setup.py file for AutomagicPythonBuildSystem so that I can work on it for quickly? [18:01] didrocks: the one on https://wiki.ubuntu.com/DesktopTeam/Specs/Karmic/AutomagicPythonBuildSystem is actually still relevant [18:01] pitti: perfect, thanks :) [18:01] didrocks: the real jockey setup.py looks slightly more complicated: [18:01] http://bazaar.launchpad.net/%7Ejockey-hackers/jockey/trunk/annotate/head%3A/setup.py [18:02] didrocks: but that's just to be nice to other upstreams who package it (distutils version check), you don't need that [18:02] and two special cases in the source layout (which are uncommon for your class of apps as well) [18:03] pitti: ok. I will just try to figure out the changes (distutils version check ;)) just for my self understanding :) [18:42] good night everyone [18:46] o/ === hsitter is now known as apachelogger === rickspencer3 is now known as rickspencer3-afk === rickspencer3-afk is now known as rickspencer3 [22:17] rickspencer3: I think you really should bzr pull now :) [22:17] * rickspencer3 pulls === jono_ is now known as jono [22:17] rickspencer3: I added support for executing any command wherever you are in the project tree [22:17] didrocks: that is soooo awesome! [22:18] rickspencer3: the good news is that it's transparent for templates command ;) You just have to act as if you are always in the directory root path :) [22:19] so, I didn't have the time to work on release command. I will do it tomorrow [22:19] time to go to bed. Have a good evening! [22:20] sweet! [22:20] it seems to work very well [22:20] g'night! [22:21] great ;) [22:21] thanks, you too! [22:23] 'night didrocks [22:30] hi seb128_ - you happy with the gnome-screensaver patch i mentioned earlier? i was going to push it to bzr before i go to sleep [22:31] chrisccoulson, where is it? I've not seen the code change but what you described made sense [22:31] urg, wrong click [22:32] seb128 - the patch is here: http://bazaar.launchpad.net/~chrisccoulson/gnome-screensaver/upgrade-fix/revision/32#debian/patches/06_unbreak_upgrade_from_jaunty.patch [22:33] looks correct yes [22:33] cool, i'll push that change now [22:33] can you push to bzr I will upload [22:33] thanks === rickspencer3 is now known as rickspencer3-afk [22:34] I have no luck getting this hardy pidgin patch to work [22:34] something wrong with ssl [22:34] did you look at what other distros do? [22:36] couldn't find it easily [22:36] i'll see if we can at least backport 2.5.7 [22:38] seb128 - i've pushed the gnome-screensaver change in to ubuntu-desktop branch now [22:43] just curious but how did you debug this one? [22:46] chrisccoulson, did you send the change to GNOME? [22:46] not yet. i can attach the patch to the upstream bug report though [22:46] uploaded [22:46] thanks:) [22:46] I see that you commented, you can as well add the patch too [22:47] thank you for the work on that, I was not sure how to debug it but I know that could be annoying issue for users ;-) [22:47] btw did you apply for motu now? ;-) [22:47] yeah, i was expecting it to take a while to debug, as i don't know the gnome-screensaver code all that well, but it only took 30 minutes [22:48] i havent got around to applying just yet. i keep getting side-tracked with other work at the moment [22:48] how did you debug it? [22:48] i need to set aside some time this week really;) [22:48] there was something interesting in the log? or stracing? [22:48] i just ran gnome-screensaver --debug and watched the console output [22:48] debug logs can be useful ;-) [22:48] the error messages from gnome-screensaver-dialog get piped back to gnome-screensaver and then printed on the console [22:49] nice and easy:) [22:49] if you don't hurry for motu you might be after the archive reorganization and have no motu team to join ;-) [22:49] i'll try and sort it this week then;) [22:50] cool [22:50] ok, enough work for me for today, good night everybody [22:50] see you tomorrow [22:51] good night seb128 [22:52] night seb [22:57] seb128 rickspencer3-afk pitti: any idea who could handle a bug like this: https://bugs.edge.launchpad.net/hundredpapercuts/+bug/389212 [22:57] Ubuntu bug 389212 in hundredpapercuts "Partitions icons on desktop and file access" [Undecided,Triaged] [22:58] I want to get some more info on it, and maybe get someone looking into the complexity of the change. [22:59] djsiegel, I would say pitti or me, the description seems misleading though [22:59] just changed it :) [22:59] djsiegel, what is an internal partition for you? [22:59] https://bugs.edge.launchpad.net/hundredpapercuts/+bug/389212 [22:59] Ubuntu bug 389212 in hundredpapercuts "Partitions icons on desktop and file access" [Undecided,Triaged] [22:59] seb128: oh, you saw the new description [23:00] an internal partition is at least a partition on the save physical disk as your main partition [23:00] djsiegel, we don't show icons for things automatically mounted at boot on the desktop, that's a policy we have since ubuntu has been created [23:00] same* [23:00] seb128: right, I agree [23:00] djsiegel, we don't display those on the desktop right now [23:00] I am just being explicit [23:00] seb128: yes, I know [23:00] what we display is ie your vista partition if you mounted it [23:00] which you disagree with? [23:00] yes, I get it... [23:00] no [23:00] ahh [23:01] so I don't understand the bug [23:01] I am saying, internal partitions should mount at boot [23:01] sounds a complicate discussion [23:01] djsiegel - that should be for the adminstrator to decide by specfying them in fstab [23:01] shouldnt it? [23:01] we explicitly don't do that [23:01] well, we need to discuss it [23:01] otherwise you quickly hit the "n mount counter" [23:01] which means long fsck on boot, etc [23:01] seems like it's something administrators wouldn't rely on as a default [23:02] we decide to mount partition "on demand" right now [23:02] and many users have multiple partitions and are confused by fstab [23:02] that avoid all the trouble about accessing something you don't use [23:02] there is a bug requesting to have an property option to select what partitions to mount on boot [23:02] seb128: yes, I see that one [23:03] I think that would be the right way [23:03] but that's not a papercut, too much work for that [23:03] right [23:03] djsiegel: no, spare internal partitions should absolutely not be automounted at boot by default; this does happen on the live CD, but that's a bug (I'm on it) [23:03] Well, I have nominated this one from a user-centered design perspective. [23:03] pitti: why? [23:03] pitti: I am here to find out why [23:03] pitti: I am representing the user only here [23:03] djsiegel: especially not if they are windows partitions [23:04] users are confused and want access to their data [23:04] they might have a hibernation image, or be unclean, and then mounting them could wreak havoc [23:04] djsiegel, for one reason you would trigger the disk check for those partitions you don't use every n boot which take a while [23:04] they can mount their windows partition in the places menu [23:04] hmm [23:04] djsiegel: also, we can't automount them anyway since that requires admin powers, thus we need to ask for the users' password [23:04] pitti: do we block them from mounting their parition if it has a hibernation image? [23:05] and "their" data is on /home [23:05] pitti, (not really true, the system could mount everything on boot) [23:05] djsiegel: no, we don't, but if you explicitly mount it, you can have an UI which points out problems, etc. [23:05] seb128: right, but we don't want that [23:05] pitti: really? [23:06] Well, I am still convinced we need to explore this further, technical problems withstanding. [23:06] I hear "absolutely not! no way! we don't do it that way!", but if it still creates a bad user experience, I am still interested [23:06] djsiegel, why do you want to mount everything on boot? [23:06] sounds like it's not a paper cut though [23:06] the computer location icons make it transparent [23:06] this was discussed several times already, actually, and IIRC the main blockers for automounting were potential data integrity breakage [23:06] pidgin hardy-backports seems to work at least [23:06] users don't have to know if things are mounted or not, they just have to use those [23:07] the system will do the mounting in an almost transparent way if required [23:07] seb128: bug reports from users who have multiple partitions indicating problems with those partitions not being mounted [23:07] trouble making them automount [23:07] etc. [23:07] let's fix those bugs? [23:07] djsiegel: "multiple partitions" in what sense? [23:07] this is one way to fix them [23:07] djsiegel: everything which has the Ubuntu system is set up by the installer [23:07] I can guaranty you you will get higher number of bugs the other way around [23:07] pitti: people with a media partition, or a windows partition, or an HFS+ partition in my case [23:07] they share music or photos, desktop wallpapers [23:08] the right way is to let you select what partitions to auto-mount [23:08] machine boots, does not mount the partition, and things misbehave [23:08] djsiegel: it would be nice to make it configurable to automount some partition by default indeed [23:08] not to auto-mount 15 partitions you don't use at every boot [23:08] seb128: do users have 15 partitions? [23:08] but I don't think we should greedily mount everything we can [23:08] triggering disk checks for those, etc [23:08] pitti: I agree [23:08] pitti: that's what I want to find out about this bug [23:08] djsiegel, it's a slight exageration but I've 7-8 of those on my old desktop install [23:09] I definitely think we should not do it if users have above a certain number [23:09] like 3 [23:09] 2 linux install with user directory, user,etc , one xp etc etc [23:09] djsiegel: so perhaps we can rename it to "offer possibility to automount a particular partition"? [23:09] pitti: so, that needs discussion [23:09] djsiegel, why would you want to automount the system partition of an another linux install? [23:09] pitti: what kind of users have extra partitions? [23:09] will they know what automount is? [23:10] where to look for it? [23:10] I am assuming very basic users who cannot figure this out [23:10] but maybe users with multiple partitions are usually advanced? [23:10] djsiegel: I don't think that "very basic" users will care at all [23:10] but a second windows partition is probably pretty common [23:10] very basis users click on the icon in nautilus and the mounting is transparent [23:10] pitti: well, below a certain threshold, yes :) [23:10] second/third linux partitions less so -> those are the power users [23:10] I don't see the issue with that [23:11] they don't even know if was not mounted [23:11] so, I used to use Mac OS [23:11] but we avoid all the technical issue [23:11] well, they get asked for their pwd [23:11] and it doesn't have this concept of present-but-not-mounted in the Finder [23:11] internal partitions are just there [23:11] just like "Macintosh HD" [23:11] your main partition [23:11] well do you know that partitions are not mounted in nautilus? [23:11] yes [23:11] how? [23:11] what's your point? [23:12] let's make those look like they are mounted [23:12] users don't need to know they are not if that happens transparently [23:12] well, part of the problem is not explicit user action [23:12] so you don't trigger unrequired mounts at boot and users are not confused [23:12] it's importing music or videos or photos from another partition [23:12] or setting a desktop wallpaper [23:12] after reboot, this content seems missing [23:13] well any application accessing one of those devices should mount it if it's not [23:13] I agree, we definitely should have automount control in something easier to use than ftab [23:13] like that other bug says [23:13] +1 from me [23:13] but I am also wondering if there is any way to make an intelligent guess for the user [23:13] there might be but that's not a papercut bug [23:14] If Ubuntu guesses I want a certain partition and mounts it at boot, that is a huge ease-of-use win [23:14] well, what about a simple heuristic? [23:14] we need to think carefully about all the bugs you will hit [23:14] like, your Windows partition? [23:14] like disk check every n mount [23:14] ah, but the hibernation problem [23:14] the hibernation issue [23:14] does mounting a hibernated disk mess it up necessarily? [23:15] you don't want to get your 300gig vista disk blocking your boot for an hour if you don't use it [23:15] not sure about the hibernation thing [23:15] that's for sure [23:15] well, we should definitely not fsck partitions we don't own by default [23:16]