=== asac_ is now known as asac === mrevell is now known as mrevell-lunch [13:56] hello [13:57] good afternoon [13:58] hi [13:58] hello [13:59] Greetings [14:00] * kwwii runs to get a coffee [14:01] oh, yes [14:01] right [14:01] meeting [14:02] *ahem* [14:02] seb128: thanks ;) [14:02] np ;-) [14:02] mvo: there? [14:02] mpt: awake? :) === mrevell-lunch is now known as mrevell [14:05] ok, the others can catch up [14:05] https://wiki.ubuntu.com/DesktopTeam/Meeting/2008-04-03 [14:05] I put together the agenda, pitti had a small bunch of agenda items - did I miss any from anyone else? [14:06] Keybuk: there are a few outstanding action points from the last two meetings [14:06] Keybuk: no but I'll have a quick one to add about the gvfs libarchive backend and whether we want to use it in hardy or not [14:07] pitti: do you have them to hand, I didn't spot one? [14:07] Keybuk: they shuold be in the 'action points' from last one [14:07] Keybuk: yes [14:07] sorry [14:07] https://wiki.ubuntu.com/DesktopTeam/Meeting/2008-03-27 [14:08] * mvo was in deep hacking mode [14:08] ah, nautilus share transition is an important one to discuss IMHO [14:08] oh yes [14:08] sorry [14:08] mvo: napping you mean? ;-) [14:08] * Sebastien to initiate a discussion about shares-admin -> nautilus-share [14:08] upgrade migration with slangasek on u-devel@ [14:08] seb128: how did that discussion go? [14:08] seb128: isn't that the same ;) [14:08] we discussed it on #ubuntu-devel [14:08] no migration possible [14:09] he doesn't think that's an issue and we will write about it in the upgrade notes [14:09] what needs migrating? [14:09] shres-admin does "normal" smb sharing [14:09] adding things to smb.conf [14:09] anything we can do in update-mangaer? [14:09] nautilus-share uses net usershare [14:09] and store things in /var/lib/samba/usershare [14:10] basically slangasek said that migration is not possible [14:10] because there is no way to know which shares have been added using shares-admin or in a other way [14:10] if they have things in smb.conf, will they still work? [14:10] and migrating those to useshare could be a security issues [14:10] because other people will have access to thing they should not [14:10] and there is no way to know which ones should be public or not [14:11] Keybuk: yes [14:11] Keybuk: we just remove the way to graphically edit this list [14:11] so it only really affects people wanting to modify shares they had in the past? [14:11] yes [14:12] can they just install shares-admin? [14:12] not now, I dropped the binary to not get duplication etc [14:12] it's not a separate package [14:12] ah right [14:12] but I think I'll do a separate binary [14:12] maybe we can just drop the .desktop file? [14:12] so was steve happy with that? [14:12] xubuntu users would like that too I think [14:12] or use NotShowIn to hide it? [14:12] pitti: either new binary package or drop nautilus integration and menu entry [14:13] pitti: I suggested that to steve but he seemed to no see the point [14:13] and I would prefer to not have to still maintain this code [14:13] it's written in perl, ugly and upstream stopped maintaining it [14:13] ok [14:13] ah, that thing [14:14] Keybuk: steve said he's happy to drop shares-admin and write something in the hardy notes about the old shares [14:14] seb128: it sounds like you've had generally good feedback from your plans and no significant concerns raised? [14:14] since we can't migrate anyway [14:14] it's either we do that or never change [14:14] and we don't want to never change so better to bite the bullet now [14:14] *nods* [14:14] Keybuk: right, I think everything is on track [14:14] sounds like a no-brainer then :-) [14:15] * Sebastien to prepare theme change back to ubuntulooks [14:15] I think I saw that get done? [14:16] yes, kwwii and mvo took care of the theme changes [14:16] * Sebastien to rename the tracker menu item [14:16] we had some random discussions on the list about menu changes, etc [14:16] and I think we should better wait next cycle to do those cleanup, that's late now to break translations [14:17] ok [14:17] * pitti raises hand [14:17] IMHO we should still fix at least "transmission" [14:17] nobody knows what that means [14:17] right, need to ping jdong about this one again [14:18] they were working on updating or backporting changes [14:18] and dunno why that has not been uploaded yet [14:18] do we need a UDS session for menu changes? [14:18] janebot noticed some of the menu inconsistencies [14:19] pitti: agreed re: "transmission" [14:19] well, I think menus changes discussions will lead nowhere [14:19] * tedg thinks Keybuk will get the room to himself ;) [14:19] we already had those [14:19] that kind of thing should really be done by usability people, or? [14:19] I think we should think about change the menu to use gnome-main-menu or consider using the gnome-control-center shell and improve it [14:19] s/change/changing [14:19] ok [14:19] let's follow up on this later [14:20] * Ken to look into updating important bits of example-contents for Hardy, primarily release name [14:20] mdz: right, we discussed it at previous meeting, there is also an inconsistancy about using or not the software names in the menus entries, that's being discussed upstream too [14:20] Keybuk: haven't gotten around to that yet [14:21] * Keybuk still has nightmares about a street of shops all names "Shoes" [14:21] ok [14:21] * (pitti) libgphoto gvfs: keep and fix, or ditch? [14:22] oh, I added that before I discussed it with Seb, sorry [14:22] I think we pretty much agreed on disabling it, since it's too buggy [14:22] ah, what was the result of the discussion ooi [14:22] what replaces it? [14:22] any objections? [14:22] Keybuk: f-spot-import [14:22] similar to gthumb import from rpevious releases [14:22] ah [14:22] the libgphoto gvfs backend is cool, but it's not ready yet IMHO [14:22] this is the thing where the camera appears accessible to the gnome vfs [14:23] instead of needing to spawn a program to deal with it? [14:23] it appears in places and computer:///, but clicking on photos doesn't work (missing eog gvfs integration), it's not automounted, and if it's mounted, it blocks f-spot [14:23] it does mount your camera [14:23] and you can use it as a standard mount [14:23] ah, and doesn't get the photos into f-spot? [14:23] Keybuk: no, you just get a file browser [14:24] (which might not be what most people want to do with their camera, dunno) [14:24] well, it makes you gphoto camera being seen as a removable disk [14:24] got it [14:24] so, I think it would be great to use in intrepid [14:24] if you're both happy with disabling for now and reviewing later, that's fine [14:24] you can copy directly from the folder if your app use gvfs [14:24] when we'll switch f-spot/gthumb to file-based import [14:24] which is not the case yet [14:24] and 'mount' the camera [14:24] right [14:24] then you can have both [14:24] *shiny* *shiny* :) [14:24] "This folder contains pictures from a camera. [ Import into F-Spot ]" [14:25] but I guess for that it needs a real fuse mount, otherwise you can just see it from gvfs-enabled apps [14:25] I half expected someone to write an LD_PRELOAD wrapper that wrapped open(), creat(), etc. with gvfs calls :p [14:25] that should already work fine if you have gvfs-fuse installed and working. [14:25] but anyway, digressing into dangerous territory ;) [14:26] * (seb128) gvfs libarchive backend, use in hardy? [14:26] there's just not a way yet to make that "apparent" to users. [14:26] is this "open tar files as a folder" ? [14:26] Keybuk: ok, where we stand now is that the current gvfs in hardy has the libarchive backend [14:26] we promoted libarchive for that [14:26] it's not used now but I've the change ready, fedora is using it [14:27] that's basically added a .desktop to nautilus which gives you a "mount archive" context menu option on zip, iso, etc [14:27] s/added/adding [14:27] and the mounted archive is listed as mount then [14:27] you can browse it, copy from it, etc [14:27] what about file-roller then? its no longer needed with that? [14:27] does it work? [14:28] Keybuk: it works fine yes [14:28] I think the "have to mount and then use" user experience is not optimal [14:28] so ideally it would work like file-roller, but look like nautilus? [14:28] yeah, I guess you just want to double-click [14:28] but it's not exposed a lot and easy to use for those who want the feature [14:29] mvo: file-roller is needed because libarchive does only cpio, zip, iso [14:29] Why isn't it the default action? [14:29] zip:///home/martin/foo.zip/path/inside/zip.txt nautilus window would be cool :) [14:29] another thing is that libarchive does reading only [14:29] Keybuk: ^ [14:29] ahh [14:29] which means you can't add files to a zip for example [14:29] any plans for it to expand its format support or be writable? [14:30] and it doesn't support rar, ace, and other formats which are used [14:30] any downsides to activating it? does it break anything else? [14:30] libarchive itself supports writing, but I doubt that it gets support for other formats anytime soon [14:30] Keybuk: Company is speaking about it, but not sure how much work that is and if that's going to happen soon [14:30] Keybuk: no, the gvfs backend is already activated [14:31] ok [14:31] do you think we should activate it? [14:31] that just adds a context menu item and a string to translate [14:31] sounds like its harmless because file-roller is still the default and its just a optional new feature [14:31] the downside is to have to support libarchive for the lts [14:31] will we have to support it for intrepid? [14:31] yes [14:31] then we'll be supporting it anyway [14:32] seb128: your decision :-) [14:32] well, I've difficulties to evaluate how many users are wanting to browse isos [14:32] _o/ [14:32] there was some users on the brainstorm website asking for a such feature [14:33] actually, I hear quite a bit that being able to use .dmg files is a killer mac feature [14:33] yeah, indeed [14:33] so taking a step towards that for us seems good too [14:33] (of course, running programs from inside isos will mean that ld.so will need gvfs support ) [14:33] ok, so my vote would be to activate it, my concern was just adding extra load on the security team for something not really required [14:33] seb128: check with kees, but if it's already in main, it sounds like he's approved it [14:34] yeah, we discussed it on friday [14:34] Keybuk: (note that file-roller browses ISOs just fine, too) [14:34] that was basically "if you really want it, alright" [14:34] :-) [14:34] * (pitti) Coordination of fixing /etc diff for upgrades (Michael started to provide diffs now, thanks a lot!) [14:35] right, so http://people.ubuntu.com/~mvo/automatic-upgrade-testing/kvm/profile/lts-ubuntu/etc.diff is a nice and handy 5.2 MB (!!) diff [14:35] between /etc/ upgraded from dapper and a fresh hardy installation [14:35] I think there are probably many bombs and bugs inside [14:35] it dosn't purge obsolete packages currently though [14:35] so we should at least go through it with a fine comb, and turn errors into bug reports [14:36] mvo: do you think you can regenerate it with purged uninstalled packages? [14:36] although I suppose stuff like the /etc/cron.daily/find bug wouldn't have appeared then [14:36] I will create a run with purge later, but my test-machine is running out of space to do too many of those in parallel (I "only" have a 320GB hdd for this) [14:36] mvo: anyway, I think we just need to check newly added files in /etc/rc*.d/ for now, and can ignore others, right? [14:37] that should make it easier [14:37] I'm willing to work on that and turn it into a bug list, but I might need some more hands to fix all of them [14:37] pitti: is that changes between /etc from dapper to hardy? [14:37] ah yes, you said that [14:37] probably, at least a exlucd list for stuff like /etc/altneratives should be added to the diff generator [14:38] Keybuk: between a dapper->hardy upgrade and a hardy fresh install [14:38] pitti: are those changes really an issue? [14:38] oh, right [14:38] yes, probably [14:38] in the last weeks I fixed a few bugs which e. g. totally broke hal on dapper->hardy upgrade [14:38] missing or wrong transitions in postinsts, etc. [14:38] seb128: some are, /etc/cron.daily/find caused trouble for example with diversion in slocate under some circumstances [14:38] or obsolete configuration files which don't match the new version any more [14:38] mvo: are you planning to review and file bugs? [14:39] so, we don't need to get the diff down to 0, but we need to watch out for real bugs [14:39] Keybuk: yeah, I plan to review them (and did to some extend) [14:39] mvo: shall we attack it together? [14:39] pitti: or are you? [14:39] sure, together is even better [14:39] mvo: we could start at opposite ends of the file and work towards meeting i n the middle or so [14:40] :) [14:40] seems that quite some issues are packages not cleaning conffiles on upgrade when they stop shipping those [14:40] dpkg should really do that automatically [14:40] we have similar diffs for gutsy->hardy and a few other cases, but I guess this is the biggest/worst one [14:40] having to add maintainer scripts to do that is annoying [14:40] seb128: yeah, the amount of copied rm_conffile is really anoying [14:41] :-) [14:41] at least if they weren't modified, yeah [14:41] just be glad we *have* a relatively standard function for it now [14:41] we should ship common snippets like rm_conffile or mv_conffile in some central package so that they can be sourced at least IMHO [14:41] I do see the point for keeping modified ones, but there is *no* reason for unmodified ones [14:41] mvo: indeed [14:41] mvo: I've been arguing for ages that they should be debhelper things [14:41] yeah, that would be fine as well [14:41] since even when people use them, they forget prep_rm_conffile, undo_rm_conffile, finish_rm_conffile in preinst and postrm :) [14:42] pitti: well, usually that's maintainers not knowing they need a snippet or not not been careful about that [14:42] TBH, most of the time I just remove them in postinst [14:42] pitti: then they show up as an obsolete file in dpkg [14:42] what I do too [14:42] erm, preinst I mean [14:42] pitti: then if the unpack fails, the file vanishes forever ;) [14:43] it's just too painful to get it 100% right without introducing other bugs FWIW [14:43] anyway, dpkg sucks [14:43] mvo, pitti: you're both happy to go through that file together and file bugs? [14:43] at least this bit in it :P [14:43] Keybuk: ack [14:43] Keybuk: yeah [14:43] I'm sure I can probably be blamed for some of them [14:44] * (pitti) next week I'll be at the LinuxFoundation collaboration summit in Austin; anything I should do/ask/tell/promote there? [14:44] I will also keep the regular kvm tests running here and report issues [14:44] ^ good to know for the default assignee :-P [14:44] liw was doing some work here too recently, that is pretty cool [14:44] oh, that's not really something to discuss here, just a general question to the team [14:44] ok, if you have things for pitti to do at the LF summit next week, please tell him ;-) [14:44] pitti: beware of "The Tilted Kilt" [14:44] buy me some hardware :P [14:45] * pitti ponders what to bring as a souvenir - a Colt, a barrel of oil, or a cactus [14:45] ok [14:45] * Final call for UDS Agenda topics [14:45] I'll be putting together the draft agenda (ie. list of topics to discuss and a paragraph summary of each) tomorrow [14:45] - rethink the menus? [14:45] Is there a list of UDS agenda topics? [14:45] you've all given me a list, but if you have anything else, please e-mail it me [14:46] tedg: that's what I'll be writing [14:46] pitti, take the colt, thats likely the most exciting at customs :) [14:47] Keybuk: Okay, I don't have any others to add :) [14:48] ok [14:48] http://people.ubuntu.com/~dholbach/sponsoring/ [14:48] mvo: bug #210173 [14:48] Launchpad bug 210173 in transmission "Transmission does not use .desktop translations from Rosetta" [Undecided,New] https://launchpad.net/bugs/210173 [14:49] will take care of it [14:49] Riddell: bug #207863 [14:49] Launchpad bug 207863 in qt4-x11 "assert in qbrush.cpp on LPIA (skype)" [Undecided,Confirmed] https://launchpad.net/bugs/207863 [14:50] hmm, proprietry software on a platform i can't test [14:50] sure, but if it's in the sponsoring queue, it means someone's already done a ptach [14:50] you just need to check that, and upload [14:50] Keybuk: "ptach" is a Klingon name-calling... [14:51] yep, sponsorship queue is on my todo [14:51] pitti: it's slightly worrying that you know that ;) [14:51] pitti, I'm amazed :) [14:51] nuQneH? [14:51] don't tell me you guy's didn't watch Star Trek TNG/Voyager [14:52] sure I did/do [14:52] seb128: bug #199402 looks like it was fixed and got re-opened ? [14:52] Launchpad bug 199402 in metacity "gnome-appearance-properties crashed with SIGSEGV in meta_theme_get_frame_style()" [Medium,New] https://launchpad.net/bugs/199402 [14:52] and waiting for the new StarTrek-movie [14:52] it just takes a special kind of person to learn klingon :-) [14:52] Keybuk: yeah, I just unsubscribed the sponsors [14:52] Keybuk: I don't think the guy who reopened has the same issue [14:53] the bug was a crash, and he has a lack of preview in the capplet [14:53] anyway nothing to sponsor there [14:53] *waits for LP* [14:53] * Keybuk sits down and starts singing about gold [14:53] mvo: bug #104553 [14:53] Launchpad bug 104553 in fontconfig "upgrade failed when fontconfig cache dirs are newer than system date" [Medium,Confirmed] https://launchpad.net/bugs/104553 [14:53] I remember that one came up before? [14:54] I'm not happy with the patch, I would rather want to fix it in fontconfig (and I believe its fixed there already) [14:55] ok [14:55] can you unsubscribe u-m-s [14:55] my bad, I need to remove it from u-m-s [14:55] ooh look one for me [14:55] :-) [14:55] mvo: also new patch on #87914 [14:56] mdz: since you're here ... bug #15051 [14:56] bug #87914 [14:56] Launchpad bug 15051 in pcre3 "grep -P is not supported" [Medium,Triaged] https://launchpad.net/bugs/15051 [14:56] Launchpad bug 87914 in update-manager "Progress bar refers to "1 minutes" or "1 hours"" [Low,Confirmed] https://launchpad.net/bugs/87914 [14:56] dholbach: could you please unsubscribe ubuntu-main-sponsors from #104553 [14:56] dholbach: likewise, please unsub from #94120 [14:56] I can add you both to the team [14:56] I don't want to be a member of the team [14:56] * mvo check 87914 [14:56] grmbl [14:57] dholbach: yeah, please add me [14:57] thanks mvo [14:57] lol [14:57] Keybuk: done (#104553) [14:57] I'm doing everything I can to cut down my bug mail to levels where I can actually read and respond to the incoming bugs [14:58] Keybuk: you're not getting bug mail - it goes to a list :) [14:58] so don't really want to join lp teams [14:58] (also it'll make people think I'm able to do sponsorship requests) [14:59] fine with me [14:59] http://people.ubuntu.com/~ogasawara/qa-hardy-list-archive/sort-by-package/desktop-buglist.html [14:59] doesn't look like there's anything new on there [14:59] or any changes from last time [14:59] pedro_: ? [14:59] no there's nothing new on there [15:00] ok [15:00] AOB? [15:00] leann add the date reported column to it and also if the report is less than 10 days old it will mark it as yellow [15:00] ah, then there are two new things on there [15:00] both for human-theme [15:01] ah, just tasks on bug #99508 [15:01] ok [15:01] Launchpad bug 99508 in compiz "Window titlebar displayed not right with compiz enabled" [Medium,Confirmed] https://launchpad.net/bugs/99508 [15:01] two of the compiz ones require a new upstream snapshot, I guess I need to talk with steve about this [15:01] change of state IIRC [15:01] the "Wrongly placed maximized window with cloned display" ones [15:01] I didn't see any additions to ReportingPage today? nothing to report for the last week? [15:01] nothing fancy from me [15:01] "bug fixing, and more of it" [15:02] upgrade test from gutsy->hardy is running, 20.000 pkgs, VM needs 2.8Gb ram to run smoothly and is 24h later at 15% [15:02] ok, adjourned [15:03] mvo, pitti, seb128: just for information I have several updated packages to be uploaded (human-icon-theme, ubuntu-gdm-themes, human-theme, and ubuntulooks) [15:03] thanks all [15:03] erm, wrong channel [15:03] thanks [15:03] thansk all [15:03] thanks [15:03] thanks [15:04] obrigado [15:07] Keybuk: what about it? [15:07] mdz: it's assigned to you ;) [15:08] you have to review it, and upload it; or unsubscribe the ubuntu-main-sponsors team, etc. [15:08] I'm almost entirely trolling [15:08] Keybuk: it is not [15:08] it has mdz in bold letters in dholbach's list [15:08] it's assigned to nobody [15:08] I just looked [15:09] I think it's too late for Hardy, unfortunately === doko_ is now known as doko