=== emgent is now known as bakunin === bakunin is now known as emgent [04:08] where does ubuntu get the logos on the main site ( http://www.ubuntu.com/ ) (e.g. the blue question mark on the left, or the cd's in the download section...) [04:16] Know tell me if my modem trustmd4050 serves to force a spitter ... because does not connect to ubuntu: ( [04:16] splitter* [04:27] TheMuso: ping [04:44] Amaranth: have you looked at james_w's patch in bug #118936, by chance? [04:44] Launchpad bug 118936 in alacarte "Alacarte does not recover deleted menu items" [High,Triaged] https://launchpad.net/bugs/118936 [04:46] yeah, that was a goofy bug on my part [04:46] but it only makes undo work, not revert [04:47] i have no idea how to make revert work [04:47] well, should we be trying to get the fix for "undo" in at least? [04:48] yeah [04:49] ... and is this patch a good one for it? :) [04:50] it looks like it, haven't tried it [04:50] will do so tomorrow [04:50] ok, thanks [04:53] slangasek, about bug #191027, i can't reproduce it on my hardy machine [04:53] Launchpad bug 191027 in totem ""Failed to connect stream: Invalid argument"" [High,Invalid] https://launchpad.net/bugs/191027 [04:54] nenolod: well, nor mine [04:54] however, i saw something else on pulse 0.9.10 which seems related [04:54] i'll try to reproduce that [04:58] slangasek, anyway, "Invalid argument" says to me that it's trying to connect to a dead socket [04:59] which I have seen before [04:59] follow sent :) [04:59] followup* === imbrando1 is now known as imbrandon [05:08] nenolod: ok, cheers [05:12] i made a scary amount of typos there [05:18] slangasek: You called? [05:19] TheMuso: yes, was hoping you could have a look at bug #191027 [05:19] Launchpad bug 191027 in totem ""Failed to connect stream: Invalid argument"" [High,Invalid] https://launchpad.net/bugs/191027 [05:19] slangasek: I'll be uploaded moblin-{applets,media} in the next hour or so if I can bug you to accept them? [05:20] Sigh. [05:20] I can type, some days [05:20] slangasek: I see its marked invalid. Is there still an issue? [05:20] StevenK: you should get it approved by motu-release so I don't have to think about it :) [05:21] StevenK: looks like i can accept them, too [05:21] slangasek: It's all in bug 181150 [05:21] Launchpad bug 181150 in moko "[FFe] moko contains a shared library with no soname/dev package handling" [Critical,Confirmed] https://launchpad.net/bugs/181150 [05:21] *and* am part of motu release [05:21] oh, meh, that's new package [05:21] That's done [05:21] slangasek: can deal with them [05:21] TheMuso: it's "invalid" for totem, "new" for pulseaudio [05:21] Hobbsee: thanks :) [05:21] slangasek: Oh ok. [05:21] This is part #2, and #3, getting the rdepends to use it [05:21] TheMuso: someone should've reassigned it to pulseaudio much, much earlier :) [05:22] slangasek: Right. Will have a look. [05:22] * StevenK wonders if he has free reign to upload the other two packages [05:33] morning guys [05:38] * TheMuso sighs. [05:38] The trouble with things like this is when you have working hardwrae, this sort of thing is hard to track down by asking the users to do things for you... [05:38] well, yes :) [05:39] I was hoping you might be more adept at it than I given this particular piece of software though :) [05:47] Right, for some of it at least, I think its to do with alsa, but I'm not 100% sure yet., [05:50] slangasek: The problem with setting up alsa apps to use pulseaudio in alsa-lib, is doing it in a way tha tdoesn't affect those desktops not using pulse, i.e KDE and XFCE. [05:51] mm, ok [05:51] yes, trickier then [05:51] And while all applications can use the default alsa device, not all apps can be set to use a custom asoundrc config, let alone choose another card to use. [05:52] Due to either their UIs, configs, or alsa support implementations. [05:52] yes [05:52] Added to that, we'd have to seed libasound2-plugins. [05:53] it may be to better effect if you were to say this in the bug log - you don't need to convince me that it's a hairy proposition, I've had enough trouble just trying to get ekiga to work with my bluetooth headset ;) [05:54] Yeah I will. Just letting you know where things stand. [05:56] remorning fabbione [05:59] hey Hobbsee === Caesar_ is now known as Caesar [06:17] In python, is it possible to wait forever, until say a file exists? As in wait for an undefined length of time. The file will exist, but I just have to wait for it to do so. [06:23] nvm worked it out === asac_ is now known as asac [07:20] good morning [07:31] TheMuso: did you check out the envyng-core update? [07:33] <\sh> moins [07:36] <\sh> pitti, when you have time, could you review bug #211326 IMHO it's good when we would go with this version for hardy, regarding sec issues in future etc. [07:36] Launchpad bug 211326 in rsync "[Freeze Exception] Please update rsync to 3.0.x for hardy" [Wishlist,New] https://launchpad.net/bugs/211326 [07:41] dholbach: It seems alright with me, but my python is not exactly strong. [07:41] dholbach: iS THERE A BUG ASSOCIATED WITH IT? [07:42] gah capslock [07:43] TheMuso: a few - hang on [07:44] bug 216303 (for one), looking for the others not referenced from the changelog [07:44] Launchpad bug 216303 in envy "EnvyNG crashes on latest hardy" [Medium,Fix committed] https://launchpad.net/bugs/216303 [07:44] ok will have a look. [07:45] bug 216009 and bug 216184 too [07:45] Launchpad bug 216009 in envyng-core "pulse.py crashed with KeyError in process()" [Medium,Triaged] https://launchpad.net/bugs/216009 [07:45] Launchpad bug 216184 in envyng-core "pulse.py crashed with KeyError in process()" [Undecided,In progress] https://launchpad.net/bugs/216184 [07:45] bug 216967 too [07:45] Launchpad bug 216967 in envyng-core "duplicate deb lines in envyng.list" [Undecided,In progress] https://launchpad.net/bugs/216967 [07:45] so 3 of 4 open envyng-core bugs :) [07:52] dholbach: If you are looking for a MOTU release ACK, then you have it. if you want me to officially do it in a bug, which one? [07:52] TheMuso: no, should be enough to let the archive admins know that you've ACKed the envyng-core upload [07:52] pitti, slangasek: ^ [07:58] dholbach, TheMuso: accepted [07:59] * dholbach hugs TheMuso and slangasek [08:01] cjwatson, evand: in an oem install is there suppoed to be a "oem-config-prepare icon"? http://wiki.ubuntu.com/Testing/Cases/OEMInstall says so in point 5 [08:08] cjwatson, evand: nevermind - my mistake [08:08] * calc hopes 3ubuntu5 is the last upload for OOo for hardy release :) [08:33] I guess it's too late for fixing grammatical mistake in the system-config-printer dialogue? [08:34] yes [08:34] string changes break translations [08:35] that's why I'm asking :) === seb128_ is now known as seb128 [08:38] hello, I could use some help with distcc here, if anyone is available this late :) === doko_ is now known as doko [08:41] asac: are there any firefox2 based browsers in hardy, do we need to register plugins for those? [08:42] there's firefox-2, if that counts as "based" :) [08:59] cjwatson: is there any way to embed environment variables in ~/.ssh/config ? I'd like to set GSSAPIAuthentication depending on my host name and location. [09:00] (and would avoid rewriting that file on login, if possible..) [09:08] mvo: morning [09:08] mvo: just FF3 on eng upgrade do you want the log files uploading to the bug I opened for comparison? [09:13] Good morning [09:14] <\sh> moins pitti [09:14] why does the encrypted lvm installation take ages to complete in the mke2fs step? [09:16] dholbach: Is it zeroing the partition/drive? [09:16] TheMuso: I created a new disk image using qemu-img before [09:16] dholbach: Right. [09:19] what exactly do I have to do when I have a debdiff for a package in main that propably needs ACKing/sponsoring? (bug 195508) [09:19] Launchpad bug 195508 in system-config-printer "applet.py crashed with AttributeError in check_for_jobs()" [Undecided,Triaged] https://launchpad.net/bugs/195508 [09:20] afflux: subscribe ubuntu-release and ubuntu-main-sponsors [09:21] does timing matter? ie. both at the same time or -sponsors after I got the ack? [09:21] davmor2: thanks for confirming this, I think its fine, just a note should be enough [09:23] zul: nagios is still problematic; can you please fix the seeds to not have nagios-plugins? (since you kept the strict dependency to -extra) [09:25] mvo: I was asked to look at bug 91814, and since lowering the priority of the question doesn't help, cjwatson suggested that perhaps it should be preseeded in the dist-upgrader. Do you have any thoughts/other suggestions? [09:25] Launchpad bug 91814 in openssl "libssl0.9.8 config asking me 'which services should be restarted to make them use the new lbraries?'" [Medium,Confirmed] https://launchpad.net/bugs/91814 [09:26] hrm, why is lowering the question priority not enough? [09:26] slangasek: I don't know at this point... was critical, I tried high, but made no difference. [09:27] mvo: no problems [09:27] TheMuso: high is the level we display, medium will not be displayed. let me have a quick look at the report [09:27] TheMuso: the default behavior is to show debconf questions of priority high or greater [09:28] dholbach: it doesn't know you zeroed it - but there should be a "Cancel" button on the progress bar to skip the erasing [09:28] Right. cjwatson also thought that if one was using apt-get on the console to upgrade, it might be a question that should be asked. [09:28] that's fair [09:28] cjwatson: ahhhhh, I thought that's completely cancel the installation [09:29] the bug was all "if I'm doing a complete OS upgrade and going to be shutting stuff down anyway" [09:29] cjwatson: that's very nice then [09:29] dholbach: nah, there's a bug about the UI I think [09:29] at the moment the text of that particular button is installer-wide [09:29] so it's a bit of work to improve ... [09:30] Mithrandir: I don't think so, sorry :( you can certainly set it in individual Host blocks, but that might not be good enough [09:30] cjwatson: if it's planned to be fixed at some stage - that's cool [09:30] TheMuso: I will add some information to the bugreport, we had a similar issue with libc6 [09:31] mvo: Ok. [09:31] cjwatson: yeah, though what I want to do is have it in revision control and having different configs for that value (and possibly others) depending on what machine I'm on. [09:31] cjwatson: and I'm trying to avoid using branches, since branches = work. [09:34] TheMuso: updated [09:36] mvo: Ok will look into that further, thanks. [09:38] dfsfs [09:39] hey !! any 1 wonna chat a litle [09:40] Xcfdj: try #ubuntu-offtopic [09:41] mvo: I new the upgrade machine now so are you sure you don't need any logs before I wipe it? [09:41] s/new/need [09:42] davmor2: hm, yeah, attach them to be on the safe side :) [09:42] mvo: np's [09:43] thanks! [09:43] so guys , I've got some stupid old machine that runs on ubuntu . and the 3d effectes are not anabled . so any one knows what are the minimal requarments for anabling the effects ? [09:46] Xcfdj: you should join the #ubuntu channel if you need support === Tonio__ is now known as Tonio_ [09:55] <\sh> if anyone is testing vmware + ubuntu-server (or desktop) please test the guided partioning step...I just had grub failing to install with this...manual partitioning with separated /boot partition works [09:57] \sh: I would appreciate the logs from your installation [09:58] (I haven't used vmware for a while, and my installation has almost certainly bitrotted given how much maintenance it needed when I was using it actively) [09:59] <\sh> cjwatson, I'll redo the installation later on, and provide it in a bug report...I'll ping you with the bug number then :) [09:59] \sh: also, what image were you installing? [09:59] <\sh> cjwatson, ubuntu-server from cdimages [10:00] <\sh> the last daily which is available up to now... [10:00] ok [10:08] hi all [10:08] does ls sort by unicode on hardy now [10:08] +? [10:08] it's very odd having dotnames in the middle of the ls output [10:08] even folders appear now in the middle [10:09] mitsuhiko: "sort by Unicode" is not meaningful; the sort order is defined by the value of LC_COLLATE in your locale, which for most languages causes dots to be ignored for sorting, yes [10:09] if you don't like it, export LC_COLLATE=C [10:10] * slangasek nods [10:10] hmm [10:11] that would affect applications too [10:12] anything that uses glibc's idea of alphabetical sorting, yes [10:30] Hmm, the last two days my system has been in a state where firefox would fail to open and clicking system->shut down wouldn't bring up a shutdown window (while System was highlighted in orange)... [10:31] Doing this also locks up the entire top bar really [10:32] doko: could you please have a look at bug #215839 ? [10:32] Launchpad bug 215839 in python-central "Hardy upgrage gedit update fails and crashes X" [High,Confirmed] https://launchpad.net/bugs/215839 [10:34] mvo: ok, trying to do so today === tkamppeter_ is now known as tkamppeter [10:47] hi tkamppeter [10:47] tkamppeter: what are foomatic-db-gutenprint and ijsgutenprint? They do not have any reverse dependencies any more, and want to go to universe [11:04] Is http://people.ubuntu.com/~ubuntu-archive/queue/hardy/new/ the correct place to check whether a particular upload is waiting for approval? [11:04] I'm looking for mobile-basic-flash 0.43 [11:04] lool: no, that would be 'unapproved', not 'new' [11:04] Thanks [11:10] lool: You want https://launchpad.net/ubuntu/hardy/+queue === evalles_ is now known as effie_jayx [11:16] * seb128_ lunch too [11:16] I'll be back in an hour [11:18] Fujitsu: Cool, thanks [11:20] Hi all, is anyone looking into the 'missing binaries from bluez-utils' bug? [11:25] slytherin: hadn't heard of it, what's that? [11:27] cjwatson: binaries like hidd, pand are removed in favour of some new dbus method. But users are not able to use newer method and complaining about this drastic change in user experience. Due to this they are not able to connect keyboards/mice or make PAN connections. [11:28] slytherin, do you have a bug about that ? [11:29] ogra: there are two, let me find them. [11:29] slytherin: oh, sorry, I thought you meant missing binary packages (which would have been an archive problem) [11:29] "some new dbus method" is a bit vague :) [11:31] ogra: yes it as vague as I am telling. Even I am not sure how it works. Tollef agreed to buy a bluetooth keyboard and look into the matter but I am not able to find him online for some time. So not sure of the status. [11:31] ogra: bug 191704 and bug 192043 [11:31] Launchpad bug 191704 in bluez-utils "hidd binary removed form bluez-utils package unable to connect as a result" [High,Confirmed] https://launchpad.net/bugs/191704 [11:31] Launchpad bug 192043 in bluez-utils "latest bluez-utils is missing /usr/bin/pand" [Undecided,Confirmed] https://launchpad.net/bugs/192043 [11:40] pitti: done [11:42] zul: thanks [11:44] Good god. [11:45] WTF [11:45] Something overwrote my LSB release details with some SkoleLinux/etch info. [11:45] Fujitsu: You're running SkoleLinux/etch now! [11:46] I do have every Gutsy package possible installed on that system, but I wouldn't have expected them to overwrite such things... [11:46] Fujitsu, likely debian-edu [11:47] have fun sorting out what cfengine changed all over your system :P [11:49] Yay, do-release-upgrade works now. [11:51] <\sh> cjwatson, re: grub phew...hating wget for not overwriting files but adding .N where N gt > 0 ;) the problem was an old -server daily iso image :( now using the correct one *grmpf* [11:52] \sh, another reason o use rsync :) [11:52] ah yes, I asked because it was a known bug in at least one image [11:52] *to [11:52] Should a package really do nasty evil things merely by being installed? [11:52] Fujitsu, in case od debian-edu thats a yes [11:52] it mangles lots of stuff in /etc via cfengine [11:52] <\sh> ogra, _if_ you can rsync ;) [11:52] <\sh> +use [11:52] Blurgh. [11:53] \sh, another reason to have zsync on the servers then :) [11:54] <\sh> ogra, well, too late..it cost me at least 4 installation tries to prove that layer 8 had a bug ;) [11:55] cjwatson: am I right in presuming that our updated kboot should be in the daily live today? [11:56] could someone of motu release team review bug 216698 ? [11:57] jeromeg: Subscribed motu-release? [11:57] You might want to poke in #-motu [11:57] Fujitsu: yes [11:57] Fujitsu: ok i'll [11:57] munckfish: only once it gets built ... ;-) [11:58] munckfish: it's release time, so images tend to be built only when necessary; but I understand there's a new batch building at the moment [11:58] cjwatson: oh I c [11:58] it's difficult to tell whether we're on it, cause [11:59] <\sh> pitti, thx :) [11:59] it's not in a package, and ofcourse otheros.bld doesn't include a version number in it's title [11:59] Can someone please remind me how to turn off SCIM ? I have created an Ubuntu chroot and for some unknown reasons SCIM appeared. Of course I don't need it as those computers will run on fr_CH.UTF-8 [12:01] pitti: so language-pack-gnome-de-base is the one i should put this mozilla thing in for the important languages, right? [12:02] (or without -base) ? [12:02] asac: it doesn't matter much whether it's -base or -de; however, my first upload will have them in -base [12:02] munckfish: yeah, I know, sorry [12:02] munckfish: the code is: [12:02] KBOOTDEB="$($BASEDIR/tools/apt-selection cache show ps3-kboot | \ [12:02] grep ^Filename | awk '{print $2}')" [12:02] ar p "$MIRROR/$KBOOTDEB" data.tar.gz | tar -xzf - -C . ./boot/otheros.bld [12:02] mv boot/otheros.bld PS3/otheros/ [12:02] rmdir boot [12:02] munckfish: so it just needs to be in the archive in advance of the build [12:02] asac: but I don't have an LP export yet, so I can't prepare new ones until tomorrow [12:02] pitti: ok ill do that too now [12:03] pitti: thats fine [12:03] ill use the ones i produced on rookery for now [12:03] asac: ok; so you'll do manual uploads of the top-5 languages or so? [12:03] pitti: right. starting with -de now :) [12:03] pitti: version will just have CURRENTubuntu1 [12:03] cjwatson: eh? you mean that's how it gets on the cd then? [12:03] in case you bother [12:03] asac: right, to pick a *completely* arbitrary language :-P [12:04] asac: version is fine [12:04] great [12:04] munckfish: yes [12:04] :) [12:06] slytherin, hmm, that intresting, nothing in any of the changelogs indicates why it was dropped [12:07] cjwatson: are the cd build daemons monitor thru LP like the normal daemons? [12:07] pitti: ok -de-base is tested and going up [12:08] munckfish: no, afraid not [12:08] ok [12:08] ogra: any chance it was dropped on debian side? [12:08] I'll just wait then [12:08] anyone can tell me why my bash completion doesn't work here sine a week or so? [12:08] munckfish: what are you looking for? [12:08] e.g. tab doesn't show any suggestions :( [12:09] slytherin: 11:55 cjwatson: am I right in presuming that our updated kboot should be in the daily live today? [12:09] cjwatson: thanks. :-) [12:10] slytherin: what he said [12:12] slytherin, probably, i dont know [12:12] slytherin, but it looks like a droped ./configure option [12:13] munckfish: When did you upload it? The build time is specified on the iso list page. I am assuming the time is UTC. You can make a guess about if it was included? [12:13] slytherin: I already dealt with the quetsion [12:14] question [12:14] pitti: in which seeds are the language packs that get on CD again? [12:14] slytherin: I indicated that it is not in the current build, but will be in the next one [12:14] slytherin: no need for further analysis :) [12:14] asac: ship and live [12:15] slytherin: yes I noted yesterday's live daily was done just a few hours before the new ps3-kboot was uploaded [12:15] asac: ship -> alternates, ship-live -> desktops, dvd -> well, dvd :) [12:15] cjwatson: Ok, I will keep my mouth shut. :-D [12:16] pitti: there are zero langpacks in ship-live [12:16] pitti: the language packs are actually in live, not ship-live [12:16] ah, right, my fault [12:16] (because they need to go on the live filesystem; ship-live is the apt archive that lives alongside it [12:16] ogra: The changelog entry for 3.26-0ubuntu4 is - enable dund, so that palm pilot ppp-over-BT works. - But it doesn't mention how it was done. I will try to check it tonight and if done then will submit debdiff. [12:16] ) [12:17] hmm we ship es xh??? and pt [12:17] whats xh? [12:17] xhosa? [12:17] why is that on CD? [12:17] slytherin, ping me if nobody reacts in time for a post RC upload, i'll see what we can do [12:17] asac, for africa [12:18] and no fr/de? [12:18] strange [12:18] its tiny [12:18] ogra: I will ping you as soon as I upload debdiff. [12:18] thanks [12:19] we prioritised xh because it was a specially commissioned translation for Ubuntu [12:20] it hasn't been maintained, and over time it has become a bit useless due to staleness, but at the same time that means the language packs are small and not much of a cost [12:20] ok thanks [12:20] so where can i find a list of most popular ones? [12:20] asac, pitti usually has one hidden on rookery [12:21] http://people.ubuntu.com/~pitti/scripts/langpacksize has the priority order up top [12:21] I believe that's still our canonical ordering [12:21] right [12:21] ok ill levave out xh :) [12:25] ogra: xh has no translation for firefox ... you think you can kick someone in africa to do that translation? or isn't it worth it? it can be done in launchpad now ;) [12:25] seb128: hm, seems the patch in bug 150193 didn't work as expected? shall I remove it from -proposed? [12:25] Launchpad bug 150193 in gdm "Remote Login via XDMCP is not working in Ubuntu Gusty/7.10" [High,Fix committed] https://launchpad.net/bugs/150193 [12:26] ogra: its just 7000 strings ;) [12:28] pitti: I don't really care either way, the patch is from upstream and fixes some issues I think but maybe not everything, but I don't expect many people will keep running gutsy and I've enough to do without chasing this one === warp10_ is now known as warp10 [12:29] seb128: a test for 'does not introduce apparent regressions' would be enough for me FWIW [12:29] seb128: ok, I'll followup on the bug again, thanks [12:30] pitti: thank you [12:30] pitti: ill do de, fr (both done) ... and now es, pt. maybe ill do chinese and jp as well [12:31] asac: you might want to check if the CJKs are on the CDs anyway, since they are very big [12:31] pitti: they are not on CD i guess. ill just do those for the funkyness :) [12:31] asac: I wouldn't mind; for network installs, people will get the fresh ones from Thursday anyway [12:32] asac, heh, ask sabdf1 :) [12:32] he might know africans to help out [12:32] s/mind/bother/ [12:32] ogra: hehe ... thought it was your edubuntu that lead to xh :) [12:32] nah [12:32] that pedates edubuntu [12:32] *pre [12:33] ok sorry for bothering you then ;) [12:37] pitti: do you remember to remove mozilla- language packs from any depends? [12:37] ogra: xh isn't a major commercial interest any more, afaik [12:37] asac: yes, that's done [12:37] pitti: those shouldn't be installed anymore. [12:37] asac: I demoted it this morning [12:37] it would obviously be good to have a community actually maintaining that translation [12:37] pitti: ok. so it makes sense that stgraber still got them installed today? [12:37] pitti: thanks [12:38] asac: hm, actually not [12:38] 'got' them installed, or 'has' them installed? they won't be removed automatically [12:38] at least not until today, since they were in main until a few hours ago [12:38] for final hardy, update-manager will uninstall them [12:38] pitti: it's in a chroot where I manually install ubuntu-desktop so I don't get exactly the same package list as on a standard system [12:39] pitti: the chroot was regenerated at ~11:00 this morning [12:39] cjwatson, well, but mark is african, he surely knows more africans than me or asac ... i didnt mean that in commercial context ;) [12:39] UTC+2 that's [12:40] erk, apparently Arne forgot to rebuild some l-support packages; doing that now [12:41] stgraber: but these only affect -be, -fy, and -mn; I doubt that you have those? [12:41] pitti: I don't [12:46] pitti: I don't see any reason why it would get installed, so probably something wrong with my chroot builder scripts ... [12:50] stgraber: can you verify please? [12:50] asac: I updated my script to fix some SCIM and locale problems, it's generating a new chroot at the moment [12:51] Riddell: dunno if you noticed bug 210443? [12:51] Launchpad bug 210443 in kmplayer "file overwrite error in hardy" [High,Triaged] https://launchpad.net/bugs/210443 [12:54] cjwatson: hmm, no, thanks will look [12:58] cjwatson, pitti: could you consider the nautilus update in the queue? === smarter_ is now known as smarter [13:10] seb128: I'm inclined to leave nautilus to post-RC, since it's largely cosmetic (well, a little more than that, but not a blocker) [13:25] If I am specifying more than one bug in changelog then is this format correct? LP: #191704, 192043 [13:30] (LP: #191704, #192043) [13:30] To make sure, you can check the resulting .changes file. [13:30] It's dpkg-source that parses the changelog entry. [13:30] It puts a Launchpad-Bugs-Fixed: header into the .changes file which launchpad in turn parses. [13:31] slytherin: ^ [13:31] soren: Ok. Thanks. [13:32] ahh, vi now shows both the bug numbers in yellow colour. :-D [13:33] slytherin: Eh? What does? [13:33] Oh, vi. [13:33] I somehow read "vi" as "it". [13:33] reading is hard! [13:36] soren: it's actually dpkg-parsechangelog, but eh ;) (dpkg-parsechangelog |egrep '^Clo') [13:37] thom: Right you are. [13:54] seb128: I'm inclined to leave nautilus to post-RC, since it's largely cosmetic (well, a little more than that, but not a blocker) [13:59] cjwatson: is there some space between rc and hardy for changes? [13:59] I think so [13:59] ok [13:59] because I would prefer to get the "cd player starts after login if you have a cd in the drive" to be fixed for hardy [14:00] but your call [14:00] after rc is fine anyway [14:09] pitti, foomatic-db-gutenprint and ijsgutenprint are not needed for standard Ubuntu. Standard Ubuntu has everything needed in cupsys-driver-gutenprint. [14:09] tkamppeter: so we can safely demote them to universe? [14:10] foomatic-db-gutenprint and ijsgutenprint are needed by Kubuntu, as the KDE Printing Manager does not support the PPD generators of CUPS 1.2.x. KDE Printing Manager needs ready-made PPDs or the Foomatic XML database, it does not look up PPDs with "lpinfo -m". [14:11] Riddell, ping [14:11] tkamppeter: ok [14:12] so they should become dependencies of something or get seeded to desktop? [14:12] Riddell, is this correct, that Kubuntu still needs these two packages? [14:12] tkamppeter: yes I expect so [14:12] adding to kubuntu seed [14:12] cheers [14:14] Riddell, did you look at my mails about the GSoC students for the Common Printing Dialog? Is it OK with you to mentor them? [14:16] tkamppeter: you split them, one for UI and one for technical stuff? [14:21] Yes, but the technical stuff needs also to be mentored by an UI expert, as it is the communication between UI apps and the printing dialog. [14:21] bdmurray: did you get a chance to retest bug 181121? [14:21] Launchpad bug 181121 in linux-restricted-modules-2.6.24 "[fglrx] gl screensavers crash on amd64" [Critical,Incomplete] https://launchpad.net/bugs/181121 [14:30] ScottK: I read you were working on a devscripts update to allow debsigning's Debian uploads; correct? O:-) [14:30] -'s [14:30] lool: It's waiting for approval by the release manager. [14:31] Cool [14:31] lool: I just grabbed the current debsign from Debian and stuffed it into our devscripts package. [14:31] Ok [14:31] Thanks [14:32] lool: No one has accepted it yet, so maybe you can help me out. The question I got, that I don't know the answer to is "do you know why the $DEBSIGN_ALWAYS_RESIGN variable is no longer honored in the new debsign?" [14:32] soren: Thanks for merging my ubuntu-jeos branch [14:32] I couldn't figure that one out. [14:32] lool: Thanks for doing it! :) [14:32] ScottK: Could it be a bug? [14:33] lool: It could be. It's not mentioned in debian/changelog. [14:33] lool: Would you have some time to look into it? [14:34] ScottK: are you a member of ubuntu-release? [14:34] ScottK: I'm still unpiling 4/5 days of email/TODOs, and I simply needed the new debsign, I don't have any strong interest in it [14:34] ScottK: I'm not particularly more powered to deal with it, I think [14:34] lool: OK. Thanks anyway. [14:35] afflux: I'm a member of motu-release (cares for Universe packages). [14:35] ScottK: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=447955 [14:35] Debian bug 447955 in devscripts "[debsign] Add "always resign" variable" [Normal,Open] [14:35] ScottK: Could it be that the DEBSIGN_ALWAYS_RESIGN was an Ubuntu specific feature that you dropped when moving to the new debsign? [14:35] lool: Thanks. [14:35] lool: I'll look into that. [14:36] ah, okay. So your ACK for 195508 is not enough for me to get this sponsored? :) [14:36] ogra: I have attached debdiff to bug 191704. [14:36] Launchpad bug 191704 in bluez-utils "hidd binary removed form bluez-utils package unable to connect as a result" [High,Confirmed] https://launchpad.net/bugs/191704 [14:36] lool: It it is, it wasn't a documented one (I always love it when that happens). [14:37] afflux: The ack I gave was an OK to upload from a distro perspective. It still needs a normal sponsors review. [14:37] I think the package is in main *confused* [14:38] Ah. [14:38] devscripts (2.10.9ubuntu2) hardy; urgency=low [14:38] * scripts/debsign.sh: implement DEBSIGN_ALWAYS_RESIGN variable to skip the [14:38] "Would you like to use the current signature?" question. [14:38] * ScottK looks at the bug. [14:38] -- Kees Cook Wed, 24 Oct 2007 15:54:58 -0700 [14:38] cjwatson: Thanks. [14:38] looks like StevenK forgot to mention that in the merge statement [14:38] Yeah. [14:39] I'll go back and fix it. [14:41] Argh [14:48] is it deliberate that a recent ubuntu alternate cd pops up the language menu just after gfxboot loads? [14:49] yes [14:50] soren: Hmm the vbox type seems broken in ubuntu-vm-builder [14:50] /usr/bin/ubuntu-vm-builder: line 880: vm_target_conversion: command not found [14:50] mv: missing destination file operand after `/root/ubuntu-vm-gutsy-amd64' [14:50] hrm, evand: is that going to stay on the final build? [14:50] hardy is on 8 days, please hurry up and fix the pink shadows & window decoration bug in the nvidia-utils drivers package [14:50] jordi: Yes, it is. Is it causing trouble for you in some way? [14:51] I've been looking but I don't see where to disable it [14:51] evand: I don't need it, and would prefer to just show the main menu on my custom cd [14:51] lool: Yeah, I got an e-mail about that a few days ago. [14:52] soren: I just installed kvm/virt-manager/libvirt-bin on current hardy (i386), and created a VM with a 5 GB image; this grinded for a while, and finally crashed with a "libvirtError: virDomainCreateLinux() failed"; is that known? [14:52] jordi: check with cjwatson as he wrote the change and knows much more about gfxboot than I currently. [14:52] soren: Hmm the code seems to be simply missing [14:52] soren: (I have the full stack trace here if you need it) [14:53] ugh, cjwatson, sorry for hammering you lately... :) I can't find where to configure gfxboot back to its previous behaviour where it didn't show the language selection screen on startup [14:54] lool: It sort of got left out in the cold during the refactoring I did. [14:54] lool: The code is sort of there. [14:54] lool: It used to do the same as either vmware or qemu. [14:55] pitti: Do you have enough space for a 5GB disk image? [14:55] soren: yes, that's fine [14:55] jordi: remove 'panel.lang' from common.inc [14:55] soren: I just noticed that I wasn't in group "kvm" (recent reinstall) [14:55] the call to it from the MenuInit function [14:56] pitti: "virDomainCreateLinux() failed" is sort of a very genereal "dear, oh dear, something went wrong". [14:56] pitti: Ah, yes, that would do it. [14:56] soren: I addgroup'ed myself and ran it under newgrp kvm, but I get the same errror [14:56] cjwatson: ok [14:56] soren: I'm trying to find out what type the .vdi files are [14:56] pitti: Try running "virt-manager --no-fork". That might give some useful info. [14:56] Interestingly, this led me to discover that vbox supports iscsi for virtual disks [14:56] jordi: do you have an /isolinux/lang file? [14:56] lool: Oh, really? That's pretty cool. [14:57] jordi: I'm wondering if we should skip language selection when that's present [14:57] lool: fwiw, libvirt will do the same in Intrepid. [14:57] Yeah, I think most people will want to use iscsi as a "lightweight" disk abstraction rather than lvm or nbd [14:57] cjwatson: just "langlist" [14:58] soren: I get three identical lines saying "libvir: QEMU Fehler :" [14:58] pitti: Don't worry about those. [14:58] cjwatson: and yes, if the user already selected a language there, I'd skip it in the installer [14:58] mm, s'what I thought [14:58] drat, getting late for hardy [14:58] soren: ah, I see I have a libvirtd process running; might be that this one isn't running in kvm [14:58] [14:59] heh [14:59] pitti: Oh, you newgrp'ed rather then logout/login. YEs, then you're probably right. [14:59] soren: hah, there we go; killing my libvirtd did it [14:59] pitti: Ok, cool. [14:59] soren: so I guess this boils down to "give a friendlier error message about kvm membership" [15:00] soren: maybe something coudl access("/dev/kvm", W_OK) and give a message box? [15:01] soren: other than that, it looks great! UI-ized image creation, hardy desktop CD boots, etc; kudos! [15:01] cjwatson: what about lang? did you mean langlist? [15:01] jordi: http://paste.ubuntu.com/7199/ ought to do that [15:01] jordi: no, langlist is the list of all available languages, but you can also stick e.g. 'ca' in a file called lang and that sets the default language [15:02] cjwatson: ooo, that's even better [15:02] pitti: :) Wait until you see the stuff that'll land in Intrepid. :) [15:02] ScottK: thanks for ACK. [15:02] cjwatson: just doing that, or adding lang *and* the change to common.inc? [15:02] emgent: You're welcome. [15:02] jordi: both [15:02] cjwatson: okay, I'll have a go now. [15:03] soren: now I just need to convince it to not use a 2 by 2 million pixel resolution, so that I can actually use the VM :) [15:04] jordi: committed for (hopefully) post-RC [15:04] cjwatson: cool [15:05] soren: drat, it just doesn't let me use the VM menu; as soon as I click on it, the VM grabs the mouse again [15:05] pitti: When you click where? [15:05] cjwatson: Thanks again for the debsign pointer. Fixed version just uploaded. [15:05] pitti: Re: insane resolutions: Talk to the X guys :) [15:06] soren: when I click into the VM window menu (Virtual machines / Display / Send key / Helf) [15:06] Help, even [15:07] Hm... Interesting. [15:07] cjwatson: any idea why the new mythbuntu tasks want to install the "wrong" packages? tasksel wants to install mythbuntu-backend-master instead of mythtv-backend-master.. http://www.pastebin.ca/987371 [15:07] pitti: I haven't seen that. WM? [15:07] soren: compiz outside, metacity inside [15:08] pitti: I'm on compiz, too. Hm.. [15:12] soren: so, I get lots of display bugs (screen update, mouse pointer, etc.), but works well otherwise [15:12] pitti: Which graphics driver are you using inside kvm? [15:12] laga: but that's correct - the *task* name (as distinct from the package name) is mythbuntu-backend-master [15:12] soren: hm, whatever the default on the live CD is; lemme check [15:12] laga: note the '^' after the task name which tells apt-get that it's a task [15:13] laga: it'll only work on CDs, not netboot [15:13] pitti: It'll probably be either vmware, cirrus or vesa. [15:13] cjwatson: it doesnt work on CDs. or on a normal gutsy box, for that matter [15:13] soren: erk, it doesn't boot from CD again [15:13] pitti: Right. Only the first time. It expects it does the installation the first time and the removes the CD after first boot. [15:13] cjwatson: err, on a hardy box of course. [15:15] soren: ok, I added a CD-ROM with the .iso, now it boots [15:15] laga: it won't work on a normal hardy box, I'm afraid (that's equivalent to netboot; mythbuntu-desktop won't work either I imagine) [15:15] laga: I'll look into the CD problem as soon as I have some bandwidth - rsyncing down a DVD right now [15:15] cjwatson: okay. i'll try again w/ the CD to make sure i'm not wasting your time. [15:16] laga: ok, something wrong with the CD, one moment [15:19] soren: yep, cirrus [15:19] pitti: Thought so. [15:19] soren: I can repair the screen corruption by moving the VM's mouse cursor over them [15:19] pitti: Changing that is more involved than you would think, so it probably won't be any different in hardy :( [15:19] yeah [15:20] pitti: For Intrepid, we'll probably be using the vmware driver. [15:20] pitti: ..or depending on how far upstream gets with it: Some sort of gl-passthrough-coolness. [15:23] laga: please apply http://paste.ubuntu.com/7201/ (fix common seed's inheritance, and make backend-master implicitly include common) [15:23] cjwatson: works great, many thanks! [15:23] jordi: cool [15:23] laga: we have a few cdimage fixes to make as well, though [15:24] but at least the second half of that patch is a prerequisite for being able to fix it [15:24] cjwatson: but that won't make backend-master install everything from common, right? [15:24] hence "a few cdimage fixes" :-) [15:25] err, backend-master *should* cause everything from common to be installed, shouldn't it? [15:25] I thought that's what you wanted [15:25] if I wanted to write a blueprint for intrepid, should I wait until the intrepid series is added to launchpad or should I tag it with [intrepid] like it's done for some specs? [15:25] twi_: is it an idea, or do you have a detailed software design document? [15:26] no. "common" is common because it's needed in live and in ship :) but backend-master should only pull in mythtv-backend-master and the mythbuntu-desktop package [15:26] cjwatson: and thanks to looking closer, I found my "black" background is actually hex 030601 [15:26] so thanks again :P [15:26] cjwatson it's only an idea for now [15:26] laga: oh, ok, skip the second bit then [15:26] laga: please do apply the STRUCTURE fix though [15:27] laga: do you want backend-slave to be on the CD too? (if so, you should have pinged me for a tasksel upload) [15:27] laga: and frontend? [15:27] pitti, ping [15:27] cjwatson: i believe tasksel has the correct tasks right now for backend-slave, too? [15:27] twi_: please use brainstorm.ubuntu.com rather than a spec, then [15:27] tkamppeter: just ask your question, please, don't ping [15:28] laga: oh, yes, it does, wonder how I missed that [15:28] laga: but do you want them on the CD? [15:28] yes [15:28] cjwatson no, I want to work on a spec. So I guess the workflow is to write a detailed spec on a wiki page and then link it later to a blueprint? [15:28] laga: then you need something like http://paste.ubuntu.com/7204/ [15:29] twi_: it's just that we try to discourage people from writing specs unless they're software engineers who are prepared to carry them through to completion; it's quite hard to find anything now because there are so many specs that are essentially wishlist bugs [15:29] pitti, I have fixed bug 213624, bug 211867, and bug 195508 in system-config-printer now. [15:30] cjwatson i am and i'm planning to go to UDS that's why I am asking [15:30] Launchpad bug 213624 in system-config-printer "system-config-printer.py crashed with error in _probe_queue()" [Medium,Triaged] https://launchpad.net/bugs/213624 [15:30] twi_: but if you are prepared to do that, then just register it at https://blueprints.launchpad.net/ubuntu/+addspec for now, and don't make it release-specific [15:30] Launchpad bug 211867 in system-config-printer "system-config-printer.py crashed with TypeError in getNPPPD()" [Medium,Triaged] https://launchpad.net/bugs/211867 [15:30] Launchpad bug 195508 in system-config-printer "applet.py crashed with AttributeError in check_for_jobs()" [Undecided,Triaged] https://launchpad.net/bugs/195508 [15:30] twi_: ok, cool [15:30] cjwatson ok thanks [15:30] twi_: sorry, we've just found we need a bit of a firewall :) [15:30] cjwatson heh yeah I can imagine :) [15:31] twi_: if you want to discuss it at UDS, there's a "propose for meeting" (or similar) link that you can use [15:31] cjwatson yeah I saw that I just wasn't sure about the release [15:31] (unfortunately that's not restricted to meeting attendees yet, but to everyone else, please only use that if you're attending the meeting in question) [15:31] pitti, they are all small patches from upstream to fix crashes. I have put the new package files and the debdiff to the ususal place. Can you please upload this to Hardy. Thanks. [15:31] but if I can not assign it i'll just do that for now [15:31] twi_: those look like Canonical staff dumping in their team roadmaps [15:32] tkamppeter: please send me a mail with the URLs to the debdiffs and packages [15:32] I see [15:32] twi_: but putting [intrepid] in there isn't required, I think it's just for tracking purposes since they have lots [15:32] tkamppeter: or, rather, just subscribe ubuntu-release to the bugs and put the URLs/debdiffs to the bug reports; that's better [15:32] (for avoidance of doubt Canonical staff don't as a rule have extra privileges with regard to blueprints, that wasn't what I meant) [15:32] cjwatson yeah I'll just keep it unassigned until they add intrepid to launchpad [15:33] cjwatson: thanks a lot, i've applied the fix [15:33] twi_: normally, the "series goal" stuff is filled in after UDS by release planners based on what seems plausible and has people prepared to work on it [15:33] twi_: it's not actually a requirement to fill that in ever in order for something to make it into the release [15:34] twi_: again it's something that's useful if you have lots and lots of specs that you're tracking [15:34] you should also have a look at brainstorm.ubuntu.com i bet there are some suggestions already that might give some additional POVs [15:34] cjwatson oh perfect then [15:52] pitti, done. [16:06] cjwatson: about UDS, is there a specific place where the planning is stored? Actually, is there a general planning at all or is it only convention between interested people around specifications? === elmargol_ is now known as elmargol === bigon` is now known as bigon [16:19] cjwatson: I'm looking into the fglrx bug now [16:22] hi all, now that OOo has built properly on powerpc, does anyone know why it is not included in alternate cd? === mathiaz_ is now known as mathiaz [16:32] Ok. I have found the problem and added debdiff to bug 164491 [16:32] Launchpad bug 164491 in ubuntu-meta "[ppc] openoffice.org not present on alternate CD" [Undecided,New] https://launchpad.net/bugs/164491 === Mez is now known as fishfinder === fishfinder is now known as Mez [17:16] Riddell, are you still around? [17:17] tkamppeter: hi [17:18] Riddell, Lars has modified his application on the Google Summer of Code web app to reflect his new project. Can you look into it and click "I am willing to Mentor"? [17:19] slangasek: There's this power saving patch for pygobject / pygtk at http://bugzilla.gnome.org/show_bug.cgi?id=481569 which has now been fixed; is this something too intrusive to push to hardy now? [17:19] Gnome bug 481569 in gtk "Calling gobject.threads_init() causes a lot of wakeups" [Normal,Reopened] [17:21] tkamppeter: "Develop a DBUS interface.."? [17:24] ScottK: erk? did debsign lose my patch? [17:24] kees: No. [17:25] kees: (See also the backlog about 3 hours ago) [17:25] ScottK: Ah, you were comparing debian debsign to ubuntu debsign? [17:25] kees: You patch didn't get documented after the last merge, so I was about to drop it, but slangesek noticed and I fixed it. [17:25] You/Your [17:25] ScottK: was that a mistake on my part or the merger's? [17:25] Mergers [17:26] okay, cool. I still need to convince the debsign maintainer better to take that change. [17:26] ssh screwless [17:26] oops. [17:26] Riddell, some form of inter-process communication is needed to couple the application with the dialog. One possibility is DBUS. As you probably have more knowledge about GUI apps, you know perhaps something better, especially as the app and the dialog are usually running as the same user. [17:27] kees: You might also look at Bug 218210 - I think a patch for the documentation (man page) might help your odds. [17:27] Launchpad bug 218210 in devscripts "debsign: $DEBSIGN_ALWAYS_RESIGN not documented" [Undecided,New] https://launchpad.net/bugs/218210 [17:28] ScottK: heh, good call. [17:28] hunh, it got missed by both StevenK and ogra. I must be very sneaky! :) [17:29] * ogra looks up [17:29] what did i do ? [17:30] kees: BTW I discovered this feature thanks to today's discussion and would be happy to use it for purely Debian uploads; it happens from time to time that I have broken checksums or similar in .changes or .dsc which I then need to resign, and I hate the confirmations [17:31] lool: please comment on the Debian bug -- I'd like to see it there too, I find it handy for sponsoring Debian uploads too. [17:32] kees: I was about to [17:33] mvo: replied in the bug [17:33] lool: cool. I'm really not clear on the maintainer's objections. [17:33] when doing sponsored uploads (debian or ubuntu) you're over-writing the sponsoree's signature. that's the whole point -- they're not in the keyring yet. [17:34] ... and they removed the 'patch' tag? pfft [17:35] what's the preferred way to submit a new package version for sponsorship? [17:35] the debdiff in this case is tiny, even with the upstream changes. [17:36] james_w: Attach the diff.gz to a bug and subscribe the appropriate sponsorship team. [17:37] james_w: At this point you'll probably need to look at a freeze exception too. [17:37] with a pointer to the upstream tarball? [17:37] is the watch file sufficient for that? [17:37] Watch file should be sufficient. [17:37] yeah, it will be a freeze exception request first. [17:37] thanks ScottK [17:37] No problem. === azeem_ is now known as azeem [17:41] * lamont wants to know which source package and where to modify to fix compiz' window focus back to his preferred focus [17:41] or if compiz is just not for me === evand_ is now known as evand [18:07] slangasek: are we still planning 20080415 as the ubuntu-server iso for RC ? [18:08] pitti, Hobbsee: It appears that Ryan is ready for you re: abiword. [18:08] mathiaz: hope not [18:08] slangasek: I've noticed a new daily build for ubuntu-server on cdimage (20080416) [18:08] mathiaz: we rebuilt everything today, and will rebuild at least all desktops again [18:09] cjwatson: ok - so there is no point in testing 20080415 [18:15] mathiaz: yes, please test 20084016 instead [18:15] cjwatson: Can we discuss issue of OOo on powerpc CD if you have time? [18:15] slytherin: not right now (I have to go), but please leave messages [18:15] slangasek: ok - could you update iso.qa.u.c ? [18:15] cjwatson: Ok. I will add comment on bug [18:16] mathiaz: now that I'm awake, yes :) [18:16] lool: eew, the pygobject/pygtk power issue didn't get resolved right the first time? [18:25] tkamppeter: please do not add milestones to all bugs you encounter/fix; milestones are for bugs which block the release, not for bugs which would be nice to fix [18:26] * pitti removes the milestones again [18:29] ogra: ping [18:29] pitti: somebody should clarify how milestones are supposed to be used but I don't think they are considered as reserved to blockers right now [18:30] seb128: hardy task > 'pool of nice to fix', milestones > blockers, AFAIUI [18:30] slangasek: ^ [18:30] pitti: "nice to fix for 8.04.1 = ???" [18:31] seb128: -> hardy task [18:31] seb128: hardy task [18:31] how do you know what is 8.04 and 8.04.1 then? [18:31] we just need to find a convention which works for everyone [18:31] if it's "nice to fix" only, do we need to distinguish? [18:31] I don't agree with that workflow [18:32] I don't see what is wrong with milestone + priority < high = target of opportunity [18:32] doesn't really make sense to take a wishlist milestoned as a blocker [18:32] seb128: we just need a way to tell apart milestone blocker bugs from 'would like to get eventually' [18:32] it would not be a wishlist if it was a blocker [18:32] pitti: that's what the importance is for no? [18:33] that clutters the +milestone list a lot, though [18:33] that's a launchpad UI bug [18:33] and it's very hard to see what's actually important on it [18:33] sort by importance? [18:33] well, I'm amenable to that in theory, but we do need to nail down a workflow, and in particular one that doesn't require the RM to manage the wishlist milestoned bugs when they miss the milestone [18:33] then you get the states all mixed up [18:34] right, and moving milestones all over is hard, too [18:34] it sounds like a greasemonkey script to hide unimportant bugs might be needed. [18:34] also, I think a different set of people can set milestones than can set bug severities, so what happens when someone raises the severity of a milestoned bug that wasn't meant to be a blocker [18:34] we should really have a team discussion about that [18:34] ? [18:34] and not have people deciding workflow and not communicating those [18:34] seb128: yeah [18:34] well, TBH I think there were several announcements about this some months ago [18:35] slangasek: that's a launchpad bug ;-) [18:35] pitti: launchpad ones or ubuntu ones? [18:35] seb128: we could also say it's a workflow bug :) [18:35] pitti: I'm using milestones to do my todos for ages and nobody ever told me otherwise until recently (one week ago) [18:36] personal TODOs are best managed with 'confirmed/assigned/in progress' IMHO [18:36] the "nominate for hardy" thing sucks imho, you can't select it in the advanced search options for example [18:36] no need to clutter the RM lists with personal TODOs [18:36] pitti: not when you have 3000 bugs on your list [18:36] seb128: it doesn't matter whether you set a milestone or set it as 'in progress' [18:36] same effort [18:36] no [18:37] "in progress" would mean I'm working on it [18:37] and I can't possibly work on all the desktop bugs myself [18:37] but having a milestoned bug without anyone working on it makes no sense [18:37] I need a way to make a list of things the desktop team should consider for hardy and for 8.04.1 [18:37] it does [18:38] that's what I can point at the weekly meeting saying "we should fix those please give a hand" [18:38] seb128: ok, seems we should have a discussion about that on u-devel@; perhaps not at this time, though [18:38] pitti: right, or at uds [18:38] slangasek: can you examine bug 218110? this seems related to the PAM changes. [18:38] Launchpad bug 218110 in samba "PAM [error: /lib/security/pam_smbpass.so" [Undecided,New] https://launchpad.net/bugs/218110 [18:38] pitti: sorry for bothering about workflow details when you have other things to do ;-) [18:38] * seb128 hugs pitti [18:38] seb128: no need to be [18:39] I'll do the way it has been decided for hardy and let's have a talk about that at uds [18:39] Can someone please sponsor debdiff attached to bug 191704? I had discussed this with Tollef sometime back, but I am not able to find him online today. [18:39] seb128: I wasn't actually aware that the current (what I regarded as) best practice was in fact still disputed [18:39] Launchpad bug 191704 in bluez-utils "hidd binary removed form bluez-utils package unable to connect as a result" [High,Confirmed] https://launchpad.net/bugs/191704 [18:39] since from my POV that didn't change for months [18:39] kees: is that not the same bug that cjwatson assigned to me? [18:40] pitti: well, I didn't read the mail about that being the official workflow apparently [18:40] pitti: or decided it didn't work for me and didn't bother ;-) [18:40] there wasn't a mail about it, totally my fault [18:40] I still think it's weird to have 2 similar things, nominations and milestones [18:40] but that's a launchpad issue [18:40] seb128: yeah, it's indeed confusing [18:41] especially that you used to not be able to nominate for the coming version I think [18:41] I think it's sensible to have two different tools, but it's not immediately obvious which one to use as a developer [18:41] seb128: you can't add tasks? [18:42] pitti: no, I mean that previous cycle it was possible to use the nomination only for stable distributions [18:42] ugh; well, *that* would be an LP bug [18:42] +not [18:42] or that's me who just associated nomination = SRU [18:43] anyway, let's sort that later [18:43] it's actually quite similar; "would like to fix in the hardy release" as opposed to "eventually" [18:43] * ScottK has too, but has seen a fair number of nominations for Hardy from end users. [18:43] slangasek: that's alright, just something to clear for next cycle ;-) [18:43] * pitti -> off to Taekwondo, cu later [18:43] pitti: have fun! [18:44] * pitti hugs seb128 [18:44] * seb128 hugs pitti [18:47] slangasek: I don't think so -- this was a private security bug until I just un-privated it. [18:47] kees: ok, then it's a duplicate, I'll fix it up [18:47] (the bug state, I mean...) [18:48] slangasek: okay, cool thanks [18:52] hi there [18:52] I want to try out Apache2 with IPv6, but I can't access to localhost [18:52] marcreichelt: Try #ubuntu-server [18:52] oh, thanks :) [19:03] mvo_: Amaranth: slangasek: I've milestoned bug 201330 as it is a regression from gutsy, a patch is available, etc. [19:03] Launchpad bug 201330 in compiz "[regression] need to whitelist multiple ATI cards, or remove blacklisting" [High,In progress] https://launchpad.net/bugs/201330 === mvo_ is now known as mvo [19:18] slangasek: hello http://pastebin.com/m2a7fa75 (fixes from regression from MIR requirements) [19:19] zul: you know you don't need to get approval from the RM before each upload? :) [19:19] zul: looks sane enough; won't be accepted now until after RC [19:19] zul: but you can still upload in the meantime [19:21] any chance that debdiff for bug 191704 will get accepted before RC? [19:22] bug 191704 [19:22] Launchpad bug 191704 in bluez-utils "hidd binary removed form bluez-utils package unable to connect as a result" [High,Confirmed] https://launchpad.net/bugs/191704 [19:22] nope [19:25] slangasek: after RC? [19:25] slangasek: thnks.. [19:26] slytherin: possibly [19:26] slytherin, i looked at the bug, i think that is rather a bug in hal than in bluez [19:27] ogra: LOL. So what do you suggest now? [19:27] slytherin, also do you have any idea how enabling the two binaries will influence the new behavior for people for whom i works now ? [19:27] it won't [19:27] ogra: AFAIK, it won't. [19:27] slytherin, well, talk to a hal person probably ... but i guess pitti is to busy with pre RC stuff to even take a look [19:28] ogra: Can you tell me how you figured out it is hal bug? [19:28] a proper fix for the situation would be to build a bluez-extra package containing the two [19:29] er, I don't see why there's anything proper about creating a new package here [19:29] me too, these binaries were present in bluez-utils previously, so creating new package doesn't make sense. [19:30] slangasek, vs unpredictable results the re-enabling of two deliberately upstream disabled binaries might cause [19:30] there's nothing unpredictable [19:30] the binaries are never executed automatically [19:30] slangasek, well, your call, i'm no RM [19:30] and there's some functionality that it's not straightforward to get using the new bluetooth "services" [19:31] i just dont know in which ways hal breaks f someting else claims the HW [19:31] if the something else claims the hardware, it's because the user asked it to [19:32] I don't see a problem there :) [19:32] slangasek, i'm by no means a bluetooth knowledgeable person, i just took a look at the bug because slytherin asked me :) [19:32] and to me it occued as if upstream had a reason (even though not well documented) to disable it [19:33] because they believed it obsolete [19:33] practically speaking, a number of people disagree [19:33] mvo: based on the bug reports, I think ati should be un-blacklisted -- more and more pci-ids are getting added to the whitelist. Also, I think we should support the open driver. [19:34] does the open ati driver support 3D yet? [19:35] xserver-xorg-video-ati does [19:35] hcoal: yes [19:35] ogra: Can you please tell me why do you think it is hal problem? [19:36] slytherin, well, as slangasek stated above 'there's some functionality that it's not straightforward to get using the new bluetooth "services"' [19:37] slytherin, there is a new way that doesnt seem to work for all yet [19:37] ah, ok [19:37] so the *right* fix is to get the new way woking for everyone instead of having two ways to do a thing in the default [19:38] for that exact reason, i agree a stop gap solution (if HAL doesn't get fixed in time for 8.04 to go gold) would be having an extra package with those binaries [19:38] Do you think I'll get better performance using the open driver rather than ati's own restricted driver? I'm using an HD 3850, performance is pretty good but video playback and full screen video/graphics are pixelated and jerky. Video looks awful fullscreen. [19:38] No [19:40] hcoal: The open driver doesn't yet support 3D for all cards [19:40] hcoal: And it is also not necessarily better for latest cards. [19:40] Thanks slytherin. That's a pity, this is the only thing stopping me from ditching Windows completely, I love ubuntu but not being to play video properly on is stopping me from making the switch at the moment. [19:42] ogra: The problem is that most of the developers working on bluetooth, AFAIK, don't have bluetooth input devices. Tollef promised he would buy one and try to fix it. But we are too late in hardy cycle. [19:42] I'll stick with restricted while I wait for now. Beryl/Compiz runs really well, I just can't figure out why video quality is so poor. [19:42] kees: its tricky stuff, Amaranth blacklisted them because we got a lot of failures for them now we get a lot of "it works" :-/ [19:42] slytherin, well, i own a bluetooth dongle i used ... umm... twice i think in two years [19:42] :-) [19:42] but only a mobile to test it [19:43] mvo: well, I'm in the list of "annoyed" ati users. we absolutely need a whitelist at the very least. [19:43] mvo: I'm currently freshening the patch and adding more IDs [19:43] kees: what ati do you have? [19:43] mvo: I think it's a mistake to blacklist the open-source drivers. the bugs should be fixed, not hidden. I have 1002:4e50 [19:44] slytherin, i actually have a BT keyboard and mouse, but i've not ever run into any issues pairing and using w/ the new services [19:46] kees: right, I agree in general - we do only blacklist ati laptop cards currently because we got a lot of failure reports for those. I'm happy to add a whitelist [19:46] mario_limonciell: Find. But when it doesn't work, users resort to using the mentioned binaries. And not finding those binaries in package is frustrating for them [19:46] slytherin, slangasek is the one to decide about your debdiff and he didnt look resistant to your debdiff :) (compared to my idea to split them out to a universe package at least :) ) [19:46] kees: the trouble is that I don't have the knowledge to fix stuff in the ati driver myself, so all I can do is add them to the blacklist [19:46] slytherin, right, i understand what you mean. [19:47] ogra: Yes, Even I don't agree with separate package because then we need to add 'Replaces' and test the upgrades. So let's wait till RC is out. [19:47] slytherin, so just wait until the RC dust settles a bit and the release team has half a hand free again :) [19:48] mvo: I'd prefer a ID-blacklist rather than a driver blacklist. :P Anyway, once I test the updated patch, can you add it? [19:50] kees: sure, the repository is in ~ubuntu-core-dev, feel free to commit it yourself [19:50] mario_limonciell: Recently many users have been complaining about obex transfers. I initially thought it was a Nokia only phone problem. But looks like the problem is with Symbian based phones. Such bugs are hard to pinpoint. So it is better to push the workaround if we know it will work for sure, than to try fixing new buggy workflows. :-) [19:51] mvo: okay -- I'm unclear if it should be done via quilt (other things quilt-patch stuff in debian/ ) or done directly. [19:51] slytherin, i've noticed obex transfers only work (for me at least) when the gnome BT applet has the setting "visible and connectable for other devices" [19:51] slytherin, so that might be worthwhile to mention in bugs you come across for that [19:52] kees: quilt please, the debian/compiz-manager script is a checkout from a git repository that we heavily patch up - that sucks a bit and will go all upstream again [19:53] mario_limonciell: Will do. Time to go to bed now. Bye. === ogasawara_ is now known as ogasawara [20:02] kees: I'd say everything from X300 Mobility and newer should stay blacklisted [20:03] jdong: charles updated the bug report about transmission's patches backport. I'm just pinging, so it doesn't get forgotten. [20:03] Amaranth: do you have a way to gauge "newer than"? :P [20:03] kees: the trick being that's a _lot_ of cards and i'd rather compiz not run at all on your machine then make it crash at login [20:04] newer then is bad, maybe "better than" :) [20:04] Amaranth: well, I'm fine with the default being disabled, but if people are already configured to use compiz from gutsy, or they explicitly try to enable it, it should work. [20:05] Amaranth: at present, people with previously-working ati-driver compiz suddenly have it vanish without notice (like me) [20:05] I'm not fine with default being disabled but if we have a regression in the driver from gusty then those people are out of luck [20:05] And actually most of these were a problem in gutsy too, we just didn't get reports for all of them [20:06] Amaranth: I'll defer to you on the general methodology, but at least the people that noticed the problem should be whitelisted. what do you think of the pciid whitelist solution? [20:06] If you can get the driver from feisty to work with the current X server we're good to go [20:07] kees: I think a whitelist on top of a blacklist sucks :P [20:07] Amaranth: hrm? I'm not following. my ati laptop worked under compiz in gutsy (and it works in hardy if I disable the ati blacklisting) [20:07] What chip do you have? [20:08] 1002:4e50 [20:08] yeah, don't plan to whitelist that one... [20:08] rv350 [20:08] The problem I see here is that is feels like blacklisting the entire open-source ati driver for compiz is a mistake -- clearly it at least works for me and all the other people complaining on that bug report. [20:09] Well, what I wanted to do was only block for r300 and newer [20:09] why shouldn't it be whitelisted? it works fine. [20:12] kees: I also have a bunch of people saying their X300 and X700 Mobility chips work fine too [20:12] But those are the main ones that seem to have problems === Whoopie_ is now known as Whoopie [20:13] Amaranth: then it sounds like some other solution is needed -- if compiz crashes, disable it. if not, let it run happily. [20:14] kees: I can't think of a good way to do that without catching other people [20:17] Amaranth: disabling compiz on the whole for an open source driver that breaks only some people is a mistake. there must be some better way to work around it. [20:19] Amaranth: how about a gconf switch to block it in the case of repeated crashes? Documentation seems to be the way to win this. [20:20] do people read documentation? [20:20] allow ati/radeon unless the gconf setting exists. [20:20] laga: not initially, but once they have a problem, yeah. [20:21] what about a popup? "there's a lot of compiz breakage, do you want to disable?" [20:21] at the very least, if not gconf, pull in /etc/default/compiz or something like that. and have BLACKLIST_LAPTOP_DRIVERS="ati radeon" by default, so I can tweak a config file. [20:22] https://bugs.edge.launchpad.net/ubuntu/+source/xserver-xorg-video-ati/+bug/108527/comments/19 [20:22] Launchpad bug 108527 in xserver-xorg-video-ati "X freezes when compiz is enabled on ATI X300" [High,Fix released] [20:22] this guy has the same chip as you, for example [20:23] Amaranth: gross. :P the problem is we don't have any indication of which is more common: working or non-working, and it seems that PCI ids aren't a way to distinguish it. [20:23] kees: a config file might be better for kde, too. [20:23] that bug status lies, people reported a million different problems in that report :P [20:23] therefore, we need a different solution to allow people to enable compiz without changing a file in /usr/bin [20:24] laga: true [20:24] kees: you can always use the generic SKIP_CHECKS [20:24] we need a way to shatter clusterbugs off [20:25] Amaranth: where is SKIP_CHECKS defined? [20:25] in /etc/xdg/compiz/compiz-manager or ~/.config/compiz/compiz-manager [20:25] kees: it isn't, by default [20:26] * Amaranth is lagging very badly [20:26] * Amaranth blames freenode [20:27] Amaranth: how about documenting it with a commented-out example in /etc/xdg/compiz/compiz-manager and making it more discoverable via messages from compiz-manager itself "your compiz is disabled since.... if you really want to use it anyway, set ..." [20:28] Amaranth: or, frankly, move the blacklist list to that file... [20:29] verbose "SKIP_CHECKS is yes, so continuing despite problems.\n" [20:29] so stick SKIP_CHECKS=yes in one of those places [20:29] slangasek, still around? [20:30] tkamppeter: yes [20:32] It is about bug 214579 as I have already packaged a new s-c-p with fixes for three other crashers (0ubuntu8) and made FF exception requests on the appropriate bugs, should I add this to 0ubuntu8 or make a 0ubuntu9? [20:32] Launchpad bug 214579 in system-config-printer "my-default-printer.py crashed with TypeError in response()" [Medium,Triaged] https://launchpad.net/bugs/214579 [20:33] tkamppeter: if the fix is reasonable, include it in 0ubuntu8 please [20:33] slangasek: Concerning the pygobject / pygtk issue: what happened was that 1) we included the patch for python in gutsy (perhaps even earlier, not sure) 2) we included the patch for pygobject/pygtk 3) we noticed it wasn't included in python/hardy and it was then added 4) we saw it caused issues to have such support and removed the pygobject/pygtk patch [20:33] ah [20:33] slangasek: We're now at the point where the pygtk/pygobject part has been fixed [20:33] slangasek, the fix is a small patch of ~10 lines, so I will add it to 0ubuntu8 [20:34] lool: so what's the current impact, though? I understood that the symptom was "landscape-client sets people's laps on fire"? [20:34] slangasek: it's a story of watching the proper file descriptors at the proper side of the patch for read or write or both type of events, I don't know what was wrong or broke expectation, but it's fixed they say [20:34] slangasek: That was the symptom at 4) [20:34] slangasek: When we decided to remove this support [20:34] ah [20:34] The symptom motivating the change is that you'll see python apps in powertop without it [20:35] without the patch [20:35] yes, yes I do ;) [20:36] lool: yes, I'll accept that patch post-RC with an appropriate amount of pre-upload testing [20:36] slangasek: seb128 was more in favor of hardy-updates [20:36] that would be fine by me too [20:36] slangasek: Let's do it post release then [20:36] I hope $someone thinks of it [20:37] slangasek: Thanks for your time [20:37] maybe card.freenode.net is just broken [20:37] anyway [20:38] <_MMA_> slangasek: You release manager ATM? [20:40] _MMA_: ah... normally? :) [20:41] _MMA_: Unless there's been some spectacular screwup I haven't noticed, I'd guess the answer is yes. ;-) [20:41] kees: i suppose we could be a bit more verbose but that only helps english speakers [20:42] Amaranth: true [20:42] <_MMA_> Ok. I thought it rotated a bit. I thought Riddell did it a time or 2. :P n. [20:42] <_MMA_> slangasek: Do I need to sign off on a image for Studio's RC? [20:42] <_MMA_> s/n/np [20:43] _MMA_: if you want me to release UbuntuStudio as part of the group RC, it needs to be validated via http://iso.qa.ubuntu.com/qatracker/build/all/all and blessed by the developers, yes [20:45] <_MMA_> slangasek: Cool. I did test todays image. Ill add my results and get others to do so. This image (20080416) will most likely be fine. [20:46] I'm trying to install a small ubuntu with a unionfs but without casper, anyone ever tried ? [20:47] _MMA_: yes, we do have some late-landing packages that we need to reroll other images for, but I don't see that they warrant a reroll of UbuntuStudio [20:47] this is mean to go on a CF card [20:47] kommander: why do you think you need caspeR? [20:47] evand: ping [20:48] kirkland: pong [20:48] evand: wondering if the OEM Kubuntu fix is in the latest iso snapshots, shall i retest bug # 217884? [20:49] laga: I googled and saw a lot of page speaking about casper [20:49] <_MMA_> slangasek: np. Should be fine. [20:49] kommander: you can use custom initramfs scripts [20:50] kirkland: it's not yet, it will be once http://cdimage.ubuntu.com/kubuntu/daily-live/current/hardy-desktop-amd64.list shows oem-config 1.35 [20:50] laga: I used debootstrap to put a small beginning on the cf [20:50] laga: do you have an example online ? [20:50] so whenever the next run of kubuntu cds occurs [20:51] evand: coolio [20:51] kommander: not off-hand, but i'm sure google will have one, or wiki.ubuntu.com or the initramfs-tools documentation. if you cant find anything, then you can look at /usr/share/initramfs-tools/scripts/nfs and at the stuff in ltsp-client for some unionfs inspiration [20:52] laga: very good, thank you :-) [20:52] slanagasek, done and all bugs updated. [20:52] kommander: have fun :) in ubuntu hardy, you can also use aufs instead of unionfs. [20:54] laga: is aufs better than unionfs ? [20:55] kommander: depends. i like it better because it works better with NFS [20:55] laga: ok, thanks for all these informations :-) [20:56] no problem :) [20:59] is it still possible to sync packages from debian? [20:59] (with the appropriate ack from the -release team of course) [21:14] reading hardy-changes I figured it's still possible :) so archive admins, please sync xpad, it's bug #216698, thanks. [21:14] Launchpad bug 216698 in xpad "Please sync xpad 2.14-1 from Debian Unstable (main)" [Medium,Triaged] https://launchpad.net/bugs/216698 [21:38] laga, did you end up getting a follow up DVD done so mythbuntu can go out with the rest of the RCs? [21:44] blueyed: hello! have you seen my comment for #63450 ? it hangs for me in a test upgrade, have you seen anything like this? [22:13] Probably a common question, but why is ubuntu's gcc spitting out "C compiler cannot create executables." messages [22:14] basslingo: thats not gcc, thats autoconf [22:14] basslingo: do you have build-essential installed? [22:14] *checking* [22:15] really simple packaging question... if my package changes versioning schemes, i.e. 1.0.2 and then 1.0.080416 , is there a way i can tell the 1.0.080416 package that it is a higher version than 1.0.2 ? i could've sworn i saw this somewhere but now that i need it i can't seem to find the link [22:16] ah, that would do it >.> spelling error = no package found === seb128__ is now known as seb128 [22:17] ceekay: that would be an epoch, use 1:1.0.080416, but it can't be undone (you have to bump the epoch) so use it carefully :-) [22:17] ceekay: 'epoch' [22:18] ah thanks guys [22:18] just as you sent that i found http://people.debian.org/~calvin/unofficial/ which mentions epoch as well === ryu2 is now known as ryu [22:35] https://a248.e.akamai.net/f/674/9206/0/www2.ati.com/drivers/linux/catalyst_84_linux.html [22:45] mvo: no, haven't seen this.. have you looked into using "hald --verbose=yes --daemon=no"? [22:46] blueyed: not yet, I have to change the package for this, I think I will do it to see what it outputs [23:52] xhaker: thanks though I won't have time to review until sometime Friday :( === MacSlow_ is now known as MacSlow