20070329
Log
tz UTC+1
10:00 mdz cjwatson,kylem,doko,rtg,tkamppeer: ping 10:00 Keybuk mvo: ping 10:00 kylem yo 10:00 rtg lurking attentively. 10:00 dholbach mdz: doko_ might be a little late too 10:00 Keybuk (one might consider the futility of pinging someone who isn't online :p) 10:00 mdz dholbach mentioned mvo might be late 10:00 pitti Keybuk: summoning powers 10:01 cjwatson here 10:01 cjwatson doko said to me he would be late due to travelling, yes 10:01 mdz cjwatson: and till? 10:01 tkamppeter I am here 10:01 mdz ok, that's everyone then 10:02 cjwatson Ben also said he might be a little late due to needing to rescue his wife from traffic 10:02 mvo hello, sorry for being late 10:02 mdz cjwatson: BenC connected a few minutes ago 10:02 cjwatson er, no, that's rescue his kids from the bus stop due to wife stuck in traffic - something coherent anyway :) 10:02 cjwatson aha 10:02 BenC I'm on, I just have to get the kids in a few minutes :) 10:03 mdz any additions to the agenda since the reports came in? 10:03 bdmurray I wanted to talk about n-m and /etc/network/interfaces 10:03 cjwatson BenC: oh, I forgot about BST and misunderstood the times you gave 10:03 bdmurray and upgrading from edgy to feisty 10:03 tkamppeter Yes, I am searching volunteers for HPLIP 1.7.3 testing 10:03 mdz bdmurray: is there a bug number for that? 10:04 mdz bdmurray: please add to the wiki 10:04 bdmurray mdz: newz2000 had a bug about it and I've seen others I believe 10:04 fabbione bdmurray: ++ 10:04 mdz bdmurray: can you add tkamppeter's item as well? 10:05 mdz ok 10:05 mdz (pitti) "Report a bug" menu item -> disable for release? 10:05 pitti my two questions are mainly directed towards the desktop team and cjwatson, since they have by far the most bug reports, I guess 10:05 seb128 I think we should stop the crash handler and let the menu item 10:05 cjwatson pitti: I don't have a problem 10:05 dholbach I'm personally happy with the menu item - we get useful information through it. 10:06 cjwatson (as far as my own bugs are concerned) 10:06 mdz I just posted something about this to the list a minute ago 10:06 pitti My feeling is that we should stop reporting signal crashes, and maybe keep Python crashes 10:06 cjwatson I'd much, much, much rather get apport bugs than "er, it crashed, I can't install Ubuntu, help" 10:06 pitti dholbach: with the per-app one, or the general one in the System menu? 10:06 asac personally i like to keep crash reports for firefox 10:06 seb128 pitti: you will make mvo cry, he gets lot of python dups 10:06 BFTD agreed 10:06 pitti cjwatson: right, that was my gut feeling too; but for the installer they seem to be more relevant to me than, let's say, gnome 2.18 crashes 10:06 dholbach pitti: both 10:07 mvo I have mixed feeling, its great to get backtrace, but I'm afraid about the flood 10:07 seb128 we will get flooded 10:07 pitti for example, we get a lot of bugs without a package through the System menu 10:07 mdz we estimate there are millions of Ubuntu users 10:07 seb128 GNOME bugzilla got over 500 dups on some bugs from bug-buddy and 90% were Ubuntu bugs 10:07 pitti and we have to consider how many crashes we realistically fix in stables 10:07 cjwatson I'll certainly get flooded, *but* I'm going to get flooded *anyway* if the installer is crashing 10:07 iwj Can we make apport only work for certain apps, easily ? 10:07 mdz if we have a crash which affects 1% of them, and 1% of those report it, that's hundreds of bugs 10:08 seb128 there were getting over 1000 edgy crashes a week 10:08 cjwatson it's either get flooded with decent-quality bugs, or with poor-quality bugs 10:08 pitti iwj: theoretically yes, but hard to decide; I'd prefer classes like sigsegv vs. python 10:08 seb128 cjwatson: well, crashes are special 10:08 seb128 cjwatson: if nautilus crash once a day due to a bug we will get hundred of people sending it 10:08 mdz I think that leaving it enabled for crashes would be fine, if we could avoid filing duplicates for the same crash 10:08 pitti seb128: how many apport and normal bugs do you get for edgy, roughly? 10:09 mdz the existing guided filing doesn't seem to do a good enough job 10:09 pitti keeping in mind that edgy's apport was much harder to use 10:09 Keybuk mdz: "me too" is planned, but not implemented 10:09 seb128 pitti: hard to tell, most bugs we get now are already feisty 10:09 Keybuk (I wouldn't hold out hope of seeing anything like that pre-1.0) 10:09 seb128 pitti: I would say 100 a week some time ago 10:09 iwj "me too" would be really nice. It would help with the false duplicates, too. 10:09 pitti right, Malone needs to become much more insistive with identical subjects, etc. 10:09 pitti but I don't bet on it for feisty 10:09 mdz Keybuk: I think that's a different problem 10:10 seb128 we have no dup finding easy capability atm 10:10 cjwatson pitti: of course, it will help even if it gets implemented after Feisty 10:10 seb128 malone search don't look in comments nor attachments 10:10 asac if we really cannot deal with crash flood we should disable imo and set high priority to get a crash database for next stable release 10:10 cjwatson changes in Malone, that is 10:10 bdmurray mdz: I added the n-m thing but don't know about the till item 10:10 mdz people who are filing duplicates are probably doing it because they either don't see the existing bug, or don't know that the right thing to do at that point is just stop 10:10 mdz bdmurray: <tkamppeter> Yes, I am searching volunteers for HPLIP 1.7.3 testing 10:10 tkamppeter See bug 98520 ubotu Malone bug 98520 in hplip "Feisty UVF ER: New HPLIP 1.7.3 release fixes lots of bugs" [Medium,Needs info] https://launchpad.net/bugs/98520 10:10 pitti right, for feisty+1 we might have a more efficient retracer, automatic retracing, and automatic dup detection; we have none of them ATM 10:10 seb128 to be honest I prefer people filling duplicates 10:11 mdz seb128: really? why? 10:11 tkamppeter This is an UVF ER for HPLIP 1.7.3, as it fixes tons of bugs 10:11 seb128 on gaim we get people attaching their crashes to some random people pretty often 10:11 seb128 dunno why 10:11 bdmurray mdz: done 10:11 seb128 but we have lot of bugs with 10 differents crashes attached there 10:11 cjwatson seb128: agreed 10:11 seb128 because people came saying "I've the same bug, here is my crash" 10:11 mvo seb128: yes, I get quite a lot of those as well 10:11 mdz tkamppeter: ok, it is on the list and we can discuss it later in the agenda 10:11 pitti seb128: NB that people won't manually attach apport reports any more 10:11 cjwatson I have made the same comment many times in the past 10:12 tkamppeter As we are shortly before release we need intensive testing on this. 10:12 seb128 I keep replying "when you don't know what you are doing open a new bug and let busquad direct you" 10:12 cjwatson tkamppeter: later in the agenda, please 10:12 pitti seb128: that was a specific problem of edgy's 'please attach me' apport, wasn't it? 10:12 tkamppeter sorry, I thought mdz was calling me. 10:12 seb128 pitti: dunno 10:12 mdz pitti: I think so 10:13 seb128 pitti: but I prefer people to open a dup than trying to figure themself if that's a dup 10:13 pitti seb128: right, the goal is that Malone figures out whether it is a dup 10:13 seb128 they get it wrong often and that's really confusing 10:13 bdmurray bughelper could help in the meantime with dupe detection right? 10:13 asac seb128: ack ... i want every crash report in its own bug ... then retrace and mark dupes afterwards. 10:13 mdz my feeling is that bug reports are not something that every user should deal with. 10:13 dholbach if people submitted their clue files :) 10:13 bdmurray dholbach: indeed 10:14 BFTD can we have fontforge a requiarment to install WINE? 10:14 mdz only users who are closer to the community and will participate in analysis should file bugs, generally 10:14 pitti mdz: mpt had a strong point about the general 'I filed a bug, and noone answered' picture 10:14 iwj I think I'm with mdz here. 10:14 seb128 bughelper is not good at dupfinding atm 10:14 pitti therefore I agree with mdz 10:14 mdz BFTD: the meeting agenda is at https://wiki.ubuntu.com/DevelTeamMeeting20070329, we're on item 3 10:14 mvo does bughelper scale well enough for this task? 10:14 BFTD mdz sorry 10:14 dholbach mvo: no 10:14 seb128 you can't look for a sequence of functions 10:14 seb128 mvo: no 10:14 iwj People who don't have the effort/knowledge/connection to get involved should perhaps be using the support system. 10:14 seb128 looking for a string is no enough 10:15 mdz if only a relatively small portion of users should file bugs, we shouldn't point every user to the bug tracker 10:15 cjwatson (bughelper can be improved post-feisty) 10:15 mdz we do want the crash reports, but until we have a better place to put them than Malone... 10:15 seb128 right 10:15 pitti mdz: I don't even thing that Malone is a bad place for it; we just need better tools to deal with them 10:15 fernando hey all 10:15 asac yeah ... imo keep crashes out of BTS ... so disable until we have crash db 10:16 seb128 pitti: it's bad, too much noise 10:16 seb128 we should have a crash collector and something making easy to move confirmed crashes to a bug 10:16 asac pitti: i think its a bad place as well 10:16 cjwatson from the disagreement in the team, I think we may need some kind of per-package solution 10:16 seb128 so we don't even have to bother looking what to dup 10:16 pitti and I think that in this cycle we were able to fix a lot of bugs in feisty with the current system; we should continue to do so in feisty+1 for feisty+1 instead of trying to fix crashes in stables 10:16 cjwatson perhaps implemented something like the existing hooks mechanism 10:17 iwj I don't know what the density of real crasher bugs in gnome is but I get a nonzero number of dpkg crash reports which are basically "my hardware is bust". 10:17 BFTD or give the user the option to install/use the reporter in the installtion 10:17 seb128 the best for GNOME would be to turn on bug-buddy 10:17 pitti seb128: ok; we have that spec already, so we need to talk about it again at UDS 10:17 Keybuk iwj: that'd surprise me, since we'd get them for everything 10:17 seb128 but backtraces are screwed on feisty, we can't send bugs upstream 10:17 Keybuk are you sure that they're not just deep dpkg bugs? :p 10:17 cjwatson Keybuk: we do get them for ubiquity ... 10:17 heno can we allow users to turn on apport manually if they see a recurring bug? that way we could ask for it from users we know will follow up etc. or ask it iof someone who has already filed 10:17 iwj I investigate them and it turns out the machine had toasted filesystems. 10:17 Keybuk I've never had them for upstart ;) 10:17 mdz seb128: why would bug-buddy be better? 10:18 pitti heno: yes, that's easy 10:18 cjwatson but that's ok, I have Tollef's greasemonkey script to help me reject them in about five seconds 10:18 Keybuk (I've never actually had a crash report at all for it) 10:18 pitti heno: well, it would be with flipping /etc/default/apport 10:18 seb128 mdz: because bugzilla.gnome has more bug triager, they have a decent dup finder, they are used to it, and 98% of crashes are due to upstream 10:18 heno pitti: so perhaps that should be our policy? 10:18 iwj They're not common but then dpkg is not the biggest memory hog out there. I just wonder how many other crashes there are that are due to slightly-dodgy machines. 10:18 pitti so if we ask an user to switch that on, we can still get selected reports for feisty 10:19 mvo iwj: I have seen quite a few in apt (well, some are real I suppose :) 10:19 heno pitti: you mean selected in a good way or bad way (skewed sample)? 10:19 iwj mvo: *snort* 10:19 pitti heno: in a way to get quantities we can cope with, together with a commitment from the reporter to give feedback 10:19 mdz heno: it would be great if we could find a way to turn it on only for responsive users 10:20 doko_ joining 10:20 iwj `To qualify for this service, you must be at least > < this competent.' 10:20 mdz we want to be able to use it to assist bug reports which would otherwise take more time to process, without getting reports which are useless 10:20 heno if we make it easy (which pitti says it is) we could request it in bug comments regularly 10:20 pitti iwj: it could check your LP karma :) 10:20 heno better than 'please upload /var/log/... 10:21 heno well, if someone has already filed a bug without the system asking them too they are more involved than the average user 10:22 mdz heno: right 10:23 mdz hmm 10:24 pitti hm, so am I right that the opinions lead towards keeping Python (for ubiquity & friends) and disabling signal crashes by default, with the possibility of enabling it again in /etc/default/apport? 10:24 mdz pitti: I assume we can have it store the crash file, but not notify the user to submit it? 10:24 pitti asac: you had two slightly different statements; what do you think about the firefox front? 10:24 mdz pitti: if so, when a user files a bug, we could give them a command to run to submit the crash to the bug, right? 10:24 asac i am unsure ... on the one hand I want to get feedback so i see if there are regressions for some users 10:24 heno (I support turning it most off too -- have no opinion about python or other categories) 10:24 pitti mdz: yes; that would require minor code changes, but no problem 10:25 asac especially since we upload new upstream reports 10:25 mdz pitti: that seems like the best of both worlds where crashes are concerned 10:25 cjwatson pitti: we could also make it behave differently on the live CD 10:25 asac otoh dealing with that flood of crashes is hard in malone 10:25 cjwatson (by tweaking stuff in casper) 10:25 mdz pitti: we wouldn't get the pre-filled summary... 10:25 cjwatson for installer crashes I see no value in giving the user a command versus the notification icon 10:26 mdz cjwatson: argeed 10:26 mdz agreed 10:26 pitti mdz: oh, I thought you meant something like 'run apport-gtk by hand' to collect and report pending stuff in /var/crash 10:26 mdz I think the installer is a special case, though 10:26 mdz because the crash reports are lost 10:26 cjwatson it is, but I need it to be addressed :) 10:26 cjwatson sorry to keep banging on about it 10:26 pitti so we would mainly turn off the auto-spawning in update-notifier 10:26 mdz cjwatson: do you want crash reports to remain on for the installer? 10:26 cjwatson mdz: yes 10:26 mdz cjwatson: I don't think anyone here would object to you receiving more reports on ubiquity if you want them 10:26 cjwatson :-) 10:26 pitti Python crash reports are really neat 10:26 mdz let's put that to rest, then, shall we? 10:27 pitti cheap in terms of bandwidth, and easily dup'able 10:27 mdz crash reports for ubiquity in 7.04 - thumbs up 10:27 pitti *idea* 10:27 seb128 well, there is the same flood with them 10:27 cjwatson the remaining question is whether that be done for all python scripts, or just for the live CD 10:27 pitti what about simply adding an option to update-notifier? 10:27 mdz seb128: but it's an isolated flood for colin, and he wants it 10:27 mvo pitti: a gconf key? easy enough 10:27 pitti turn that off by default, and provide a moderately hidden switch to turn it on again? 10:27 seb128 mdz: I was speaking about python crashes in general, not only ubiquity 10:27 seb128 I think ubiquity is ok 10:27 mdz seb128: ok 10:28 cjwatson hence my suggestion of doing it in a package hook or similar 10:28 seb128 I've looked at all the unconfirmed apport bug in malone last week 10:28 seb128 and we have hundred of them already 10:28 cjwatson look, we seem to be going round in circles, and we're 28 minutes in 10:28 pitti cjwatson: casper could flip that gconf key for the live session, too 10:28 seb128 and a good bunch of them were python crashes 10:28 cjwatson can we all agree to thrash this out on ubuntu-devel? :-) 10:28 seb128 ok 10:28 pitti ACTION: pitti to summarize discussion in #u-devel and continue there 10:28 janimo pitti: the python backtraces for system-config-printer were all very helpful FWIW 10:28 asac maybe we can move things to mailing list instead of irc? 10:28 seb128 all the crash files are really useful 10:28 mdz I think we should briefly discuss the menu item in the meeting, since that's a separate issue 10:29 pitti so we covered my second TOP now 10:29 cjwatson pitti: erm - mailing list rather than IRC? 10:29 cjwatson or maybe IRC would work, dunno 10:29 pitti cjwatson: erm, of course 10:29 seb128 the thing is that after some time you get the sames crash again and again and not a lot of new 10:29 seb128 and the noise increase 10:30 mdz crash reports -> ubuntu-devel@ 10:30 pitti bdmurray: I take it you get a lot of bug reports without a package through the System 'report a bug' menu? 10:30 mdz menu item -> here 10:30 mdz bdmurray: yes, tell us about what you see as a result of that menu item 10:30 bdmurray pitti: I'm not sure what would be the give away that it was from that menu item? 10:30 mdz bdmurray: it should include a footer 10:30 pitti bdmurray: the apport info in them 10:30 pitti about Ubuntu version, uname, etc. 10:31 pitti of course the user can select a package in malone, but many of them don't 10:31 pitti and mpt raised the point that getting tons of unanswered bugs neither helps the reporters nor us 10:31 bdmurray I haven't counted that much but don't think there are a lot and the Ubuntu version is helpful because people do forget that a lot. 10:31 mdz especially since it's called "report a problem" 10:31 mdz and users have many problems which aren't bugs 10:32 mdz including problems which have nothing to do with Ubuntu 10:32 mdz bear in mind that OEMs will preinstall Ubuntu in mass quantities 10:32 mdz users who get Ubuntu and don't even know what it is could end up reporting this as a "problem" with their computer 10:33 pitti I for myself get more and more 'unqualified' reports now, without any answer in many cases 10:33 pitti the menu item in the application's help menu (lp-i) might be better, it already has the package and such, and more likely to get a proper description 10:34 seb128 right 10:34 mdz pitti: and it can call package hooks 10:34 pitti right 10:34 mdz I propose that we disable the item in the System menu 10:35 bdmurray and it is slightly more obscured too 10:35 mdz does anyone feel that we'd be losing out on better bug reports by doing that? 10:35 bdmurray so people would have to be looking for help 10:35 pitti we should rather improve the Firefox start page or help center and have them point to wiki, etc. 10:35 pitti and the more community-oriented people will find Launchpad just as well as they did so far 10:36 cjwatson that would be a pretty small change to the Firefox start page, I think, and that's something OEMs can and probably (without proof) do customise 10:36 cjwatson I'm with mdz here 10:36 seb128 mdz: the problem is that only a small set of app are patched for lpi 10:36 pitti right, you cannot use it to report a postgres bug, but *shrug* 10:37 seb128 maybe moving the panel one to applications, accessories? 10:37 asac btw, the firefox startpage does not ship in firefox package 10:37 seb128 or system, admin ;) 10:37 pitti seb128: you hit me when I did that :) 10:37 mdz seb128: are you getting useful bug reports from it? 10:37 cjwatson asac: indeed, ubuntu-artwork et al 10:37 dholbach ubuntu-docs 10:38 pitti anyway, any strong opposition to the current proposal? 10:38 seb128 mdz: no, but the apps I'm working on are using lpi 10:38 mdz seb128: we do have ubuntu-bug now, though it needs to be more widely advertised 10:38 seb128 fair enough 10:38 pitti we even have a command-line apport frontend for servers now ;) 10:39 mdz consensus to disable System->Report a problem, then? 10:39 seb128 I think having the command line command and lpi is good enough 10:39 mdz pitti: oh, neat 10:39 seb128 +1 10:39 asac +1 10:39 Mithrandir mdz: yes, I think that makes sense. 10:39 pitti +1 10:39 mdz ok 10:39 bdmurray mdz: I'm running a bughelper query for those bugs and a lot are showing up. Ones report by that menu item that is. 10:39 kwwii yippee! that fixes my one bu for me 10:39 pitti ACTION: pitti to disable system apport menu entry 10:39 mdz as I wrote on the mailing list, we should revisit it when we have better infrastructure in place for guiding the user, and handling the reports on the server side 10:40 bdmurray but it would me more useful if they had an application in them rather than going "nowhere" per se 10:40 pitti sorry for the long discussion, but I felt it is important to collect info === dholbach hugs pitti 10:40 mdz pitti: can you add this to the agenda for apport refinements at UDS? 10:40 pitti mdz: definitively 10:40 mdz (bdmurray) What is the best / most effective way to communicate bugs I think might be important? e-mailing the distro team? 10:40 mdz pitti: thanks === cjwatson proposed milestoning for things we need to look at before release 10:41 pitti mdz: 'that' -> bdmyrray's last comment, I take it? 10:41 cjwatson we can always unmilestone if need be 10:41 Mithrandir bdmurray: milestone + assign. 10:41 mdz pitti: no, my last comment (and the corresponding post to -devel) 10:42 mdz pitti: a discussion about generally how the user should report bugs, when, which tools, etc. 10:42 pitti right 10:42 mdz bdmurray: I see this as a release management function 10:42 pitti bdmurray: more email--; bug assignment + IRC pings work fine for me, personally 10:43 bdmurray IRC can be challenging with the time difference though 10:43 mdz bdmurray: so if you feel a bug is important enough to be tracked separately, it should be escalated to Mithrandir 10:43 Mithrandir bdmurray: we should ship you to Europe. 10:43 iwj irc pings aren't always very reliable. 10:43 mvo bdmurray: I like milestone + subscribe 10:43 bdmurray indeed 10:43 pitti bdmurray: but we already have bug email for the asynchronous discussions 10:43 Mithrandir iwj: for others, it's much more reliable than anything else. 10:43 iwj Mithrandir: That's scary to me. 10:44 cjwatson the problem with IRC pings for anything that needs attention urgently is that if the problem is actually urgent then they can be quite disruptive 10:44 cjwatson you need to stop what you're doing in case you forget 10:44 seb128 I like IRC pings 10:44 Mithrandir iwj: why? it feels like I'm subscribed to half of the packages in the archive; there's no way I'll never miss a thing there. Pings show up in my session. 10:45 seb128 I set the milestone, bug settings, etc as required and switch back to what I was doing usually 10:45 iwj Maybe your network/client/stuff is more reliable. I find I occasionally miss them. 10:45 seb128 it's not too disruptive and I know about the bug then 10:45 seb128 iwj: well, if the other side doesn't reply I consider he didn't read it and try again later ;) 10:45 Keybuk heh, I have to agree with IRC here 10:45 pitti bdmurray: I think we recently got better bug mail headers, so that it should be possible to separate out the subscriptions from other people from the mass of bug mail from +packagebugs 10:45 Mithrandir cjwatson: so it depends on the person. I don't develop very much software any more, I spend a fair amount of my time talking to people on IRC doing release management. 10:45 Keybuk I just about don't read bug mail anymore 10:45 iwj seb128: Ah, IC. That makes much more sense now :-). 10:45 Keybuk since it takes longer to read it than it takes for more to arrive 10:45 mdz seb128: are you needing a standard method which works for the entire team, or are you asking what to do with bugs where you don't already know who to talk to and how? 10:45 cjwatson Keybuk++ 10:46 Keybuk I've been unsubscribing from things as much as I can, but haven't yet got it down to a level I can actually *keep up* with 10:46 Keybuk (so if you've subscribed/assigned a bug to me ... I don't actually know about it yet <g>) 10:46 pitti mdz: that was meant for bdmurray, I take it === Keybuk is somewhere in February right now, and that's even after heavy purging 10:46 mdz pitti: er, yes 10:46 mdz bdmurray: ^^ s/seb128/bdmurray/ 10:46 Mithrandir Keybuk: the fun starts when teams are subscribed to bugs, and they should be so you can't just unsub them. 10:46 bdmurray mdz: more the latter 10:46 Keybuk Mithrandir: this is part of my problem 10:47 mdz ok, so let's not discuss personal preferences then 10:47 bdmurray for example bug 95821 seems pretty important ubotu Malone bug 95821 in mysql-dfsg-5.0 "Replication failure with auto-increment and on duplicate key update" [Undecided,Unconfirmed] https://launchpad.net/bugs/95821 10:47 mdz the question is what to do about orphan bugs 10:47 bdmurray but I have no idea where to go with it 10:47 iwj I think we may need to hurry this along a bit ? 10:47 mdz I've suggested that they go to the release team 10:47 pitti bdmurray: that sounds like upstream is the single appropriate target 10:47 cjwatson if there's nobody obvious to go to, then set the milestone and the RM can figure it out :-) 10:48 Mithrandir bdmurray: as the rest of the people here says; if you're unsure, milestone it. unmilestoning is easy. 10:48 mdz Mithrandir: does that seem sensible to you? 10:48 seb128 milestone them and encourage people to look at the milestoned bug every now and then? 10:48 mdz seb128: Mithrandir will do that regularly 10:48 seb128 or let RM find the people ;) 10:48 seb128 ok, works for me 10:48 bdmurray sounds good 10:48 Mithrandir I have said it before and I'll probably say it again: Please milestone bugs which you believe or think may be RC. If you're unsure, ask me, or milestone anyway. 10:48 mdz Mithrandir: maybe the QA docs should say something about this if they don't already 10:49 mdz (bdmurray) bug 63365 and bug 76901 ubotu Malone bug 63365 in xorg "[feisty] xorg incorrectly detects "lv3:ralt_switch" option in xorg.conf (dup-of: 76901)" [Undecided,Confirmed] https://launchpad.net/bugs/63365 ubotu Malone bug 76901 in console-setup "right Alt key not recognized by default (lv3:ralt_switch shouldn't be set by default for US keyboard layouts)" [High,Confirmed] https://launchpad.net/bugs/76901 10:49 Mithrandir mdz: I don't know if they do; I'll make sure they do. 10:49 bdmurray mdz: that's not so important more just a reminder 10:49 Mithrandir ACTION: tfheen to make sure QA/bug triage information includes information about when to milestone. 10:49 iwj bdmurray: Err, so it doesn't need discussion here ? 10:49 cjwatson indeed, the first is a dup of the second and the second is milestoned 10:49 bdmurray iwj: no not really my mistake 10:50 cjwatson no discussion here needed I think unless there's anything further on it 10:50 iwj Fair enough. 10:50 Mithrandir malone's handling of duplicates is slightly icky, but that's mostly my problem. 10:50 mdz the right alt key on US keyboards is labeled "Alt" and should be equivalent to the left Alt key 10:50 iwj I like quick agenda items. 10:50 mdz it's assigned to cjwatson 10:50 mdz done 10:50 mdz (bdmurray) network-manager and /etc/network/interfaces when upgrading from edgy to feisty for example bug 96521 ubotu Malone bug 96521 in network-manager "network manager won't let me connect to wifi" [Undecided,Rejected] https://launchpad.net/bugs/96521 10:50 mdz bdmurray: is this the same issue that SteveA brought up? 10:51 bdmurray mdz: I'm not sure I saw thought. But I believe n-m wasn't the default in Edgy. So people may have stuff configured in /etc/network/interfaces that confuses n-m. 10:51 mdz SteveA pointed to https://beta.launchpad.net/ubuntu/+source/network-manager/+bug/87941 10:51 bdmurray after upgrading ubotu Malone bug 87941 in network-manager "network-manager does not take into account static configurations *properly* (dup-of: 82335)" [High,Confirmed] ubotu Malone bug 82335 in network-manager "network-manager should not set offline mode when it manages no device" [High,Fix released] 10:51 Mithrandir bdmurray: s/confuses/makes NM ignore the interface/, yes. 10:51 mdz I guess that's a different issue 10:52 fabbione and n-m breaks ipv6 10:52 bdmurray and does the upgrade process take this into account at all? 10:52 iwj Is there something controversial here ? (Still loading bugs pages over gprs ...) 10:52 mdz bdmurray: I think we need to take this to ubuntu-devel 10:52 Mithrandir fabbione: that's something I don't really worry or care about release-wise. 10:53 fabbione Mithrandir: it's a regression from edgy so you do care.. 10:53 Mithrandir fabbione: no, I don't. Just don't use NM if you're one of the ten people in the world using IPv6. 10:53 mdz bdmurray: I think the short answer is that the UI should make it obvious when NM is ignoring the interface due to manual configuration, and allow the user to get quick access to the checkbox to set it back to automatic 10:53 fabbione Mithrandir: i can't remove NM without removing ubuntu-desktop.. 10:53 bdmurray mdz: why is that? I'm trying to understand what is appropriate for the distro-team vs ubuntu-devel 10:54 Mithrandir fabbione: you can disable it. 10:54 Mithrandir mdz: yes, I'm going to tweak NM to do that. 10:54 cjwatson bdmurray: distro-team => private 10:54 mdz bdmurray: distro-team = Canonical, ubuntu-devel = Ubuntu developers 10:54 fabbione Mithrandir: last time i asked you how to, you told me you couldn't.. 10:54 Keybuk Mithrandir: 10 is on the generous side 10:54 Mithrandir fabbione: rm /etc/dbus/event.d/25NetworkManager 10:54 mdz bdmurray: it's not that it's inappropriate for this meeting, it's that we only have 5 minutes left and can't discuss it in that time 10:54 bdmurray mdz: okay, that's fair 10:55 Mithrandir bdmurray: I'll be happy to discuss it further with you, I don't think it's important for the whole team? 10:55 mdz ACTION: bdmurray to summarize network-manager/ifupdown issues on ubuntu-devel for further discussion 10:55 mdz (cjwatson) Bug day on Friday: distro team participation 10:55 cjwatson just a reminder to everyone, since last Friday was beta day 10:56 cjwatson it's not actually been announced on u-d-a apparently - dholbach? 10:56 mdz cjwatson: reminder to what? join #ubuntu-bugs? 10:56 dholbach I thought it was... bdmurray? 10:56 cjwatson but anyway, distro team staff should make an effort to turn up on #ubuntu-bugs and be a bit more vocal about what you're doing to stimulate discussion and educate people 10:56 cjwatson and generally help out 10:57 cjwatson done 10:57 mdz the community QA team seems to be growing and healthy 10:57 iwj I could volunteer our poor tender bug helpers to help with bug 75681. ubotu Malone bug 75681 in mdadm "boot-time race condition initializing md" [High,Confirmed] https://launchpad.net/bugs/75681 10:57 iwj (ducks) 10:57 mdz I'm seeing good actions from new members 10:57 mdz iwj: talking about problems like that during bug days will increase awareness of the problems and how to debug them 10:57 mdz ok 10:57 cjwatson iwj: you could certainly talk them through your thought processes trying to debug it, although they might run away :) 10:57 dholbach cjwatson: apparently it was not announced, I'll whip something up 10:57 mdz (tkamppeter) I am searching volunteers for HPLIP 1.7.3 testing 10:57 iwj cjwatson: :-) 10:58 mdz tkamppeter: my OfficeJet is in a shipping crate somewhere between here and London 10:58 tkamppeter See bug 98520 ubotu Malone bug 98520 in hplip "Feisty UVF ER: New HPLIP 1.7.3 release fixes lots of bugs" [Medium,Needs info] https://launchpad.net/bugs/98520 10:58 Keybuk my LaserJet doesn't appear to work with HPLIP 10:58 bdmurray I have one here I can test 10:58 tkamppeter HPLIP 1.7.3 has many fixes to bugs reported to us. 10:59 bdmurray in my spare time 10:59 tkamppeter As we are shortly before releasing, we need test reports from as many people as possible. 10:59 mdz tkamppeter: send mail to ubuntu-devel-discuss 10:59 mdz that's where you can reach people who are running feisty 10:59 doko tkamppeter: 1.7.3 works for me, scanning included 10:59 mdz adventurous sorts 10:59 mdz Mithrandir: beta blockers? 10:59 mdz (last week's actions) 10:59 tkamppeter So please do the tests described in my 2nd and 3rd posting and post your results to the bug report. 10:59 cjwatson if that's not done, it's irrelevant :P 10:59 mdz bdmurray: wiki documentation cleanup? 11:00 Mithrandir mdz: too many, but going down, slowly. I have utterly forgot to send out the list, so I'll do that tomorrow. 11:00 cjwatson unless s/beta/release/ 11:00 tkamppeter Thanks for your help. 11:00 mdz cjwatson: yes, we still have blockers :-) 11:00 bdmurray mdz: I slipped on that but am working on it actively now 11:00 mdz ok 11:00 mdz Mithrandir: please send that out tomorrow 11:00 mdz any other VERY BRIEF business? 11:00 Mithrandir mdz: will do. 11:01 mdz adjourned 11:01 mdz thanks, all
MeetingLogs/UbuntuDev/20070329 (last edited 2008-08-06 16:41:22 by localhost)