2008/05/11/#ubuntu-bugs.txt

[00:01] <secretlondon> should anything that crashes with a segfault give a core dump in /var/crash? All I can find is a mention in syslog
=== gnomefre1k is now known as gnomefreak
[00:05] <hggdh> secretlondon: if apport is active, it should -- see /etc/default/apport
[00:06] <secretlondon> hggdh: apport has been disabled on hardy afaik
[00:06] <hggdh> secretlondon: yes, but if you want/need it, you can re-activate it by editing /etc/default/apport
[00:07] <secretlondon> hggdh, ah thanks! So I need that to get core dumps?
[00:07] <hggdh> secretlondon: that would help, yes ;-) the good part of using apport is that it will try hard to give you a meaningful bt
[00:08] <hggdh> (by a selective use of apport-gtk and apport-retrace)
[00:09] <secretlondon> hggdh, thanks, I've re-enabled
[00:09] <hggdh> secretlondon: welcome
[00:10] <secretlondon> I think a segfault bug without a trace at all is pretty useless
[00:10] <hggdh> well, er, yes I would say. Kind of difficult to zero in the failure... :-)
[00:11]  * secretlondon is almost tempted to close her own bug with Please Try Harder
[00:11] <hggdh> noelferreira: if you have not done so, please open a bug in launchpad (https://bugs.launchpad.net)
[00:12] <secretlondon> hggdh, we've tried talking to him, he doesn't respond, he's been posting the same thing repeatedly for several days
[00:12] <hggdh> noelferreira: Falas Portugues?
[00:12] <noelferreira> i tried but my message from register won't come. :( hggdh
[00:13] <noelferreira> estou à espera do email do registo hggdh
[00:13] <noelferreira> hggdh, mas este bug está a dar.me cabo do juizo
[00:13] <noelferreira> já desactivei o acpi mas acho que não é isso
[00:14] <noelferreira> hggdh, às vezes estou a escrever e quando vou a ver metade das letras não estão lá.
[00:14] <noelferreira> outras parece que as teclas ficam presas
[00:14] <noelferreira> não sei se é do pgdown pgup ou do shift
[00:14] <noelferreira> é estranho
[00:14] <noelferreira> agora só acontece as vezes, por exemplo
[00:15] <hggdh> noelferreira: hum. está o teclado limpo?
[00:15] <noelferreira> sim
[00:15] <noelferreira> portatil, amd64 com compiz e ati
[00:15] <noelferreira> mas metacity é o mesmo, nada a ver com compiz
[00:16] <noelferreira> tb já andei a mexher no xorg.conf mas nada
[00:16] <noelferreira> até já alterei os locales
[00:16] <noelferreira> já fiz 1001 coisas
[00:16] <hggdh> noelferreira: vosso pastebin não me mostra muito... qual tecla é a tal de keycode 240?
[00:16] <noelferreira> parece que fica bom mas derepente o problema volta :)
[00:16] <noelferreira> hggdh, não sei, como vejo isso?
[00:17] <hggdh> hah... talvez melhor se voltarmos ao Ingles -- não me recordo de como verificar a maldita tecla
[00:18] <hggdh> I should know, but I forgot -- anyone: what programme we should use to return the keycodes?
[00:18] <hggdh> My memory has gone south, I am afraid
[00:19] <secretlondon> its on the wiki
[00:19] <secretlondon> somewhere
[00:19] <hggdh> xmodmap?
[00:20] <secretlondon> https://wiki.ubuntu.com/DebuggingKeyboardDetection
[00:20] <hggdh> secretlondon: as usual IOU
[00:20] <secretlondon> https://wiki.ubuntu.com/LaptopTestingTeam/HotkeyResearch is better
[00:21] <hggdh> noelferreira: please look at the above link ^^
[00:21] <noelferreira> ok thanks
[00:22] <hggdh> and noelferreira: por favor, assim que possivel (e se necessario for)... https://bugs.launchpad.net :-)
[00:22] <noelferreira> é agora hggdh já recebi o email
[00:22] <noelferreira> :)
[00:23] <secretlondon> hggdh, you rock!
[00:23] <hggdh> thanks, noelferreira. With a bug opened thinks will be much more easier
[00:23] <noelferreira> sure
[00:23]  * hggdh blushes 
[00:31] <noelferreira> hggdh, done
[00:31] <noelferreira> https://answers.launchpad.net/ubuntu/+question/32704
[00:31] <noelferreira> maybe i put there to much paste :)
[00:33] <noelferreira> hggdh, i used to have other bug bigger than this one. and i think they are related. that's why i thought it was a acpi issue hggdh.
[00:34] <hggdh> noelferreira: noprob. just remember, one bug per report, one report per bug. Your pastebin showns a keypress in the beginning, and no keyrelease. It might be a stuck key indeed
[00:36] <noelferreira> hggdh, didn't analyse the output deeply. you are right
[00:36] <noelferreira> but it is related so....
[00:48] <secretlondon> hggdh, this desktop doesn't get the e-d-s bug btw
[00:48] <secretlondon> my laptop gets it 90% of the time..
[00:48] <hggdh> secretlondon: this is the bloody hell of it! What is the difference?
[00:48] <secretlondon> hggdh, this doesn't have my google calendar set up in it
[00:49]  * secretlondon shrugs
[00:49] <secretlondon> both 1 cpu
[00:49] <secretlondon> 32 bit i386
[00:49] <hggdh> secretlondon: I set up a google calendar for me. No problems so far...
[00:49] <mrooney> secretlondon: any reason you marked bug 229105 as a duplicate, instead of leaving it Invalid?
[00:49] <ubottu> Launchpad bug 229105 in update-manager "Updating to 8.04 from 7.04 and it said "The upgrade aborts now" (dup-of: 229106)" [Undecided,Invalid] https://launchpad.net/bugs/229105
[00:49] <ubottu> Launchpad bug 229106 in update-manager "Updating to 8.04 from 7.10 and it said "The upgrade aborts now"" [Undecided,Confirmed] https://launchpad.net/bugs/229106
[00:50] <hggdh> OK. out for dinner now. I will be back.
[00:52] <secretlondon> mrooney, because I didn't see that it had been set as invalid
[00:53] <mrooney> secretlondon: is one more appropriate than the other? I have seen cases where it is obvious the person made a typo and refiled the bug, should the first be Invalid of marked as Dup?
[00:53] <secretlondon> I generally set them as duplicates
[00:54] <secretlondon> so that they are linked together
[00:54]  * sectech takes today off for bug triaging
[00:54] <mrooney> it seems to me Invalid is more appropriate as there is no need to make the bug appear more prominent by essentially marking it a duplicate of itself
[00:54] <sectech> from rather
[00:55] <mrooney> but if dup'ing is the accepted thing I can start doing that I guess
[00:55] <secretlondon> I don't know what the policy is
[00:55] <secretlondon> I think by settingas invalid you lose the connection to the master bug
[00:56] <mrooney> well yeah I put in the comments that it appears to be the same as #x
[00:56] <mrooney> by the way I am not intending to argue with you, just discussing :)
[00:57] <secretlondon> I don't know tbh
[00:57] <secretlondon> I'm not strongly fighting my side, that's just how I do them
[01:01] <hggdh> if 229105 was filed in error, and the reporter then filed in 229106, then indeed there is no need of marking it as a duplicate
[01:01] <hggdh> not that it will make much of a difference, anyway
[02:19] <zyx386> what is happning with firefox
[02:19] <secretlondon> you got problems with firefox?
[02:19] <zyx386> with ff3 is clear type active?
[02:20] <secretlondon> I don't actually know what clear type is
[02:21] <zyx386> that is exampe with ff2 http://img410.imageshack.us/img410/3976/ff2yy7.png
[02:21] <zyx386> and this with ff3 http://img152.imageshack.us/img152/1408/ff3zg5.png
[02:23] <zyx386> that is deference
[02:29] <zyx386> ?
[02:33] <secretlondon> it's a problem with our arabic support?
[02:33] <zyx386> no
[02:34] <zyx386> is cleartype von ie7
[02:35] <secretlondon> I honestly don't know
[02:35] <zyx386> von=from
[02:35] <secretlondon> zyx386, asac is the expert on firefox
=== gnomefre1k is now known as gnomefreak
=== wolfger__ is now known as wolfger
[08:04] <Iulian> G'morning
=== dodger_ is now known as Hurtz
=== fdd-0 is now known as fdd
[10:46] <afflux> morning
=== narcan_ is now known as narcan
[11:32] <zyx386> hi
[11:33] <zyx386> why you removed unixtime format uunder Hardy?
[11:33] <zyx386> you help say right click  clock aplet 12,24,unixformat
[11:33] <zyx386> but is't so
[11:34] <zyx386> is just 12 and 24
=== _neversfelde is now known as neversfelde
[11:54] <qense> hello
[13:26] <Laney> Do I need some kind of permission to set bugs to "Won't Fix"?
[13:29] <afflux> yes
[13:30] <afflux> Laney: this is restricted to members of ubuntu-bugcontrol. You still can ask here if you need to set this status, someone else may do it for you.
[13:30] <Laney> afflux: Right, thanks. I was just wondering why it wasn't coming up, that's all
=== asac_ is now known as asac
[13:53] <afflux> sudo currently takes the hostname from /etc/hostname and searches for an IP for it in /etc/hosts
[13:53] <afflux> if someone now sets a "domain" in networkmanager, the hostname stays the same but the entry in /etc/hosts gets adjusted so sudo won't find an entry
[13:53] <afflux> is this a known bug (or feature?)
[13:56] <persia> afflux: Are you sure it only looks in /etc/hosts, and doesn't use /etc/nsswitch.conf to determine how to get the hostname?
[13:56] <afflux> hum, not sure
[13:57] <afflux> but I have a user in #ubuntu-de complaining about having a $hostname and a $domain set, and sudo not working in hardy because it can't determine the $hostname
[13:58] <persia> My first guess would be that they set a domain, but that their host wasn't registered to that IP address in that domain according to the nameservice from which they collect information (likely global DNS, but there are other options, including /etc/hosts)
[13:59] <afflux> hm, I'll be back in 30 mins ;)
[14:00] <persia> Alternately, I could misunderstand, and they could just not have a "search" entry in /etc/resolv.conf.  Unfortunately, I'm not sure where that would need to be added to be automatically included in that file.
[14:32] <afflux> persia: I don't really know how sudo works and where it gets hostnames/ips/whatever from, but the thing is that a user changed his "domain" field in NM after an installation and now sudo complains about not being able to resolve the host
[14:33] <afflux> persia: I just checked and I can confirm that
[14:33] <afflux> I put "foo" as hostname and "bar" as domain, running "hostname" prints "foo", and /etc/hosts contains "127.0.1.1 foo.bar" (but no entry for "foo" alone)
[14:33] <persia> afflux: Right.  That's a bug.  What I'm not sure about is whether the issue is that NM isn't altering the resolver configuration, or sudo isn't checking the domain.
[14:34] <persia> Alternately, it could be that sudo isn't actually using the standard resolver, and looking in /etc/hosts, but I'd find that least likely.
[14:34] <afflux> true
[14:35] <afflux> should I file a bug?
[14:35] <persia> In summary, while I agree it's a bug, I don't know whether it ought be assigned to NM or to sudo.
[14:35] <afflux> hm okay
[14:35] <persia> It makes sense to file a bug.  If you can track down which is the right package, you'd get extra points.
[14:38] <albert23> persia, afflux: the sudo bug has been fixed
[14:38] <albert23> the fix is in sudo (1.6.9p10-1ubuntu3.1) hardy-proposed
[14:38] <persia> albert23: Ah.  Excellent.  Thanks.
[14:38] <afflux> albert23: you mean it's bug 32906?
[14:38] <ubottu> Launchpad bug 32906 in sudo "sudo fails if it cannot resolve the local hostname and no MTA is installed" [High,Fix released] https://launchpad.net/bugs/32906
[14:39] <albert23> afflux: right
[14:39] <afflux> I'm running exactly that version
[14:40] <afflux> albert23: If I read that one correctly, it's not about the domain field of the fqdn but rather only the hostname
[14:41] <albert23> afflux: At  least sudo should not fail anymore if the hostname cannot be found
[14:41] <afflux> hum
[14:41] <afflux> I'm gettint the impression it's an NM bug
[14:42] <afflux> because "hostname --long" currently fails with: hostname: Unknown host
[14:42] <afflux> not entirely sure if that's related, though
[14:43] <afflux> albert23: ah, right, sudo does not fail, but it still prints the message. It failed for the user I spoke with, because he had a clean install
[14:43] <persia> afflux: Reading the the 32906 traffic, I'm inclined to agree with you.  There appear to be two bugs: sudo was failing when trying to send the admin a note about being unable to resolve the hostname due to the NM bug.
[14:47] <afflux> hum
[14:48] <afflux> just noticed that I don't use the NM but the network-admin for setting the domain
[14:48] <afflux> thats gnome-system-tools, right?
[14:49] <persia> afflux: In which case, it might be network-admin rather than NM.  Whatever is setting the domain isn't also telling the resolver to search the domain, which is confusing, at best.
[14:52] <elmargol> HI I can not find this window http://www.ubuntugeek.com/images/au/6.png is this a regression?
[14:52] <afflux> elmargol: that's in nautilus, somehow
[14:52] <afflux> at least I know that I saw it there
[14:53] <afflux> just check the nautilus options
[14:53] <elmargol> afflux, found it...
[14:53] <elmargol> stupid change!
[14:57] <afflux> persia: reported as bug 229256
[14:57] <ubottu> Launchpad bug 229256 in gnome-system-tools "setting the domain in network-admin breaks sudo" [Undecided,New] https://launchpad.net/bugs/229256
[14:57] <persia> afflux: Well, except that, as albert23 pointed out, it doesn't break sudo any more.
[14:58] <afflux> right, forgot to mention this
[15:00] <persia> afflux: For best likelihood of getting someone to fit it, I'd recommend retitiling/redescribing the bug to complain about the resolver specifically.  For best results, find a small utility that calls gethostbyname(), and fails after you've changed the domain.  Describe this as a test case.
[15:01] <persia> Having the test case also makes it easy for someone else to confirm the issue.
[15:03] <afflux> persia: added a testcase, but I'm not sure how exactly the new title should look like
[15:05] <persia> afflux: I'd suggest something like "Adding a domain name with network-admin prevents local lookup of the hostname".
[15:08] <persia> Interestingly, it appears I get different results from `hostname` (matching /etc/hostname), and `hostname -s`.  Something fairly low level is behaving oddly.
=== _stink__ is now known as _stink_
=== sebner_ is now known as sebner
[16:13] <rudybee> I would like to file a bug regarding kernel modules when installing Hardy using the installer in hardy/main/installer-i386/current/images/hd-media/
[16:14] <rudybee> But I cannot find a bugzilla or a similar system. How do I file a bug?
[16:16] <rudybee> It looks like the kernel version of hardy/main/installer-i386/current/images/hd-media/vmlinuz is different than the one in the installation ISO images.
[16:17] <rudybee> The kernel modules fail to load when using hardy/main/installer-i386/current/images/hd-media/
[16:17] <rudybee> All forks fine with 7.10
[16:17] <rudybee> Can someone look into it?
[16:18] <afflux> rudybee: not sure about your bug, but generally, the ubuntu bugtracker is at bugs.launchpad.net, make sure to target the bug against the ubuntu project
[16:19] <rudybee> Thanks!
[16:19] <rudybee> I appreciate it
[16:19] <rudybee> Bye bye
[17:27] <jarlath> A bug reporter has had his issue fixed by an update. Do I mark the report as invalid / fix released or let someone from bug-control take care of it?
[17:28] <afflux> if you know which specific version fixed it, set it to fix released and make sure that it is set to the correct package
[17:28] <afflux> otherwise use invalid
[17:30] <jarlath> Thanks afflux.
[17:38] <greg-g> anyone available who is running the DE locale? can you install "stellarium" and run it from the command line and tell me if it segfaults?
[17:40]  * afflux <-
[17:40] <afflux> looking, wait a minute
[17:41] <afflux> downloading half of qt4 now :)
[17:44] <afflux> greg-g: my screen goes white (I can alt+tab to my desktop though), but it does not segfault
[17:52] <greg-g> afflux: thanks
[17:52] <afflux> you're welcome
=== fdd-0 is now known as fdd
[19:42] <emgent> hello
[19:42] <emgent> My membership in ubuntu-bugcontrol is about to expire
[19:42] <emgent> someone can review it? :)
[19:53] <hggdh> emgent: bdmurray, ogasawara, or pedro_ can do it
[19:53] <emgent> hggdh: ok thanks :)
[19:54] <hggdh> emgent: bad news is they are probably enjoying the weekend somewhere else ;-)
[19:55] <ogasawara> emgent: I'll fix it
[19:56] <emgent> thanks ogasawara ;)
=== apachelogger is now known as rpmlogger
=== rpmlogger is now known as apachelogger