/srv/irclogs.ubuntu.com/2010/03/21/#ubuntu-bugs.txt

=== anzenketh|brb is now known as anzenketh
dupondjeIf a bug is reported in another language, do we mark it as invalid or incomplete ?00:14
kklimondadupondje: invalid, ask the reporter to open a new one in english (or not at all)00:19
chrisccoulsonor find someone to translate it ;)00:20
hggdhpace_t_zulu: looking for me?00:20
chrisccoulsonsometimes it's possible to work out what a reporters problem is by using an online translator00:20
kklimondachrisccoulson: right - and it's not like we have enough on our plates already :)00:21
hggdhanyway, not being able to translate is not a valid reason to close invalid00:21
dupondjejust checking some thunderbird bugs, as its a mess right now :)00:21
chrisccoulsonagreed, but some reporters may not be able to re-submit in english00:21
hggdhdupondje: what is the bug #? Perhaps I know (or somebody here knows) the language00:23
dupondjehttps://bugs.launchpad.net/ubuntu/+source/thunderbird/+bug/45636900:24
ubot4`Launchpad bug 456369 in thunderbird (Ubuntu) "Ligthning no agrega nueva tarea desde barra de texto (affects: 1)" [Undecided,New]00:24
dupondjefor example ...00:24
hggdhlooks spanish00:24
dupondjehttps://bugs.launchpad.net/ubuntu/+source/thunderbird/+bug/242265 => and this is a bug in 'english' ;p what should we do with such thing for example ?00:24
ubot4`Launchpad bug 242265 in thunderbird (Ubuntu) "This is not a bug (dups: 1)" [Undecided,New]00:24
hggdheeeek. Here goes a rather bastardised translation00:25
hggdhin the window that opens when one presses the button below and to the left "Tareas" (tabs?) there is a text area whe it should be possible to create new "tareas"00:27
hggdhnow wtf is tareas...00:27
hggdhbelow there is a grille with the existing tabs00:28
chrisccoulsonhggdh - my translation says it is "tasks"00:28
chrisccoulsonwhich makes sense00:28
hggdhoh. Now it does, indeed. Thanks, chrisccoulson, this is what happens when I translate from Spanish-looking languages ;-)00:29
hggdhtareas == (Portuguese) tarefas == (English) tasks. Heh00:30
dupondjeseems like half of the thunderbird bugs are garbage. 'Hi, I can't sent emails' ... :p00:34
dupondjeanywya going to sleep00:36
dupondjenite00:36
hggdhdupondje: I added a comment on the bug. You can always use https://wiki.ubuntu.com/Bugs/Responses#Not%20reported%20in%20English00:39
alex_mayorgaanything I need to add for Bug #543114 ?01:23
ubot4`alex_mayorga: Bug 543114 on http://launchpad.net/bugs/543114 is private01:23
nigelbalex_mayorga: what bug is it?01:24
nigelb(I can't see it since its private)01:24
alex_mayorganigelb: try once more01:41
alex_mayorgaBug #54311401:41
ubot4`Launchpad bug 543114 in gnome-screensaver (Ubuntu) "gnome-screensaver-gl-helper crashed with SIGSEGV in _fini() (affects: 1)" [Undecided,New] https://launchpad.net/bugs/54311401:41
alex_mayorgathis is a bug on Lucid01:41
nigelbinformation is enough, changed back to private01:42
nigelbneed to wait for apport retracing01:42
=== yofel_ is now known as yofel
ddecatormicahg: have you seen bug 531987 ?06:34
ubot4`Launchpad bug 531987 in firefox (Ubuntu) "firefox displays extension page on start (affects: 1)" [Undecided,New] https://launchpad.net/bugs/53198706:34
micahgddecator: sorry, not in the bug mindset tonight06:39
ddecatormicahg: that's fine, i'm just not sure if it's something you've seen before (didn't find any dupes) and if the error message has enough info for you. no rush though, you can look at it later06:40
micahgddecator: BTW, openjdk is registered in my browser on upgrade to lucid06:49
ddecatormicahg: you mean that java is recognized by firefox without needing to change anything?06:50
micahgddecator: right06:50
* micahg hunts for a java site06:50
ddecatormicahg: well that's good news =)06:51
micahgddecator: ugh, the verify java test on java.com crashed my browser :(06:52
ddecatormicahg: darn...06:53
ddecatormicahg: in better news, i've gotten sqlite to build and so far the build is going further than it has before =)06:55
dbwalsh72boo08:22
dbwalsh72any one home?08:22
ddecatorwhatcha need?08:23
dbwalsh72I was trying to work out how to file a bug08:23
dbwalsh72on a failed installation08:23
ddecatorwhen did it fail?08:23
dbwalsh72install worked well08:23
dbwalsh72but on reboot, there is a glitch with booting using UUID08:24
dbwalsh72had to force it to /dev/sda1 to boot up08:24
dbwalsh72then of course X wont start cause of a segmentation fault with the nouvea driver08:24
dbwalsh72so thought Id try and log a bug and help out08:24
ddecatorhm, so the first problem was with grub?08:24
dbwalsh72but all the online documentation says about opening the help menu, which of course I cant get to08:25
dbwalsh72yeah, 1 with grub08:25
dbwalsh722nd on nouvea08:25
ddecatoralright, have you searched to see if they have been reported yet?08:25
dbwalsh72so I guess the question Im hoping for help is08:25
dbwalsh72where do you go to log a bug if you can't get to it in the menu08:26
dbwalsh72"help and support"08:26
ddecatorbugs.launchpad.net08:26
dbwalsh72ok going there now08:26
ddecatoryou may want to do a google search of "site:bugs.launchpad.net <search terms>" to see if those bugs have been reported yet =)08:26
micahgdbwalsh72: ubuntu-bug08:26
ddecatoroh yah, i keep forgetting about that command...08:27
dbwalsh72thatnks you two08:27
dbwalsh72Ill search first08:27
ddecatordbwalsh72: you can also run "ubuntu-bug <package>" to use apport for reporting a bug =)08:27
dbwalsh72that works from cli?08:28
ddecatoryes08:28
dbwalsh72(no X due to nouvea bug)08:28
dbwalsh72cool thats exactly what I needed08:28
dbwalsh72thanks08:28
ddecatornp =)08:28
ddecatordbwalsh72: have you tried manually starting X?08:29
micahgdbwalsh72: I think that it might work on lucid in cli, just not exactly sure how08:29
dbwalsh72yes.08:29
dbwalsh72get a segmentation fault on nouvea08:29
dbwalsh72so next step was to reconfigure X and use a different driver08:29
ddecatorhm, didn't know if it just happened automatically or if you tried manually too08:29
dbwalsh72but thought perhaps I might help out with submitting the bug first08:29
ddecatordbwalsh72: well we appreciate the help =)08:30
dbwalsh72I been using linux long enough that perhaps Its a way  I can help for a change :)08:30
dbwalsh72it works from lucid in cli08:34
dbwalsh72gives you lists to choose from08:34
ddecatorperfect =)08:34
dbwalsh72hmmm wants me to reproduce the crash and produce a backtrace, this is interesting :)08:35
ddecatoryup, that'll help apport look for possible dupes and, if it's new, will be information that devs can use to help debug the problem08:36
dbwalsh72ooops08:42
dbwalsh72crashed the whole systme hehehehe08:42
ddecatorhow did you install lucid?08:42
dbwalsh72usgin latest beta cd08:42
dbwalsh72the alpha3 worked fine on this other box08:42
dbwalsh72thought Id install the beta08:42
dbwalsh72gotto reboot it08:43
dbwalsh72didnt like gdb08:43
ddecatorhuh, ok, just wanted to make sure you had the latest release08:43
dbwalsh72gotto go give it the old three fingered salute08:43
dbwalsh72bbl08:44
edakiriI'm trying to debug a Firefox segfault on Karmic. I saw a new graphical debugger recently but have lost it.  Anyone know what it might be?   recommendations?08:47
Damascenehow many bugs about empathy in Lucid?08:50
Damasceneit's really strange for me that I need to open and close the program more than 3 times if I'm lucky for it to connect08:51
ddecatorDamascene: https://bugs.launchpad.net/ubuntu/+source/empathy08:52
ddecatoryou'll have to search for lucid-specific ones08:52
DamasceneI see08:52
ddecatorDamascene: a lot of times they'll have a 'lucid' tag =)08:53
Damascenehttps://bugs.launchpad.net/ubuntu/+source/empathy/+bugs?field.searchtext=&orderby=-importance&search=Search&field.status%3Alist=NEW&field.status%3Alist=INCOMPLETE_WITH_RESPONSE&field.status%3Alist=INCOMPLETE_WITHOUT_RESPONSE&field.status%3Alist=CONFIRMED&field.status%3Alist=TRIAGED&field.status%3Alist=INPROGRESS&field.status%3Alist=FIXCOMMITTED&assignee_option=any&field.assignee=&field.bug_reporter=&field.bug_supervisor=&field.bug_commenter=&fi08:54
Damasceneeld.subscriber=&field.tag=lucid&field.tags_combinator=ANY&field.status_upstream-empty-marker=1&field.has_cve.used=&field.omit_dupes.used=&field.omit_dupes=on&field.affects_me.used=&field.has_patch.used=&field.has_branches.used=&field.has_branches=on&field.has_no_branches.used=&field.has_no_branches=on08:54
Damascenesorry08:54
Damascenethat was the advanced search08:54
edakirihow do you add a directory heirarchy in gdb?09:20
edakiriyou add 1 dir with the 'directory' command.  I don't see how it finds a heirarchy.09:20
=== kermiac is now known as kermiac_
edakiriare there any more appropriate channels to help with these questions?09:26
edakirimore appropriate --> more likely to get answers09:26
nigelbprobably #gnome on gimpnet09:26
edakiriGot my resolution to the Firefox bug: the developers are uninterested because the cause is being Out Of Memory.11:47
* edakiri sighs11:47
vishnigelb: for the cheese apport hook , you are doing it like the rb one , asking questions of the problem before collecting info , right?12:33
nigelbvish: nothing to ask12:33
vishk..12:33
nigelbI'm collecting hardware info, gstreamer info, and, xorglog12:34
vishrighto.12:35
nigelbvish: running lucid?12:35
vishyup12:35
nigelbaha, perfect test candidate12:35
* vish should probably have thought before answering :p12:36
nigelbokay, so create a file in /usr/share/apport/package-hooks/ called source_cheese.py12:36
nigelbvish: paste this content into it http://pastebin.com/WPk3VQmz12:39
nigelband run ubuntu-bug cheese from alt+f212:39
nigelbsee if it opens cheese. if it does not, I'm doing something wrong12:39
vishnigelb: cheese doesnt open , let me upload the bug report it generates12:48
nigelbdid you get warning?12:48
nigelberr, information pop up12:48
nigelbthe "Cheese will now open.  Please wait till you see a video or an error message before closing cheese" message?12:49
vishnigelb: you got those12:49
vishbut cheese didnt actually open12:49
nigelbvish: yep12:50
vishs/you/yup*12:50
nigelbah12:50
nigelbI hate that command_output thing12:50
nigelbits always a little messy12:50
nigelbvish: try changing to "report["Cheese-Debug.log"] = apport.hookutils.command_output(['GST_DEBUG=*cheese*:3', 'cheese -v'])"12:50
vishnigelb: nope12:56
nigelbvish: last try (['GST_DEBUG=*cheese*:3 cheese -v'])12:56
vishnigelb: hehe , tried that too :)12:56
vishdidnt work12:56
nigelbokay, try (['cheese', '-v'])12:57
vishnigelb: \o/12:59
vishthat works12:59
nigelbwhich means I need to talk to pitti about apport dealing with debugging symbols13:00
nigelbvish: okay one more favor13:00
vish...?13:00
nigelbtyping13:03
nigelbvish: http://pastebin.com/6bbtfEQk13:03
vishnigelb: GST_DEBUG=*cheese*:3 cheese -v 2>&1 | tee cheesedebug.txt13:05
vishthat works^13:05
nigelbvish: ah, well whatever ;)13:05
vishsimply doing > debug.txt didnt work13:05
nigelbvish: you wont see output, but it should work13:06
vishnigelb: yup , i mean the output doesnt get written in the .txt , it stays only in the terminal13:06
nigelbvish: worked for me.  anyway, nonissue ;)13:07
nigelbvish: done yet?13:14
vishnigelb: cheese -v give nothing except the version and no debugging13:15
nigelbvish: ugh :(13:15
vishnigelb: only this line > Cheese 2.29.9213:15
vishnigelb: its the pre Cheese 2.29.92  and post difference i guess13:16
nigelbvish: I'm on 2.28, so its very differnt from that13:16
vishexactly :)13:16
nigelbvish: this means, hook not ready yet.  Need to talk to pitti13:16
nigelbvish: i'm not sure debugging needs to be done in the hook yet though13:17
nigelbmost packages require uses to debug only when the debug output is really necessary13:17
vishnigelb: i dont think we need debugging in the hook always..13:18
vishnigelb: but would be easier if it was there ;)13:19
nigelbvish: I did that for rhythmbox, but seb told me not required13:19
nigelbif we want it, we can ask it13:19
vishnigelb: why not have it as an option?  when debug required we can ask user to select the debug option..13:20
nigelbvish: basic UI design.  Present user with least amount of choice.13:21
nigelbit is very difficult for user to judge whether required or not13:21
vishnigelb: not necessarily ..13:21
vishnigelb: when we have the bug report , we know what option to ask the user.. if only hard info needed we can say collect info.  if the user says  video crappy/no audio recorded/ crashes , we can say select the option 2 and it will collect the info13:22
vishhardware info*13:23
nigelbah, apport doesn't yet do that difference13:23
nigelbit presents the same ui with apport-collect and ubuntu-bug13:23
nigelbthe same choices.  maybe something that needs to be added to appotr13:24
vishnigelb: what wrong in having same choices for both?  we need the proper wording , thats all... like how the RB one is done13:24
nigelbvish: In RB different things can go wrong.  Here there are only 2 and either ways we need both13:25
nigelbone is hardware and other is gstreamer13:25
vishyes , so lesser choices ;) .. i'm  still not sure why it is a problem asking for the option13:26
nigelbvish: we can't identify whats wrong with symptoms13:26
nigelbwe need hard information13:26
vishnigelb: then we need debug always.. which you mentioned it was not needed.. ;) hence i mentioned the options ;p13:27
nigelbvish: no no, for debugging gtreamer trouble, gstreamer log would be sufficient13:27
vishnigelb: hmm... then when is it we /need/ the cheese debug?13:29
nigelbvish: unsure.  its when something breaks and you want to figure out waht13:30
nigelbvish: lemme ask upstream dev anyway13:31
vishneat.. :)13:31
nigelbvish: just a doubt13:32
nigelbGST_DEBUG=*cheese*:313:32
nigelbvish: ^ runs cheese and gives debugging info?13:33
vishnigelb: nope13:33
vishGST_DEBUG=*cheese*:3 cheese -v13:33
nigelbtry removing -v?13:33
vishthat works too , but we need to start "cheese"13:34
vishnigelb: lemme compare the two debugs13:34
nigelbvish: without -v cheese doesn't start?13:35
vishit starts13:35
nigelbthis could fix the trouble13:36
nigelbany difference between debugs?13:36
nigelbvish: need to step out at 7:30.  we need to wrap up soon :)13:36
vishnigelb: diff of cheese -v and cheese > http://paste.ubuntu.com/398765/13:38
nigelbvish: only one difference, it doesnt report version number, I can live with that13:39
vishyea13:39
nigelbvish: try playing with the hook with (['GST_DEBUG=*cheese*:3', 'cheese'])13:39
nigelband (['GST_DEBUG=*cheese*:3 cheese'])13:39
vishnigelb: nope nothing starts cheese :(13:41
nigelbugh13:41
nigelbvish: last try (['GST_DEBUG=*cheese*:3')13:42
vish:( no..13:42
* nigelb thinks in a cooler mind it might be something to with python13:43
nigelbvish: hacking apport source code now at frustration :p14:10
vish:)14:10
Lion_My sound not mute external audio when input my headphone, it's a bug or how can I do this with configuration?14:34
vishnigelb: just mention to fillipo , you meant "for all the bugs " :)15:00
=== ccooke_ is now known as ccooke
=== yofel_ is now known as yofel
nigelbvish: around?17:31
nigelbvish: test hook with (['cheese', '--gst-debug=*cheese*:3']) and let me know what happens17:33
* vish TRIES17:43
vishbad caps lock :/17:44
vishnigelb: \o/ works17:45
nigelbvish: success.  I'll take your choice for menu options.17:46
vishnigelb: the wording for the first dialogue needs a bit of change too17:47
vishnigelb: since we wont be using that always17:47
nigelbvish: your call.  you're the designer.  I'm only the developer17:47
vish;)17:47
vishnigelb: can cheese be restarted if we use the Help> report a problem?17:48
nigelbvish: we can ask that cheese be closed if its already running :)17:49
vishnigelb: right , so we should detect if it is running first . then ask for close if the debug option is selected17:51
nigelbvish: not really.  Just modify the current info pop up.  Please close cheese if its already running.  Cheese will be restarted in debug mode to collect more information.17:52
vishnigelb: yes , but asking for close if the user wants to select the second option would not be required17:52
vishnigelb: or we should ask for user to close before proceeding when debug is selected17:53
nigelbvish: I'll check with some pythonf folks abot it17:53
nigelbvish: thts what I meant.  when debug is selected.  we ask user to close cheese17:53
vishnigelb: hmm , right. but the dialogue show appear only when already running17:54
vishnigelb: if they run apport-collect or ubuntu-bug it would be misleading17:54
nigelbvish: I think it is entirely possible from code.  Have to check with python folks17:54
vishnigelb: neat thanks..17:54
nigelbvish: the gz is not my doing18:15
nigelbits thanks to LP18:15
nigelbvish: could you take a look at bug 501054 if it comes under papercutters purview?18:20
ubot4`Launchpad bug 501054 in gcalctool (Ubuntu) "gcalctool 5.29 hides switch for display format in its settings window (affects: 4)" [Low,New] https://launchpad.net/bugs/50105418:20
OrbordeI have a bug that looks like a kernel bug, but I'm not sure.19:03
OrbordeWhat can I do to isolate useful information on it?19:03
OrbordeWhen I create a process that uses enough memory that it forces a large swapping operation, the system hangs with no disk activity for several minutes, before commencing the expected thrashing and eventually becoming responsive again.19:04
OrbordeWhat do I need to include in a bug report to maximize the odds that it will be successfully reproduced?19:04
OrbordeI have a C program that triggers it pretty consistently, as well as an equivalent Python snippet.19:05
crimsunvmstat, iostat19:07
crimsun(yes, I realise there will be non-trivial interaction)19:07
Orbordecrimsun: What precisely do I need to do?19:10
crimsuncontinuous output from both those19:12
Orbordecrimsun: As in do iostat 1 > iostat.log, vmstat 1 > vmstat.log, and run the triggering program?19:17
crimsunthat's one approach, yes19:19
Orbordecrimsun: Did you have another one in mind?19:19
crimsunthere are a couple of increasing complexity and resourcefulness (no pun intended)19:19
OrbordeShould I file this against package "linux"?19:25
crimsunit really isn't very useful at this stage.19:25
OrbordeWhat can I do to make it more useful?19:25
crimsune.g., you haven't demonstrated that it isn't related to possibly-broken hw19:25
OrbordeSuch as bad HDD sectors?19:26
crimsuncorrect19:26
crimsun(assuming you have rotary, of course)19:26
OrbordeIt seems like it would be logical to find someone else running Ubuntu64 and see whether the same behavior happens when they run the code snippets.19:28
Orbordecrimsun: I don't know how I would rule out a bad HDD. I checked smartctl, and no errors or sector relocations have been reported.19:32
Orbordecrimsun: Is there any special formatting on swap partitions? I am thinking of swapoff'ing and zeroing the partition to see whether it chokes.19:35
crimsunOrborde: none, really.19:36
crimsunOrborde: if you've run the thorough test in smartctl, you could file a bug report and attach all the info.19:37
Orbordecrimsun: smartctl -t long ?19:37
OrbordeThat runs fine in the background even when the machine is in use, correct?19:38
OrbordeOK, here's some interesting stuff from dmesg: http://pastebin.com/AeGktjza19:50
OrbordeI have a bunch of entries like this.19:50
OrbordeFrom various random programs.19:51
crimsunhmm.19:52
crimsunOrborde: check bug 26284319:52
ubot4`Launchpad bug 262843 in linux (Ubuntu) "[2.6.27-2.3] (sometimes temporary ?!) system deadlock with io_schedule (affects: 3)" [High,In progress] https://launchpad.net/bugs/26284319:52
OrbordeCould be related, I guess.19:54
OrbordeDoesn't seem like they have a specific set of steps to reproduce19:55
crimsunscheduling is hard </barbie>19:55
crimsunif you're running -generic, can you reproduce the symptom using -preempt? (You have to be using amd64.)19:55
Orbordecrimsun: Are you asking me to change a boot flag, install a package, or what?19:57
Orborde*install a kernel image package19:58
Orbordecrimsun: linux-image-*-rt?19:58
crimsun-preempt. It's available for amd64.19:59
Orbordecrimsun: I can't find it. It's not in a weird repo, is it?20:01
crimsunare you using lucid on amd64?20:02
crimsunsorry, I should have made that explicit20:02
Orbordecrimsun: 9.1020:03
Orbordekarmic20:03
crimsunright, so that is insufficient for -preempt20:03
crimsuntry reproducing it in 1) a mainline 2.6.33.1 kernel, 2) a mainline 2.6.32.10 kernel, 3) a mainline 2.6.31.12 kernel20:04
OrbordeEuuuuuugggghhh.20:04
crimsunhttp://kernel.ubuntu.com/~kernel-ppa/mainline/20:05
Orbordecrimsun: Where did you get those kernel versions you wanted me to try?20:07
OrbordeAlso, I don't see a 2.6.33.1 on here, just a bunch of RCs20:07
crimsunbah, apw ^^^  (no 2.6.33.1)20:08
crimsunOrborde: those kernel versions are the latest stable point releases for upstream20:09
crimsuni.e., one of the first things that will be asked is, "Can you reproduce the symptom using a mainline kernel?"20:09
OrbordeW: Failed to fetch http://ppa.launchpad.net/kernel-ppa/ppa/ubuntu/dists/karmic/main/binary-amd64/Packages.gz  404  Not Found20:17
Orbordecrimsun: ^20:17
crimsunOrborde: it isn't a ppa.20:18
crimsunOrborde: I specifically pointed you to the directory because it *isn't* a ppa.20:18
Orbordecrimsun: Oh, I should be downloading the .debs directly, then?20:18
crimsuncorrect.20:18
Orborde2.6.32.10 doesn't seem to have any debs -> http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.32.10-lucid/20:19
crimsunit failed to build from source (ftbfs)20:20
crimsunsigh, it'll probably be addressed this work week20:20
crimsunsorry, I can't push magic buttons,since I'm not a Canonical employee20:20
OrbordeHaha, I understand20:20
=== FlannelKing is now known as Flannel
=== anzenketh_ is now known as anzenketh

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!