[01:54] <nigel_nb> hey, anyone here?
[01:57] <jtniehof> yep. what's up?
[02:06] <nigel_nb> does open office support docx editing?
[02:07] <nigel_nb> jtniehof: check out bug 494813, as far as I know it doesn't
[02:07] <ubot4> Launchpad bug 494813 in openoffice.org "problem with subscript/superscript font while editing .docx files" [Undecided,New] https://launchpad.net/bugs/494813
[02:08] <jtniehof> last I checked, it tries to handle it, but poorly
[02:08]  * jtniehof grubs around a bit
[02:08] <nigel_nb> I remember being able to open it
[02:09] <nigel_nb> but being able to save.... thats news to me
[02:10] <jtniehof> yep, right there on the save as dropdown
[02:10] <jtniehof> and sure enough, superscript doesn't stay
[02:11] <nigel_nb> jtniehof: thank you, I'll confirm it, but I guess its upstream
[02:11] <nigel_nb> another new sign up
[02:12] <jtniehof> hmmm. I'll try it in Windows and see
[02:28] <nigel_nb> jtniehof: reported upstream :)
[02:28]  * nigel_nb finds it irksome joining so many bug trackers 
[02:31] <nigel_nb> interesting bug 494758
[02:31] <ubot4> Launchpad bug 494758 in evolution "Evolution incorrectly reports dates in the near future" [Undecided,New] https://launchpad.net/bugs/494758
[02:31] <nigel_nb> I dont know about everyone else, but I dont generally get mails in the future
[02:36] <jtniehof> nigel_nb: did you check the upstream build of OO? not seeing the docx saving on Windows, going to pull the upstream linux build next
[02:37] <nigel_nb> jtniehof: I reported the bug upstream, so they'll take care
[02:54] <nigel_nb> When a bug is reported against the current release, but is fixed in the dev release (not reproducible), is the bug considered fixed?
[02:55] <micahg> nigel_nb: if you can reproduce it in the reported release or there is a changelog/upstream changelog entry for the devel release
[02:56] <nigel_nb> bug 493250, I could reproduce it before the last update to ubuntu (now I realize) in karmic
[02:56] <ubot4> Launchpad bug 493250 in software-center "Ubuntu Software Center doesn't list GoldenDict when searching for 'dictionary'" [Undecided,Incomplete] https://launchpad.net/bugs/493250
[02:57]  * micahg sees it in karmic
[03:00] <micahg> nigel_nb: maybe ask to try again...I see it down the list
[03:00] <nigel_nb> micahg: yeah me too, I remember checking it out and not seeing it.
[03:03] <nigel_nb> micahg: check out bug 494758
[03:03] <ubot4> Launchpad bug 494758 in evolution "Evolution incorrectly reports dates in the near future" [Undecided,New] https://launchpad.net/bugs/494758
[03:04] <nigel_nb> I dont even know if its worth wishlisting, I mean, that is almost impossible to reproduce (and isn't it worthless to fix)
[03:07] <jtniehof> nigel_nb: docx saving is not present in upstream, so it's not an upstream bug (maybe deactivated in standard build b/c they're still working on it)
[03:08] <nigel_nb> jtniehof: oops
[03:08] <nigel_nb> jtniehof: which means its only on the repos?
[03:08] <jtniehof> yeah, checking Debian now, to see if it's in the debian build
[03:08] <nigel_nb> it should be
[03:09] <nigel_nb> because the build is 5ubuntu (I think, lemme check)
[03:09] <micahg> nigel_nb: the evo bug isn't a bug, the date displayed was correct based on the header information
[03:09] <nigel_nb> I'll close it as invalid then
[03:09] <jtniehof> it's 5ubuntu1, so there are changes between debian and ubuntu
[03:09] <micahg> nigel_nb: be sure to explaiyn wh
[03:09] <nigel_nb> jtniehof: openoffice.org-core 1:3.1.1-5ubuntu1
[03:10] <nigel_nb> micahg: I'll paste what I'm going to say first
[03:10] <nigel_nb> dont wanna get flamed to the sky
[03:10] <micahg> nigel_nb: pastebin
[03:10] <nigel_nb> sure :)
[03:10] <nigel_nb> micahg: as such, we've been getting flamed to much ;)
[03:11] <dtchen> eh, come talk to me about being flamed after you've done Ubuntu audio for a few years
[03:12] <nigel_nb> dtchen: hehe, I forget we've got some veterans at getting flamed ;)
[03:14] <dtchen> I've been very close to writing off Ubuntu completely because of it
[03:17] <nigel_nb> micahg: whoa that was one hell of a flame
[03:17] <nigel_nb> "The Ubuntu bug triage “process” is worse than useless" --- whoa :O
[03:18] <jtniehof> dtchen: I found pulse a royal pain for several revisions, but there's no excuse for the stuff people said when you'd try to post information
[03:18] <jtniehof> at this point, it's settled down to something very nice
[03:24] <nigel_nb> micahg: http://mdzlog.alcor.net/2008/10/29/ubuntu-quality/ is the best I've seen as a reply to a flame :)
[03:24] <nigel_nb> oh about the evo bug, how is it not a bug?
[03:25] <jtniehof> nigel_nb: docx is in Debian sid, oo.o-writer 3.1.1-8, and the super/sub bug is present there. So might be worth filing on the Debian package
[03:25] <micahg> nigel_nb: evo and every other client determines the time from the header in the email which the sending client adds
[03:25] <nigel_nb> jtniehof: will do :) thanks for tracking it down
[03:26] <nigel_nb> micahg: okay
[03:26] <nigel_nb> the current date is 12/9/09
[03:26] <nigel_nb> the date in the header is 12/11/09
[03:26] <nigel_nb> and its saying tomorrow
[03:26] <nigel_nb> so isn't it wrong?
[03:27] <micahg> oh...
[03:27]  * micahg needs to read...
[03:28] <micahg> nigel_nb: I would ask for the date line from the header...I"m guessing it's a timezone issue
[03:28] <jtniehof> anyhow, g'night folks
[03:28] <nigel_nb> oh, will do
[03:29] <micahg> nigel_nb: if said friend was +12 from user, time would be correct
[03:29] <micahg> nigel_nb: also ask for user's timezone
[03:29] <nigel_nb> micahg: so I'll ask him to paste the time from the header?
[03:29] <nigel_nb> okay, will do
[03:55] <LimCore> I found incredible stupid bug in kerneltop application, how to get it's fixed version to ubuntu?
[04:02] <dogatemycomputer> micahg:  Greetings.  Do you have a moment?
[04:03] <LimCore> I sent a patch to fix https://bugs.launchpad.net/ubuntu/+source/kerneltop/+bug/486218  please apply it
[04:04] <ubot4> Launchpad bug 486218 in kerneltop "kerneltop reports wrong map line on 2.6.31-14-generic" [Undecided,Confirmed]
[04:05] <micahg> LimCore: we don't apply patches in here...someone in -motu would
[04:06] <micahg> dogatemycomputer: sure
[04:07] <dogatemycomputer> micahg:  I am looking at bug #494810 .   S/he is trying to view a specific movie that uses Windows Media Player.   I don't know if their use of a proprietary format constitutes a bug?
[04:07] <ubot4> Launchpad bug 494810 in firefox-3.5 "I can't see the video stream in the site" [Undecided,New] https://launchpad.net/bugs/494810
[04:08] <LimCore> micahg: can you set bug priority? the application seems to be useless - it does not work at all with ubuntu kernels due to bug in application
[04:11] <micahg> LimCore: done
[04:12] <micahg> dogatemycomputer: I'm not sure, the user has the totem plugin for x-ms-asf
[04:12] <dogatemycomputer> micahg: so do I..  but it never actually plays..
[04:13] <micahg> dogatemycomputer: well, I can test from home later
[04:13] <LimCore> thanks micahg
[04:13] <micahg> LimCore: np
[04:14] <dogatemycomputer> micahg: I believe it is actually a website coding error.  The reason I believe so is because it fails to specify the FQDN of the movie in the link.  If you attempt to view it then it dumps out an error indicating the location of the video is invalid (not the video itself).
[04:15] <micahg> dogatemycomputer: if that's the case, then the user should report it to the website, but I can't test that now
[04:16] <dogatemycomputer> micahg: I'll leave the bug alone for the moment.   It plays fine on Windows with WMP but that still does not mean it is a bug.
[04:22] <dogatemycomputer> micahg:  thank you for your time!
[04:22] <micahg> dogatemycomputer: np
[04:53] <dogatemycomputer> greg-g:  do you have a moment?
[04:56] <dogatemycomputer> I am new to triaging bugs.   Could someone tell me if it is appropriate for multiple bug reports to be submitted on the same bug?
[04:56] <dogatemycomputer> ..  multiple un-related bug reports submitted on the same bug.
[06:35] <^arky^> When is the ubuntu bug hug  day starting ?
[07:01] <Wimbuntu> hi guys i need some help on ubuntu 9.10 , ubuntu error main process 735 , x doesn't startup , anyone able to help out there
[07:02] <nigel_nb> Wimbuntu: if u are facing a problem and you need help, please ask in #ubuntu
[07:03] <Wimbuntu> Hi Thanks mate , sorry newbie
[07:03] <nigel_nb> Wimbuntu: no problem, I've been there too :)
[07:14] <thekorn> good morning
[07:58] <fujimitsu> alright, kernel updates ready .. time for bacon
[08:01] <nigel_nb> thekorn: good morning :)
[08:02] <nigel_nb> though its kinda like afternoon here ;)
[08:02] <thekorn> hehe, good "kinda like adfternoon" to you, nigel_nb
[08:02] <nigel_nb> hehe
[08:07] <nigel_nb> thekorn: can you check out a bug for me? donno what to do next
[08:07] <nigel_nb> bug 411229, can't reproduce with my camera, so what do we do? a trace?
[08:07] <ubot4> Launchpad bug 411229 in f-spot "f-spot import folder structure wrong time zone" [Low,Incomplete] https://launchpad.net/bugs/411229
[08:08] <thekorn> nigel_nb, sure
[08:12] <thekorn> nigel_nb, first of all I would ask the reporter to run  apport-collect 411229  to make it a proper apport styled bugreport
[08:13] <thekorn> let me start f-spot to try to understand the reporter's problem
[08:13] <nigel_nb> there is no hook for f-spot :(
[08:13] <nigel_nb> we donno if its camera specific, f-spot specific, or camera+f-spot specific
[08:18] <thekorn> nigel_nb, hmm, maybe I don't completely get what this report is about, but maybe it is a duplicate of bug 175191
[08:19] <ubot4> Launchpad bug 175191 in f-spot "f-spot changes timestamp in an incorrect way" [Unknown,Confirmed] https://launchpad.net/bugs/175191
[08:19] <thekorn> this bug also has a upstream bugreport linked
[08:21] <nigel_nb> thekorn: there is a patch
[08:21] <nigel_nb> I could try and pester someone to get a debdiff for lucid or teach me how to do that, does that help?
[08:24] <thekorn> nigel_nb, looks like the author of this patch already has build a package with this patch in his PPA
[08:24] <thekorn> https://edge.launchpad.net/~smcgrath23/+archive/f-spot/+packages
[08:24] <nigel_nb> so, what would be the ideal next step?
[08:25] <thekorn> ask the reporter to install the version from this PPA, ask him to confirm if this package fixes his issues
[08:25] <thekorn> if so, mark your bug as a dupolicate of the other one
[08:26] <nigel_nb> I already asked him for apport-collect
[08:26] <thekorn> super
[08:26] <nigel_nb> shall I tell him to forget it or say this after the apport-collect?
[08:31] <thekorn> nigel_nb, tell him to use the PPA after running apport-collect, running apport-collect is not a big thing and nice to have
[08:32] <nigel_nb> thekorn: I'll not mark it dup now, coz it will flood the other bug with a lot of noise, I'll just leave a comment
[08:32] <nigel_nb> what if he doesn't know how to do a ppa, I'll just tell him its a dup and there is a ppa version and test it
[08:32] <thekorn> nigel_nb, right, mark it as dup if the patched package fixes is issue
[08:33] <nigel_nb> thekorn: how do we get the ppa into the actual release?
[08:34] <thekorn> you mean the package from this PPA, if it is confirmed it morks better than the version we have in karmic?
[08:38] <nigel_nb> thekorn: yep
[08:39] <nigel_nb> thekorn: a, it would be nice to learn how to do it.  b, it would be nice to get it fixed
[08:41] <thekorn> nigel_nb, we have the SRU (stable release update) process to get bug fixes into stable releases
[08:41] <thekorn> you can read more about it at https://wiki.ubuntu.com/StableReleaseUpdates
[08:41] <nigel_nb> thank you
[08:42] <thekorn> nigel_nb, but there are two more important steps: try to convince upstream to either adopt this patch, or provide a different solution
[08:42] <thekorn> and fix this bug in lucid first
[08:42] <nigel_nb> thekorn: so that we dont branch from upstream
[08:43] <thekorn> right, we should try to keep the diff to upstream as small as possible
[08:43] <nigel_nb> I'm talking about fixing it on lucid, if at all
[08:43] <thekorn> which makes the product easier to maintain on our side
[08:43] <nigel_nb> yea, I read up the process some time back
[08:44] <nigel_nb> but with main software, getting it working is an important factor too isn't it
[08:45] <thekorn> nigel_nb, I'm sorry, I have to do some other stuff right now, will be back in a few hours,
[08:45] <thekorn> keep on asking questions, I'm sure others are able to answer them too
[08:45] <nigel_nb> thekorn: I'm off to catch some sleep
[08:45] <nigel_nb> catch ya later :)
[08:45] <nigel_nb> thanks a lot for the help thekorn :)
[08:46] <thekorn> no problemo, sleep tight ;)
[13:04] <TLE> hey guys
[13:05] <TLE> I was wondering about bug 246205
[13:05] <ubot4> Launchpad bug 246205 in ca-certificates "Wordcommunitygrid Message "Peer certificate cannot be authenticated with known CA certificates."" [Medium,Fix released] https://launchpad.net/bugs/246205
[13:05] <TLE> it has been fixed for intrepid and jaunty, but the problem exist now again in karmic, I thought that those fixes would automatically trickle down
[14:48] <bddebian> Boo
[15:33] <greg-g> heya dogatemycomputer, Sorry. I was with out power last night (my IRC session is on a server I connect to). Just saw your message. To answer your question: no. Un-related bug should be submitted in separate bugs. Otherwise it would be next to impossible to actually track the progress on anything.
[15:33] <greg-g> dogatemycomputer: so, if someone did that, politely ask them to open separate bugs for each issue.
[15:33] <greg-g> dogatemycomputer: and of course, if they sound like support requests, have them open them on answers.launchpad.net instead
[16:17] <malev> hi! how are you foks! I'm comming for the hug day :D I'm gonna read a little about it and the I wanna help here
[16:21] <thekorn> what, we have a bugday today? /me goes getting the mail
[16:22] <thekorn> oh, it's on compiz
[16:22] <thekorn> happy hugday everybody!
[16:22] <malev> haha!
[16:23] <malev> hey, how can I find if a bug is duplicated?
[16:26] <malev> I wanna learn about the work at the bugsquad, can anyone help me. for example I'm looking at the bug https://bugs.launchpad.net/ubuntu/+source/compiz/+bug/492271
[16:26] <ubot4> Launchpad bug 492271 in compiz "ubuntu 9.10 totally freezes when compiz is enabled" [Undecided,New]
[16:27] <thekorn> malev, to be honest, I think finding out if a bug is a dup of another one is one of the hardest things to start with,
[16:27] <thekorn> because it needs some understanding of how the software works, at least in some cases
[16:27] <malev> thekorn, oks! thanks for that!
[16:27] <thekorn> malev, looking at this bug now
[16:27] <malev> thekorn, yes i think like you about that
[16:28] <malev> thekorn, you are investigating this bug too? excelent! maybe you could teach me! :D
[16:28] <hggdh> malev: also, you do not need to ask for help -- just ask your question. We are here to help triagers
[16:29] <thekorn> hey hggdh
[16:29] <hggdh> hi thekorn hope life is good for you. Here life is extremely cold (-20C)
[16:29] <thekorn> hggdh, och, are you at the noth pole?
[16:29] <malev> hggdh, hi! -20C where are you?
[16:29] <thekorn> north
[16:30] <nperry> That is cold :)
[16:30] <hggdh> that:-) Chicago area, US, right now
[16:30] <thekorn> hggdh, just wanted to complain about the weather here +1C and *alot* of rain
[16:30] <thekorn> hggdh, ah, have heard about it in the news today
[16:30] <hggdh> ah, at least no rain today :-)
[16:31] <malev> hggdh, well Buenos Aires is 26C :D :D :D
[16:31] <hggdh> and, since my hotel is about 750m from  the customer, I decided I would walk (instead of renting a car). Real bad choice
[16:32] <hggdh> but -- back to business -- malev: tahnk you for helping. We appreciate it.
[16:33] <malev> hggdh, don't worry.
[16:33] <thekorn> malev, ok, back to your bug. WeatherGod has been very active lately traiging bugs, and he has commented on the bugreport, saying he is waiting for similar reports
[16:33] <thekorn> I think it is still on his radar
[16:33] <hggdh> (and a bit of OT) I was down nearby Rio the week before... 40C!
[16:33] <thekorn> so maybe you are safe to just skip this one and go to another one ;)
[16:34] <hggdh> yes. And chat with WeatherGod on questions about this (and similar) bugs
[16:34] <malev> excelent! now I'm with https://bugs.launchpad.net/ubuntu/+source/compiz/+bug/491770 and it also has apport. but of course, I can't reproduce it in my machine
[16:34] <ubot4> Launchpad bug 491770 in compiz "compiz cube rotate mouse pointer lost" [Undecided,New]
[16:35]  * hggdh does not deal with X, and compiz...
[16:35]  * thekorn neither
[16:35] <thekorn> malev, but this one is easy,
[16:35] <thekorn> it should be fixed by now
[16:36] <malev> ... why?
[16:36] <malev> easy to ...?
[16:36] <thekorn> there was an issue in kubuntu with two x-server running
[16:37] <malev> so it's duplicated?
[16:37] <thekorn> hmm, I'm unable to find the related bugreport right now
[16:37] <malev> thekorn, I'm gonna look for it, don't worry
[16:38] <thekorn> malev, not sure if this is a duplicate, but there might be the same cause
[16:38] <thekorn> malev, in such cases it makes sense to ask the reporter to update the system, and check if this issue is still happening
[16:39] <malev> thekorn, oks! well I'm gonna ask that!! thanks
[16:41] <thekorn> malev, I *think* this is the bug I mean, bug 491483
[16:41] <ubot4> Launchpad bug 491483 in gdm "Since failsafe-x was enabled in karmic it starts if gdm is disabled and kdm is used. (low graphics mode error)" [High,Fix released] https://launchpad.net/bugs/491483
[16:47] <malev> thekorn, what's about this answer to bug 491770    http://pastie.org/737492
[16:47] <ubot4> Launchpad bug 491770 in compiz "compiz cube rotate mouse pointer lost" [Undecided,New] https://launchpad.net/bugs/491770
[16:48] <nperry> malev: I would still mark it as a duplicate - imho
[16:48] <thekorn> malev, super, change it to [...]fixed care (bug 491483)[...] - this will make it a clickable link in your comment
[16:48] <ubot4> Launchpad bug 491483 in gdm "Since failsafe-x was enabled in karmic it starts if gdm is disabled and kdm is used. (low graphics mode error)" [High,Fix released] https://launchpad.net/bugs/491483
[16:49] <malev> nperry, thekorn oks! thanks you two! I'm gonna mark it as duplicated too :D
[16:49] <thekorn> my personal opinion here is: only mark it as duplicate if you are sure, to avoid additional bugmails to the subscribers of the potential master and its dups
[16:51] <malev> ready! now i guess i have to paint green the row in the wiki :D I'm really happy this is my first bug work :D
[16:52] <thekorn> malev, there are three things which should also be done by you: you should subscribe to the bug you commented on (to get notofied when the reporter answers your question)
[16:53] <thekorn> set the status to "incomplete"
[16:53] <malev> oks!
[16:53] <thekorn> and mark the entry on the hugday wiki page in green
[16:53] <hggdh> malev: you may want ot look at hugday-tools also
[16:53] <thekorn> haha, yeah
[16:54]  * thekorn is suprised this little tool is still working
[16:55]  * hggdh is happy it still works ;-)
[16:57] <malev> so, I have to run:  hugday init --user malev --cookie /home/malev/.mozilla/firefox/zhucgzuk.default/cookies.sqlite     ?
[16:57] <malev> and :  hugday close 491483   ?
[16:59] <thekorn> malev, exactly
[17:00] <thekorn> maybe you have to also run   hugday current --remember  after running the hugday init command
[17:01] <thekorn> otherwise you have to always run   hugday close 123456 --date 20091210
[17:01] <malev> I'm going for launch, i come back in a while
[17:02] <thekorn> super, thanks for your help
[17:45] <malev> hi! is anyone here?
[17:47] <persia> malev: Lots of people, but you'll get a more specific answer to a more specific question.
[17:49] <hggdh> heh
[17:50] <malev> well here is the thing. I've been working on bug 491770  and I'm getting: RuntimeError: You are not allowed to change the content of 'https://wiki.ubuntu.com/UbuntuBugDay/20091210'. Plase run 'hugday init' and try again.
[17:50] <malev> but I thing is something about "persmisos" I can't remember how to translate that word :S
[17:51] <ubot4> Launchpad bug 491770 in compiz "compiz cube rotate mouse pointer lost" [Undecided,Incomplete] https://launchpad.net/bugs/491770
[17:52] <nperry> Hummm 4171 incomplete bugs alot of them are old, going to do some housekeeping by invaildating them
[17:56] <hggdh> malev: permisos?
[17:56] <hggdh> (or permissos, don't remember)
[17:57] <pedro_> hggdh, i'm trying to solve the issue with malev by priv
[17:57] <pedro_> spanish speaker, so might be faster that way
[17:57] <hggdh> pedro_: OK. But it sounded liker permissions
[17:58] <pedro_> hggdh, yeap that's the translation
[17:58] <pedro_> looks like he wasn't logged into the wiki
[17:58]  * pedro_ reading the instructions
[17:58] <malev> hey! i wasn't that difficult persmisos = permissions :D
[18:00] <pedro_> will update the doc, there's nothing about logging into the wiki first
[18:02] <malev> YESS!! I make it!! haha I'm happy :D :D :D
[18:02] <malev> thanks pedro_ and hggdh
[18:02] <pedro_> malev, rock on! if you have any further question just ask here at the channel
[18:03] <malev> pedro_, I will don't worry :D
[18:04] <pedro_> ;-)
[18:08] <thekorn> ohja, right, of course, you need to login to wiki.ubuntu.com first, before being able to use the hugday tool
[18:12] <malev> Me again! I'm looking at bug 451974 I can't reproduce it, but I've tried. It has a lot of information and it has apport. And I've seen other bugs about totem and compiz, but they are focus con crashing when returning from full screen mode.
[18:12] <ubot4> Launchpad bug 451974 in fglrx-installer "Black video minimizing Totem window" [Undecided,New] https://launchpad.net/bugs/451974
[18:12] <malev> the question: what can I do?
[18:12] <malev> :D
[18:17] <malev> the bug was reported in october, maybe there is a solution if he upgrade.... I don't know
[18:21] <thekorn> malev, I'm not sure about this bug. But to give the reporter the feeling of working on this bug, and given the fact that he used a pre release version of ubuntu karmic, I think it makes sense to ask if this is fixed on an up-to-date system
[18:21] <malev> thekorn, thanks! I'm taking care of it right now!
[18:21] <thekorn> super
[18:45] <nperry> Is there a patricular way on triaging ubuntu-one bugs?
[18:47] <persia> nperry: Upstream is active enough with Ubuntu that it's probably worth creating upstream tasks fairly quickly (but you might want to confirm that with upstream first)
[18:49] <nperry> How would i go about that persia?
[18:50]  * persia checks the upstream docs to find the right channel
[18:50] <thekorn> they have an rc channel
[18:50] <thekorn> #ubuntuone
[18:50] <thekorn> I think
[18:50] <Pici> Yes.
[18:50] <persia> That's what I think too, but I wanted to check their website to confirm.
[18:51] <persia> OK.  Three of us is enough to skip checking :)
[18:51] <thekorn> hehe, yeah
[18:51] <persia> nperry: Go check on #ubuntuone if they want us to do anything special, and let us know :)
[19:11] <malev> well my friends. I'm taking of for now!
[19:11] <malev> thanks to everbody and have a nice day
[19:13] <nperry> #ubuntuone said most of the bugs they've had in the last couple of days are dups
[19:14] <nperry> https://wiki.ubuntu.com/UbuntuOne#Bug%20Triage apport uploads everything the devs need :)
[19:15] <persia> nperry: Cool.  Thanks for checking and passing on the pointer.
[19:16] <hggdh> hum. This link should be in our wiki (cannto check now, real slow wiki connection)
[19:17] <nperry> Would you like me to add hggdh as im there now?
[19:17] <hggdh> nperry: yes, please
[19:22] <nperry> Done :)
[19:25] <hggdh> nperry, thank you
[20:53] <WeatherGod> hey, is anyone here triaging compiz bug reports related to systems hanging?
[21:31] <WeatherGod> bug 487519 can probably be marked as "Triaged"
[21:31] <ubot4> Launchpad bug 487519 in compiz "“window previews” plugin leaves garbage on screen" [Undecided,New] https://launchpad.net/bugs/487519
[21:31] <WeatherGod> the OR gives very detailed description of the bug and screenshots to supplement it
[21:32] <WeatherGod> there really isn't anything more someone can do, except confirming it
[21:33] <hggdh> ?
[21:33] <hggdh> WeatherGod: did you reproduce it (just curious)
[21:34] <WeatherGod> I don't use Compiz
[21:35] <WeatherGod> my graphic cards are too weak
[22:09] <WeatherGod> hggdh, do you think I should flag bug 487165 as a possible security issue?
[22:09] <ubot4> Launchpad bug 487165 in compiz "screensaver unlock dialog under other windows" [Undecided,New] https://launchpad.net/bugs/487165
[22:10] <WeatherGod> he can't seem to reproduce it yet, but the fact that it happened once shows that it is possible
[22:11] <hggdh> looking at it
[22:11] <WeatherGod> it should at least also be linked against the gnome-screensaver
[22:14] <hggdh> WeatherGod: please wait. It takes for even and ever, hallelujah, etc, for a page to load here
[22:14] <WeatherGod> heh
[22:16] <WeatherGod> is nperry around?
[22:17] <persia> WeatherGod: /names will give you a complete list (not right now that I see)
[22:17] <WeatherGod> sometimes people have different sigs here than they do for launchpad
[22:18] <WeatherGod> so I just wanted to double-check
[22:18] <hggdh> well. There is a clear security/privacy issue here -- whatever windows that are open on the foreground would be visible. The severity could be discussed, though
[22:18] <hggdh> add to it the fact that has not been repeated...
[22:19] <hggdh> WeatherGod: you can always ask at #ubuntu-hardened
[22:19] <WeatherGod> maybe the gnome-screensaver people can come up with some other possible ways to cause this?
[22:19] <WeatherGod> you mean, ask if they think it is a security issue?
[22:21] <hggdh> persia: WeatherGod yes. I consider it a potential one (at the minimum, affecting privacy)
[22:21] <hggdh> persia: heh You been here the whole day? don't you sleep ;-) ?
[22:21] <persia> Wait.  What!
[22:21] <hggdh> what?
[22:22] <persia> hggdh: I have discovered the secret of ubiquitous presence: bip.
[22:22] <hggdh> LOL
[22:22] <persia> What's the potential security issue with gnome-screensaver?
[22:22]  * persia was looking through that code just in October
[22:22] <WeatherGod> bug 487165
[22:22] <hggdh> not with g-ss (which I do not even know it is involved), but with the unlock dialog being *under* active windows
[22:23] <ubot4> Launchpad bug 487165 in compiz "screensaver unlock dialog under other windows" [Undecided,New] https://launchpad.net/bugs/487165
[22:23] <WeatherGod> ok, maybe not gnome-screensaver... what ever it is that locks screens
[22:23] <hggdh> :-)
[22:24] <WeatherGod> I am used to my screen locking to be tied to my screensaver, so it was the first thing I thought of
[22:24] <hggdh> WeatherGod: the point is we do not know what Chris is using for that. It is *probably* g-ss, but not necessarily
[22:24] <WeatherGod> well, he said he suspended his laptop (presumedly by closing the lid)
[22:25] <hggdh> so, before we add a task for g-ss, we need to ask Chris what is it s/he is using
[22:25] <WeatherGod> so, power-manager?
[22:25] <hggdh> e.g. kubuntu, Ubuntu, Xubuntu, etc
[22:25] <WeatherGod> oh, well, the picture shows Gnome
[22:26] <WeatherGod> well, an odd config, maybe
[22:26] <hggdh> I did not even try to open it. It would take 30 min to load the beast :-(
[22:26] <WeatherGod> dunno...
[22:26] <WeatherGod> the top toolbar looks like gnome... but everything else isn't what I am used to seeing
[22:27] <WeatherGod> ok, I will ask
[22:28] <persia> RIght.
[22:28] <persia> It is gnome-screensaver that locks the screen.
[22:28] <persia> But it is *supposed* to raise the screensaver window over all other windows.
[22:28] <persia> And it grabs *all* XInput events
[22:29] <WeatherGod> well, he said that he could not interact with any of the windows
[22:29] <WeatherGod> so that is still good
[22:29] <persia> And then it passes some selected XInput Events to other modules linked with XEmbed (e.g. keyboard, memo pad, unlock dialog).
[22:29] <WeatherGod> but, what if the screensaver fails?
[22:29] <maco> crappy that his stuff's visible though
[22:29] <persia> WeatherGod: Supposedly it won't fail
[22:29] <WeatherGod> or even crashes?
[22:29]  * persia looks for the relevant rant
[22:29] <maco> then it unlocks
[22:29] <WeatherGod> lovely
[22:29] <persia> screensaver is architected not to be able to crash
[22:30] <maco> i have on quite a few occasions unlocked my screen by going into a tty and kill -9'ing the screensaver
[22:30] <persia> xscreensaver almost really can't ever crash
[22:30]  * WeatherGod headdesk
[22:30] <maco> i use killall on kscreenlocker pretty often since kscreenlocker often rejects correct passwords
[22:30] <persia> (well, an individual screensaver will crash, but that just drops back to the controlling process which either dispatches another one, reloads the crashing one, or does a blank screen, depending on config
[22:30] <persia> gnome-screensaver uses more libraries, so might be a bit more vulnerable.
[22:31] <persia> Right.  See http://live.gnome.org/GnomeScreensaver/FrequentlyAskedQuestions#Have_you_read_and_understood_http:.2BAC8ALw-www.jwz.org.2BAC8-xscreensaver.2BAC8-toolkits.html.3F
[22:32] <WeatherGod> personally, I would program it to stop the screen saver (if one is on)... and take control of all X display... making them black
[22:32] <WeatherGod> then only display the dialog when needed
[22:32] <persia> Anyway, I'm unsure of the right place for the bug.  It sounds like a problem with the master window.  Either it's compiz not respecting the application's request for a full-screen top-level window and grab of *all* XInput events, or it's gnome-screensaver failing to make the call in a failsafe way.
[22:32] <persia> WeatherGod: IT's stacked.
[22:32] <persia> So it takes control of the display and makes it black.
[22:33] <persia> And then it runs (as an embedded process with trapped crashes) the screensaver as yet another window over the blocking controlling window.
[22:33] <WeatherGod> ah, I see...
[22:33] <persia> This way if the actual screensaver display code crashes the underlying desktop isn't (supposed to be) exposed.
[22:33] <WeatherGod> I was getting a bit mixed up in my thinking
[22:33] <persia> But clearly there's some issue with the interaction there that means that jwz's rant applies.
[22:34] <hggdh> but persia is right, anyway -- this is proably not a screensaver issue but an X one (somewhere)
[22:34] <persia> hggdh: No.
[22:35] <persia> It is *definitely* a bug in one of: gnome-screensaver or compiz.
[22:35] <hggdh> you state compiz due to the transparency?
[22:35] <persia> There's a small chance it might be a bug with X, but that would be a protocol-level bug, which sort of thing tends to be glaringly obvious in lots of ways.
[22:36] <persia> I don't know if compiz entirely follows the normal manner of window managers
[22:36] <hggdh> oh beauty. My PuTTY download is going at 56 bytes per second
[22:36] <persia> So I can't rule out that it might be permitting transparency of the full-screen gnome-screensaver window that's supposed to hide the screen.
[22:37] <WeatherGod> interesting thought
[22:37] <WeatherGod> however, how did the dialogue box get behind the terminals?
[22:38] <persia> But it could also be an issue with gnome-screensaver having some bug in an underlying library somewhere (although I'm more inclined to suspect compiz given the typical paranoia of screensaver developers)
[22:38] <persia> Weather Window Manager sorting issue?
[22:38] <persia> Appearance of the windows is supposed to be controlled by the window manager
[22:38] <persia> Full screen applications still respect that, but are supposed to be given the full screen.
[22:39] <persia> Going full screen and then grabbing all XInput Events *should* mean that it's impossible to access the screensaver.
[22:39] <persia> s/screensaver/window manager/
[22:39] <persia> Now, if one were to implement a window manager such that it grabbed the full screen and drew a GL environment there
[22:40] <persia> And grabbed all the XInput events to pass to the virtual windows in this GL environment
[22:40] <persia> and provided a workaround to let applications appear to be full screen and appear to grap input while still keeping them within the GL display, it might not work as expected.
[22:41] <persia> But I don't really know how compiz works: I'm just speculating about a design that would be able to circumvent gnome-screensaver.
[22:41] <persia> I certainly hope that's not how compiz works.
[22:42] <WeatherGod> the security people are a little exasperated about another "compiz breaking screensaver" issue
[22:42] <WeatherGod> hehe
[22:43] <persia> It happened before?
[22:43] <hggdh> weird is docky, shown apparently as having foreground
[22:43] <WeatherGod> "kees: is it for an older release?  there were similar bugs fixed in intrepid, IIRC"
[22:44] <WeatherGod> isn't docky supposed to do that?
[22:44] <persia> Yes, but it's not supposed to be able to override the screensaver :)
[22:44] <WeatherGod> key word: *supposed*
[22:44] <persia> Indeed.
[22:46] <kees> WeatherGod: namely: http://www.ubuntu.com/usn/usn-537-2 and http://www.ubuntu.com/usn/usn-688-1
[22:47] <WeatherGod> lastly, before I leave for the day... I have noticed some bug reports for bug day seems to have already been dealt with by Pedro or Neil Perry, but was not closed on the hug day list
[22:47] <WeatherGod> should I do something about that, or what?
[22:47] <persia> Ugh.  With the way those are written, compiz does appear to be architected in a way at least potentially similar to that I outlined above :(
[22:53] <WeatherGod> ok, gotta run... probably will be on later tonight
[22:54] <chrisccoulson> compiz contains a hack to unredirect the screensaver window unconditionally on NVIDIA hardware. i wonder if that should just be extended so that it's unredirected unconditionally for everybody
[22:54] <chrisccoulson> it would probably stop things like that happening
[22:55] <persia> chrisccoulson: I think there's a *very* strong possibility that there should be such a hack.
[22:55] <persia> The difficulty then lies in identifying when a given process is a screensaver.
[22:55] <chrisccoulson> persia - i will discuss this with Amaranth
[22:55] <chrisccoulson> identifying the screensaver window is easy, as it sets WM_NAME
[22:56] <chrisccoulson> the hack was implemented to work around a specific bug in the nvidia driver, but it might be beneficial to just unredirect the screensaver for everybody
[22:56] <persia> chrisccoulson: So, if I have a program called slkdfj, is it a screen saver?
[22:57] <persia> I think it would be beneficial, because the screensaver tries to do an unconditional grab on a top-level full-screen window to be secure.  Any way around that is a security issue.
[22:57] <persia> (and it's likely that the compiz code paths are less well audited than any screensaver codepath)
[23:11] <chrisccoulson> persia - sorry, i had to disappear
[23:11] <chrisccoulson> the current patch in compiz identifies any window with WM_NAME=gnome-screensaver
[23:12] <chrisccoulson> that is set in GDK when the window is created, based on the string returned from g_get_application_name()
[23:12] <chrisccoulson> but applications can override that
[23:12] <persia> chrisccoulson: That's probably sufficient, as compiz mostly only gets used for GNOME.
[23:12]  * persia thinks
[23:13] <chrisccoulson> it might be beneficial for screensaver windows to identify themselves by setting a property on the window, which screensaver developers could agree to between themselves
[23:13] <chrisccoulson> rather than needing to identify every possible screensaver by their name
[23:13] <persia> I can see an alternate DoS attack using that hack, but I think the risk of needing to reboot is less than the risk of exposing information.
[23:14] <persia> That's a nifty idea, but not something that likely can be done in the short-term, as it requires lots of people to agree.
[23:15]  * chrisccoulson wonders what happens if gnome-screensaver calls gdk_window_set_override_redirect()
[23:15] <chrisccoulson> i don't know if compiz will then unredirect it
[23:16] <persia> I think a real solution needs to sort out the architectural issues.
[23:16] <persia> If compiz is grabbing the entire display and mitigating everything, it ought take responsibility for locking.
[23:17] <chrisccoulson> yeah, possibly
[23:17] <persia> Mind you, it might not be so well audited, but I think trying to apply workarounds is a recipe for falling behind.
[23:17] <chrisccoulson> we've had quite a few screensaver issues this cycle
[23:17] <persia> It doesn't really need screensaver code: it could just run the screensaver as an embedded process.
[23:17] <chrisccoulson> one of them being a crash when entering your password incorrectly 5 times
[23:18] <persia> which means that any password works if tried 6 times?
[23:18] <chrisccoulson> persia - it didn't crash every time, but i could reproduce it
[23:18] <chrisccoulson> but it's fixed now anyway:)
[23:18] <persia> Did you happen to compare the behaviour with metacity vs. the behaviour with compiz?
[23:19] <chrisccoulson> for the crash?
[23:19] <persia> Yeah.
[23:19] <chrisccoulson> the crash wasn't related to the window manager. it was a race between the lock dialog process and the main gnome-screensaver daemon
[23:20] <persia> Um, that shouldn't be able to happen because the lock dialog is an embedded process of the main gnome-screensaver.
[23:20] <persia> Or else my understanding of gnome-screensaver is completely wrong.
[23:21] <chrisccoulson> the lock dialog process is a completely separate process. gnome-screensaver reparents the lock dialog drawable in to the main window with XEMBED
[23:21] <chrisccoulson> the crash was triggered because gnome-screensaver tried to create a pixmap derived from the lock dialog drawable, which had just been destroyed because the lock dialog process exitted
[23:21] <chrisccoulson> and that triggered an X error
[23:22] <persia> Ugh.
[23:23] <persia> Given that bug, I'm less sure of the response to jwz's rant I pointed to earlier.
[23:23] <chrisccoulson> which one was that? (i've not looked at the whole scrollback)
[23:24] <persia> http://live.gnome.org/GnomeScreensaver/FrequentlyAskedQuestions#Have_you_read_and_understood_http:.2BAC8ALw-www.jwz.org.2BAC8-xscreensaver.2BAC8-toolkits.html.3F
[23:25] <chrisccoulson> yeah, we actually handled the issue in gnome-screensaver, by adding some synchronization between the 2 processes to avoid the race
[23:26] <persia> which isn't ever supposed to happen :)
[23:26] <chrisccoulson> but it's actually been fixed also in GTK now, by not creating pixmaps derived from foreign windows in the code path where the bug occurred
[23:26] <persia> "... This way there is no transfer of grabs to cause a race condition. " :)
[23:27] <chrisccoulson> yeah, i slightly disagree with the idea that running things in separate processes eliminates race conditions :)
[23:28] <chrisccoulson> but it does mean that the main gnome-screensaver daemon doesn't use much GTK really
[23:28] <persia> Which is actually a good thing, just because GTK is 1) complex and 2) subject to GTK extensions
[23:29] <persia> (mind you, these are *good* things for non-screensavers)
[23:29] <chrisccoulson> it is. i find GTK issues quite hard to track down ;)
[23:29] <chrisccoulson> like the screensaver crashes
[23:30] <persia> Indeed.  I like to think I'm fairly good with stacktraces, but the GTK ones nearly always baffle me.
[23:30] <chrisccoulson> yeah, me too. but in the case of X errors, xtrace makes it a bit easier :)
[23:30] <persia> Yep :)
[23:31] <chrisccoulson> it's often obvious why the error occurred without trying to interpret the stacktrace