/srv/irclogs.ubuntu.com/2009/03/25/#ubuntu-desktop.txt

BojhanHey02:07
BojhanI am wondering who I should contact regarding usability/ux projects/efforts in ubuntu?02:08
pittiGood morning07:27
pittiasac: ah, I think I identified the culprit: it seems to happen while trying to apply patches to generate the sourceful stack trace; debian/rules setup seems to trigger starting xulrunner07:51
mvois the retracer running again (I guess it takes a bit of time to catch up again). bug #347602 is why I ask07:52
ubottuLaunchpad bug 347602 in update-manager "update-manager crashed with SIGSEGV in debPackagesIndex::FindInCache()" [Undecided,New] https://launchpad.net/bugs/34760207:52
pittimvo: it got stuck again07:53
pittimvo: I just killed them, and will now work around it by disabling sourceful stack traces07:53
mvothanks pitti07:54
tonyyarussoAny chance I could get someone to take a look at a small pet bug for my dad?  (It's the last thing keeping him from using OOo apparently.)  He filed a report at https://bugs.edge.launchpad.net/openoffice/+bug/27567608:22
ubottuUbuntu bug 275676 in openoffice "[upstream] date text in autofilter header is pushed out by pulldown arrow" [Unknown,Confirmed]08:22
asacpitti: huh?08:26
asacpitti: isnt it just ".postinst"?08:26
asacwe run xulrunner --gre-version08:26
pittiasac: haven't looked at the package yet, just applied the workaround for now (disabling sourceful stack trace)08:27
pittiasac: I'll try to get back to this in the afternoon and take a closer look08:27
asacpitti: well. if its the same issue from yesterday i can confirm that simply installing xulrunner-1.9.1 in fakechroot hangs08:27
pittiah, okay08:28
pittiasac: I have a handful of such cases, I usually just dpkg-divert them away and replace them with a no-op shell script08:28
asacoh so xulrunner isnt alone? thats good to know ;)08:29
asacits the postinst you would need to prevent somehow i guess08:29
pittiasac: gconf-schemas and polkit-auth also give trouble, yes08:30
pittiand ucf08:30
pittiasac: as long as the number of affected programs is so small, it's easier to divert them away than trying to fix fakechroot for each and every corner case08:31
asacpitti: thats good. as long as that doesnt mean we wont get backtraces for xul 1.9.1 at least08:32
seb128good morning there09:05
seb128pitti: hello09:05
didrocksgood morning seb128 :)09:06
seb128lut didrocks09:06
didrocksseb128: are you enjoying being in UK?09:06
pittibonjour seb12809:06
seb128didrocks: yes, being in London and at the canonical office is always nice ;-)09:06
didrockshi pitti09:06
didrocksseb128: great! ;)09:06
seb128pitti: did you read the xulrunner discussion yesterday?09:10
pittiseb128: yes; I think it hung on the "get sourceful stacktrace" step, when trying to apply patches09:10
seb128pitti: hum, what we figured with asac is that it hangs on the xulrunner --gre-version09:11
pittiseb128: I disabled that for now and restarted them09:11
seb128ok that was my question09:11
seb128thanks09:11
seb128I close IRC and went to the pub just after asac suggested that09:11
seb128so I didn't know what happened next09:12
asacpitti: so you disabled xulrunner-1.9.1?09:12
* seb128 hugs pitti09:12
pittilet's see how it goes09:12
pittiseb128: I'll look into a more proper workaround this afternoon09:12
pittiseb128: I think I can just dpkg-divert the xulrunner binary09:12
asaci just want to know if there is action required from my side if i want to get backtraces for xul 1.9.1 or not ;)09:12
pittiasac: not yet, working on some other stuff still09:12
asacok let me know. we can probably eliminate the xulrunner --gre-version invocation if thats needed.09:13
pittiasac: you should get them again09:13
pittiit shouldn't run the postinst in the first place09:17
asacah ok.09:19
crevettehello09:23
asachi crevette09:24
crevettehey asac09:24
asacis there a "advanced new bug" page or something for bgo where i can do everything on one page?09:40
asachmm i cant find gtkhtml3 in bgo09:41
seb128asac: it's gtkhtml there09:42
seb128asac: and what do you mean, there is everything on the same page in bgo09:42
seb128asac: http://bugzilla.gnome.org/enter_bug.cgi?product=gtkhtml09:42
seb128I guess you opened the simple form ui rather than this one09:42
seb128no?09:43
asacseb128: could be ... i didnt expect that simple form is decided on first page ... let me check09:43
asaci am blind yeah09:43
asacstill odd that i only could find gtkhtml2 on the simple form ;)09:43
seb128asac: it's written (simple form), did you click there? ;-)09:43
asacseb128: its not even on this page: http://bugzilla.gnome.org/enter_bug.cgi09:44
asaconly gtkhtml209:44
seb128asac: it's listed there09:44
asacseb128: i clicked on "New bug" on the top ;)09:44
seb128asac: GtkHtml09:44
seb128in the desktop section09:44
asacsigh09:44
seb128" GNOME's lightweight engine for the rendering, printing and editing of HTML."09:44
* asac puts new eyes on his wishlist09:44
seb128asac: you should get some coffee ;-)09:44
asacseb128: so is gtkhtml we have mostly trunk or i should i look if the patch is relevant on trunk before submitting?09:50
seb128_re09:53
seb128_<seb128> asac: we have mostly trunk, ie the tarball is a week old and they are not that active on it09:53
asacouch09:54
asaca new brain too.09:54
* crevette wishes gtkhtml will be replaced by something better (webkit)09:54
asacright i have to check webkit too09:57
asacwhat is using webkit?09:57
crevetteasac, some projects are migrtating like lifearea10:02
crevettesome Novell developer involved in evolution are rewriting a lightweight mail client for small form factor called anjal using webkit too10:03
crevetteasac, http://blogs.gnome.org/sragavan/2009/03/18/announcing-anjal-the-new-mail-for-netbooks/10:04
seb128_crevette: it's not a rewrite, it's a new UI but it's still using evolution10:07
crevetteseb128_, yeah it on top on eveolution10:08
andreasncrevette, there is a devhelp branch using webkit as well10:11
seb128_asac: did you consider webkit 1.1 for jaunty btw?10:12
crevettehey andreasn,10:12
crevetteandreasn, I was searching the one which were switch from gtjhtml to webkit, devhelp use moz I guess10:12
andreasnyes, it does10:12
crevetteasac, the patch in http://bugzilla.gnome.org/show_bug.cgi?id=576694 seems broken10:15
ubottuGnome bug 576694 in Miscellaneous "evolution treats some pixelsized fonts as point sized" [Normal,Unconfirmed]10:15
crevettes/broken/not a patch/10:16
seb128_it's a patch with noise too ;-)10:17
mvopitti: what is the best way to debug http://people.ubuntu.com/~mvo/tmp/Fehler.png ?10:18
pittimvo: f-spot remembers the previously used directory and selects it by default; if it doesn't exist, it brings up that error message10:18
pittimvo: it should just check if it doesn't exist any more, and silently select the default folder instead IMHO10:19
mvopitti: this was a fresh instrepid install with upgrade10:19
mvopitti: ok, I can have a look at the code10:19
asaccrevette: thanks indeed10:19
pittimvo: if you find a bug, I heard that robert_ancell is interested in the mono stack10:19
pittimvo: LP bug report, I mean10:19
crevetteasac, you're welcome I'm so bored at work I read patch incoming for gnome :/10:20
* robert_ancell regrets this already...10:20
seb128_crevette: you could triage bugs ;-)10:20
crevetteI could ... but this is much work10:20
asaccrevette: thats indeed interesting ;)10:20
pittirobert_ancell: cowtrading season is open10:20
asacdont let your boss find the channel log10:20
asacwe need text file preview in gtk file selector ;)10:23
seb128_pitti: who wants to trade when you can give things away? ;-)10:23
crevetteasac, I'm not a slacker I do some work though10:23
=== seb128_ is now known as seb128
seb128crevette: you could triage bugs!10:23
seb128and prove you are not slacking ;-)10:23
crevetteseb128, it was about my day work, but i'm not slacking for ubuntu, I raise my karma from 1000 to 5000 damn it !!10:25
crevette:)10:25
seb128good work ;-)10:26
seb128it's nice that you work on the bluetooth satck10:26
seb128stack10:26
crevetteyeah, I'm not that happy, I don't understand that much bluetooth10:26
seb128good chance to learn ;-)10:27
crevetteyeah :/10:29
didrockscrevette: I'm sure seb128 will be delighted if you put some work on Compiz as well :p10:31
crevetteI don't use compiz and I don't wanna have to learn what it is10:32
crevette:)10:32
didrocks:)10:32
seb128didrocks: ok, you can get compiz if you really insist!10:33
didrocksseb128: let's say I'm not using it too ;) And I know that you are so eager to maintain the associated package that it will be a crime to remove this pleasure to you :)10:34
didrocksseb128: and first, I'm going deeper in GNOME technology first :) reading http://library.gnome.org/devel10:35
crevettedidrocks, and not being cano I can easily refuse any assignment on compiz :)10:35
didrockscrevette: yeah, no tag to push for us ;)10:36
pittiplease note that compiz has a very responsive upstream10:36
didrockspitti: I think it's not a problem of responsive upstream, but more to go deep in the code, managing graphics layers, etc. I think that getting into the code might be more than time-consuming10:37
didrocksit's a really good piece of work, but I don't know how new people easily can get into it10:38
pittididrocks: I don't think that mvo did much hacking on the code iself either10:38
pittidon't get me wrong, I'm not trying to push the package on you10:39
pittibut from my perspective it seems that maintaining compiz is more a "keep good relations to upstream" and "do good bug triage" problem than hacking on the code yourself10:39
pittimvo: ^ what are your experiences so far?10:39
pittimvo: i. e. what would you hand out as a 'job description'?10:39
didrocksyeah, I think that having mvo point of view is the right way :)10:40
seb128ok, I already asked that before10:40
seb128but where would you put a list of packages <-> usual maintainer10:40
mvopitti: yes, that is a good description, preparing updates, keep an eye on the git tree and be nice to upstream (they are always very nice to us)10:41
seb128I was thinking desktop-bugs or ubuntu-desktop team bzr on launchpad10:41
pittiseb128: you want to keep that as a static list?10:41
seb128I'm open to suggestions10:41
pittiseb128: or something like "set of uploaders of last 10 uploads"?10:42
seb128what I see is a list of packages the team maintain10:42
seb128and who is usually doing updates10:42
mvoI don't mind doing updates, compiz is fun really, the problem is that I suck at bug triage10:42
seb128ie if somebody usually does ask him or her before starting on an update10:42
seb128I'm usually doing the gvfs and nautilus updates for example because I track upstream changes closely etc10:43
asachmm ... is there a way for me to grep through the full gnome svn ?10:43
seb128so I don't want somebody to jump on those because I'm not around10:43
pittiseb128: my initial gut feeling is that this could just be a dynamic list10:43
seb128but nobody is working actively on, let's say gucharmap and that is free to claim by contributors10:43
seb128ie first to come = first served10:43
pittiseb128: ah, I see; so for gucharmap the changelog scanning wouldn't say "free for all"10:44
asacseb128: how about setting Maintainer to seb128@ubuntu.com for those packages that you care really hard about10:44
asac;)10:44
pittiand for the others, ubuntu-desktop@10:44
asacyes.10:44
didrocksI saw that in some packages already10:44
asacwell at lesat mozillateam sets it to mozillateam to indicate that folks should talk to us first before touching.10:45
asacso if ubuntu-desktop is too broad you can narrow it down by your own name10:45
seb128asac: well what we do know is that I dispatch updates on IRC10:45
seb128but I'm not always around and I don't want to block work10:46
seb128and it would be better to have a public list of what is being worked by who10:46
asacyeah. thats a workflow system ;)10:46
asacyou could auto create a batch of Needs packaging bugs10:46
asacwhen tarballs get rolled upstrream10:46
asacand then use the assignee notion10:46
seb128right, I don't fancy opening 70 bugs every new GNOME week10:48
seb128it's lot of "paper work"10:48
seb128and waiting on launchpad10:48
crevetteseb128, back to bluetooth, the problem is bluez doesn't have a BTS which is really boring10:48
asacseb128: autocreating bugs would be ok though? or even too heavy weight?10:50
seb128it feels like lot of paper work to me10:51
seb128we could have a table which lists upstream versions and current jaunty ones10:51
pittiseb128: what about a social solution to just tell every contributor to check the changelog and Maintainer: field first?10:51
seb128and say that anybody is free to claim anything not update10:51
pittiseb128: http://www2.bryceharrington.org:8080/X/PkgList/versions_current.html is pretty neat10:51
seb128so we would move the maintainer info to individuals?10:51
asacso you want merge-o-matic for upstream with work assignment feature ;)10:52
seb128pitti: right, something around those lines10:52
seb128asac: not really mergo-o-matic is about producing diffs no?10:52
seb128asac: well something around that for updates yes10:53
asacseb128: yeah i referred to the "newer version in upstream part"10:53
seb128ie listing things not uptodates as we list merges to do would be nice yes10:53
asacmaybe talk to scott ... gnome tarballs should be as debian packagesjust that they dont have a debian/ dir ;)10:53
seb128we have plenty of similar pages already10:54
seb128see the ubuntu-x page pitti just gave10:54
seb128the pkg-gnome debian team has one too10:54
didrocksseb128: this one: http://www.0d.be/debian/debian-gnome-2.26-status.html ?10:55
seb128yes10:55
asacbut if you just that that and you cannot claim a "tarball" it doesnt solve your initial problem?10:55
asacor is your initial problem that you have to hand out tasks ?10:55
seb128right we want:10:55
seb128- a list10:55
seb128- an easy way to claim updates on this list10:55
asacyeah. claiming updates should be like making a timed lock10:55
seb128which are somewhat orthogonal things10:55
asace.g. for 36 hours or so10:55
asacor whatever time seems reasonable10:56
seb128I was thinking just have a bzr where you add a line "source name"10:56
seb128to claim that you work on "source"10:56
seb128and have the page code looking that that bzr10:56
seb128we could also have a small command list tools10:56
seb128ie, list-to-do10:56
asacseb128: so if we want to look for the future claiming could simply be creating a new10:57
seb128which would list available updates and being worked ones just by using watch files and this bzr todo list10:57
asacUNRELEASED changelog entry10:57
asacin bzr10:57
asace..g10:57
pittiseb128: on a tangent, retracers are grinding10:57
seb128well, that's not because you fix a typo in the control description and don't want to upload that you want to keep a lock on the source10:57
asacdch -i -DUNRELEASED "*new gnome tarball x.x.x"; then commit means its claimed10:57
seb128pitti: excellent10:57
asaci mean if you could just express that you work on it in the bzr tree where you will work on directly it makes most sense, no?10:58
asacinstead of maintaing a different file that everybody has to remember to use and so on10:58
seb128right10:58
asacso how about the idea of claiming through UNRELEASED changelog entry11:01
seb128<seb128> well, that's not because you fix a typo in the control description and don't want to upload that you want to keep a lock on the source11:01
asacand a website that parses bzr trees for versions currently worked on11:01
asac(and how long idle)11:01
asacseb128: hmm11:01
asacvalid argument.11:01
seb128my feeling is that bugs are too much of a constrain11:01
seb128and that changelog updates are racy and don't work in all cases11:01
seb128we used the wiki for a while but it takes ages to load and commit changes11:01
seb128I leaning toward a simple bzr list and a wrapper tool11:01
seb128desktop-todo claim source11:01
seb128desktop-todo info source11:02
seb128that sort of thing11:02
seb128and use a bzr for infos storage11:02
LaneyCouldn't you have a page like Bryce's with a "Claim" button that uses LP auth?11:02
seb128well if somebody wants to do that11:02
pittiCLI++11:02
didrocksseb128: I can work on that if you wish, once jaunty released (too much work atm, as you know)11:03
seb128but I've no clue about lp authentification and I don't want to spend time on a website11:03
seb128didrocks: would be nice, we can discuss it at uds if you will be there11:03
Laneyi'll see11:03
seb128dunno if invitations have already been sent11:03
Laneythe code is probably stealable from bryce and REVU11:03
didrocksseb128: I hope I will be there, still waiting for sponsoring results :)11:03
asacdidrocks: maybe make claiming so that it either shows how old the claim is11:04
asacor that it gets automatically unset after a while11:04
asaci guess showing  how old is good11:04
didrocksasac: yes, timeline is import there, for not having deadlock11:04
seb128asac: right11:04
didrocksimportant*11:05
andreasnmpt, ping11:07
seb128asac: you already got commit approval for your evolution change ;-)11:12
seb128the evolution guys are reactive nowadays that's cool11:13
crevetteyep M barnes is really working hard on evo11:13
crevetteuntil he'll be relocated on other things11:14
crevette:/11:14
seb128does redhat relocate people often?11:14
seb128mchra is doing most of the bug fixing work I notice11:14
crevettemccaan seems to be a moving target11:14
seb128mbarnes is rather doing the switch away from bonobo etc changes11:15
seb128right11:15
seb128but mchra and mbarnes are on evolution for several cycles now11:15
crevettebut I guess now mchra and barnes are really involved in evo so I hope they'll continue to work for some time11:15
crevetteLaney, hey11:16
Laneyyo11:16
crevetteLaney, upsating libgupnp fixed the crash in nautilus-sendto-universe11:16
Laneyexcellent11:16
crevetteI did a FFe request to have it updated11:16
Laneyseb can approve that11:17
crevetteeven universe packages ?11:17
* Laney thinks so11:17
seb128crevette: only universe packages rather ;-)11:17
seb128I'm not approving packages for main11:18
seb128what is the bug number?11:18
Laneybug 34812211:18
ubottuLaunchpad bug 348122 in gupnp "FFe: Sync gupnp 0.12.6-3 (universe) from Debian unstable (main)." [Wishlist,New] https://launchpad.net/bugs/34812211:18
asacseb128: great. can you commit ;)?11:19
seb128asac: you don't have commit access to the GNOME svn yet? you should ask for that ;-)11:19
asacseb128: good. whats the process?11:20
seb128asac: http://live.gnome.org/NewAccounts11:21
crevetteLaney, I didn't had to answer on the bug report but the upgrade seems to be really safe11:23
seb128nothing out of nautilus-sendto-universe is using the lib11:24
crevetteyep11:24
crevetteand all other gupnp packages are developped by the same people11:24
crevetteI don't understand why this one was not updatd11:24
Laneybecause nobody did it11:24
crevettethere others packages were synced and not this one11:24
Laneyit was updated in Debian after FF11:25
Laneyafter DIF even11:25
seb128asac: mbarnes is commiting the change now, you should still apply for svn access ;-)11:25
asacseb128: are you voucher?11:26
seb128asac: yes11:26
seb128asac: you can probably list danw too11:26
asacseb128: i have to select a module11:26
seb128I guess you worked enough with him for that no?11:26
asaci think so11:27
seb128hum11:27
seb128asac: select nm, maybe just ping danw before ;-)11:27
asacyeah11:28
seb128vuntz: ^11:28
asacseb128: you are not a module owner right?11:28
seb128vuntz: is there any way to apply for a svn account without being active commiter on a specific component?11:28
seb128asac: well I'm one of the gnome-control-center ones but it would be weird if you applied for this one since you didn't contribute on it11:28
asacheh11:28
asacok11:29
asaci will wait till dan pops up11:29
asaci can ask chpe too11:29
seb128just ask danw and pick nm I would say11:29
crevetteseb128, this active component is not really relevant, I didn't touch specific modules neither11:29
seb128oh right11:29
asaclet me check if he is online ;)11:29
seb128crevette: what component did you pick on the mango page then?11:29
asacnot either ... i will decide based on who pops-up first ;)11:29
asacseb128: do the evo guys also have a hand on gtkhtml?11:30
seb128yes, mbarnes does gtkhtml too11:30
crevetteseb128, I don't remember, dodji was the coucher for me, and I set hadess also11:30
asacseb128: ok. i marked the gtkhtml patch as a depends of the evolution bug so i guess they will see11:30
seb128asac: I expect he will review your other patch soon too, if he doesn't I'll ping him about it11:30
crevettes/coucher/voucher/11:30
asacseb128: no hurry.11:31
asaci am sure tat most patches get in in time for .111:31
seb128right11:31
asacgnome folks have always been reponsive ... except maybe at-spi ;)11:31
asachmm ... just found out that pidgin isnt even gnome ;)11:32
seb128no it's not11:36
seb128empathy is the GNOME im client11:36
seb128pidgin is only a GTK application ;-)11:36
asacseb128: re webkit. i have to check why gwibber doesnt work. i asked gwibber folks to investigate let me check11:39
seb128asac: don't bother, it's important in no way for jaunty11:43
seb128asac: I would just push epiphany-webkit to universe if we have 1.111:44
seb128but a ppa will do11:44
asacseb128: how about syncing the webkit stuff to -desktop ppa? and then seeing how well it works?11:45
asaci dont want to put it into my ppa because its currently used for "font" topic ... unfortunately there is no sub-ppa support yet11:46
asacso if we can use -desktop for that it would be great11:47
seb128asac: there is a webkit ppa11:50
seb128asac: https://edge.launchpad.net/~webkit-team/+archive/ppa11:50
asacoh good12:00
seb128pedro_: hey12:04
pedro_salut seb12812:04
seb128pedro_: I think mvo could use some help triaging his components12:05
seb128pedro_: is there anything scheduled for the new bug day?12:05
seb128I was just running accross https://bugs.edge.launchpad.net/ubuntu/+source/software-properties12:05
seb12882 bugs, 46 new12:05
seb128could that be a bug day target?12:05
asaci think mozilla components need a massive mass and flash hug ;)12:05
pedro_seb128: yep totally, there's nothing for tomorrow though and next week we are having a xorg one to help bryce with the bugs12:06
seb128pedro_: perhaps you can quickly announce software-properties for tomorrow? ;-)12:06
seb128well there is no hurry12:07
seb128I just came across it12:07
pedro_m i think we're too close to the date, we announce those to UWN as well12:08
mvoI would not mind help with hat12:08
pedro_seb128: i'll add it to the planning for April though12:08
seb128pedro_: thanks12:08
pedro_and will do some triaging there today in the meantime12:08
seb128you rock!12:09
* seb128 hugs pedro_ mvo12:09
* pedro_ hugs both 12:09
seb128pedro_: maybe you can manage to tackle that backlog by your own today who knows ;-)12:09
seb128I marked some duplicate right now but that's a bit too many bugs to triage right now for me to go through the list12:09
asaci dont even know the magnitude of my backlog12:10
pedro_heh yeah maybe, let's see how it goes :-)12:10
pedro_grgr i am the only one getting a lot of spam lately?12:11
asacdefine "a lot" ;)?12:11
pedro_well at least 40 emails a day12:12
asacpedro_: with or without spam filtering?12:12
seb128lol12:12
seb128pedro_: I'm going some hundred of spams through my spam filtering a day12:12
seb128and some thousand in the spam filter too12:13
asaci think i get a couple hundreds each day and also a bunch of legitimate mails i never see because my spam filtering is too strict12:13
pedro_asac: well I'm not filtering those locally12:13
pedro_mm i might consider that option, i'm getting tired of those12:13
asacso 40 mails get through your provider filter?12:13
pedro_yeah12:14
kenvandine_wkpitti: can you confirm you can login to ekiga.net with 3.2?12:15
pittikenvandine_wk: ack; was broken this morning, seems to work now12:16
pittikenvandine_wk: @ekiga.net's echo test still doesn't work for me, though12:17
pittikenvandine_wk: oops, it does work for me now12:17
kenvandine_wkit works for me with 3.0.1, but not 3.212:17
kenvandine_wki get this request terminated message, which there has been at least one other report of on the mailing list12:18
* kenvandine_wk is debugging it12:18
pittikenvandine_wk: ekiga.net seems to be very brittle for me, too12:18
kenvandine_wkweird it works for you12:18
pittitried it again, and it's silent12:18
* pitti wonders whether so many people complain at ekiga, the application, because ekiga.net, the service, is totally unreliable12:19
kenvandine_wklikely :)12:19
kenvandine_wkthe error code i am getting is what happens when you try to register/auth and before it answers you cancel it12:19
pittiboth my Canonical as well as my diamondcard.us account have always worked perfectly12:19
kenvandine_wkbut it happens in like a second12:19
kenvandine_wkrick said ekiga.net has been reliable for him, with 3.0.x12:20
kenvandine_wkit works amazingly well with our canonical setup12:20
* kenvandine_wk thinks it is better than skype :)12:20
pittican't compare, I have never used skype12:21
kenvandine_wki've only used it a few times... but i always had little weird problems...12:21
crevettethanks seb128 for the ack for gunp12:28
crevettegupnp12:28
seb128crevette: you're welcome12:28
pedro_asac: do you know if there's anybody working on bug 341684 (in order to assign it to and set the right status) ?12:36
ubottuLaunchpad bug 341684 in network-manager-applet "nm-applet notifications should have more information for disconnected states" [Low,New] https://launchpad.net/bugs/34168412:36
asacpedro_: commented12:39
pedro_asac: great, thanks you12:40
kenvandine_wkpitti: now ekiga.net is working for me in 3.2... but i get a segfault on exit12:42
kenvandine_wkpitti: do you get that?12:43
pittikenvandine_wk: no, I don't12:43
kenvandine_wkrunning it from a terminal?12:43
kenvandine_wkit doesn't seem to trigger apport12:43
* kenvandine_wk doesn't know what triggers that12:43
kenvandine_wkwell the segfault isn't a regression... it is happening on 3.0.1 as well :)12:44
seb128you should not spend too much time on debugging ekiga I think12:45
seb128if the new version is mostly working push it after beta and send bugs upstream when they come12:46
kenvandine_wkyeah12:48
kenvandine_wknow that i see it is working with ekiga.net i will get it ready to push12:48
kenvandine_wki didn't want to push it if it didn't work with ekiga.net12:48
kenvandine_wkwhat triggers apport of a crash?12:48
kenvandine_wki am curious why this seg fault doesn't trigger it12:49
pittikenvandine_wk: if it's a real SIGSEGV, you should get one12:49
pittikenvandine_wk: have a look at /var/log/apport.log ?12:49
kenvandine_wkok12:49
kenvandine_wkapport (pid 11235) Wed Mar 25 08:31:50 2009: called for pid 11212, signal 1112:50
kenvandine_wkapport (pid 11235) Wed Mar 25 08:31:50 2009: executable: /usr/bin/ekiga (command line "ekiga")12:50
kenvandine_wkapport (pid 11235) Wed Mar 25 08:31:50 2009: this executable already crashed 2 times, ignoring12:50
pittithere you go12:50
kenvandine_wk:)12:50
pittikenvandine_wk: rm /var/crash/*12:50
pittiand try again12:50
kenvandine_wkok, now it is triggering apport12:51
asacKeybuk: could you look into /lib/udev/rules.d/77-nm-probe-modem-capabilities.rules ... do you see anything obviously wrong with the $attr{idVendor} $attr{productId} subsitution?12:51
Keybukasac: where is that rules file?12:52
asacKeybuk: /lib/udev/rules.d/77-nm-probe-modem-capabilities.rules ?12:52
Keybukasac: other than the fact you probbaly mean $attr{idProduct} ? <g>12:52
asacyeah12:52
asacsorry. typo12:52
Keybukerr12:52
asacKeybuk: bInterfaceNumber works ... udevadm showed me that idProduct is in parent12:52
KeybukI don't have any content like that in my file12:52
asacKeybuk: oh then you are not running latest12:52
asaclet me psate12:52
KeybukI updated a few days ago ;)12:53
asacKeybuk: http://pastebin.com/f2ae57f1512:53
Keybuknm-probe-modem is in /lib/udev ?12:53
asacKeybuk: yes12:54
KeybukI'm not sure $attr{driver} will work12:54
asacKeybuk: that works12:54
asacsame for bInterfaceNumber12:54
Keybukoh, neat12:54
Keybukwhich doesn't work then?12:54
asacKeybuk: idProduct and idVendor12:55
Keybukhmm12:55
Keybuk$attr{} should walk up the parent devices12:55
asaci thought its either a udev bug not looking high enough in parent12:55
Keybukcan you show me the udevadm info -a for this device?12:55
asacor its because its a feature and i have to match something in subsystem usb12:55
Keybukcould be both12:55
asacKeybuk: yeah ... wait a sec12:55
asacudevadm info -a -p $(udevadm info -q path -n /dev/ttyUSB0)12:56
asachttp://pastebin.com/f434327a812:56
Keybukhah12:56
Keybukyou know you can use -n right? :P12:57
Keybukudevadm info -a -n ttyUSB012:57
asacheh ... no ;)12:57
* asac happy now12:57
Keybukok13:03
Keybukso this matches the SUBSYSTEM=="tty" (top-level)13:03
Keybukit matches the KERNEL name too13:03
Keybuk(though that's pretty redundant given the subsystem match <g>)13:03
asacyeah13:03
asachmm ... have to run to lunch ... back in half an hour13:04
Keybukso I think this should work13:04
Keybukit's probably a udev bug that it doesn't13:04
mvoseb128: do you know about http://people.ubuntu.com/~mvo/tmp/err1.png ?13:08
mvoseb128: that is with brasero13:08
seb128mvo: no but I'm not watching this product bugs, maybe better to ask pedro_13:08
mvooh, sorry13:09
mvopedro_: have you seen http://people.ubuntu.com/~mvo/tmp/err1.png before? happend on a fresh jaunty upgraded system when burning13:09
seb128mvo: did you get a successful record?13:11
mvoseb128: yes, that was fine13:18
pedro_mvo: if the disc was burned ok that's probably bug 29445513:26
pedro_(sorry for the delay was on a call)13:26
ubottuError: Could not parse data returned by Launchpad: timed out (https://launchpad.net/bugs/294455/+text)13:26
mvothanks pedro_13:27
pedro_i truly hate you ubottu13:27
pedro_mvo: no problem :-)13:27
seb128what's up with this bot13:29
seb128it keeps timeouting since yesterday13:29
asacKeybuk: any idea where the bug could be or do we need to workaround this for now?13:50
Keybukasac: haven't looked yet13:51
Keybukit'll just be in the bit of udev that applies $attr{}13:51
pittirickspencer3: good morning13:55
rickspencer3pitti: hi13:56
Keybukasac: I have another udev problem to debug at the same time ;)14:14
asacKeybuk: so you are looking at something related now? (/me just did a dbg build to see more log output)14:15
Keybukasac: will be in a few minutes14:15
Keybukjust clearing through my bugs to line them all up :)14:16
asacKeybuk: ok. this one is 34683514:17
pittiseb128: ah, I finally found out why consolidating the dup db takes two hours, and why we got all those "inconsistency detected: bug #336952 does not appear in get_unfixed(), but is not fixed yet" issues which slowed it down so much14:37
ubottuLaunchpad bug 336952 in gnome-applets "stickynotes_applet crashed with SIGSEGV in g_slist_remove_all()" [Medium,Invalid] https://launchpad.net/bugs/33695214:37
pittiseb128: it's fixed now; the next run will throw out all these invalidated bugs, and from then on it should be a matter of 5 minutes14:38
seb128pitti: excellent! what was it?14:38
pittiseb128: the function which checked the bug status still checked for "Rejected", instead of "Invalid"14:40
pittithat must have been changed ages ago14:40
seb128oh ok14:40
asacseb128: is there a fundamental reason why autologin with keyring cannot work?14:49
seb128asac: if you set a password for gnome-keyring? something needs to provide it the password14:50
seb128asac: when you type the password on the login screen we can unlock the keyring using that14:50
seb128if you do autologin how do you want to unblock it?14:50
asacseb128: i am looking at the -devel thread ... i just think we should fix the autologin problem instead of starting to move wifi keys to plain text14:50
seb128if the password was possible to read automatically by a software it would not be secure storing14:50
asacyeah. but making access to keyring insecure on systems that have autologin sounds more sensible than starting to move apps away from keyring14:51
seb128asac: right, see my reply on the list about setting a blank password when selecting autologin15:00
pittiI actually liked mdz's idea of just prompting for your own password15:03
Keybukthat kinda defeats auto-login though15:09
Keybuksince you still have to type a password15:09
rickspencer3I don't think we should set a blank password by default15:10
rickspencer3that seems very insecure to me15:10
rickspencer3don't people store quite sensitive info in the keyring?15:10
asacKeybuk: ubuntu branch mentioned in control doesnt have 140 (udev)? ... last log is 139-215:13
Keybukasac: pushing now15:14
asacjust fyi15:14
asacah cool.15:14
asaci wanted to diff 139 to 140 ... thats handy then15:14
Keybukasac: could you run15:15
Keybukudevadm test $(udevadm info -qpath -n ttyUSB0) for me15:16
asacKeybuk: http://pastebin.com/f4f8cf1115:17
Keybukasac: are you on i386 or amd64?15:18
asaci38615:19
Keybukasac: grab http://people.ubuntu.com/~scott/udevadm15:21
Keybukthen run15:21
KeybukUDEV_LOG=debug ./udevadm test $(udevadm info -qpath -n ttyUSB0) > udevadm.log 2>&115:21
Keybukudevadm.log will be quite large15:21
asacKeybuk: http://people.ubuntu.com/~asac/tmp/udevadm.log15:23
Keybuklet me check with kay on this one15:29
Keybukthe code is deliberately not recursing up the parents15:31
asacyeah. i saw that it somehow only does two/three steps15:31
Keybukit looking at the event device15:32
Keybuk(ttyUSB0)15:32
Keybukand the device matched (the one that DRIVERS matched I think)15:32
Keybukwhich is the usb interface15:32
Keybukthe code matches the comment15:34
asaccode == case SUBST_ATTR in udeve-event.c, right?15:35
Keybukright15:35
asacutil_resolve_subsys_kernel -> /* handle "[<SUBSYSTEM>/<KERNEL>]<attribute>" format */15:36
Keybuk"the parent device, other matches have selected"15:36
Keybukwhich is not "all parents"15:36
asacdoes that mean we have to specify something like XXX/idProduct ?15:36
Keybukohh15:37
Keybukno, this is deliberate15:37
Keybukthis is because you can do DRIVERS=="foo", ATTRS{...}=="xxx"15:38
Keybukand they both HAVE to match the same device15:38
Keybukie. it finds a single parent15:38
Keybukit's consistent15:38
Keybukthe manpage is just wrong15:38
Keybukwhich was your rule again?15:39
asacyeah15:39
asachttp://pastebin.com/f2ae57f1515:39
asacKeybuk: ^^15:39
KeybukDRIVERS will match the usb interface I think15:39
Keybukcan you confirm that for me15:39
Keybukls /sys/bus/usb/drivers/option/*15:39
asachttp://pastebin.com/f672c0f3d15:40
pedro_mvo: may you give your opinion on bug 184226 later?15:40
ubottuLaunchpad bug 184226 in software-properties "Feature Request: Install all updates without confirmation." [Wishlist,New] https://launchpad.net/bugs/18422615:40
Keybukasac: great15:41
asaccan you think of a trick how we can still get the product and vendor id for that IMPORT command? or do we need to dig that out on our own in the prober?15:41
mvopedro_: sure, thanks15:41
Keybukasac: try this15:41
Keybuk$attr{[usb/]idProduct}15:42
Keybuk$attr{[usb/]idVendor}15:42
asacdamn15:42
asaci tried usb/idProduct ;)15:42
Keybukif you change the rules, then you should be able to run the udevadm test command I gave you (using the debug binary)15:42
KeybukI'd like to compare output15:42
asaci should have read the parser more carefully ;)15:42
asacchecking15:42
asachmm15:44
asacdoesnt work i think15:44
Keybukgot the udevadm test output to compare?15:45
asaci still get15:45
asacMar 25 16:43:45 tinya udevd-event[31220]: '[usb/]idProduct=(null)' added15:45
asacMar 25 16:43:45 tinya udevd-event[31220]: will substitute format name 'attr'15:45
asacin syslog15:45
Keybukyeah, need to look at the udevadm test output15:45
asacKeybuk: you wnat the long one?15:46
KeybukI'm not quite sure it's right yet15:46
Keybukbut need to do a process of elimination to get there - since I don't have your dongle :p15:46
asace.g. with UDEV_DEBUG ... i guess so15:46
asachttp://people.ubuntu.com/~asac/tmp/udevadm2.log15:47
asacKeybuk: ^^15:47
dobeypitti: ping15:48
Keybukasac: huh15:49
asacKeybuk: did i do something wrong?15:49
pedro_mvo:  thanks you!15:50
* mvo hugs pedro_15:50
* pedro_ hugs mvo back15:50
Keybukasac: $attr{../idProduct} works though, right? :P15:51
Keybuknote no []s15:51
asacKeybuk: you mean usb/idProduct?15:51
asaci tried that yesterday after i found that code in udev_util15:52
asacor really ".."?15:52
asaclet me try ../idProduct15:52
Keybukreally ..15:53
asacKeybuk: yeah. that did the trick. thats simple ;)15:53
Keybuk.. from a usb interface kobject is the usb device kobject15:53
seb128rickspencer3: the keyring on a normal install has basically your evolution and network manager passwords15:53
Keybukok15:53
Keybukgreat15:53
seb128rickspencer3: ie wireless, email accounts, ldap and calendars server passwords too15:53
asacKeybuk: yeah ... is that the right fix ... is that how its ment to be or a trick?15:53
Keybukthat's supposed to work15:54
asacok. so its not something that will suddenly will be unsupported ;)15:54
Keybukie. it's not a hack or a trick15:54
asacgreat15:54
Keybukright15:54
asaccool. thanks for your help15:54
asacthats fixed then15:54
KeybukI'll talk to Kay later though, because I'm not entirely convinced that this attr expansion being limited to the current device is the right behaviour15:54
KeybukATTR matches, sure15:54
Keybukeven ATTRS matches15:54
Keybukbut once you've matched, you should be able to use $attr{something-in-a-parent}15:54
Keybukat least I think so :)15:54
Keybukhmm, but then I suppose $env{} doesn't behave that way15:55
Keybukat the very least, this type of thing should be documented with examples <g>15:55
Keybukand the manpage fixed15:55
pittidobey: hi16:01
dobeypitti: hey. i found some issues with the check command, and just pushed some fixes to my branch16:02
dobeypitti: and just proposed a merge for it :)16:03
pittidobey: ah, nice; thanks16:04
asacKeybuk: i liked the manpage behaviour ;)16:07
Keybukasac: it doesn't quite make sense though with the way udev tries to parse its rules16:07
Keybukit's been gradually moving to a very well defined narrow behaviour16:08
Keybukto avoid unexpected issues16:08
Keybukesp. importance one everything currently in HAL fdi has to be rewritten as udev rules16:08
asacKeybuk: yes true. it was just confusing for me as an outsider reading the manpage and then not being able to access any attr from parents further up16:11
pedro_MacSlow: may you look at  344888 if you have a time later? it was reopened a few hours after you closed it16:14
pedro_MacSlow:  bug 34488816:14
ubottuLaunchpad bug 344888 in notify-osd "fade when mouse cursor in/out" [Wishlist,New] https://launchpad.net/bugs/34488816:14
MacSlowpedro_, *sigh* who did that? I wonder if they read this comment from Mat "Update - this won't be fixed for Jaunty (not enough time), but we'll factor that in for future development."16:16
MacSlowpedro_, I would set it to confirmed16:17
pedro_MacSlow: Mat reopened it. Ok, I've set it to Confirmed, thanks!16:19
Keybukasac: -> #udev?16:41
seb128re17:20
seb128ok, so how do I valgrind X? ;-)17:21
seb128valgrind doesn't like the X binary setuid17:21
seb128ups17:23
seb128wrong channel17:23
pittiseb128: bug 328035 has a recipe for strace, which certainly works for valgrind as well17:47
ubottuLaunchpad bug 328035 in xorg-server "X server crash: *** glibc detected *** free(): in valid next size (fast)" [High,Fix committed] https://launchpad.net/bugs/32803517:47
seb128pitti: cf #ubuntu-devel17:47
dobeypitti: hrmm. are you/glatzor planning to release a new version of distutils-extra soon?17:56
pittidobey: yes, I'll upload it to debian tomorrow morning17:57
pittidobey: and then we can sync it to jaunty after the freeze (on Friday)17:57
dobeypitti: ok. i was going to build a package and stick it in ppa, so i can start using it in code.17:59
dobeypitti: and then we can just use the jaunty version after you sync it17:59
pittidobey: sure, sounds good17:59
dobeypitti: awesome! thanks again18:00
pittidobey: uploaded to Debian now18:23
dobeypitti: awesome18:24
dobeypitti: i've put a version in my ppa too :)18:24
jcastroseb128: are you guys (desktop people) staying for the full desktop summit or just the first few core days?18:41
seb128jcastro: dunno yet18:42
jcastroseb128: when you book travel can you ping me? I want to travel with the cabal.18:42
seb128depends of the actual scheduling details18:42
seb128ok18:43
seb128I think we will discuss it during the desktop team meeting next week18:43
jcastroah great, I'll just follow along the notes then18:43
seb128is the detailled schedule available now?18:43
dobeyah crap, i need to do that18:43
* mvo hugs pedro_ for his software-properties bug triage19:39
* pedro_ hugs mvo back19:40
pedro_mvo: btw could you look at bug 308920 ? ;-)19:40
ubottuLaunchpad bug 308920 in update-notifier "Add option to not check/download updates automatically when using mobile broadband" [Wishlist,New] https://launchpad.net/bugs/30892019:40
* seb128 hugs mvo pedro_19:40
* pedro_ hugs seb12819:40
=== jono_ is now known as jono
cjawwww22:55
=== bluesmoke_ is now known as Amaranth
=== asac_ is now known as asac
Laneybratsche: What should be done for the gnome-user-share bug?23:39
LaneyIs the patch right?23:39
bratscheLaney: Let me take a look.. do you have the url handy?23:50
bratscheI found it23:51
bratschehttps://bugs.launchpad.net/ubuntu/+source/gnome-user-share/+bug/337352  ?23:51
ubottuUbuntu bug 337352 in gnome-user-share "gnome-user-share notification changes" [Medium,Confirmed]23:51
bratscheLaney: If that's the one you mean, then no.. I think this patch is incorrect.  I'll try to post a new one.23:54
bratscheSorry this one kind of fell off my radar.  I'll try to implement the new specification soon.23:55
Laneybratsche: (bug 337352)23:56
ubottuLaunchpad bug 337352 in gnome-user-share "gnome-user-share notification changes" [Medium,Confirmed] https://launchpad.net/bugs/33735223:56
LaneyI'll unsubscribe the sponsors then. Please resubscribe when ready23:56
bratscheThanks.23:57
bratscheSorry for the trouble.23:57
Laneyno problem at all23:57
bratscheI'm still kind of new at Ubuntu, figuring how how to work in Launchpad still. :)23:57

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