[13:24] b'ah, this KDE patch is totally screwed for firefox 10 [14:41] micahg: Hi. I'm sure you know this but I see Seamonkey tagged 2.6.1 this morning. I guess there are issues with 2.6, so we should wait off on that one. [16:12] m_conley_away, have you seen where the next UDS is? [16:12] http://uds.ubuntu.com/ [16:12] oakland :) [17:20] * timeless pokes alex_mayorga [17:57] joelesko: right, there was a respin [18:39] * alex_mayorga acknowledges timeless poke [18:40] ok [18:40] did, you set the breakpoint on fputs? [18:41] timeless: no, I didn't [18:43] can you please? :) [18:44] i'd like to get stacks done so that you can write a howto :) [18:45] timeless: I can do it when I get back home, drop the instructions on the piratepad [18:46] i did [18:46] they're below your "this didn't return anything" [18:47] timeless: OK would check that out later, BTW the actual bug causing all my tribulations is https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/907012 [18:48] Launchpad bug 907012 in indicator-datetime "indicator-datetime uses 80% of RAM or 3.1g" [Undecided,Incomplete] [18:48] that thing is eating all the RAM and making Nightly act wonky [18:48] cute [18:48] well, aptitude install valgrind :) [18:49] or something [18:50] timeless: the problem is that it is a service and I'm befuddled on how to valgrind such thing [18:50] * alex_mayorga is learning to debug ubuntu [18:51] pstree [18:51] if the service has a parent, then find the location of the parent's script [18:52] edit the script to call valgrind w/ the service + args as args [18:52] or just replace the service app on disk w/ a shell script that calls valgrind + renamed service + args [18:52] either way [18:52] it's probably easier to rename the service on disk [18:52] e.g. mv /usr/sbin/stupid /usr/sbin/stupid-bin [18:52] cat > /usr/sbin/stupid [18:52] #!/bin/sh [18:53] /usr/sbin/stupid-bin $* [18:53] err [18:53] /usr/bin/valgrind /usr/sbin/stupid-bin $* [18:53] hah, indicator-datetime leaking memory. why am i not surprised? [18:53] you can even make the wrapper clever and have it look for a file to decide if you it should use valgrind or not [18:54] if [ -e /usr/sbin/stupid.valgrind ] ; [18:54] DEBUGGER=/usr/bin/valgrind [18:54] else [18:54] DEBUGGER= [18:54] fi [18:54] $DEBUGGER /usr/sbin/stupid-bin $* [18:54] something like that :) [18:54] then you can touch /usr/sbin/stupid.valgrind if you want the service to be valgrinded and rm it if you don't [18:54] one day I'd get there ;-) [18:55] (you'd of course need to restart the service, which you could probably do by shooting it [kill!] [18:55] anyway, for valgrind, you /may/ or may not want additional args or output redirectors... [18:56] chrisccoulson: it leaks really bad on my laptop, but I asked on #ubuntu+1 and no one else seems to have that [18:56] chrisccoulson: are you using pangolin? [18:56] alex_mayorga, yeah [18:57] but i have 8GB of ram and a SSD. i don't always notice such things ;) [18:57] chrisccoulson: perhaps you could try to repro ;-) [18:57] alex_mayorga, do you use evolution? [18:57] chrisccoulson: no that I'm aware [18:58] it's only using 1.7MB here. the biggest consumers on my machine are thunderbird (420MB) and firefox (380MB) [18:59] and zeitgeist (270MB). yikes [18:59] * chrisccoulson kills that [18:59] zeitgeist? [19:00] timeless, it's for event logging [19:00] unity uses it for prioritizing most frequently used apps and most recently/frequently opened files in the dash [19:00] but it's written in python [19:02] python doesn't have to be bad [19:02] i consider it bad to write long running daemons in python :) [19:03] would you rather java? :) [19:03] hah [19:04] * micahg wonders if chrisccoulson is asking for more erlang daemons [19:06] lol [19:06] that's probably as bad as java! [19:07] :D [19:19] is there a new blocker bug? aurora publishing is disabled again [19:20] xjjk, no [19:20] why disabled? [19:21] it's disabled until the new 11.0a2 builds are working [19:22] oh, OK. so there isn't a blocker bug like there was w/ 10.0a2 [19:22] no, but i don't recall there being one before either [19:23] for 10a2 there was something about Firefox disabling distribution-provided extensions [19:23] in particular, all of Ubuntu's [19:24] no, that was for the 8.0 release [19:24] bug 888307 [19:24] Launchpad bug 888307 in thunderbird "Bundled Firefox extensions disabled on upgrade to 8.0" [Critical,Fix released] https://launchpad.net/bugs/888307 [19:25] ah [19:30] nice, all the builds are failing [19:45] on a side note… is there any chance of KDE integration coming back [19:46] I think it was dropped a while ago, and I miss it a lot =/ [19:54] * alex_mayorga hat tips to chrisccoulson in case he fixed the dailies [20:00] alex_mayorga, yeah, those should be fixed now [20:00] by dropping the KDE integration ;) [20:00] the KDE integration is completely broken from firefox 10 onwards, and i don't really have time to fix that [20:03] chrisccoulson: if debfx fixes it, could you reenable? [20:03] micahg, not for precise. and it looks like it's pretty difficult to fix without reddesigning the nsKDEUtils interface [20:04] i've already spent a fair bit of time trying to fix it up, but it's impossible to link libbrowsercomps.so [20:13] hasn't opensuse ported the patches to ff 10 yet? [20:13] debfx, no. and for the same reason that we haven't [20:17] I wish they had spent their time working on a kde integration that is actually upstreamable :/ [21:00] micahg, are you taking care of 9.0.1? [21:02] chrisccoulson: yes, you want me to upload to firefox-stable also? [21:02] micahg, i'll just copy the package from security-proposed [21:02] that's what i did before [21:02] chrisccoulson: ok, that might not be up until late tonight, I want to add the ia64/sparc patches before I upload those [21:03] micahg, don't worry about those. i'd rather not increase risk any further for 2 architectures that nobody is even using. i'd rather those waited for another release tbh [21:03] sparc isn't even bootable, and ia64 only exists in server hardware afaik [21:04] ok, I can upload 9.0.1 then and pick up those fixes later, will do in a bit [21:04] thanks [21:04] it's a regression over 3.6 though, so before the official testing starts, I wanted to get the patches in, but I have 1.5 weeks for that :) [21:05] we've had sparc / hppa failures in the past (when we did the 3.0 -> 3.6 transition), and we never really worried about those [21:05] the number of users affected by that is pretty much zero, seeing as sparc users can't even boot lucid ;) [21:09] I thought sparc couldn't install, but if you upgraded it would boot [21:09] anyways, Debian has the patches, so it's not a lot of work [21:14] i'd still prefer not to add anything to the build that adds risk for next to zero return [21:16] * micahg will look at the patches, but guessing there's almost 0 risk [21:17] chrisccoulson: BTW, thunderbird.oneiric is messed up, seems like my changes didn't get in the branch before you merged, I'll work it out, I have my original one locally [21:21] which changes? [21:24] changelog stuff [21:24] i added it already didn't i? [21:24] yeah, it was for the tracking bug I think, not your fault, I think I might have forgot to push