[10:38] <didrocks> JohnLea: hey, on the user test results from the previous unity, I got those feedbacks on unity:
[10:38] <didrocks> people complained on wm/menu_basic that right clicking in the title bar works and bring a menu, but not when the application is maximized (as the title bar is in the panel). If we don't want to change that design-wise, we need to rephrase the test.
[10:52] <JohnLea> didrocks; the ultimate plan is to remove this menu, but we need some new features to land first.  Was going to remove the menu in 11.10 but this would have been premature.
[10:54] <didrocks> JohnLea: ok, thanks for the summary :)
[12:45] <mhall119> didrocks: JohnLea: FYI, we're trying to move all the Unity conversations to #ubuntu-unity
[12:46] <didrocks> mhall119: yeah, I just talk to him here as he's not on #ubuntu-unity!
[15:06] <smoser> greyback, just commented on bug. it most certainly seems to be "fixed" if i don't run indicator-multiload
[15:14] <didrocks> smoser: ok, so either indicator-multiload issue or on the libappindicator level
[15:14] <didrocks> tedg: do you remember about this?  ^ memleak when an indicator changes a lot of the image to show
[15:15] <tedg> didrocks, Uhm, perhaps.  There was a ref_sink() one a while back.
[15:16] <didrocks> tedg: do you think the issue is in the appindicator part then?
[15:16] <didrocks> (libappindicator, isn't it?)
[15:17] <tedg> didrocks, Uhm, it could be.  I mean, I have no reason to say it isn't, but I also don't really have any proof to say it is :-)
[15:17] <tedg> didrocks, Where is the leak showing up?
[15:17] <didrocks> tedg: smoser got high mem leak using indicator-multiload (I think it's in the unity-panel-service)
[15:17] <didrocks> smoser: ?
[15:18] <didrocks> tedg: cjwatson was complaining about the same thing not that long ago
[15:18] <didrocks> with indicator-multiload as well
[15:18] <greyback> smoser: great, thank you
[15:18] <smoser> unity-2d-panel is the program that shows the memory usage
[15:19] <smoser> and it gets *really* fat over time. i'm sure that valgrind woudl show the leak pretty quickly.
[15:19] <smoser> it leaks fast
[15:19] <smoser> ie, hundreds of MB in 5 days or something
[15:19] <didrocks> tedg: can be 2 issues then, we had the same in 3d
[15:19] <tedg> Ah, so then it's not libappindicator, that's in the program's memory space.
[15:19] <didrocks> smoser: ^
[15:19] <didrocks> yeah
[15:19] <tedg> It would have to be in indicator-application.
[15:20] <tedg> smoser, Run valgrind please?  :-)
[15:20] <smoser> tedg, i do have other things to do
[15:20] <smoser> :)
[15:20]  * tedg is not going to install multiload on principle
[15:20] <seb128> didrocks, is that specific to precise?
[15:20] <didrocks> seb128: no, it's not, cjwatson had that on oneiric
[15:20] <didrocks> but some part was fixed
[15:20] <didrocks> hence, I'm wondering if it's now -2d only or still due to service
[15:21] <seb128> ok, I though it was fixed in oneiric, I wondered if that regressed in unity in precise
[15:21] <didrocks> some part, IIRC, the leak was never fully fixed
[15:21] <didrocks> so I guess that maybe greyback is leaking as well :)
[15:22] <greyback> smoser: yes I read that in your report, it's on the todo list:)