[00:35] <Fudge> attente:  are you ears on?
[00:36] <Fudge> bug #1282782, was wondering if that would also in your opinion cause Orca to read the title of the active window everytime the 'alt' key is touched?
[00:36] <ubot2> Launchpad bug 1282782 in compiz (Ubuntu) "Newest compiz/unity has decided that it owns my Alt key" [High,Triaged] https://launchpad.net/bugs/1282782
[00:43] <attente> Fudge: not really sure tbh
[00:43] <Fudge> attente:  thanks anyway mate
[00:43] <Fudge> I'll keep looking
[00:43] <attente> holding the alt key does show the menu bar
[00:44] <attente> but i'm not really familiar with how orca works
[00:44] <attente> it definitely sounds plausible though
[00:46] <Fudge> :) basically if you had Orca running, push ALT a bunch of times and Orca will on each keypress read the title of the window. e.g On thunderbird just now, alt=orca saying inbox unified folders mozilla thunderbird frame then if I push it again I hear it again.
[00:46] <attente> Fudge: do you know how long that bug has been happening?
[00:47] <Fudge> quite a while attente , I've been on Trusty from the start, I would have to say confidently at least a few months.
[00:48] <attente> Fudge: so that bug that you linked to here started around late Jan, early Feb
[00:48] <attente> i mean the regression was introduced around that time
[00:50] <Fudge> is it specific to jsut gnome terminal?
[00:51] <attente> Fudge: no, but holding the alt key in general causes the menu bar to become visible
[00:51] <attente> Fudge: sorry, it was actually introduced on Feb 14
[00:51] <Fudge> ah the penny drops, so it is hidden before then is it?
[00:51] <attente> Fudge: before alt had no effect on showing the menu bar due to a bug
[00:51] <Fudge> how about tapping the alt, does it start to appear?
[00:52] <attente> Fudge: tapping alt would cause the hud to appear
[00:52] <Fudge> I have hud disabled
[00:52] <attente> but i don't know why it would cause the screenreader to read the window title
[00:52] <attente> hmm.
[00:53] <Fudge> it could be since the bar is appearing that Orca is reading the whole thing and not just the menu focus?
[00:53] <attente> Fudge: if you remember having the bug older than Feb 14-ish, i'd say it's caused by something else.. sorry :(
[00:53] <Fudge> I'll ask Themuso since you've given me better understanding of it now
[00:53] <attente> Fudge: good idea
[00:53] <Fudge> you've still helped me so thank you mate
[00:53] <attente> Fudge: Trevinho would also be helpful i think
[00:54] <attente> Fudge: good luck, sorry i could help more :)
[00:54] <attente> *couldn't
[00:54] <Fudge> thank you attente. Trevinho  if you followed our conversation maybe you might have some ideas?
[01:13] <darkxst> attente, The keygrabber moved into Unity around that time
[01:13] <darkxst> prior to that key grabs were handled by gnome-settings-daemon
[03:39] <attente> darkxst: yep, i just wasn't sure if that was the cause since Fudge said he had the problem for a few months now
[23:46] <robert_ancell> seriously, how long does unity take to compile?
[23:51] <ochosi> hey robert_ancell
[23:51] <robert_ancell> ochosi, hey
[23:51] <ochosi> robert_ancell: i quickly wanted to ask you about that gtk-greeter bug we're seeing where the greeter doesnt exit and appears in the session (if you have time)
[23:52] <ochosi> (if you dont have time, no worries)
[23:52] <robert_ancell> ochosi, sure
[23:52] <robert_ancell> I have some time
[23:52] <ochosi> robert_ancell: so as i said last time, i can hardly reproduce the bug and so far i've only seen it once (didn't think it would be a greeter bug back then and hence didn't save the logs)
[23:53] <ochosi> any idea how to even get to reproduce this? basically the gtk-greeter still exits the same way it always did
[23:53] <ochosi> so i have no idea what could go wrong there
[23:54] <ochosi> the only thing that's new is support for gtk3 indicators (started via upstart and then killed, same as in unity-greeter)
[23:54] <TheMuso> robert_ancell: Yeah it does take a while to compile.
[23:55] <robert_ancell> ochosi, I'm kind of at a loss too without being able to reproduce and check what's running post login
[23:56] <ochosi> robert_ancell: i'm wondering whether there is any kind of debug output i could add to help with this, but i guess then i'd already have to know where the problem lies..
[23:57] <ochosi> robert_ancell: so i suppose there weren't any changes in lightdm that could cause this behavior?
[23:57] <robert_ancell> ochosi, it's possible the signal handling got messed up and lightdm just thinks the greeter quit when it didn't?
[23:57] <robert_ancell> But I'm pretty sure it wouldn't do anything until the process actually quit
[23:59] <ochosi> robert_ancell: well what we added and what wasn't there previously (although that rather should fix stuff than break it) is gtk_main_quit(); in the sigterm_cb before exit (0);