On Sun, May 24, 2009 at 11:25 PM, Dwight Schauer <[email protected]> wrote: > If I move from an xpra window to another xpra window the focus changes > correctly (and cursors change to indicate that focus change). > > However, if I move from xpra window A to non xpra window B then xpra > window A still beleives it has focus (cursor still indicates focus) > even though non xpra window B really has the focus. None of the > keypresses actually go to xpra window A, even though it still beleives > it still has focus. This behavior is consistent regardless if I'm > using focus follows mouse or click to focus. If I move from non xpra > window B back to xpra window A, then window B shows loss of focus and > keypresses now go into window A again. > > The above behavior is consistent when using xpra with xfwm4, metacity, > and compiz. > > I'm wondering if this somehow gets xpra's window focusing confused and > causes xpra windows to lose focus altogther.
I can't see how the behavior you describe could be causing any extra problems, but it is certainly not how things are supposed to work (and is not how it works here. drat.) When a client window loses focus, the underlying window on the server is supposed to lose focus; when a client window gains focus, the underlying window on the server is supposed to gain it. We knew that latter didn't work for some people; apparently the former doesn't either. Here's something to try: could you use 'xpra attach ... -d xpra.protocol' to attach, and then watch the scroll in the terminal you used? I'd like to know: -- what gets printed when you move the mouse from A to B -- what gets printed when you move the mouse from B to A (ideally, both in the case where it works and the case where it doesn't, if it's intermittent, but definitely in the case where it doesn't) -- what gets printed when you move the mouse from one xpra window to another In particular, I'm wondering about lines like "sending ['focus', ...]". -- Nathaniel _______________________________________________ Parti-discuss mailing list [email protected] http://lists.partiwm.org/cgi-bin/mailman/listinfo/parti-discuss
