Nathaniel Smith spake unto us the following wisdom: > On Fri, Aug 15, 2008 at 3:35 PM, Ethan Blanton <[EMAIL PROTECTED]> wrote: > > Here's the xprop for some xpra litter: > > > > http://elb.elitists.net/elb/litter-xprop.txt > > > > Looks like it doesn't have another window associated with it, maybe? > > If I xprop its parent (a firefox window), I get a bunch of junk about > > firefox. Interestingly, if I xprop that window when it's "on top of" > > another window, I get the information for the window below (at least > > some of the time). > > I'm a bit confused here -- that looks like xprop run on the window > popped up by the xpra client, which is fine, but I suspect you don't > mean that the xpra client's window is a child of a firefox window?
Sorry, I should have been more explicit.
This is xprop run with the xpra Xvfb as its DISPLAY, clicking on the
litter left behind by firefox. The firefox process, in this case, has
already exited. The window behind the litter on the "real" X server is
firefox, but I don't know how the xpra X server would know that (?).
> Could you arrange that there are no other windows (esp. xpra windows)
> overlapping a piece of litter, and then send, clearly labeled, the
> results of:
> xprop when run against your real X server
> xwininfo -all when run against your real X server
> xprop when run against xpra's Xvfb
> xwininfo -all when run against xpra's Xvfb
> ?
I will do this the next time I get litter. I'll try to induce it this
afternoon, in fact.
> > Firefox seems to litter pretty regularly, by the way...
>
> Even though I don't understand everything you said, the mention of
> some window having Firefox as a parent has gotten me wondering if this
> is some weird interaction with Firefox plugins. Say, our getting
> confused by them popping up on the root window, then getting
> reparented into firefox. I even remember there being some weird issue
> with firefox plugins before... but when I try now, firefox seems to
> behave fine under xpra. It might be because my previous tests were
> with firefox 2, whereas now with firefox 3 I can't seem to trigger any
> bugs.
No plugins should be involved here, at least directly. This litter
appeared over the menu/tabbar area of a firefox window which was
displaying a rather simple (plugin-content-free) web page. I'm pretty
sure all of my litter problems have been caused by tooltips, though.
I've seen litter on firefox, azureus, and various Gtk+ applications.
My xpra server is running on a not-so-fast machine (dual PIII/500MHz),
and my xpra client is connected to it over a not-so-fast connection
(802.11g with a fair amount of interference). It's my *feeling* (no
hard evidence to back this up) that most of the litter occurs when a
tooltip is posted by the X client within xpra, but my mouse cursor has
moved on (and thus dismissed it) by the time the window is rendered by
the xpra client.
I'll collect more data as I can.
Ethan
--
The laws that forbid the carrying of arms are laws [that have no remedy
for evils]. They disarm only those who are neither inclined nor
determined to commit crimes.
-- Cesare Beccaria, "On Crimes and Punishments", 1764
signature.asc
Description: Digital signature
_______________________________________________ Parti-discuss mailing list [email protected] http://lists.partiwm.org/cgi-bin/mailman/listinfo/parti-discuss
