Georges Ibrahim writes:
 > My problem is that sometimes (happens about 10% of the time) the click
 > on the screen is only partially acted on: the button gets depressed, but
 > is not released, and therefore the signal doesn't get generated. It is
 > only after the user clicks somewhere else on the window where the button
 > is located that the button is no longer pressed and the signal gets
 > sent.

I don't have any help, but I can say that I have seen remarkably
similar behavior with XFree86 and the touchscreen on the Compaq
iPAQ. The button release event seems to be held up in the X server
until some other event causes it to be flushed. I've found that either
tapping the touchscreen or pressing a button, (which causes a KeyPress
event), will cause the ButtonRelease to finally go through.

I have no idea what code (if any), is common between the processing of
these two touchscreens, but if this is the same problem, maybe
this will help point someone down the right track.

-Carl

PS. I can't absolutely rule out a hardware problem on the iPAQ, (or a
software problem in the touchscreen device driver). I keep meaning to
sit down with my iPAQ sometime with some other machines handy and wait
for the ButtonRelease to get stuck. Then I could cause some X events
over the network without touching the touchscreen or any physical
buttons. If that causes the ButtonRelease to flush then we could be
quite certain this is a software problem in X somewhere.
_______________________________________________
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert

Reply via email to