dwilbert@bbn.com (Deborah Wilbert) (12/07/90)
I have been experiencing a problem with my software which I am at a
loss as to how to track down. I am using the following configuration of
hardware and software:
Sun SPARCStation1
Dual heads, one with the GX, one with the CX
SunOS 4.0.3
Sun OpenWindows 2.0
ICS Motif 1.0.3
The problem is with the mouse pointer. Everything will be operating
smoothly for long periods of time. Then, seemingly at random, the mouse
pointer will start to act flaky. This has occurred both when someone was
actively using the program and when someone returns to the program after
a period of inactivity.
The exact nature of the mouse pointer's flakiness has to do with tracking
from one monitor to the other. Under normal circumstances, as the user moves
the mouse from left to right, the mouse pointer will move from the
left-hand-side of the left-hand monitor to the
right-hand-side of the left-hand monitor to the
left-hand-side of the right-hand monitor to the
right-hand-side of the right-hand monitor.
When the problems start to occur, moving the mouse from left to right causes
the mouse pointer to move from the left-hand-side of the left-hand monitor to
the right-hand-side of the left-hand monitor. When it reaches the
right-hand-side of the left-hand monitor, it disappears. Eventually, if one
continues to move the mouse to the right, the mouse pointer will reappear at
the left-hand-side of the left-hand monitor and move to the right. It
finally "hits the wall" about 1/5th of the way across the left-hand monitor.
Reversing the mouse movement back to the left produces the reverse behavior
and so on. It seems as if MWM or the XServer has been affected, because
none of the mouse button events has any effect on my program's widgets or
on the MWM window frame widgets. On the other hand, once my program is
killed (one must remotely login and use "kill" to do this), the mouse
pointer works fine and mouse button clicks regain their effect.
Does anyone have any insights as to what is happening or how I could go
about diagnosing the problem? Please email responses to me directly.
-------------------------------------------------------------------------------
Deborah Wilbert (Alien examining Pioneer 10)
Bolt, Beranek & Newman "Looks to me like what we got here is a race
dwilbert@bbn.com of hydrogen-obsessed pervert science wimps."
Luckily, my last name's not Ilbert. -Dave Barry