GX/cursor problem in SunView

[David L. Markowitz]

In comp.sys.sun you write:

>I've got a weird problem with a piece of SunView software that seems to be
>correlated with the availability of a GX accelerator.  Basically, as I
>move the (crosshair) cursor over a canvas containing rapidly changing
>graphics, bits of graphics don't get erased properly and portions of the
>cursor drop out.  If I refresh the screen everything is okay again until I
>start moving the cursor.  It looks like there is some cursor/graphics
>interaction which is screwed up, but since this has worked fine until the
>last year or so and still works great on machines without a GX, I'm not
>convinced that the problem is in the software.  Are there any known bugs
>involving the GX or SunView cursors?

I have also recently discovered a "Sunview crosshair cursor on the GX"
bug.  This is on a SS1+ running 4.1c.  It is similar:

I have a crosshair over a window with graphics that are COMPLETELY
rewritten about 5 times a second, so it is hard to determine if any data
under the cursor is being corrupted.  The bug manifests itself by making
the crosshair dissapear when the mouse is not moved (i.e.  the window is
rewritten, but the cursor isn't refreshed afterwards).  As soon as you
move the mouse, it reappears (until the next stop).

Personally, it doesn't affect my application much, and I am planning a
port to XView soon anyway, so I consider the bug almost moot.

	David L. Markowitz
	Genisco Technology Corporation
	dav at gtc.com