On Mon, 3 Jul 2000, X-Odus wrote:
> Chris Emerson wrote:
>
> > I say stay compatible with current viewers. What's the actual
> > problem?
>
> the RFB protocol sends the exact X,Y position of the mouse, where as
> plex86 emulates a PS/2 mouse and sends movement (ex: up 1, right 1).
> Its not difficult to get how far the mouse has moved by comparing
> the X,Y's but its hard to tell where the mouse in the emulated os
> starts out, so the initial mouse coordinates has to be guessed
> (although i assume it to be center).
That's a bit of a problem. Could it emulate something like a
tablet-type input instead, which uses absolute co-ordinates? Or
perhaps some sort of hack like sending "up 16000, left 16000" at the
start to force the mouse in the guest to the top left of the screen.
> Also, the vncviewers only caputure mouse movements while the mouse
> is over the window, so if you move off the window on the right side,
> then come back onto the windows from the top or some other side, it
> will screw up where the mouse is.
If it's sending absolute X,Y coordinates anyway, isn't that fine?
It'll just jump to whichever edge it comes on from, which is what you
want it to do.
> > I have no problem with supporting a "better" way of doing it for
> > those with suitable viewers, but I think it's important to be
> > compatible with current viewers. I doubt you could modify them
> > all.
>
> I don't think I would do all the modifying... and even if so, the
> modifications arent that big. Just make the viewers capture all
> mouse movements instead of just when the mouse is over the window.
> That should fix the problem at least.
I don't see how that helps. When the mouse isn't over the window, the
guest shouldn't be getting mouse events. When it comes back on, it
should get a jump to the new location, which you can do anyway, can't
you? This still doesn't solve the problem about where the guest
cursor starts off.
Let me know if I'm being stupid. :-)
Chris
--
Chris Emerson, obsessed Cambridge juggler
E-mail: [EMAIL PROTECTED]
Web page: http://www.chiark.greenend.org.uk/~cemerson/
