Robert Comer wrote:
>> Okay, that's good to know. So we're left just waiting for MSFT to
>> accurately pass the monitor metrics through to the guest VM, then. :-(
>
> I wouldn't know, I never use extended desktop like that, however I don't
> think it's passing parameter that's the problem, it's probably the drawing
> technology. In VPC2007 it was a directdraw problem, but I don't know about
> WVPC.
I could find out definitely enough, I guess, by just writing up a quick
little monitor enumeration applet and running it in that mode. But I
think, from the time I've spent working with the monitor API, that it's
pretty clear what's happening. The seamless apps are simply being told
there's one monitor, and its dimensions are given as that of the entire
virtual monitor which is the accumulamulated area of all physical
monitors combined.
IOW, say you have two monitors, side by side. Both running at
1280x1024, with the 0,0 point at the upper-left of monitor 1, which is
on the left. The coordinate space for the twp physical displays would
be something like:
0,0 1279,0 1280,0 2559,0
*1* *2*
0,1023 1279,1023 1280,1023 2559,1023
(Hope you have a fixed-width font! <g>)
But monitor "0" is the virtual display, that covers the whole thing.
It's coordinates would be:
0,0 2559,0
*0*
0,1023 2559,1023
That's the one being given to the seamless apps, when they inquire
about monitor dimensions. This also explains why their popup dialogs
(msgbox's, etc) so often appear split between the two monitors.
They're simply centering themselves on what they see as "the" monitor.
--
..NET: It's About Trust!
http://vfred.mvps.org