The screen resolution switching is simpler (I feel). You just tell it what
you want, and you get the best it can do. If it can’t switch down to a
lower resolution, it puts a window in the middle of the screen and blacks
out the rest. It would be difficult to simulate your code with this.
I agree. I’ve seriously considered doing this, but I want to give flexibility
of enumerating the available video modes to the user…
Actually, you could do both:
Allow enumeration
In video mode set:
If video mode is one of those enumerated, set it
else set closest mode
I like that. The only problem I can see is when you are setting a small
mode in 8 bpp and then do palette changes. The background color will change
drastically, although if we punt like some video cards and just set it to
zero, that might make it easier…
I’d also like to provide enumeration of audio formats and rates, but
very few audio APIs support this.
The screen resolution switching is simpler (I feel). You just tell it what
you want, and you get the best it can do. If it can’t switch down to a
lower resolution, it puts a window in the middle of the screen and blacks
out the rest. It would be difficult to simulate your code with this.
I agree. I’ve seriously considered doing this, but I want to give flexibility
of enumerating the available video modes to the user…
If you do this, there will be a problem with direct memory access, I mean there
will be RealWidth-FakeWidth pixels per scanline that we have to skip over when
doing memcpys or putpixels.
If you do this, there will be a problem with direct memory access, I mean there
will be RealWidth-FakeWidth pixels per scanline that we have to skip over when
doing memcpys or putpixels.
This is already true with many hardware surfaces where
surface->h != surface->pitch;
This can happen often in DirectX, and sometimes in X11 DGA, so
you should never assume that you can write more than one scanline at a time.
If you do this, there will be a problem with direct memory access, I
mean there will be RealWidth-FakeWidth pixels per scanline that we have
to skip over when doing memcpys or putpixels.
Is this a problem though? skipping pixels is only an add… should be
buried by the memory operations.