On Fri, 13 Apr 2001, Christoph Egger wrote:
> a) do nothing => blit-operation fails
>
> b) stretch or scale the source during the blit down to a size that
> fits
>
> c) Just do the blit as it would fit and cut off the part that doesn't
> fit
> Problem 1: Method b) and/or c) might not be supported by some targets.
Method b) should require the user to use a StrechBlit API function;
it should never be default behavior.
Method c) should be emulated if it is not supported. Cases where it
is not will be rare. It should be the default.
There is another option for the small->large op, that being tesslating the
source to the destination. That should also only be done with a special
function, or, if the ROPs I sent are what we will be using, setting the
ROP bit for pattern fill.
--
Brian
