> The problem is that namely HTML rendeing is very complicated and my
> knowledge of ASM is poor. But to start, someone may wish to write some
> kind of ultra fast HTML tag analysis (I am currently doing switch based on
I will send you some routine witch I have already done for windowing
system I am curently working on.
> So I think I will keep current virtual screens with software scrolling.
> (I am sure there must be way to accelerate it, as there are pretty fast
> demos which use same modes as Arachne...).
Well If I understud right what virtual screen is(part of mem in witch
you render html page at oance and then just copy part you need to video
memory) then routine for drawing virtual screens should be simply mem to
screen copy. And that shouldnt be some cind of philosophy. Maby the
problem with your code is that it uses only 16bit instructions and you can
copy only words and not dwords ? I woud like to see your routine drawing
VR sreen.
