Re: Request a discrete refresh circuit schematic

["sicklittlemonkey" <Nick.Westgate@gmail.com>]

heuser.marcus@freenet.de wrote:
> I don't see how non-linear video memory actually saves memory. AFAIK it
> saves address decoding circuitry.

In "Undestanding the Apple II" (Appendix I) Woz says he could have
added linear addressing with 2 more chips. I guess we can blame his
chip count fetish.

> But it would've helped tremendously if the GS would've been clocked
> faster for this fourfold increase in video memory - from a game
> programmers perspective the linearity doesn't really help here.

On modern platforms it does make a difference because you can simply
add/subtract the screen width in bytes to a pointer in a register to
draw the next row of a sprite etc. On the Apple II you're probably
correct, as a CLC, ADC #$40 has to interact with a pointer in memory.

Still, I'm guessing a few cycles might be saved, which could be
relevant.

Cheers,
Nick.