Re: Request a discrete refresh circuit schematic

["Bryan Parkoff" <nospam@nospam.com>]

<heuser.marcus@freenet.de> wrote in message 
1158500306.606660.269840@k70g2000cwa.googlegroups.com">news:1158500306.606660.269840@k70g2000cwa.googlegroups.com...
> Knut Roll-Lund wrote:
>>
>> To save a little logic he made all graphics programming a nightmare...
>> :-)
>
> The addressing of the video memory isn't that bad, IMHO.
> For anything speedy you need an address table anyway and it does not
> matter what addresses are in it.
>
> Or do you really want a (slow) multiply by 40 for each line?
>
> In any case you actually *have* lines with 40 bytes in the standard
> hires graphics mode - which can't be said of other popular computers of
> the time *cough* C64 *cough*...
>
> The color generation and its effect of the video memory organization
> (two bytes to form seven pixels, the color bit etc.) is an entirely
> different case but AFAIK completely independent of the memory refresh.
> But you can do a lot with masks and tables, too.
>
> The 8-bit Ataris for example have cleaner video modes (linear video
> memory, "better" color generation via color palette registers) but they
> are not necessarily faster when doing the same type of hires graphics
> and certain things are even beyond them.
>
>
>> I was wondering why, but assuming it saved some IC's, and could until
>> now not see what would make it worthwhile.
>
> I guess even Woz didn't foresee how fast hires graphics would
> eventually be when he made his concept.

Hello,

    It is clear that Apple II's TEXT/LORES and HIRES/DHIRES are not linear 
memory for saving memory reason, but Apple IIgs' super high resolution is 
linear memory in bank $E0/$E1.  Correct?

Bryan Parkoff