Defining Disk Image Dump Standard

From: allisonp_at_world.std.com <(allisonp_at_world.std.com)>
Date: Fri Jun 2 07:26:29 2000

> Allison wrote:
> > Cant speak for the 6502, but NS* did it with minimal hardware a few years
> > earlier and it was a respected design. Not a whole lot of ttl on the board
> > either considering it was S100, Boot prom and FDC.
>
> Yes. But the part that makes Woz's design interesting is that he did
> it *without* an FDC, and it only ended up as seven TTL chips *total*
> (eight if you include the boot PROM).

Well the NS* controller does NOT use an FDC. IF you kick out the bus
interface (apple for the most part has none) and the fact that apple used
a tweeked disk drive too the NS* controller is really simple with the core
being less than 8 or so chips. Of course back in '77 current ttl s100
interfaces usually added 5-7 chips to any IO design.

> With an FDC chip, there's certainly a range of design styles available;
> I've seen simple and reasonably elegant designs, and gross disgusting
> kludges. But the most elegant FDC-based design is still not even in
> the same class as Woz's design.

I don't know, I"d done a few 765 designs (especially 37c65) that were low
on parts and programatically one heck of a lot lighter on the cpu and memory.
Though the best one was my first as that was only 11 chips for the whole
S100 board and did all FDC standards for 8", 5.25" and 3.5" (excluding off
data rate 1.2mb). Stll use that one.

> Note that I'm not claiming that Woz's method would be the right thing
> to do today. But at the time he did it, it was awesome.

Yep, it was cheap.

Allison
Received on Fri Jun 02 2000 - 07:26:29 BST