Path: utzoo!attcan!uunet!mcsun!unido!balu!tilmann@cosmo.UUCP
From: tilmann@cosmo.UUCP (Tilmann Reh)
Newsgroups: comp.os.cpm
Subject: Re: floppy drive hardware
Message-ID: <5715@balu.UUCP>
Date: 26 May 90 19:37:08 GMT
References: <XX00011533@pallio.UUCP>
Sender: news@balu.UUCP
Reply-To: tilmann@cosmo.UUCP (Tilmann Reh)
Organization: CosmoNet, D-3000 Hannover 1, FRG
Lines: 36

dg@pallio.UUCP (David Goodenough) writes:
> So does the 1793. At least the 1793 in my Televideo 803 seems to talk quite
> happily to all four drives attached.

Of course it does. The hardware developers had to do by extra hardware what
the chip didn't support. As a result, it is impossible to step four drives
simultanously, while always watching all four 'READY' lines (just an example).

> I've worked with both the 1793, and the 765, and the
> software overhead for the 765 is way higher than that for the 1793.

Unless you try to achieve the overall comfort and performance of the 765,
I agree... (Just think of the 'READY changed' interrupt.)

> Fooling with the bits _DOES_ have it's advantages:

And it's disadvantages. All the troubles with different disk formats are
caused by people who wanted to create 'their own format', incompatible to
everything else (or to get slight features, thereby nevertheless becoming
incompatible). And now you try to argue for a format with different sector
sizes on one track. I strongly recommend that NOONE should ever use such a
format. This method, together with irregular sector (or track!) numbering
should be forbidden anyway. Not for the 765, but for some basic rules.

High disk capacities, OK. But only the legal way, s'il vous plait.

> I'll agree with that when I see it put 440K on a DS DD floppy.

Independent of the previous: please show me a member of the x79x family
with the whole floppy interface (as mentioned earlier) on a SINGLE CHIP !
Just fits in the gap between CPU and FDD. No glue logic, no clocks, no PLL,
even no drivers!

(You see, I'm hardware designer)

Tilmann Reh