Path: utzoo!attcan!uunet!jarthur!spectre.ccsf.caltech.edu!tybalt.caltech.edu!toddpw
From: toddpw@tybalt.caltech.edu (Todd P. Whitesel)
Newsgroups: comp.sys.apple2
Subject: Re: Fast reading of floppies...
Message-ID: <1990Mar15.142012.15985@spectre.ccsf.caltech.edu>
Date: 15 Mar 90 14:20:12 GMT
Sender: news@spectre.ccsf.caltech.edu
Organization: California Institute of Technology
Lines: 32

cwilson@NISC.SRI.COM (Chan Wilson) writes:

>Say, is the unidisk faster than the appledisk in reading things, or is
>just my overworked imagination?  

Its your imagination. The UniDisk has its own controller (a 2 mhz 65c02)
and uses the SmartPort packet protocol to communicate with the host. The
packet protocol is slow because

	(a) it works at disk ][ speed which is half as fast as raw 3.5
	(b) it has to read the disk block and then transmit it through
		the port which takes even more time

so you literally cannot get respectable speeds from a unidisk. 4:1
interleaves are optimal, and that is pretty sad...

Why didn't Apple

	(a) make smartport as fast as the 3.5 instead?
	(b) give the unidisk controller enough RAM to do 1:1 interleave
		track reads and caching? (come on, 16K (2 6264's) is cheap!)

Because they haven't bothered to improve anything once it is out the door...
EXCEPT the 1 meg GS. The Unidisk could have become the 3.5 of choice but
no... it is still an inefficient double-buffering drive controller. If I could
figure out how to modify one and make it killer for cheap, I'd make gobs of
bux selling upgrade kits.

except I shouldn't be doing that, Apple should...

Todd Whitesel
toddpw @ tybalt.caltech.edu