Path: utzoo!attcan!uunet!zaphod.mps.ohio-state.edu!ncar!cruff
From: cruff@ncar.ucar.edu (Craig Ruff)
Newsgroups: comp.sys.nsc.32k
Subject: Re: Problems working with SCSI
Message-ID: <9472@ncar.ucar.edu>
Date: 10 Dec 90 19:11:11 GMT
References: <9012100048.24380@otto.bf.rmit.oz.au> <10427@ucrmath.ucr.edu>
Reply-To: cruff@handies.UCAR.EDU (Craig Ruff)
Organization: Scientific Computing Division/NCAR, Boulder CO
Lines: 19

In article <10427@ucrmath.ucr.edu> rhyde@ucrmath.ucr.edu (randy hyde) writes:
>The only problem I have with working with SCSI is
>that you've got to design a whole 'nother computer
>system to do the interface.  Considerably more work
>than adding a latch to obtain a parallel port.

Well, maybe not.  I read an article in Circuit Cellar, Ink (I think)
recently that showed how one might get away with less.  Basically, you
fake enough of the interface so that all real SCSI devices know the
bus is busy and stay off while you access your pseudo-SCSI device.
Then, you could use the SCSI bus however you liked (within some
restrictions).  As long as BSY is asserted, you can use the other
control lines to do things like register select, read/write, etc.
I can check for the exact article and issue if anyone is interested
in the details.
-- 
Craig Ruff      	NCAR			cruff@ncar.ucar.edu
(303) 497-1211  	P.O. Box 3000
			Boulder, CO  80307