Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!sun-barr!newstop!texsun!pollux!attctc!chasm
From: chasm@attctc.Dallas.TX.US (Charles Marslett)
Newsgroups: comp.sys.ibm.pc
Subject: Re: ESDI Compatibility (was Re: ESDI controller/disk recommendations wanted)
Summary: Actually, reformatting is the rule not the exception
Keywords: ESDI disk controllers standards formatting
Message-ID: <10514@attctc.Dallas.TX.US>
Date: 7 Dec 89 14:46:39 GMT
References: <259@cica.cica.indiana.edu> <52@qmsseq.imagen.com> <6468@tekgvs.LABS.TEK.COM>
Distribution: na
Organization: The Unix(R) Connection, Dallas, Texas
Lines: 35

In article <6468@tekgvs.LABS.TEK.COM>, keithe@tekgvs.LABS.TEK.COM (Keith Ericson) writes:
> In article <10869@thorin.cs.unc.edu> davis@cs.unc.edu (Mark Davis) writes:
> <In article <52@qmsseq.imagen.com> pipkins@qmsseq.UUCP (Jeff Pipkins) writes:
> <>...anytime you change ESDI controllers, you have to low-level
> <>format again.
> 
> <Pardon my skeptism, but is this really true?
> 
> In playing with (switching between) the Adaptec 2322B and the WD 1007 ESDI
> controllers we've found that the drive must be reformatted (low level) to get
> it to talk to the new controller.

The data format (byte stream or whatever) is usually (almost always?) portable
between one controller and another, even if they use different controller
chips -- that's because the definition of MFM and RLL define the bit stream on
the media (the stuff going into the controller card), and the others define
the interface equally well.  All but SCSI drives pass a byte (or lower) level
data stream to the controller card, however, and the header information, the
CRC and the ECC information are not standardized.  For example, one or another
of the controller chip sets use 48, 56 or 64 bits of ECC data -- possibly
compounded by the ECC algorithms being different for the same number of bits
(here I'm really in over my head!).

So portability between controllers is pretty much between two of the same model
(or two WD/IBM compatible MFM controllers) or between two SCSI controllers.
(And here, take compatibility with a small grain of salt. ;^)

None of this may apply to the IPI standards (I don't know anything about them).

> kEITHe


Charles Marslett
chasm@attctc.dallas.tx.us
<< corrections gratefully accepped, and perhaps ignored :)) >