Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!samsung!brutus.cs.uiuc.edu!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: <1990Mar12.101903.13038@spectre.ccsf.caltech.edu>
Date: 12 Mar 90 10:19:03 GMT
References: <14075@fs2.NISC.SRI.COM> <14076@fs2.NISC.SRI.COM> <90069.141109BRL102@psuvm.psu.edu> <7400@latcs1.oz.au>
Sender: news@spectre.ccsf.caltech.edu
Organization: California Institute of Technology
Lines: 28

stephens@latcs1.oz.au (Philip J Stephens) writes:

>  This is just the tip of the iceberg in regards to how Locksmith copies disks
>so fast.  It actually pretends that the sector has been stored in a different
>(but similar) format to the 6&2 encoding actually used.  I can't remember
>exactly how it interprets it as it's been ages since I peeked at the raw code.
>However, it does compress it back into 256 bytes (which is why it can copy so
>many tracks at one time), but the data is scrambled.
>   It turns out that this scrambled data can be re-converted back into the
>350 odd disk-bytes _quicker_ than the 6&2 encoding scheme; in fact, it can be
>done on the fly.  Hence Locksmith can write at maximum speed.
>   Why then, don't we use this alternative scheme rather than the 6&2 encoding
>scheme?

Got me. I remember figuring out something similar to it trying to write an
untra fast disk reader, and wondering the same thing... it could be done if
we reformatted all our disks with it but that's a pain in the neck. (new FST!)

Actually, scrambling/unscrambling in memory doesn't hurt as much as waiting for
sectors to come by, so if you read in the whole track fast and then unscrambled
it you'd still get a nice speed improvement. if you did the unscrambling while
waiting for the head to step to the next track then you get even more speed.
Whether you can hit the physical limit of the disk drive I don't know though.

Would make an interesting demo...

Todd Whitesel
toddpw @ tybalt.caltech.edu