Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!apple!motcsd!starnet!sschaem
From: sschaem@starnet.uucp (Stephan Schaem)
Newsgroups: comp.sys.amiga.programmer
Subject: Re: Bizarre Fantasy: RLL Flopies?
Keywords: MFM, RLL, Floppies, Blitter, Clueless, Fantasy, etc.
Message-ID: <1991Apr3.165223.12953@starnet.uucp>
Date: 3 Apr 91 16:52:23 GMT
References: <1991Apr2.061633.17980@kessner.denver.co.us> <1034@cbmger.UUCP>
Organization: Starnet-Public Access UNIX-Los Altos,CA 415-949-3133, login:info
Lines: 32


 I wont say mutch, but with special coding you can get 2 bit out of 3.
 To be realable you need 3 0 bit set. resulting in all the 89 sifted.
 code the 2 bit with 3 only when the previous bit is a 1.
 To get the most out of it, you need to check you data first to see if
 you need to code 1->0 or  1->1.
 This should be reiable enouhgt for some use but slow coding.
 But you can get more out of it with tables, with innertracks and outer
 tracks 1/0 bit max.You will find that you need to change your table
 because you can write more or less of both.
 I already try the above, work but is not reliable.And you will find
 that you need multy read to get your data right.
 Also I dont know done that already, but preparing data for MFM.(a
 compactor creating MFM or MFM+:-)
 The only problem with 2 bit coded with 3 bit is that 4489 is not valid
 anymore.

 So by using 3 0 bit you have a 50% waste only on 50% or more instead of
 100%.(I'm not think to clearly right now...)
 That mean Variable track size ALSO:-) But around 250K+ per disk.

 And dont forget to check you evaluation on track 0 and 79...

 Another thing you could do is write data to each track, every
 posibilities.And create table from that.
 A fun thing to do also is have a counter on the posibilities and than
 play with the motor speed (can do mutch else:-(.

 If people played with aplle II drives you will know that it depand alot
 on the drive quality and media.

							Stephan.