Date   : Fri, 21 Nov 1986 14:03:31 EST
From   : "John S. Fisher" <FISHER%RPICICGE.BITNET@wiscvm.ARPA>
Subject: Request for help.

Does anyone out there know the details of the compression algorithm(s) used
by the CRUNCH/UNCR programs?  I would like to incorporate UNCR functions
into my IBM (mainframe) utility, but have been unable to decode the crunched
files.  The older versions of the program pair claimed to used LZW, but
its a completely different variation from ARC.  The latest version of CRUNCH
uses GEL(?) compression.

Perhaps someone can send me the source for CRUNCH/UNCR formerly distributed
with version 1 I believe (and no longer distributed with version 2, sigh.)