From walter.felscher@uni-tuebingen.de Sun Dec 17 07:24:33 1995
Date: Sun, 17 Dec 1995 13:22:40 +0100 (MEZ)
From: Walter Felscher <walter.felscher@uni-tuebingen.de>
To: Lee Dickey <ljdickey@math.uwaterloo.ca>
Subject: Re: Extended precision

MPASPAC

A.

MPASPAC.ATF  will work under each of IBM's APL2/PC , APL2/OS2 , APL2/6000 ,
running under DOS or OS2 on PCs or under AIX on RS6000 work stations.
The file should be loaded with the command ")IN MPASPAC" and can,
if so desired after having made changes, be saved either as an ATF-file
with the command ")OUT newfile" or as a processor dependent APL-file
with the command ")SAVE newfile" .

B.

With a few exceptions, the functions in MPASPAC are variants of functions
coming from

 C.C.Sims: A classlib user's manual.
             Wiley 1985

which is a companion to

 C.C.Sims: Abstract Algebra. A computational approach.
             Wiley 1984 .

In particular, the functions beginning with MPZ...  are directly from
there, as is ZNDET .

C.

For input and output, multiprecision integers are treated as character
vectors of digits to base 10 : C-10-vectors.

For computation, these are transformed into numerical vectors of
numbers which are digits to base G=1E6 : N-1E6-vectors.

Functions transforming input are MPZUNF and MZCNCO .

Functions transforming output are MPZFORM and REEM, REEN .

The C-10-vector functions MZ___ use for computation analogously named
N-1E6-vector functions MG___ .

*

Errors and improvements may be reported to :

Walter Felscher , felscher@maplog.informatik.uni-tuebingen.de