This is version 6.0 of the Imperial College Exact Real Arithmetic
Library. It provides support for lazy real and lazy boolean types
callable from C plus a suite of primitive functions on reals, predicates
and boolean operators.  See the separate copyright file for conditions of
use and distribution.

To install, you must have already installed the Gnu Multiple Precision
Arithemtic package (GMP) Version 3.0.1. Edit the Makefile in this
directory and change:
  REALDIR to point to this directory and
  GMPDIR to the root directory of the GMP tree.
Then type make. Further details and an explanation
of the compile flags can be found in doc/manual/manual.{tex,dvi}.

The library makes trivial use of the utility strsep(3). This may not
be available on some platforms but is easily obtained from the web.
For example, the GNU version is available from:
http://www.gnu.org/software/libc/libc.html
Install strsep.c in the directory base and add the target to the Makefile.

The library has compiled successfully under Linux and
SunOS 5.5.1 (Solaris 2.5.1?)

Lindsay Errington, June, 2000
lindsay@kestrel.edu

--------

The library was written by Lindsay Errington with contributions from Marko
Krznaric and Reinhold Heckmann. It is based on theory developed by Abbas
Edalat, Martin Escardo, Reinhold Heckmann, Peter Potts, Philipp Sünderhauf
and Lindsay Errington (lazy booleans).

--------

When viewing the source files, set tabstops to 4.
The directory has the following structure.

base - the low-level routines used in the representation of the reals
  including those for lfts and booleans.

math-lib - a collection of analytic functions.

doc - see manual/manual.tex for a description of the types and
  functions provided by the library.

icons - used when the tool is connected to daVinci

tests 

Makefile

-------

Known bugs:

sqrt_R diverges for 0. This is because sqrt is not continuous at
0. Various options are being considered for this. We may choose to make
sqrt continuous or introduce complex numbers.