Changes in release 1.7.5, 2018-09-10:

 This is a portability release.

 * Fix compilation errors for NTL version >= 9: don't depend on NTL_STD_CXX macro

 * Fix runtime errors for NTL builds with NTL_RANGE_CHECK=on and for NTL version >= 11


Changes in release 1.7.4, 2018-09-03:

 This is a bug fix and portability release.

 * Fix compile errors with clang.

 * Make it an error, not a warning, when cddlib is not found at configure time.

 * Make test suite pass when configured without 4ti2.

 * Relax test for location of gmp.h, needed for Ubuntu


Changes in release 1.7.3:

 * count and integrate commands now correctly return a non-zero error code on failure
	
 * Fixes problems when building on Macs

Changes in release 1.7.2:

 * Maple versions of some LattE commands are now available; see code/maple/README.
	
 * Fixes problems when building on Macs


Changes in release 1.7.1:

 * Fixes problems when building on Macs
	
 * Improved the top-ehrhart-knapsack command


Changes in release 1.7:

 * New program for computing the highest coefficients of Ehrhart
   quasipolynomials for knapsack polytopes.  


Changes in release 1.6:

 * New program for computing top Ehrhart coefficients (see manual). 

 * Fix problems with whitespace in filenames or installation directory names.


Changes since release 1.4:

 * New program for computing volumes and integrating polynomials or powers of linear forms over polytopes.
 
 * Computing volume by triangulating 
 		
        integrate --triangulate --valuation=volume

 * Computing volume by cone decomposition
         
        integrate --cone-decompose --valuation=volume
        
 * Computing integrals by triangulating 
         
        integrate --triangulate --valuation=integrate [--monomials=file | --linear-forms=file]

 * Computing integrals by cone decomposition
         
        integrate --cone-decompose --valuation=integrate [--monomials=file | --linear-forms=file]
        
        
Changes since the official release 1.2:

 * Use GNU Autoconf and Automake.

 * Portability fixes.

 * Code clean-up.

 * Some performance improvements.

 * New exponential substitution 

        count --exp

 * New primal irrational decomposition 

        count --irr

 * New non-unimodular enumeration

        count --maxdet=N

 * New all-primal irrational decomposition 

        count --all-primal

 * New method for computing Ehrhart polynomials of integral polytopes 
   in the primal space 

        count --all-primal --ehrhart-polynomial

 * The homogenized method now allows V-representation input 

        count homog vrep

 * The functionality of `ehrhart' has been merged into `count':
 
	count --ehrhart-series FILENAME

	      (replaces: ehrhart FILENAME)

	count --simplified-ehrhart-series FILENAME

	      (replaces: ehrhart simplify FILENAME)

	count --ehrhart-taylor=N FILENAME

	      (replaces: ehrhart N FILENAME)

   (The `ehrhart' program is still available, but it does not accept
   the new command-line options of `count'.)

* The 4ti2 program (http://www.4ti2.de) can be used instead of cddlib and CDD+
  to compute the vertex cones of polytopes and triangulations and duals of cones.
  In many cases, 4ti2 is faster.

        ./configure --with-4ti2=PATH-TO-4ti2

	count --triangulation=4ti2 --dualization=4ti2 --compute-vertex-cones=4ti2

* New command-line option --redundancy-check={none,cddlib,full-cddlib}.
  
  - "full-cddlib" (the default) uses cddlib to compute an irredundant
    system of linear equations and inequalities describing the
    polyhedron.  This corresponds to the traditional LattE behavior;
    it can be expensive.

  - "cddlib" (used to be the default in the 1.2+mk-0.9.x series) uses
    cddlib to compute some implicit linearities only; it often fails
    but is faster than full-cddlib.

  - "none" does nothing, the input description of the polytope should
    be irredundant.

* New programs, `latte2ine' and `latte2ext', are built that perform
  the trivial conversion from LattE-style input files to CDD-style
  input files.

  	latte2ine < FILENAME.latte > FILENAME.ine
	latte2ext < FILENAME.vrep.latte > FILENAME.ext

* New command-line option --multivariate-generating-function.

  It allows to compute the multivariate generating function of a
  rational polyhedron.  For unbounded polyhedra, one needs to combine
  it with --compute-vertex-cones=4ti2, since the traditional method in
  LattE (using CDD+) currently refuses to handle unbounded polyhedra.

  	count --compute-vertex-cones=4ti2 --multivariate-generating-function FILENAME

  writes the multivariate generating function (in Maple notation) to "FILENAME.rat".

