CHANGES.HAPcryst

0.1.11:

- HAPcryst now calls polymaking with keywords "FACES" and "DIMS" to create
  a resulution of a Bieberbach group. This requires a "next generation"
  version of polymake.
- One minor change that should reduce the number of calls to polymake by one.
------------------

0.1.10:
- dependencies changed: AClib and Polycyclic added

------------------


0.1.9:
- package moved to Evansville

------------------

0.1.8:
- IsFundamentalDomainBieberbachGroup added 
- FundamentalDomainBieberbachGroup now tests for Bieberbachness in the 
   non-pcp case as well.
- FundamentalDomainStandardSpaceGroup now uses the same algorithm for all cases
- HAPcryst now requires polymaking version >=0.7.6
- polymake files generated now know about their data type
- Package moved from Galway to Kaiserslautern


------------------

0.1.7:
- TranslationOnRightFromVector now uses a ShallowCopy of the given vector as 
   the last row of the returned matrix (rather than the vector itself)
- Resolutions now pretend to have length <dim>+1.
- FundamentalDomainBieberbach doesn't test VOLUME any more. This is far too
   expensive as a standard test.
- The "examples" directory is now included in the package archive and a bit 
   more user-friendly
- Test file added
- Some changes in README.HAPcryst
- HAPcryst is now a deposited package
------------------

0.1.6:
Changed requirements:
--------------------
- HAPcryst now depends on GAP version >=4.4.10
- HAPcryst now depends on Hap
- HAPcryst now suggests GAPdoc

New functionality:
-----------------
- HapLargeGroupRep introduced
- experimental framework for HAP datatypes implemented and documented.
   Book name: HAPprog  living in lib/datatypes/doc
- some functions to generate chain complexes from resolutions added
- WhichSideOfHyperplane now has an NC version.


Changes/improvements
--------------------
In FundamentalDomainBieberbach:
-  now evaluates the option ineqThreshold.
   This value triggers convex hull computations if "enough" new inequalities 
   are known. Default is 200.
-  now tests the volume of the initial 
   polyhedron. This is experimental and may be too slow as it involves 
   calculating a triangulation via beneath-beyond.
-  now uses randomness.
   vertices are chosen in a random order to test if they can be mapped into
   the approximation for the fundamental domain.
-  now writes the inequalities for polymake in 
   random order. This seems to improve speed.
-  should now be a little bit faster.
   (Use of WhichSideOfHyperplaneNC,
    Set->List modifications in the main procedure and newInequalities
    the elimination of a unnecessary polymake call)

In FaceLatticAndBoundary:
- doesn't end with a large demand for memory any more.
- the hasse diagram now contains pairs [gen,groupringel] of integer <gen> and 
  group ring element <groupringel> to say that the <gen>th entry of the boundary
  vector is <groupringel>. This is then turned in a vector by the boundary 
  function. So no change in behaviour for the user.

In ResolutionFromFLandBoundary:
 - the function now copies the .elts list of the face lattice,

In other places:
- Replaced some AddSet with Add commands in the orbit-stabilizer procedures.
- TranslationsToBox now returns an iterator rather than a list 
   (->RDS needed for CartesianIterator. This will probably vanish when the
     next GAP version is released).
- ResolutionBieberbach now removes the temporary file it creates.
- The example program now uses FACE_LATTICE rather than HASSE_DIAGRAM
   (output type of HASSE_DIAGRAM has changed in polymaking)
- some functions to generate chain complexes from resolutions added
- some very experimental code for contracting homotopies added (undocumented)
- test file added.
------------------

0.1.5:
- Changed bug in FaceLatticeAndBoundary the output is now HAP compliant again.
  (does not make a difference for `Homology').

- Changed PackageInfo.g to point to the new homepage


------------------

0.1.4:
- Fixed an embarrassing bug in StabilizerOnSetsStandardSpaceGroup

- introduced new methods OrbitPartInFacesStandardSpaceGroup and
  OrbitPartAndRepresentativesInFacesStandardSpaceGroup

- FaceLatticeAndBoundaryBieberbachGroup is now a good deal faster 
  in some hard cases.

- removed unused variable HAPCRYST_TMPDIR. POLYMAKE_DATA_DIR 
  should do the job.

- corrected a few errors in the manual

- README written (thanks to David Joyner).

------------------

0.1.3:
- Changed package dependencies (Cryst >=4.1.5)
- Added description for PointGroupRepresentatives to manual.
------------------

0.1.2:
Changes affecting behaviour of HAPcryst:

- Removed LatticeTransAndPointGroupRepsOnRight.

- Added PointGroupRepresentatives (attribute)

- Fixed a bug in FundamentalDomainBieberbachGroup which could lead to wrong 
  results in the non-standard case.

- Fixed a minor bug in FundamentalDomainAffineCrystGroupOnRight which could lead
  to calculating "fundamental domains" in cases where this is not possible...

- Added StabilizerOnSetsStandardSpaceGroup 

- Added IsFundamentalDomainStandardSpaceGroup 

- All OrbitStabilizer algorithms now return the stabilizer rather than just generators.

- FundamentalDomainBieberbachGroup now checks if a group is Bieberbach or outputs a
  line of information, if no check is performed.


Minor changes:

- Added LICENSE.txt.

- Added "examples" directory. This currently contains a program to 
  calculate colored fundamental domains of 3-dimensional Bieberbach groups.

- Some changes in documentation

------------------

0.1.1:
- First version as a package.------------------