| AArch64 | |
| ppc64le | |
| s390x | |
| x86-64 |
- Fix "env-script-interpreter" rpmlint warning
- Clean spec file.
- Build and package python3 bindings. - Switch to cmake build, autoconf will be removed in the next version. * Add 0001-Fix-CMakeLists.txt.patch
- Update to version 1.14.1
* Version number increment just to release pypi package.
- Update to version 1.14
* Improvement against cyclic point change over applying
standardizaion repeatedly.
- Update to version 1.13
* Fix Issue#57 of SeekPath: Standardizations of choices of a,b,c
axes for Cmma (67), Ccca(68), Ibca(73), Imma(74) were fixed.
For monoclinic systems, lengths of non-unique axes are now
checked by an exhaustive way.
* Issue#79: Treatment of ir-kpoint search for special cases with
three fold rotation (e.g., hexagonal or primitive FCC).
- Update to version 1.12.2
* Minor release for memory leak fix (PR#71).
- Update to version 1.12.1
* Bug fix of PR #70. #include <stddef.h> is inserted. In
principle, this line is necessary to compile.
- Update to version 1.12.2
* After this minor version, addition of new features will be
stopped until version 2.0 will be released.
* Experimental support of site symmetry symbols.
* Comparisons of basis vector lengths are more carefully done
for letting them work for different computer architectures to
treat very small numerical different that is close to the
machine precision.
* Similarly, MOD-Z operation (bring values x inside the rage of
0<= x < 1) is now done considering the difference of machine
precision.
* To enable dense k-point mesh sampling, integer types of
variables that represent grid point indices were changed from
int to size_t.
- Update to version 1.11.2
* Bug fix. Again, origin shift was still not correctly computed
when the standardized basis vectors are rotated in this method
to find most similar choice of basis vectors to the idealized
ones.
- Update to version 1.11.1
* Bug fix. Origin shift was not correctly computed when the
standardized basis vectors are rotated in this method to find
most similar choice of basis vectors to the idealized ones.
This bug affected to 'origin_shift' and 'std_positions' in
spglib dataset and the functions 'spg_standardize_cell',
'spg_find_primitive', and 'spg_refine_cell'.
- Update to version 1.11.0
* This looks for a similar choice of starndardized basis vectors
to that of idealized starndardized basis vectors among those
generated by proper symmetry operations.
* This computes a rigid rotation matrix introduced
in the idealization step. The rotation is made to agree a-axes
and a-b planes between systems before and after idealization.
- Update to version 1.10.4
* Collection of minor fixes.
A text on spglib algorithm is prepared for this version. This
text will be available soon at arXiv. To be consistent between
text and code, the code refactoring was made.
- Update to version 1.10.3
* New algorithm to search translation part of space group
operation is implmented by Michael Lamparski to improving the
search speed especially for large unit cell systems. For this,
new files 'overlap.*' are added and so the version number is
incremented.
- Update to version 1.10.2
* Automake didn't work due to the lack of determination.{c,h}
lines. This was fixed.
- Update to version 1.10.1
* A large refactoring was made to replace the outer-most
iteration loop of symmetry search algorithm to a new file
'determination.c'.
* These routines had accessed directly to a function close to
spacegroup.c. Now these are replaced by passing through
'get_dataset'.
- Update to version 1.10.0
* 'mapping_to_primitive' and 'std_mapping_to_primitive' in the
spglib dataset are put available on the spglib
document. 'spg_get_hall_number_from_symmetry' function
('get_hall_number_from_symmetry' method for python) is also
- Update to version 1.9.10
* Pure translation search is made faster using a property of
group, which is significantly effective for supercells with
large multiplicity.
- Initial package based on 1.9.9