----------------------------
* First major release
* Python wrappers for three implementations ('brute', 'kd_tree', 'buffer_kd_tree')
* Performance improvements for both kd-tree based implementations
* Multi-OpenCL-Device support for 'buffer_kd_tree' implementation
* Large-scale construction for 'buffer_kd_tree' implementation
* Multi-OpenCL-Device support for query phase (queries are processed in chunks)
* Added Sphinx documentation