BibAuthorID: new major release
- Tortoise: The clustering is separated from the matrix creation which results into huge scheduler improvement.
- Tortoise: The matrix cache is no longer stored on the database but on the file system. Save and load methods are much faster and reliable.
- Rabbit: The algorithm can now optionally store parts of the database in the memory for a performance boost.
- Rabbit: INSPIRE ids are included in the decisions.
- A lot of interface improvements and fixes.
Co-authored-by: Samuele Carli <samuele.carli@cern.ch>