Currently, users have to make educated guesses regarding compression buffer sizes. This information should be provided automatically by the library. Add support for 64-bit integers