Basic building blocks like the BLAS, the BLACS and the PBLAS have been made publically available. At the time this paper was written, the current version of the PBLAS was being extended by removing alignment assumptions made on the operands. Moreover, the PBLAS package is being internally restructured to facilitate its maintenance and reinforce its robustness. Concurrently, many of the LAPACK functions missing in ScaLAPACK are being assembled and integrated. These include condition estimation, iterative refinement of linear solutions and linear least square solvers. We are planning improved versions of the symmetric eigenvalue routine. SVD and nonsymmetric eigenvalue routines are also in preparation. More elaborate testing and timing programs are being developed to ensure the robustness and the efficiency of the library. Finally, banded, general sparse, and out-of-core prototype routines are being investigated.