The SYMBOLICDATA project evolved as a permanent interplay between its two facets: collecting data and extending/improving concepts, design, and tools.
As of today, the SYMBOLICDATA contributors collected more than 1100 sd-records, wrote 40 Perl modules with more than 15000 lines of code, and implemented 22 actions for the standard interface program symbolicdata. The following short alphabetical overview of tables which currently exist may give the reader a feeling about the overall structure of the data that was collected so far.
Collects bibliographical information in BibTeX format, short abstracts, and relational information to the GEO, INTPS, and PROBLEMS tables.
Collects information about the address, author, email, url etc. of the software, and also a short description.
This may be a problem description, a pointer to the origin of the problem, related CAS code, and/or certain key words.
We started first benchmark computations on Groebner bases, using various coefficient domains and monomial orderings. These benchmarks have been (and are) run on the more than 500 INTPS records using 10 versions of different Computer Algebra systems. Other benchmark computations on polynomial systems (like ``solving'', real root isolation, syzygy/resolution computations) are in preparation.
www.SymbolicData.org will soon become the central site of the SYMBOLICDATA project, containing its WWW-pages, and its CVS and FTP repositories. It will be related to the MEDICIS project  that ``can be used by anybody to solve scientific calculations with the tools of computer algebra and symbolic computation. It can, in effect, put at your disposal hardware resources, software and expertise.'' (from their web pages).