LAPACK Archives

[Lapack] LAPACK/BLAS

And Douglas W Cameron writes:
 - Are there any procedures in place to insure that no one contributes
 - code to LAPACK/BLAS who is not entitled to?

There is a draft policy that requires major contributors
to submit a signed copyright assignment form which also
requires that the contributors attest that the
contributions are clean.  This policy will be used to
handle all outside contributions.  Contributions within
the "core" group are handled by things like employment
contracts and Ph.D. thesis originality requirements.

Also, original authors almost always submit papers
describing major contributions to widely read, peer-
reviewed journals.  The journals have somewhat similar
originality requirements, so the authors are attesting
that they can claim the work that way, too.

For existing code, well, remember that some of it pre
-dates current copyright and patent law.  I'm fairly
sure that much of the rest was produced via Ph.D.
theses...

Also, the current LAPACK is widely used by commercial
products (e.g. MATLAB(tm)).  So it has been checked by
a few other companies' lawyers.

 - IBM is thinking of using LAPACK/BLAS v3, [...]

Um, the IBM contributors may be surprised that they're
not currently using it...

Jason

P.S.  How much has the SCO lawsuit cost IBM so far?  ;)


<Prev in Thread] Current Thread [Next in Thread>


For additional information you may use the LAPACK/ScaLAPACK Forum.
Or one of the mailing lists, or