[gmx-developers] future of installed headers and libraries

[Peter Kasson]

I would also advocate that we keep a public API (and can help maintain it
if needed--let me know).  I think a Python API will be super-useful and
would become the most common means of scripting Gromacs.  Nonetheless,
having a C/C++ API, even if not a super-slick one, is quite helpful as
Christoph mentions.  I've used it in a number of pieces of un-released code
(things not in the Gromacs source tree that build with Gromacs).  One could
argue that un-released code could hack in hooks, but that's kind of ugly.

What are the major reasons to remove the API?  Ugliness, install bloat, the
fact that every feature has non-zero maintenance effort?  Other things I'm
missing?

Thanks,
--Peter

----------------------------------------------------------------------
Peter Kasson, MD, PhD
Assistant Professor
Departments of Molecular Physiology and Biological Physics
and of Biomedical Engineering
University of Virginia
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-developers/attachments/20140708/dba4258e/attachment.html>