[SciPy-dev] Pyd

Kirk McDonald kirklin.mcdonald@gmail....
Thu Apr 24 20:02:22 CDT 2008

Neal Becker wrote:
> Is pyd still actively developed?  Seems to be rather quiet for a while.

I haven't worked on it in months, it is true. I ran into certain
issues with D and became discouraged (Pyd pushes D to its
metaprogramming limits, and D's compile-time introspection is still
slightly less than perfect). After that, I started working on some
other projects. However, I don't consider the project dead, and I've
had a number people request better interoperability with NumPy. It's
certainly a feature I would like to add.

However, I don't think I'll be getting back to work on Pyd in the
immediate future. I have a new job, and I am moving, so things may
take a while to settle down.

In the future, I may be reworking Pyd's API (again) into a more
SWIG-like model. (That is, Pyd would become a utility which generates
code, rather than a library which generates code entirely with D's
templates.) This would neatly sidestep many of the current weaknesses
of Pyd, and it might even motivate me to do more serious work on it.

--Kirk McDonald

More information about the Scipy-dev mailing list