[BioPython] Lisp in bioinformatics(!)

[Antoine van Gelder]

Johann Visagie wrote:
> 
> Cayte on 2000-08-21 (Mon) at 14:13:54 -0400:
> >
> >   As an ultimate goal, I think we should aim for integrated package, with each
> > language doing what it does best and not a patchwork quilt.
> 
> A libbio implemented in C++, with stubs to create wrappers in HLLs?
> 
> I'm serious...

One will tend to run into semantics issues with such an approach, what
expresses naturally in C++ becomes a kludge in Python and what flows
with zen-like elegance in python will make g++ explode.

Which is not to say that compute intensive code cannot be written in C,
but I believe that this is already taking place ?

I think that, at the end of the day, it is less important to have one
world, one API as it is critical for specific _services_, implemented in
languages that are natural for the particular type of service, be
exposed and made accessible to other languages and API's.

Access of this type can be easily accomplished using either Corba, or
SWIG without having to sacrifice the expressiveness of the
implementation to a small subset of language common across all target
platforms.

If one were to try utilising a common C++ libbio underlying everything
it would not be long before you're going to start getting confused
between implementation semantics and functional semantics - triggering
flame wars resulting in the total eradication of all intelligent life in
the universe.

 - a