[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 970121: Fusin fusion possible collaboration

>To: address@hidden
>cc: address@hidden
>From: Curt Bolton <address@hidden>
>Subject: possible collaboration
>Organization: Office of Fusin Energy Sciences, DOE
>Keywords: 199701211650.JAA13452

Hi Curt,

> THe fusion program is interested in netCDF, but have found that it 
> does not work well with object oriented languages adn data 
> structures. Are you interested in working together to develop the 
> extensions necessary to make it fully compatible with F90 and other 
> object oriented languages.

Yes, we're certainly interested in those goals.

We have developed a C++ interface for netCDF which is already in use by
several projects and is part of the current netCDF 2.4.3 release; for
more information, see

   http://www.unidata.ucar.edu/packages/netcdf/cxxdoc_toc.html

We have developed another experimental C++ implementation of netCDF that
is not a veneer on top of the C interface, to see how it would work to
base the C and FORTRAN implementations on top of a C++ implementation.
We're still evaluating this.

As part of netCDF version 3 (a pre-release of which is now available),
we have developed new C and FORTRAN type-safe interfaces; the new
FORTRAN interface works with FORTRAN 90 (we have an interface block for
it).  I am currently working on a User's Guide for that interface, but
the ideas behind the new type-safe interfaces are described in the
netCDF-3 prerelease document at

   http://www.unidata.ucar.edu/packages/netcdf/3.1a.html   

We have a Java read-only interface for netCDF based on work by Joe
Sirott when he was with the University of Washington.  As I get time,
I'm preparing the Java read-only implementation for release.

We also have plans for an object-oriented "Geoscientific Information
Framework" that is closely related to our efforts in this direction.
For details, see the *draft* document

  http://www.unidata.ucar.edu/staff/dave/in_progress/Proposal2003/GeoF.html

We have another collaboration underway with NCSA to evaluate whether we
can develop a common format for the next versions of netCDF and HDF,
taking into account planned object-oriented extensions.

So from the above you can see that any future work we do on
object-oriented extensions to netCDF would have to be consistent with
the efforts and collaborations we already have underway and constrained
by the limited resources we have for additional development.

I'd be interested in what you have in mind for object-oriented
extensions.

--Russ

_____________________________________________________________________

Russ Rew                                         UCAR Unidata Program
address@hidden                     http://www.unidata.ucar.edu