Re: rmicdf

Hi Tom,

>       I was just curious how the rmicdf development/debugging was going.
> When does Unidata plan on releasing the software?

I'm the only one working on it, and I took a week of vacation followed
by a week to catch up with other assignments.  This morning I think I
got the problems involving scalar variables solved; at least the dump
method now works on a netCDF file containing a scalar variable.  If
that's holding you up, I can send you a patch file.  If you've found
other problems, it would help to have a small example that demonstrates
the problem.

I also took a detour trying to use the JWS debugger, resulting in
diagnosis of debugger problems rather than solutions to the problem I
was trying to debug.  But we're putting up JWS 1.0 that was released
today, so maybe those problems will be fixed ...

I converted the RMI client/server package to the new (Alpha2) version of
RMI, but more work is required to incorporate other changes discussed
with Joe Sirott before that is available.  I think I'll release the
netCDF package earlier than the RMI client/server package that depends
on it, because the former is further along.

Before releasing the netcdf package, I'd still like to 

 - make sure the interface will work in a future read/write
   implementation;
 - make name changes: the "Ji" prefix should be eliminated
   for something we maintain and other changes are necessary to avoid
   confusion with CDF; and
 - add attributes, instead of treating attributes as another kind of
   variable.

I can only spend about 1/4 time on this right now, so it's not going as
fast as I'd like.

--Russ