We talked a bit more about the Kepler Project.
It's very Java-centric.
We had much discussion about steering
clear of tying ourselves to only IDL or only
ENVI or only XXX.
Next step:
I'm going to look through the KeplerProject source code... working to understand it better. To focus that looking, I'm going to have an eye on how it could be used to monitor and leverage the Shipman/Kremins data, etc.
Afterthoughts:
I think it makes the most sense to model what we want the architecture to do... how we want it to act... before we worry about binding that architecture to a technology: CORBA, XML-RPC, SOAP, ???
For data access, we can go a long way toward making data available in every application under particular OSes if we go for a LarchFSIdea.