Time for a mobile service approach?

Jon Udell writes in Indexing and searching Outlook email

As I learned this morning, my closing lament -- that the CPython/MAPI and Jython/Lucene halves of this project do not communicate directly -- is somewhat mitigated by the existence of Lupy (1, 2), a Python port of Lucene. But I think the general point still stands. Must every component be rewritten in every language? Let's not go there.

No lets not go there - lets go to a service based approach - if Jon makes this into a service we could all use there is no need for anyone to have to decide among Python and Java.

Nice idea but the issue comes down to the data and how to move it around if I need to send my message store to a service every time I want to do a search the performance will be awful.

Keeping the constraint of I want to use a service and I want good performance - so lets move the mountain. Can the service become mobile and move closer to my application?

Building distributed applications requires balancing the amount of data moved, computation efficiency and granularity of the computational unit. Making common web services mobile and putting intelligence into the network for optimization of distributed applications may solve the problem