Strategy

To expand into network space

ChemSymphony has a good client side data model.
Daylight has extremely powerful server side functionality.
Clients generally have less powerful software/hardware & slower connections.
Minimize the data stored on client & data transfer.
Perform computer-intensive operations on the server.
Also maximize the use of server-side functionality for arbitrary data (originating elsewhere).
Example: Daylight (CLogP, substructure & similarity searching).

Compared to Servers, Clients are usually characterized by less powerful software & hardware and are usually connected by relatively slow network. Thus it is desireable to minimize the amount of data stored on the Client, minimize the amount of data transfer and as much as possible perform any comput-intensive operations on the Server.

We have implemented our system using these principle and also we have attempted to maximize use of server-side processing for arbitrary data, which may not even have originated there.
For example using the functionality & speed of the Daylight engine on its own as well as on the data stored in its databases. For example: CLogP calculations, substructure and similarity searches.