Architecture suggestions?

[Jonathan Tillack]

Hello All,

I am looking for some guidance on the best way to set up a database... but first the background.

I have created an electrical engineering component database for my company. The database contains fields like component parameters, data sheets, links to simulation files, links to PCB footprints etc. The database also has numerous fields that are used for "filemaker internal" functions like searching, scripting, and container fields for icons etc.

My EE program (altium designer)uses ODBC to access the database, but unfortunatily, it isn't very smart about it. If you access a component in the database, it brings over all of that component's fields (including globals), even if it doesn't use them (in other words, rather than looking up the fields that have mapped to component properties, it brings over all the fields and then throws away ones it doesn't need)

The overhead in network traffic is making things VERY slow - especially since the some of the un-needed fields are containers for PDFs.

The number of fields that the CAD program actually uses is small (20 or so), and the fields are text with a max size of 128 characters.

What would be the best way to create a database that only mirrors these important fields from the main database? I invision accessing this "mirror" database over ODBC so my not-so-smart CAD program doesn't pull over everything in the main component database.

Any ideas on how I should set this up?

Your help is greatly appreciated!