Monday, September 17, 2007

The $100MM Canonical Model

A co-worker of mine in Indianapolis, IN David Christiansen has a good post with The $100,000,000 Canonical Model.

Lest we forget, there are no sliver bullets.

There are two extremes in my mind when it comes to data formats and integration architecture:

  1. Big Ball of Mud
  2. A canonical model / data exchange architecture will solve everything
The pragmatic approach, IMHO, is to start small and build these assets over time via projects. The wrong approach is to go off for months/years to create it. Clearly, an asset like this needs to be as simple as possible, embrace change, be accessible, and based on Atom. (ok the last bit was bias).

No comments: