At 03:17 AM 12/1/2007, Jeff Davis wrote:
>The impedance mismatch has more to do with the fact that the meaning of
>an application's internal data structures changes frequently (through
>revisions of the code), while data in a database needs to be consistent
>across long periods of time. So, a well-designed database will hold
>facts that have meaning in the real world and from which inferences can
>be made. Mapping application data structures (which contain context-
>sensitive information and implementation artifacts) to real-world facts
>is the impedance mismatch.
The people who try to make a database that maps so well with the
objects in a single particular program are solving a very different
problem from those of us who use a database partly as a "lingua
franca" (or "vehicular language") for many different programs and people.
The "impedance" then is unavoidable. It's not going to be easy to
change a hundred other programs anyway - probably some unknown (till
they inconveniently stop working because someone decided to "match
the impedances" with some pet program ;) ).
But anyway, I guess postgresql's "table inheritance" thing isn't
broken then just misunderstood...
Link.
One man's impedance mismatch is another man's layer of abstraction or
"comms protocol" :).
