[
Lists Home |
Date Index |
Thread Index
]
> > schema documents using key, keyref and/or unique: 53
> > uses of unique: 105
> > uses of key: 179
> > uses of keyref: 229
> Are you sure that these schemas are from real world i.e. these are
> actually used in the application.
A survey of Henry's data and repositories such as XML.org will provide a
snapshot of current adoption, but not as a predictor of future use. It's too
soon to predict what schema use patterns will emerge over the long term.
If you look at language technologies such as Java and SQL, there's typically a
lag between introduction of a feature and a significant level of adoption.
For example, we heard about EJB container-managed persistence long before we saw
significant availability and adoption.
There was a significant lag between standardization of the SQL core language and
refinements such as integrity constraints, and an even longer cycle for
widespread adoption of those features. The first SQL standard arrived in 1986,
but it was three more years before unique and referential constraints became a
part of the standard. Add another six years for other constraints (min, max,
ranges) and even more years for features such as identity columns.
|