|
I am a bit disappointed with this article. A reader of just this part could easily come away with the impression that DAML is just another web schema language that happens to be based on RDF[S]. As someone unfamiliar with DAML, I was hoping to see a discussion of those features that set it apart from languages for information modeling. Hopefully, the next two parts of this series of articles will focus on this.
There were a few other questions and comments that I had about Part I.
1) The discussion of Unique properties was confusing. This would have been better left to a time when the cardinality could have been constrained to one-to-one, so that productNumber could have been an identifying property for the class Product.
2) The issue of extensible enumerations has come up again and again in different contexts. I think many would disagree with the assertion that these are rarely desirable.
3) Where are the restricted types which were defined at the top of the second page, reused within the DAML below?
4) Can restricted type definitions such as these use compound constraints (something like a minExclusive value AND maxInclusive value constraint)?
5) Doesn't the definition of the ObjectProperty "usedFor" have its domain and range switched? Shouldn't it be a Product is usedFor an Activity?
Evan K. Wallace
Manufacturing Systems Integration Division
NIST
ewallace@nist.gov
|
Previous Message
