changes to be made to 5 Mar version of Semantics

Changes for FPWD

- my affiliation is Nuance Communcations nuance.com

- remove "in V" from simple interpretation

- value space of datatypes - change to conform with Concepts
  - remove sentence about the value space

- well-typed literal - can literals from unrecognized datatypes be
well-typed?
  A literal whose datatype is recognized but is not ill-typed is well-typed.


Changes to be made (not necessarily for FPWD):

- blank nodes - add at the end of the second paragraph
  Two graphs not in the same blank node scope do not share any blank nodes.

- instance lemma - add in definition of instance
- anonymity lemma - add in definition of lean graph and proper instance

* datatype map - not useful here
  , where D is a set of IRIs that constitute the recognized datatypes.
  IRIs listed in Concepts Section 5 must be interpreted as described there,
  and the IRI ...
  When other datatypes are used, the interpretation of the recognized
  datatype IRIs must be specified unambiguously.

- datatype extensions
  - need to add something about if the datatypes are interpreted the same
way

- status of rdf:langString

- I think that the RDFS semantic conditions should be made  less redundant
  - remove IL(E) is in LV, for every well-typed literal E
  * currently unrecognized literals are well-typed, which makes the above
    condition non-redundant AND a change to the semantics AND destroys
    semantic extensibility
  - status of rdf:langString
  - remove "x and y are in IP" for rdfs:subPropertyOf
  - remove "x and y are in IC" for rdfs:subClassOf

* value space of datatypes
  - remove "ICEXT(I(aaa)) = { ... }"


Changes to be made to Concepts

- definition of rdf:langString as a datatype with a trivial L2V

- remove datatype maps - not used in semantics