[sc34wg3] TMCL Strawman and Updated Draft of TMCL Requirements and Use Cases

[Martin Bryan]

Mary

Some comments on the TMCL specification draft text:

Re 1.2 "Constrainer A constrainer is a TMQL expression that identifies one
or more topic map data model information items that MUST or MAY exist
(depending on validation type) given a selector." constraints must also be
able to state that an expression MUST NOT exist.

Re 2.1 "This expressive model is called TMCL." it is a bad idea to call the
expressive model by the name of the standard. It should be something like
"TMCL metamodel".

Re 2.2 [ED note 1: Does ConstrainerExpression need to be a collection?] the
answer must be Yes. It should not be necessary to repeat SelectorExpressions
when multiple constraints apply. (There is a good case for forbidding the
same selector expression to occur more than once so that all
ConstrainerExpressions assigned to a particular pattern are defined at a
single point in the schema.)

Re 2.2 "[ConstrainerExpression]: A string that contains a valid TMQL
expression." I am not convinced that TMQL expressions are sufficient to
define constraints. At the very least you need some form of description that
explains why the constraint exists, which can be used as part of an error
report when the constraint fails. You also need some form of unique
identifier to indicate which constraint has failed and, in the case of
constraints which contain OR groups, the part of the constraint that failed.

4 References. "It is unclear why [OWL] Web Ontology Language (OWL) Guide
Version 1.0, W3C Candidate Recommendation. Michael K. Smith, Chris Welty,
Deborah McGuinness, World Wide Web Consortium. 18 August 2003." has been
referenced


Martin Bryan
IS-Thought: Thinkers for the Information Society
29 Oldbury Orchard, Churchdown, Glos. GL3 2PU, UK
Phone: +44 1452 714029 Fax: +44 1452 859991
E-mail: martin@is-thought.co.uk

Thought of the week:
Philosophy is good advice
Seneca, Letters XXXVIII