- From: Morris Matsa <mmatsa@us.ibm.com>
- Date: Tue, 19 Dec 2000 18:27:39 -0500
- To: www-xml-schema-comments@w3.org
The term "finite-length" is used many times in part 2 of the spec, but never defined. It seems implied that "finite-length" means "of a length which is any non-negative integer." If you look up the mathematical definition of "finite" (http://www.dictionary.com/cgi-bin/dict.pl?term=finite) you find many options which seem to indicate ("1. Having a positive or negative numerical value; not zero.") that zero is not included. This would imply that lists, strings, decimals, binary values, etc. are not allowed to be empty. Furthermore, some types (e.g. IDREFS http://www.w3.org/TR/xmlschema-2/#IDREFS) specify that the value space is a finite-length sequence of elements, but the lexical space has no such constraint (being a "set of whitespace separated tokens"). It would seem that this term should be used either for both value space and lexical space, or neither, in the case of a list type. My question is what the actual meaning in the spec is for "finite-length" and I suggest that it is defined in the spec. A related question: Lists are a "finite-length" sequence of values. (2.5.1 http://www.w3.org/TR/xmlschema-2/#atomic-vs-list and 3.1 http://www.w3.org/TR/xmlschema-2/#namespaces), alternatively their value space is composed of "finite" sequences of values (5.1.2 http://www.w3.org/TR/xmlschema-2/#derivation-by-list). This implies that there is no difference between a "finite-length sequence" and a "finite sequence". Am I correct? If so, why are they worded differently?
Received on Tuesday, 19 December 2000 18:28:40 UTC