[Bug 5079] Type tables and substitution groups

http://www.w3.org/Bugs/Public/show_bug.cgi?id=5079


cmsmcq@w3.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED




------- Comment #1 from cmsmcq@w3.org  2007-12-27 00:45 -------
Thank you for the comment.

There seem to be two possible answers to your question.  The first is:  no,
there is no problem.  At least, there is no logical contradiction here.

The rules for substitutability require a particular relation between the 
declared type of the substitution-group head and the declared types of the 
members of its substitution group.  It is true that particular instances of the
head and particular instances of the substitution-group members may have 
different types -- in 1.0 this is possible by means of xsi:type, and in 1.1
it is possible both by means of xsi:type and by means of conditional type
assignment.  The guarantee that schema validity gives relates to 
the declared type; this has consequences for the governing types of the
element instances involved, but it does not have the consequence that there
is any substitutability relation between the governing types of different
element instances which match the same particle, only that there is a
substitutability relation between the governing types and the declared type
given in the particle.

A second answer would just turn the question around:  the schema spec makes
a given guarantee.  Is that guarantee sufficient for users' needs (in
particular, yours)?  Or is it a problem?  If it's not a problem for users
of XSDL, then it's not a problem.  If it's a problem for a sufficiently
large number of them, then it may be a problem.  But that is for you to say,
not for the WG.

Received on Thursday, 27 December 2007 00:45:49 UTC