[tei-council] solving the Birnbaum Biznai

[Sebastian Rahtz]

Two issues have been raised in regard to this, and we need to take
a strategic decision.

*Firstly*, do we want to change the basic idea of what a model class is,
to allow
for the notion that it may encompass such things as a,b,c  or a?,b?,c?
or a*,b*,c*
or a+,b+,c+ as well as the current a|b|c? If we accept the idea, it
allows us
to deal with the requests of msdescription using the class system. If we
do not
allow it, we must find another way.

So if we accept that msdesc is currently broke, we have a choice:

 1) uses classes, and introduce very relaxed content models which permit
      elements to repeat in unwanted ways
 2) use classes, with the proposed new meaning of what a "class" is
 3) introduce and implement module dependencies, and accept that
     it will be harder to guarantee schemas which don't have dangling links

Each of these routes is possible. The simplest is 1), which requires no
new work.  I have implemented 2), but only partly (see below). For 3),
no work has been done, but it is not that hard to make Roma follow
"suggests"
links and tick an extra box by default. However, it will require serious
work to look at guarenteed removal of dangling links.

Personally, I think that route 2) is most in line with where the TEI
ODD system has been going. But YMMV.

*Secondly*, if we do take route 2), it has serious implications
for the non-ODD extension mechanism, ie using schema/dtd
module fragments and combining them in a DTD subset or
hand-written RelaxNG schema. Lou and I had a long (and probably
incoherent) discussion about this at Kansai airport and I think
we came up with solutions. But it raised the question as to
whether we definely want to carry on with this whole
method of working.

Anyone got thoughts in this?

-- 
Sebastian Rahtz      

Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431

OSS Watch: JISC Open Source Advisory Service
http://www.oss-watch.ac.uk