[tei-council] customization in P5

[Sebastian Rahtz]

I don't agree with Syd's hard line; though I do find it sympathetic.

My reasons are simple:

 a) We have no mandate to undo the widely published and unchallenged
     declaration that we will support DTD and W3C Schema.
 b) As things stand, supporting DTD and W3C Schema is not an impossibly
     big burden (that's probably the most controversial part of this, of 
course;
     I know Syd would make interesting changes to content models if he 
could,
     but is prevented from doing so by DTD constraints)
 c) W3C Schema and DTD are, for good or bad, the constraints systems
     supported by a majority of tools. If we said we didn't care about them,
     we'd be isolating ourselves.
 d) If we dropped XSD and DTD, and gave ourselves the freedom to
     use Relax in an idiomatic way, we'd risk delaying P5 by another year.
     Is this the moment to throw things into flux again?

My feeling is that our direction is clear, to work on additional modules,
make increased use of the class system, write better tools, do I18N,
improve documentation etc. Rethinking decisions about constraint languages
is, to my mind, a low priority.

But I say again, my heart is with Syd on this; just not my head.....

Let me note that if we do only support ODD-generated schemas,
we have to drop the DTD and Relax NG parameterized schema fragments
entirely. We can't say "here they are, but you mustn't use them". It
plainly _is_ possible today to use P5 in the same way as we used to use
P4, with DTD subsets.


-- 
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