[tei-council] Conformance .... the continuing saga

[Syd Bauman]

LB> 1. Revise MD to make explicit that there is more than one way of 
LB> building a schema from the TEI Guidelines. In fact there are three:
LB> (a) write an ODD and process it with a conformant ODD processor
LB> (b) write a DTD subset using parameter entities and ting and ting
LB> (c) write a Relaxng schema which combines TEI RelaxNG modules ad lib

CW> My course would be: do (a), forget about (b) and either mention
CW> (c) in a footnote or point to some white paper on the TEI website
CW> that explains how to do that, clearly marked as advanced
CW> material.

So far I agree with you completely, Christian.


CW> This is what we invented ODD and Roma for. Whats the point in
CW> showing people how to do this on foot?

But now you've lost me. Are you thinking that "stitching together" a
Relax NG Schema for validating TEI instances should do nothing more
than module selection? I (apparently incorrectly) took Lou's
"combines TEI RelaxNG modules" to mean the same thing I have been
talking about, which includes the usual kinds of customizations users
need (without the documentation and perhaps some of the rigor ODD
provides, and thus a non-conformant methodology).

I think module selection alone would be insufficient. The point of a
schema is to provide constraint -- the user has to mold it so that it
provides the constraint she wants, else it is not a very useful
feature.


I think it may be a good idea to discuss/show/demonstrate how to do
this because, as has been pointed out
* user may not have easy access to roma or webRoma
* it's rude to insist people use a system for which the only supplier
  of software is us

In P4 the user didn't need a tool at all. Just a text editor and a
copy of the Guidelines would do the trick. (Although I have to admit,
a copy of Goldfarb is a good idea, too :-)