[tei-council] numbered divs

[Syd Bauman]

> it's the thin end of the wedge, of course; how do we decide what
> constraints to delegate? of course, this

It is not so obvious to me that this particular constraint should be
delegated to Schematron, as opposed to schema-enforced, at least not
at the moment.


> Its an interesting idea. The flaw in practice is that the
> simplification of smoothing out things to clean up after element
> removal is done before odd2dtd kicks in. 

Oh, blimey.


> Wouldn't your dummy element be offered all the time by the editor
> as a possible choice?

Depends on how smart the editor is, but yes, many would do that.
(Which is why there should be only 1 such dummy GI, and its name
should make it obvious it is not a real element.)