[tei-council] tables

[Syd Bauman]

> I agree that (d) is a good option, because I think this is an error
> on the part of OASIS and they should fix it.

Check. Will do so, probably tomorrow.


> For the record though, I believe one good solution could be encode
> the FPI as a URI as described in RFC3151.
> ...
> This would give a namespace URI of 
> "urn:publicid:-:OASIS:DTD+XML+Exchange+Table+Model+19990315:EN"

Jolly good! I like that as an interim solution. Thanks, Conal!


> > In any case, I have an ODD that works, but I ended up extracting
> > a table DTD from http://www.oasis-open.org/specs/tm9901.html,
> > putting it in my local filesystem, converting it to RNG, adding
> > an ns= attribute to <grammar> in the RNG, and then pointing to
> > that from my ODD. I'm not sure that's entirely kosher.
> 
> Syd, do you feel it's necessary to pull that RNG apart and embed it
> in TEI tagdoc elements?

Do you mean re-encode the RNC (or perhaps transform the RNG) of the
external schema into ODD language? My instinct is no, esp. since for
some schemas it would be impossible. (RELAX NG can express
constraints that ODD cannot.)


> Cheers!

G'night!