James Clark wrote:
>It's certainly much weaker and in my view much more ad-hoc.
I think yours is already very weak and ad-hoc. Rather than pretending
to have a generalized solution, I think that we should honestly admit
that our mechanism in RELAX NG is very weak and ad-hoc and is only
slightly better than ID/IDREF.
> To make a fair comparison of the relative complexity you need to do a
> formalization using inference rules that prove a keyAmbig() judgement. If
> you do this, I think you will find your proposal is a little simpler, but a
> not lot.
I will extend my rewrite, but I don't think inference rules are absolutely
required. 7.2 does not have any inference rules and this is just fine.
Cheers,
Makoto
