Re: Issue: shall we restrict subordinates of <oneOrMore>

James Clark wrote:

> I would propose instead the following restriction (also in terms of the
> normalized form):
> 
> If a <oneOrMore> element has an <attribute> descendant, it must not have a
> <group> or <interleave> descendant.

This looks fine to me.

Intuitively, this restriction allows (@att | elem | #data | EMPTY)+ only.
It can be always decomposed into ( elem | #data | EMPTY)+ and
(@att | EMPTY)+  .  Thus, GNF normalization always works.

Cheers,

Makoto