[Om3] OMS and OMV contents?

[James Davenport]

On Wed, 11 Jul 2007, David Carlisle wrote:
> > On Wed, 11 Jul 2007, Michael Kohlhase wrote:
> > it, but with the alignment goal this question raises it's ugly head
> > again.
>
> I think not, and probably not in "canonical mathml" either.
<snip>
> Thus I think it's sufficient to align on csymbols with text. Of course
> even for text we have the problem that OMS names are XML Name tokens but
> csymol can take arbitrary XML data. <csymbol>2-norm</csymbol> for
> example doesn't translate too naturally to OM2.
This is what I was saying (I think) only much better expressed.

> James wrote
> > OMV: Here the question is more subtle - does the presentation create a
NEW
> >      OMV, or not.
> For MathML, it does, that is if you use presentation in a ci in a bvar
> construct it is only supposed to bind ci using the same content
> (presentation and all). But you can still, as above, generate a unique
> name for the bound variable that takes account of the presentation
> but is itself just some flat string.
In that case, David's construction works well here.
In cases where the presentation is purely that, e.g. "If e substitute for
the red x only, we get ..." then OMATTR seems like the right answer.
James