[Om3] OMS and OMV contents?

[James Davenport]

On Wed 11 Jul 2007, Michael Kohlhase wrote:
> one thing we have to think about wrt. OM3/MathML alignment is whether
> we want to applow presentation MathML in the body of OMV and OMS.
DavidC and JHD are disucssing this now.
> James wrote
OMS: here the addition of presentation can only change the presentation.
<snip>
     James and David still agree with this, as per David's mail.
     This would mean (David says) that MathML should say that
     'presentation inside csymbol is not canonical'. This poses the
    question "what is the canonical form". This is NOT an OpenMath
    problem. The MathML solution MIGHT be:
    "If the csymbol does not contain a NCName, then its canonical form is
    a csymbol with an NCName as content, and an atribution which supplies a
    presentation MathML rendering. Within a given scope, the same NCName
    should be used consistently and uniquely."

OMV - changing the presentation only: here the issue, and solution, both
      Openmath and MathML, is the same.
OMV - a mathematically new variable.
      Here we need a new name for the new OMV.
      <OMATTR>
        <OMATP>
          <OMS name="presentation" cd="mathml"/>
          <OMFOREIGN>
            <mi mathvariant="fraktur"/>  A </mi>
          </OMFOREIGN>
        </OMATP>
        <OMV name="gothicA"/>
      </OMATTR>

      We could look at syntactic sugar such as
      <OMATTR>
        <OMATP>
          <OMS name="mathvariant" cd="mathml"/>
          <OMSTR>fraktur</OMSTR>
          <OMSTR>A</OMSTR>
        </OMATP>
        <OMV name="gothicA"/>
      </OMATTR>

      A more radical suggestion would be
      <OMV name="A" mathvariant="fraktur"/>
      but this would be a more fundamental change to OpenMath.
      However, neither David nor I would be strongly against this.
      We would need to be very clear in the standard that this is
      DIFFERENT from <OMV name="A"/>, i.e. it is the entire body of the
      OMV that makes for mathematical uniqueness.

James/David