Re: Parallel markup: Why do the links point from content to presentation?

On Tuesday 20 January 2009 13:44:54 David Carlisle wrote:
> > could you probably tell me a use case that justifies the direction of the
> > links in parallel markup?  They point from the content markup to the
> > presentation markup.

David, I think I understood your explanations, but now it's not clear to me
what you personally support or advocate.

> Actually I think it's usually more useful to put the content mathml in
> the base of the semantics, and annotate it with presentation mathml if
> you need to override the display.

So you argue in favour of a revision of the <semantics> spec?

> Also, links work better in that direction, the mathml2 chapter 5 example
> "simplifies" things a bit by only having two terms in the plus, so you
> only get one "+" term, but in general
> if you have
>
> <apply><plus/><ci>a</ci><ci>b</ci><ci>c</ci><ci>d</ci></apply>
>
> rendering as
> a+b+c+d
>
> using simple links you can make each of the <mo>+</mo> link to the same
> <plus/> but you can't make the <plus/> link to each of the +.

This example is convincing, but having each of the <mo>+</mo> link to the
<plus/> is not possible in MathML 2 or 3, is it?

> In absence of other criteria, the first branch of the semantics element
> is a sensible choice to contain the id attributes. Applications that add
> or remove annotations will then not have to re-assign attributes to the
> semantics trees.

So by "first branch" you mean the branch that comes before the <annotation>
and <annotation-xml> children?

Cheers,

Christoph

-- 
Christoph Lange, Jacobs Univ. Bremen, http://kwarc.info/clange, Skype duke4701