Re: XS: Add element-with-id or map-constructor as non-core query procedure

Subject: Re: XS: Add element-with-id or map-constructor as non-core query procedure
From: Paul Prescod <papresco@xxxxxxxxxxxxxxxxxxxxxxxxx>
Date: Sat, 24 May 1997 21:10:05 -0400
James Clark wrote:
> That's an argument in favour of allowing links to be to named flow objects,
> which can easily added: eg add an address flow object and
> flow-object-address procedure.  It's not an argument in favour of
> disallowing links to nodes, which would require style-sheets to continually
> invent names for flow objects, 

Well, if we're talking about links to IDs, then choosing the name is
easy. I can see that there are cases where it would be easier to link to
elements though.

> and wouldn't work for links outside the
> current flow object tree.  

Links between flow object trees seems beyond the scope of XS, unless we
are talking about links to *urls* which cannot really be accomplished by
node-linking either (can it?).

> Not to mention the serious backward compatibility
> problems that would cause.

True, but we are removing features and adding features so XS doesn't
seem like it will be a true superset or subset in any case.

*If* we choose to have a single linking mechanism, to flow objects or
nodes, I think that linking to flow objects is more general and is often
no harder -- especially if an "element-number" function is provided. I
can see that it might be convenient to have both, however.

 Paul Prescod



 DSSSList info and archive:  http://www.mulberrytech.com/dsssl/dssslist


Current Thread