Re: Using HyTime Arcform stuff (was Re: Ephemeral XML?)

dgd@cs.bu.edu (David G. Durand) wrote:

> ><!DOCTYPE MyDoc [
> > <?ArcBase XML-Link XML-MyArc XML-InfoMaster>
> >Architecture//EN" >
> > <!NOTATION XML-MyArc PUBLIC "-//ME//NOTATION My Architecture, XML
> >Profile//EN" >
> > <!NOTATION XML-InfoMaster PUBLIC "+//ISBN 0-189773::IBM//NOTATION
> >InfoMaster Architecture, XML Profile//EN" >
> >]>
> I strongly urge that we abandon any attempt to include these additional
> markup bytes in XML linking. I'm not sure that we need notations at all, we
> simply need a way for XML to declare what AFs are in use. If we decide to
> allow _user defined_ AFs we might need notations.
> In any case, rigamarole declarations like the above should not be required
> for XML linking: the use of the XML-linking architecture can imply _all_
> the declarations required by HyTime.
> [...]

I agree with this -- the AFDR facilities for which <!NOTATION...>
declarations are used are probably not needed for XML, and the
less authors are made to type the better.

> I'm worried that they will be antsy at a declaration like
> <?XML arch xml-linking xml-forms>
> which will be required by nearly every document.
> I don't think we can sell too many lines of magic startup incantation, when
> we're competing with no lines of magic incantation to get going with HTML.

I don't see how we can make the magic incantation any
shorter than that (and I firmly believe that *some*
incantation is necessary).

But if this is really seen as a problem, how does this
sound as a compromise solution:  In the absense of any
<?XML-arch ...> declarations, an XML user agent shall
process all architectures that it is capable of processing.
If one or more <?XML-arch ...> declarations are present, an
XML user agent shall not process any architecture that is not

IOW, an author can use <?XML-arch...> to explicitly specify
which architectures the document uses, or can leave it out
to indicate that it (at least potentially) uses every predefined
XML feature.  [Since the latter option is IMHO the moral equivalent
of using a one-size-fits-all DTD, I _personally_ would not want
to go this route, but there are plenty of current HTML users who
would no doubt prefer it.]

This is of course all based on the assumption that there
will be more XML-based architectures in the future; if
hyperlinking is the only semantic we plan to predefine,
I'd be perfectly happy with "use any GI as long as it's not

--Joe English