Simple API

[Peter Murray-Rust]

[announced on comp.text.sgml]
Henry Thompson has posted an impressive picture of what a grove looks like:

http://www.cogsci.ed.ac.uk/~ht/grove.html

It describes the grove for a simple document (2 element types, 2 elements)
and it's sufficiently complex that only *part* of it is shown.  

[I make it clear that I'm impressed by this, but that personally it would take
too much effort to implement for the benefit I would get.  Many other readers
of xml-dev will probably find it's exactly what they want].

-----

It highlights for me that the spectrum of possible approaches to the API is
too large to pick an approach that suits everyone.  The grove has obviously
enormous power if you take the time to learn it but it is not trivial.  
Henry's diagram is much more reader-friendly than 10179, but confirms that
this isn't just a problem of terminology - it's an extra level of complexity.

My own suggestion is that we should produce a ReallySimple API independently
of the grove approach.  I'm sure this won't cause a schism - we need something
to test out the language, build simple trees for trying out TEI pointers, etc.

IMO most of the things that are bugging us at the moment are not conceptual but
	- how do we implement this bit of the spec?
	- how do we read in both Files and URLs (a Java problem)
	- how do we cater for applets and applications
	- what structure do we hand over at the end?  Can it be subclassed?
	- how do we get at the DTD?  (from a validating parser).
	- how do we treat parameter enetities :-)

P.

-- 
Peter Murray-Rust, domestic net connection
Virtual School of Molecular Sciences
http://www.vsms.nottingham.ac.uk/

xml-dev: A list for W3C XML Developers
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To unsubscribe, send to majordomo at ic.ac.uk the following message;
unsubscribe xml-dev
List coordinator, Henry Rzepa (rzepa at ic.ac.uk)