Re: anti-goals

Peter Murray-Rust (peter@ursus.demon.co.uk)
Tue, 06 Jan 1998 21:31:36

At 14:13 06/01/98 -0500, David Megginson wrote:
[...]
>
>Here are some additional examples:
>
>- most transformations
>- producing a rendered version of an XML document (electronic or
> paper)
>- context-sensitive searching
>
>Your example, extracting data, itself covers a wide range of
>applications:
>
>- database import/export
>- online transactions
>- configuration information
>- meta-data exchange
>- general client/server protocols
>
>Essentially, SAX should cover most general requirements (most
>transformations occur as part of a processing chain, and do not need
>saved comments, internal entity references, etc.). Those applications
>that do need access to lexical information -- mainly authoring tools
>and repositories -- will, of course, need to use a different or
>extended API.

I find this list very useful. I'd add to this that SAX should cover the
functionality of the XML-related languages XLL and XSL. (Neither of these
require reconstruction of the original documents - unless you want to
*edit* the *.xsl and preserve comments, etc.). If - as has been suggested
in some camps - namespaces and 'schemas' in XML become common, then there
is an even greater potential use. [Of course these may be represented
*internally* as trees, but (if I'm right) these can be built from the SAX
interface (as JUMBO does/will_do) without requiring the full DOM model.]

When XML fully takes off there will be a requirement for a lot of software
to process the complex mixture of *.xml, XLL and *.xsl that will be
received client-side, and I would have thought that SAX was exactly what
most people want.

P.

Peter Murray-Rust, Director Virtual School of Molecular Sciences, domestic
net connection
VSMS http://www.nottingham.ac.uk/vsms, Virtual Hyperglossary
http://www.venus.co.uk/vhg