Storing Lots of Fiddly Bits (was Re: What is XML for?)

[Mark D. Anderson]

Very interesting discussion. Let me try to "echo back" what I'm hearing.

On one side, there are those who think of XML primarily as a
serialized representation of The Real Thing.
They decry notions of programming via the DOM for two reasons:
- (eliot kimber) The XML data model that you would use
programmatically may not actually match the the data model
of The Real Thing.
- (paul prescod) Using the DOM or other XML-level manipulation
is a level of indirection from manipulating The Real Thing, and
is hence more obscure and inefficient.

Then on the other side (not that I'm trying to make this a battle),
there are those who say:
um, ok, but trust me, there are cases where I really don't want
to deal with The Real Thing because in my application there are
actually a lot of heterogenous Real Things. Or maybe just one,
but it is unbelievably painful to deal with (say, IBM CICS).
Or maybe some demented person is keeping the Real Thing itself in xml.
And I would like to have a "standard" way to address that model --
meaning some way to read and possibly write it from a programming language.
(Does this include simon sl, tim b, david m?)

Then on the other side (I'm just reflecting how these discussions
go :)) there are those who say well, ok, but it is futile to
try to find such a mapping that works on all data models
(including both tabular style and document style) from our many
varied programming languages. And to the extent that it is possible,
it already exists (list your favorite set of perl grove utilities here).

Just to add my own $.02, I feel that it *is* possible to arrive at
a common API, with automatic transformed equivalence among
programming languages. It isn't the DOM, or just the DOM, because
the DOM forces programming based on things like "getAttribute",
not "employees.next().name". Sort of like doing lisp programming
entirely with "car" and "cdr".

However, I think that it can't be done until things like XLink
and DCD-the-next-generation are squared away, and it will be mostly
useful to scripted languages that can do such dynamic binding --
though i expect that a gen utility could make .h files to allow
similar programming from C/C++, in much the same way that 
some object-relational gateway products work.

-mda






xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)