xml-dev - RE: Who will maintain SAX?

[Michael Brennan <Michael_Brennan@Allegis.com>]

> From: Paul Tchistopolskii [mailto:paul@qub.com]
> [...]
> There is some simplification in my statement, but 
> I think it is clear that DOM is *not* a first-class 
> API, when SAX  *is* a first class API.

DOM serves it's purpose, but has issues. The popularity of JDOM in the Java
world is a testament to the fact that developers want simpler APIs that more
consistently abide by the conventions of their chosen programming language.
Plus, there is a major gap in DOM in that is does not address how to convert
a byte stream or character stream into a DOM tree. SAX addresses this.

> The entire XML world is built on SAX. I think it is 
> just funny that such an important API is produced
> in unpredictable fashion.

And it's remarkable that it has achieved such broad compliance in the
industry. This is truly a testament to the tremendous value it has brought
to the XML world. And even though it started out as a Java API, it is
routinely supported in C++ and proprietary 4GL tools, as well.

> Either W3C should drop DOM or it should take SAX.

I don't think it should drop DOM. There is some value in DOM, as long as
developers understand that there need not be religious adherence to DOM.
Alternative APIs that are more consistent with the conventions of a
particular programming environment are a plus, so long as they do the right
thing in terms of producing well-formed (and valid, where appropriate) XML
for interoperability. But adopting SAX as a complement would be very
advantageous. Of course, the points raised by others regarding the speed in
which the W3C could address this (or whether they would have the resources
to address this at all) begs the question of whether this is a viable
course, in spite of the fact that I believe it would make great sense.