Re: Collected Works of SAX

[David Brownell <david-b@pacbell.net>]

> I've summarised the recent 'SAX, missing features?' thread 
> in this weeks XML-Deviant article [1]. Both to introduce the 
> discussion to a wider audience, and hopefully provide a 
> useful feature list if the suggested work goes ahead.

Applause!


> Is there any interest in progressing this further? Where 
> do we go from here?

I think there's interest.  I think part of the answer is
clearly http://sax.sourceforge.net ... plus, another 
question needs to be "how do we go there".

That didn't exactly get resolved the last time it came up;
there was a stall around the time Oasis got mentioned,
as I seem to recall.  There's an issue that Len has been
mentioning lately:  privatizing public resources.  I'm not
keen on giving SAX to any consortium, myself.

I keep thinking in terms of two parts:  the core, and the
layers on top of it (including translations to other languages
like Pascal and Python).

As for core:  I plan to put a few simple proposals
together, crossposed to the relevant SourceForge
lists and (of course :) xml-dev, much like David
Megginson did originally.  If they fly, they can go
into the sourceforge project.

As for layers on top, I confess that initially I had
thought of just a web page to collect links.  (Hence
"collected works". :)  Heck, maybe a Wiki would
be a good thing to have around in any case!  We
don't know all that's available right now, so it's
hard to know what parts are really missing.

Eventually issues of architectural compatibility start
to get important, but to start with it'd be good to
just have a more complete catalog of interesting
components.  A "SAX2 Standard Library" isn't
quite the place to start, I suspect -- though I agree
that having more streaming validators (RelaxNG,
W3C Schema, etc) should be a goal.

- Dave

p.s. Perhaps a SAX2 BOF at the Open Source
    conference next week in San Diego would be
    a good place to discuss some of this?  :)