[Xml-bin] Some central design issues

[Al Snell]

On Thu, 12 Apr 2001, Stefan Zier wrote:

> > are nearly good enough, we should try to influence these, to get the
> > remaining 5% instead of starting all over.
> 
> I think we all agree on that one. Has anyone gotten any URLs or papers von
> BiM yet?

I'd be keen!

BiM may be a stream-focussed protocol with no seekability, is my main
worry.

The ASN.1 project seems to be based around the idea of using a schema to
encode data with no self-description information in it, and requiring the
same schema at the other to make sense of it. Although there are
applications for this, I feel a representation that you could round trip
XML through without reference to any schemas or DTDs will be more
generally useful and will not suffer from not being self-describing.

We can probably suggested an ASN.1 based representation for a
self-describing XML representation as a logical extension to the ITU's
work, and get them to accept it as part of the same recommendation... not
sure about the politics, though.

> > binXML should be an alternate external representation of the (PSV) XML-
> > Infoset, so all XML Recs are readily applicable and it is possible to
> > roundtrip binXML -> textXML -> binXML.
> 
> I have to agree on this one, too. Every well-formed XML doc should be
> representable in the binary notation and vice versa.

Definitely, otherwise we raise barriers to adoption...

> There is another requirement that might need a different orders of things:
> fast parsability through SAX or DOM (where either of those two might have
> different orders).

Mmmm; I have a solution for those two requireements :-)

ABS

-- 
                               Alaric B. Snell
 http://www.alaric-snell.com/  http://RFC.net/  http://www.warhead.org.uk/
   Any sufficiently advanced technology can be emulated in software