Re: Request: Techniques for reducing the size of XML instances

[Paul Cody Johnston <pcj@stanford.edu>]

I agree with Mark's statements about gzip: it's standard and works
great.  Any hopes for Binary XML should forget about the compression
angle -- there's no need to invent a binary XML standard to solve
compression problems, it's already been tackled.

However, I do think binary XML has promise in line with the notion of
"compiled data", especially as it relates to validation.  It's natural
for me to imagine a sort-of "expendable" or "temporary" pre-validated
binary rendering that can be ripped from the "master" XML document and
used as a cached or internal format for certain applications.  Seems
like that would be a useful tool for application builders that
warrants general research.  Not currently popular though! Many people
seem to have a visceral (irrational?) opposition to the very thought
of binary XML.

Just my two cents,
Paul

+---------------+---------------+------------------+
| Paul Johnston | pcj@inxar.org | http://inxar.org |
+---------------+---------------+------------------+