A binary representation of XML?

I have mixed feelings about the discussion of binary encoding for information in XML documents. It took place
in late September in Santa Clara, California, under the auspices of the W3C. It was no casual get-together,
judging from the representatives of Oracle, IBM, Sun, Microsoft and others who attended (via root.cz, September 30, 2003, Akta X 0309).

What could binary XML encoding offer? Probably quite a lot to those companies. XML data
is now used at every conceivable level of information systems. XML has become a standard for web services,
application and system configuration, data storage and, last but not least, message exchange.
I disagree with the idea that XML is a marketing fad. It has a purpose.

I can imagine complex systems producing huge quantities of XML data for their other
operations. In that case, frustration over the lack of a binary encoding standard is justified. But then
I have to ask whether the data needs to be kept in XML in the first place. Maybe it does, maybe it doesn’t. There’s no simple answer. It should be judged
in context, and so should the need for binary encoding of XML data.

If I change “could” in the original question to “should,” then what I’d want for portability is an unambiguous standard
that would let us separate XML itself from its binary representation or representations. Then we could, as the saying goes,
have our cake and eat it too.