I was willing to consider XML, but the following persuades me that
we would be ill-advised at this time:
1. XML 1.0 was designed for representing documents, not attributes.
While there are some in the W3C XML group that want to extend XML
for representing attributes, subsetting the parts of XML not needed and
adding data types, it is not clear whether they will prevail on the
part of the W3C XML group that considers XML fine as it is for representing
documents.
2. If the IPP WG were to use XML 1.0 now, we would make decisions about
using XML for representing attributes that would likely be different
than the XML group will do, if they do at all. For example, the
representation of dates, ranges, multi-valued attributes, data types,
data type names, etc. (I had hoped that a draft of IPP in XML would
have been forth coming during our review that just used basic XML, but
that doesn't seem possible, now that we understand the current capabilities
of XML a little better).
3. If the IPP WG were to wait for the W3C XML to support attributes
in an approved XML version, IPP would be delayed at least six to nine
months, if not longer. The prototyping work of the last year
would need to be largely repeated. Printer, network, and OS vendors would
likely deploy their own proprietary versions in the meantime, making IPP
largely irrelevant.
4. One of the concerns that the IESG has had about IPP using HTTP is that
HTTP has lots of other requirements that pull it in directions that might
not suit IPP. The same is likely to be true for XML (e.g., representing
documents vs. representing attributes).
5. The prototyping work that we have done during the past year shows that
the current protocol is workable.
6. IPP has had a lot of significant involvement and contributions from
printer vendors, NOS vendors, and OS vendors. This is a historic
achievement! I have never seen a standard in this area in which all the
major players have been involved. Lets not lose this "window of opportunity"
in the name of waiting for something better. When that comes, if ever,
there will be something coming down the pike after it, that is even better.
That is the nature of our business. Lets ship what we have now, since
we have shown that it meets our requirements and has been designed to
be extensible to meet our future requirements.
Tom Hastings