XML Conferences #2: Open Standards, Sydney

Related link: http://www.open-standards.com/papers

A good little conference, for OASIS/messaging/web services people.
(Conference attendees can view papers with a password mailed to them
here.)
I presented a short introduction to ISO Schematron
and gave a half-day tutorial on XML Schemas.

The standout paper was by Brett Jackson
from Fairfax Digital, detailing how they moved to validated XHTML with CSS.
They deliver 164,000,000 page impressions per month
for the major dailies
The Sydney Morning Herald and The Age. They do day parting: different kinds of stories are
emphasized at different times of day. Later in the day people want gossip,
analysis and entertainment but in the morning they want breaking news and weather.
He reported they save a million dollars a year on bandwidth costs by moving
to validated XHTML with CSS!

Brett said there is a benefit of being first, because you can
suck up the available XHTML/CSS aware human resources and so starve the
competition. But you must have or make a culture where
"designers are coders":
use text editors not QUASIWYG.
Announcing a staged migration away
from supporting generation 4 browsers (but still providing a dull text page for
them) is essential: "It is ridiculous to support
old browsers". Validation is a big gain for quality control. But there
still needs a final layer to serve different CSS for different browsers.

The move CSS, plus design changes, let them move from
38 second delivery of their front page over modems to about 3 seconds. Faster pages translates into more pages being
viewed at the same site.

Handsome lawyer Kieren Power gave a good paper Are we free to use open
standards? which I found infuriating, though I am undoubtedly kicking
against the pricks. His points were that open licenses have not been tested
in case laws, that there is an awful lot of IP around, that providing an
open source license to your IP today may not prevent you from
turning around tomorrow once the technology is being established,
taking it back and demanding license fees. He was trying to
state fairly the status quo, and did mention that despite these dangers
patents had not hindered the progress of open standards, in the sense
that he did not know of a case of a company pushing its technology
as an open standard then, subsequently, turning around and demanding
fees. (GIF was not proposed as an open standard.). He could not
see that there was any reason why not to have software patents
if you have hardware patents.

My response would be a couple of points. First, an international
public policy concern: the grab for IPR by rich countries/organizations just robs
less rich. Second, the criteria for obviousness are way too lax: evidence
of this is that there are hundreds of thousands of patent applications.
A non-obvious idea comes to a few people once or twice in a lifetime;
surely the sheer volume of applications and granted patents shows that IP is being granted to people who just
come to a problem first and come up with the obvious answer.
Why should anyone have any special rights for discovering
something obvious to the next person who comes along?

Also, had a very interesting time with W3C's accessibility/semantic bod Charles Macathy-Neville,
who was arrested in Italy for carrying a broadsword.
He mentioned RDF is developing better capabilities
to name collections of RDF statemtents, for example
so that you can express statements of trust or
assign an ontology to them.