[
Lists Home |
Date Index |
Thread Index
]
Good enough. I took it to be a set of easy
to do items that would make interoperability
possible. I note the restrictions (eg, don't
use container elements) and they don't seem
that bad. Items like not using mixed content
or container elements are *usually* decent advice
in most schema designs.
I remember a thread on this list in which
someone said it was easy to make an XML
document "RDF friendly" and I asked how.
John replied with a concise answer and no
one seemed to object. As a result, I took
that topic back to the HumanML group and we
agreed that given the nature of the kinds of
data our language user would collect, this
was a good idea. Manos agreed to take on
this task when the HumanML schema reached a
further stage of completion. I didn't view
this as "supplication". I am a Democrat.
Supplications are not in my repetoire.
Transforms are harder to master. Even if it
is more powerful, it ain't that friendly. ;-)
As to the backlash, that sort of thing was mild
compared to standards wars I've witnessed. Standards
making is always political. One gets used to it
even if never learning to like it. That is why
the one piece of advice they gave that would
seem to be uncontroversial (use namespaces to
get pieces of other ontologies) actually can
be very controversial. It depends on the business,
the locale, and the state of standardization and
specification. Ambitions...
len
From: Dare Obasanjo [mailto:dareo@microsoft.com]
Mike has already expressed some of the thoughts that led to my original
comment. Looking at the article, it seems the intention is to state that
making an application RDF-friendly involves overly restricting ones
markup and supplication to RDF. If I had written such an article, I'd
probably have leaned more towards showing people how to design their XML
documents so it is easy to transform them to RDF using XSLT or some
similar mechanism.
At least that's what I learned from the backlash against RSS 1.0.
|
