Lists Home |
Date Index |
10/8/2002 4:52:03 PM, "Bullard, Claude L (Len)" <firstname.lastname@example.org> wrote:
>REST must die. What we really need is a weaker UDDI. ;-)
More cryptic than usual today, Len. I'm missing the connection
with RSS. Or maybe that's the point somehow?
I found the article Mr. Winer linked to quite interesting in
that it applied Shirky's thought-provoking parallels
between evolutionary theory and the Web to XML and RSS.
I was intrigued, however by the rhetorical
device of calling the alternative to full-blown namespace wellformed
and RDF compliant RSS 1.0 "RSS 0.9x/2.0". They seem to be two
different beasts: "RSS 0.9x" is explicitly "XML" in appearance
only, i.e., is clearly processable by the proverbial Desperate Perl
Hacker that uses XML's syntactic patterns but not its "draconian
error handling" philosophy. It has been evolved from all sorts of
work at Netscape, Userland, and elsewhere (I will leave the job of
sorting out the highly controversial details to future Ph. D students
in Techno-History) and clearly has succeeded in its
ecological niche. "RSS 2.0" on the other hand, is all of a month
old, is controlled by Mr. Winer alone, and a casual reading of Mr.
Winer's esteemed weblog leaves one with the impression that
it has caused considerable challenges and inconveniences to its
I'm no fan of the complexity in XML (and namespaces in practice) but
it seems obvious that if one is going to claim compliance with a
spec, one should comply with it. If you're going to ransack it for
ideas, say you're ransacking it. RSS 0.9x does this explicitly,
whereas Mr. Winer's charmingly informal specification implies that
it is well-formed, namespace-compliant XML.
If indeed it is the hardy cockroaches that survive the flame wars
of the Web, then the half-joking proposal for "RSS 3.0" will end up
on top. XML is overkill for Really Simple Syndication formats that
don't need the hierarchical and recursive structure that XML supports,
because without the draconian error handling and namespace
well-formedness constraints, (not to mention the array of additional
tools such as XSLT that can be brought to bear on "real" XML)
XML is just a verbose way of labelling text values.
If one is not interested in leveraging the parsers that
actually implement XML, a 1-line Python program is a more
evolutionarily stable "meme" than reams of Desperate Perl Hackery.