W3C home > Mailing lists > Public > www-style@w3.org > August 2002

Re: Serving generic XML (was: storing info in XSL-FO: new issue?)

From: Albert Lunde <Albert-Lunde@northwestern.edu>
Date: Tue, 20 Aug 2002 08:59:54 -0400 (EDT)
Message-ID: <20020820125949.GE7638@futomaki.tss.northwestern.edu>
To: www-style@w3.org

It sounds like the state of the art for attaching "meaning" to
XML documents is agreement out-of-band, whether that takes the
form of a W3C document, an entirely private agreement, or something
in between. CSS can provide presentation information, and DTDs
and schemas can provide syntax constraints, but there's no
"semantic markup language" we can dust on an XML document
to make it meaningful. 

I tend to agree that while class or element names may be suggestive,
they don't convey meaning in an unambiguous, interoperable way.

It seems like this discussion is losing track of the likely
role of XML dialects not derived from (X)HTML.  There are two
related points: (1) Not everything served up with HTTP will
be intended for human readers (2) Deriving an XML vocabulary
from XHTML isn't the only way to make it meaningful.

It seems to me that the most likely way we will get a "semantic
web" as opposed to tag soup, is the development of a lot of
specialized XML vocabularies for particular purposes,
rather than one big omnibus language. It's not likely that
a single generic "web browser" client will understand and
present them all, even for non-disabled users.

As another example, I'd offer "xCal" an XML dialect for
interchanging calendaring information, featured in a current
internet draft:


(This is a representation in XML of objects previously defined
in other terms.)

The utility of this XML vocabulary, should it come into use,
is not that I could point a web browser at it, and read it, but
that could point some sort of calendaring client at it, and have
that understand it.

It maybe that web gateways with XHTML user interfaces are the
best hope for accessibility for specialized vocabularies/protocols
like this. I don't see a way we can give them enough intrinsic
information (other than the out-of-band agreements that define a 
protocol) to make them "accessible" or particularly meaningful
as stand-alone documents, taken out of context.

    Albert Lunde          Albert-Lunde@northwestern.edu (new address)
                          Albert-Lunde@nwu.edu (old address)
Received on Tuesday, 20 August 2002 12:23:12 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:27:03 UTC