[
Lists Home |
Date Index |
Thread Index
]
9/27/2002 1:40:49 PM, "Dare Obasanjo" <dareo@microsoft.com> wrote:
>Simple question: Does HLink fit into the rest of the XML architecture that is currently being built
by the W3C?
>
Uhh, nowhere? Your personal opinion is shared unanimously by the TAG.
The TAG's opinion is not shared unanimously by the XHTML folks, however,
which is why www-tag is so lively these days.
The basic ideas behind HLink are similar enough to a SGMLish design pattern called Architectural
Forms (which was apparently considered for the Namespaces mechanism, but then exorcised) that the
"name remapping" ideas are getting some traction again.
I opened this thread because, to steal Arjun Ray's phrase,
lots of things happened for fear of being Left Behind in the late '90s,
and some "technology audits" might be in order now that the bubble has
burst. I don't have strong opinions on what will survive the audit, but
let's re-examine what colonified namespaces are and aren't good for
before they become one of those things that everyone uses, muttering
quiet curses against their anonymous creators, simply because
everyone else uses them :-) [Sorry, but I do believe the phrase
"Who the F*** invented this S**t" is frequently heard in most development
shops, at least those I've worked in. ]
Likewise, HLink, AF, and name remapping MAY have more utility than
they had been given credit for, and I'd like to hear the arguments,
thoughts on how to use it in actual software. Someone made the point that
it doesn't work with XSLT, but would be easy with SAX ... that's the
kind of thing I was hoping to see. So, is HLink name remapping a
design pattern that should be in our toolboxes along with colonified
namespaces, the WXS spec that seems to be heavily dependent on that paradigm,
and XSLT? Or is it just a side passage in the XML mine that can be
ignored by most? Or can it replace the dominant paradigm in a majority
of cases?
|
