[
Lists Home |
Date Index |
Thread Index
]
It is a good paper. This documents the results that the VRML
community made note of during the X3D design effort. The term
"impedance mismatch" was used widely there as well and described
on this list. The chilling effect of the network effect is that
VRML doubled back on a clear and clean object description to
use XML. We get the tools but at the cost of clarity in the
model by having to boot it up a level of abstraction to
accomodate two data models (XML and VRML97) and two syntaxes.
While the use and convenience of this is obvious, there is something
spooky about making concepts such as compositors first class
data objects. It has the same uncomfortable feel to it that
WYSIWYG does but I can't quite say why.
So the future of Longhorn is XML as first class objects and XUL?
The MID lives.
Commodity marketing: the best shall be worst because the least
are the most. The success of the web architecture is that it
mandated that all applications should be equally uncaring and
dumb by removing the possibility of agressive local evolution.
As was hoped for, XML and its attendant processes have hogtied
the dominant vendors, successfully slowing their rate of innovation,
and constrained the web mavens into debating the details of an
architecture already moribund by the time it became dominant.
Now the new predators will emerge from the niches.
The revenge of the forty-somethings is complete.
len
From: Joe English [mailto:jenglish@flightlab.com]
Dare Obasanjo wrote:
> [Simon St.Laurent]:
> > If I was interested in building type systems around XML documents, I
> > think I'd be even more frustrated with W3C XML Schema than I am today.
>
> You're not the only one. See
> http://www.research.microsoft.com/~emeijer/Papers/XML2003/xml2003.html
This is the most promising approach to "data-oriented" XML
that I've seen. I think Meijer's on to something here.
|
