[
Lists Home |
Date Index |
Thread Index
]
Title: But they cheated(was RE: [xml-dev] misprocessing namespaces) (was Re: [xml-dev] There i s a meaning, but it's not in the data alone)
That's
the beauty of being near the throne and the trouble with autocracies
but no
one has a better organization, so we let that dog hunt.
However, I think namespaces reflect the dominant
concerns of the applications
those
who proposed them were most familiar with at the time. If they work
for
other applications, that's gravy. AFs also reflect the limits of the
designers'
visions based on the particular applications they were
most familiar with at
the
time: document processing using DTDs. DTDs have mono-namespaces
and
have to punt to SUBDOCs for integration of a compound document, or
flatten out into a more comprehensive DTD. AFs
made it possible to get
around
that with mapping rules. That solves some problems but still
leaves
the semantics (the S word again) largely an issue of documentation
if
better referenced. I'm not sure right now that compound
documents
can be
reliably processed and as such, other than applying them to
the
relational systems (joined tables, views, etc.), they can easily
force
us toward more centralization.
So we
are not much further down the path toward a solution of behavioral
specifications are we? Steve?
(Gad! We're doing the MID thing again!)
len
If namespaces have won, it's only because they changed the
rules to suit themselves. Mind you, it wouldn't be the first example in the IT
world.
Nice thing about DTDs, of course, is that the rules seem
slightly less open to interpretation (or misinterpretation, depending on your
viewpoint).
| 