[
Lists Home |
Date Index |
Thread Index
]
This is precisely what the Unisys managers told
us during the MID project. It interested them
precisely because we turned our backs on passing
complex objects and passed only data objects. If you have
to scale outwards, you have to reduce vertical
complexity (depth). It comes down to why
the mail sends you letters in envelopes (only
handle the address information and size issues),
why children are taught to read with phonetics,
and composers are best taught harmony at a gross
level over counterpoint first.
This may be news to the web, but hardly anywhere
else. OOP was meant to solve programmer problems:
maintenance and reuse. Relational solves problems
for the data owner: maintenance and reuse. Hypermedia
solves a problem of creating relationships among
data that has different owners but does not necessarily
help maintenance and reuse. It's easy to design
hypermedia systems that fail at that and harder to
design ones that don't. Putting controls in
the text is one part of the problem. Systems
that point in are easier but only if the change
control is absolute.
At the end of it all, hybrids are the solution,
so no, this is not the beginning of a new meta
model for large scale architecture. Not at all.
It's the same stuff with hypermedia being brought
forward as mainstream rather than lunatic left
wing fringe.
len
-----Original Message-----
From: Mike Champion [mailto:mc@xegesis.org]
Are we seeing the outlines of a new meta-
model for large scale software architecture that challenges a lot of the
received wisdom of the OO paradigm?
That is, perhaps "encapsulation" of data is not such a good thing after
all ... or at least it's not scaling well to the Web. Perhaps it makes
more sense to expose the data via generic operations rather than
exchange objects that encapsulate data behind customized operations. I'm
not sure, but things are getting interesting ....
|
