[
Lists Home |
Date Index |
Thread Index
]
On Wed, Dec 01, 2004 at 08:49:41PM -0500, Rich Salz wrote:
> Why expect a globally consistent message for how to do distributed
> computing over the Internet?
There are obviously many different ways (i.e. in many different
architectural styles) that one can build Internet scale systems. So why
haven't any of them been described? All that's been done so far is to
describe the lowest common denominator(*) of all of them, which doesn't
get us very far, since the LCD is also a base from which a lot of
unworkable Internet scale systems can be built.
> I suspect your post was a stalking horse for a REST discussion...
Nope. I thought you would have realized by now that subtlety's not my
style. 8-) REST's special in this context (in the sense that the Web
is special - a system of systems), but not that special..
(*) I'm thinking of Don Box's "tenets", but even those were ambiguous
as I mentioned to Dare. Dave Orchard's "SOA constraints"[1] was far
more ambitious, but nobody picked up on it like they have with Don's
tenets. The Web services architecture document[2] could be seen to
describe different styles, but unfortunately wasn't nearly rigorous
enough for anybody to use effectively, since it didn't describe the
styles in terms of constraints.
[1] http://lists.w3.org/Archives/Public/www-ws-arch/2003Feb/0055.html
[2] http://www.w3.org/TR/ws-arch/
Mark.
--
Mark Baker. Ottawa, Ontario, CANADA. http://www.markbaker.ca
|
