Re: [TCLUG:18319] "Web Architecture"

[Philip C Mendelsohn <mend0070@tc.umn.edu>]

On Tue, 30 May 2000 Nick.T.Reinking@supervalu.com wrote:

> In addition to opinions on where "web architecture" will be headed in five
> years, some ideas on what "web architecture" _actually is_ would be
> appreciated.  I've come up with (so far):

I think the most accurate answer is that "web architecture" is
non-existant.  I'd have to look up a formal definition to make my argument
stand (and it may really miss the intent of your project), but an
architecture is specified top down.  The DEC Alpha and VAX were real
specified architectures, where the processor was not designed, but rather
the things it needed to do were laid out.  The x86 family is not -- things
were cobbled together as things went from design to design.

My point is that the web is more that sort of arrangement.  A country wall
pieced together from available rocks found on site, rather than built to
designed specs using materials specified in advance.

> Any further ideas/comments would be appreciated.  And I've got to try
> to make this exciting, so speculation on cool things coming up the pipe
> would be appreciated.

The most exciting thing I saw was the article in the business section of
the STrib where M$ is announcing their vision of the future, which happens
to be the things Sun, Oracle, et al, have been establishing themselves in.

I would look for tighter server communication protocols -- maybe ways to
try and do things like wide-area clustering, though real clustering has to
be supported at the processor level, and I don't know that Intel has
enough on the ball to reinvent something done 20 years ago.

But definitely, think about things on the server side and what that could
do for people.

Pardon the bandwidth -- just my $2 x 10^-2

Phil M

-- 
Lottery:    a tax on people who are bad at math