Re: Evolutionary question

Russell Holt ([email protected])
Wed, 22 Nov 1995 12:44:19 -0500


>Now, most of the other protocols are being supplanted by HTTP and it's
>roots as a unification of all protocols is less noticable, yet I'd
>hesitate to exclude smtp, nntp, telnet, ftam, sql, x500 or any other
>protocol in use.

Yes, these are all part of the "online universe". Don't forget commercial
online services merging into the Internet.. Except of course for MSN,
which is going to obsolete the Internet, right right? In which case,
all this talk is academic anyway, because there can be only one! And
it is called "Blackbird"! ;-)

>What WWW is today may not be very clear, but it's future, I believe,
>lies in HTTP++, i.e. providing connectivity amoung all our objects.

Really? That's interesting. I certainly think that HTTP or HTTP++
will be a major part, and it may become the future, but I'm not
convinced that it's the best or most desirable Future of The Web,
if we really consider that The Web is "everything". HTTP
is far too limiting, in my opinion - at least in its present form
(or any proposed changes I've seen). Distributed objects via ORBs
and real embedded document systems such as OpenDoc make HTTP + CGI
look like the stone age.. even with Netscape 2.0's frames, "live
objects", and kludgy (albeit interesting) session support --- but less so.
I'm more interested in a dynamic environment, which excludes "forms"
(and therefore CGI) in favor of event-driven back-and-forth client/
server interaction, as implied (to me) by a distributed object system.
In that scenario, there isn't really a "web server", and there are no
"cgi programs", etc - that's handled in the object system.
But the infrastructure for that on a scale comparable to the Web today
simply doesn't exist, and it's a lot easier to do HTTP/CGI/... probably
more efficient as well.

-Russell

=(Russell Holt)====-===-==-=-=--=---=---- --- -- - - - - -
[email protected] http://www.destinyusa.com/russ/
Destiny Software Corporation http://www.destinyusa.com/