Subject: Re: Web Interactivity From: allaire@worldmedia.com (Jeremy Allaire) Date: Mon, 3 Jul 1995 01:51:40 -0500

Cc: online-news@marketplace.com
Sender: owner-online-news@marketplace.com
Precedence: bulk
Status: RO
X-Status: 

Most of the responses to the post:

What are people doing about Web Interactivity?

prompted some interesting, though I think
problematic responses.

The problem of Web-Interactivity is not about
'how to do chat well', nor is it about environments
that are so distant and so unusable by significant
portions of the online community (more on this later).

The problem is about developing basic applications
using existing Web technology, most of which have
been extremely underutilized because of the UNIX
centric nature of Web development (until recently,
at least).

For instance, the Web provides extremely rich ways
for developers to create dynamic on-the-fly Webs,
based off of relational database systems.  Many
major services use such technologies.  For the
most part, use of such things has been (1)
cost prohibitive, (2) difficult given the current
state of server-side development tools, and (3)
mostly on UNIX -- which, of course, contributes
to both (1) and (2).

The response from several list members has been
to either (1) shout Java, or (2) propose even
loftier and more distant technologies such
as Sun's DCE, or more generally the Standards
based CORBA.

First, as Web publishers, we must be asking
what we can do to enhance and develop our
technologies to be more dynamic (literally,
no static pages) and user-driven -- more
interactive, not necessarily in the user-to-user
framework (though this is a big part of the issue).

We also need to be asking how we can do this NOW,
how we can leverage existing software talent as
opposed to development languages and distributed
technologies that are barely tested and hardly
accessible to the vast majority of Web developers.

We need to be able to tap into the massive technical
knowledge created by the desktop PC revolution --
namely, database and visual programming on PCs, mostly
in the relational database world, but also in the
emerging convergence of relational databases and
distributed applications -- e.g. client/server computing.

We need to use affordable and easy-to-use tools that
are scalable, that can integrate the massive pool of
existing database development talent, and that allow
for the rapid development of interactive applications --
and crucially, which don't assume some significant
and myopic leap in the technology available to the
average PC user.

There is enormous potential to develop the Web into
a significant force in client/server distributed
database applications -- existing standards provide
enough of a shell to do such a thing.  There is no
compelling reason (at least as far as I've seen --
someone speak up if they can come up with one) to
pitch one's dreams in technologies which DONT EVEN
RUN on _ANY_ desktop computer systems today (e.g.
HotJava -- runs on Sun, WindowsNT and is in Alpha
stage, has NO development tools, NO support, and
NO testing on the OS's that sit on 95% of the world's
computers).

While the world of distributed object computing is
on the horizon, it is mostly in the lab's of major
technology firms such as Sun, Apple, and Microsoft
(though, if folks read last weeks Info-World --
http://www.infoworld.com -- they might learn some
interesting and frightening things about Java's
history).

And, incidentally, the Web is clearly positioned
to evolve to be the core interface to the world
of distributed objects, perhaps even displacing
Microsoft as the leading OS vendor in the future --
see my article "WebOS:  The Future of Distributed
Computing?"  -- http://www.worldmedia.com/webos.htm.

Where does this lead?  I think it is somewhat clear
that we need systems that allow small to large
publishers created database applications using
existing Windows talent on significantly less
expensive hardware and on significantly less
expensive software -- and, incidentally,
much easier to use systems.  

Large-scale vertical solutions such as Netscape's
Community Server are both over-priced and
inflexible -- e.g. lack significant customization.

Systems such as NaviSoft's NaviServer provide
limited database functionality -- e.g. poor
query capabilities, and extremely limited
output systems -- and only high-end interfaces
for additional development (e.g. Tcl).

Thoughts on solutions . . . I have a few . . . 

Jeremy Allaire


From owner-online-news@marketplace.com Mon Jul  3 13:26:17 1995
Received: from marketplace.com (majordom@marketplace.com [199.45.128.10]) by cnj.digex.net (8.6.12/8.6.12) with ESMTP id NAA01083 ; for ; Mon, 3 Jul 1995 13:26:15 -0400
Received: (from majordom@localhost) by marketplace.com (8.6.12/8.6.12) id IAA01316 for online-news-outgoing; Mon, 3 Jul 1995 08:45:30 -0600
Received: from server.nww.com (nww.server.wing.net [204.57.38.17]) by marketplace.com (8.6.12/8.6.12) with ESMTP id IAA01310 for ; Mon, 3 Jul 1995 08:45:27 -0600
Received: from adam146.nww.com ([204.57.53.146]) by server.nww.com (8.6.5/8.6.6) with SMTP id KAA05157; Mon, 3 Jul 1995 10:42:56 -0400