W3C home > Mailing lists > Public > www-talk@w3.org > March to April 1998

Re: hmmm

From: Marc Salomon <mes@slip.net>
Date: Thu, 12 Mar 1998 10:39:07 -0800 (PST)
To: www-talk@w3.org
Message-ID: <Pine.GSO.3.96.980312084302.14803B-100000@slip-3>

1. Although the W3C's non-user-centered membership policy has been
troubling since day one, I don't see the W3C as a bottleneck for the
advancement of http-ng. 

2. Once upon a time, folks who would contribute the technical kernel to a
nascent Netscape wrote a GUI browser called Mosaic that was also
source-code free for personal use (on Unix at least, and if not encumbered
by Motif). The cathedral method of development was used and
not-invented-here user-contributed patches/enhancements rarely made it
into the distribution.  This list was once largely a discussion list for
Mosaic development in progress.  Just as it appeared that the going was to
get better, NCSA pulled the plug on the project as it feared competing
with Netscape. 

3. Those same folks at Netscape, at a time when the internet still held
the status of "not-invented-here" at Microsoft and before Netscape decided
to position itself as a tech white knight fighting the for their right to
party on your desktop, had no compunctions about manipulating
<BLINK>interoperability</> by fiat of implementation and universal
distribution for their own commercial benefit.

4. It might be comforting to take the leap of faith that all good ideas
will get incorporated into a future navigator release.  But based on
history and a fuzzy quasi-death-row corporate conversion away from the
private cathedral model of aggressively accumulating intellectual property
through software development and towards the more open agora model, I am,
to say the least, not optimistic.

5. We have never had it so good cross-platform interoperability-wise. 
This is because the transport (TCP/IP) and application (HTTP/HTML) layers
of the WWW were not developed by scrappy entrepreneurs risking their own
capital and engaging in ruthless competition to make the world a better
place while enriching themselves. This all came about--insulated from
profit pressure--as a result of the taxpayers of the US and Europe funding
permanent militarization of the US economy and high-energy physics
research respectively.  The first widely-distributed, multi-platform GUI
web application, Mosaic, was also developed at a state University's US
Federal government-funded research center, although by that time most US
state universities had atrophied from financial neglect to the point where
all discoveries, patents and products funded by several flavors of public
dollars were up for private license, as was Mosaic.

6.  Indeed, Java excepted, the pace of evolution and deployment of
interoperable standards has slowed to a crawl since it largely moved
behind closed doors three or four years ago.  The cold corpses of cool
ideas proposed and nurtured by the best minds in committees over those
years yet ignored by the big implementors have filled rooms.  But the
physicists whose need for fast pre-print academic communication brought us
this medium still can't represent their mathematical formulae on the web
because nobody found it profitable to agree on a math standard, if even an
intermediate first cut.  Note there were plentiful resources available to
extend the infrastructure for publishing the advertising and pamphlet
genres on the web whose only need for mathematics was the decimal point
and currency symbol. 

7.  I don't see how, with other potential options out there for testbed
browsers, Amaya, Mnemonic and the ashes of Mosaic 2.x/3.x or Arena, that
the release of source code for Netscape, however interesting the
dissection and hack session will be, can provide critical mass for
accelerating http-ng implementation and deployment without a committment
by Netscape to a framework that will vet, incorporate and maintain
contributed code that delivers functionality accepted by the community.

8.  Although its nice to see Netscape's new found committment to casting
the widest possible net towards keeping web standards open, I can't help
but recall the day that Microsoft started sending smart folks to IETF to
participate in the development of open standards. I was pleasantly shocked
when that started (and would still take bets that it will all end quickly
as soon as they get 50% + 1 of stable market share).  I am more changrined
now at watching Netscape fighting for its life at the hands of others
practicing its own selfish tactics of coercing open standards towards
proprietary that they thought were so successful early on

9.  This squandering of resources and time competing to end competition is
why it is essential for there to be a space for interested third parties,
not content providers nor user agent vendors but end users (Participants,
Not Consumers) to have a substantial influence in the development of
interop standards.  Many had hoped that the W3C would fill this role.  The
evolution of the WWW is nothing short of the unfolding of the
communication medium that will be a major forum for human interaction over
our lifetimes; an impact that will be too pervasive and important to risk
leaving to corporate interests alone.  Inertia unchecked is moving the web
towards the consolidated one-way toll road that is television, and its up
to us to ensure interoperability *in our interests* by deflecting the
trajectory under which web evolution unfolds away from a
corporate-centered and towards a people-centered model.

10.  http-ng in its various stages of proposal is a great idea.  The
question is: Given the constraints that exist, will we see it in our
lifetimes?

-marc

 ---------- Forwarded message ----------
Date: Thu, 12 Mar 1998 02:20:41 +0000
From: Martin Hamilton <martin@net.lut.ac.uk>
To: www-talk@w3.org
Subject: hmmm
Resent-Date: Wed, 11 Mar 1998 21:21:06 -0500 (EST)
Resent-From: www-talk@w3.org

-----BEGIN PGP SIGNED MESSAGE-----

At the end of the month, Netscape will be releasing (most of) the
source code to their WWW browser, under a relatively "free" copyright
which (amongst other things?) is intended to support ongoing Internet
community based development of their product.

When you factor in Apache's market share, this means that hacker
powered "products" will account for a significant proportion of both
client and server installations.  According to the February 1998
Netcraft survey, Apache is on around half of the WWW servers Out
There.  I'm missing a recent statistic on browser use (anyone else got
one?), but even given the recent encroachments by Internet Explorer,
it seems reasonable to assume that Netscape still have a very large
number of users.

Why is this interesting/important ?

I think it's both of these things, because the upshot is that hackers
(as opposed to marketing/PR departments, middle managers, and big
business - or waffly academic - oriented "standards" groups) will be
in a position to make an impact on the future development of the WWW.

In particular, I'd like to suggest that now might be a good time to
start thinking about what a next generation "HTTP replacement"
protocol should look like.  I'm not sure whether this list is a good
place to have this discussion, but we should find out pretty
quickly... :-)

The "secure shell" protocol being promulgated through the IETF's
"secsh" working group looks very interesting.  Go to your local
Internet Drafts server and check out draft-ietf-secsh-*.  At a loss ?
Check out <URL:http://www.ietf.org/> for more info.

Ciao!

Martin

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3i
Charset: noconv
Comment: Processed by Mailcrypt 3.4, an Emacs/PGP interface

iQCVAwUBNQdGb9ZdpXZXTSjhAQGVXgP+JrKdlGslZJfE9z5GO46xDVup83sbDnSa
Fm0lRFCAQEPjO8rJH+m4KOsG5xBUA9Q61fhvTvEZVDBlPIJUrcIXXXbiPw4RH7oa
OCM/y95gpY8ruHrwNakTpfn3a+iV603lsk0vk+fr5XduEPvYyONsi8okZRhFYw+O
IPrKo9SNWM4=
=OXL4
-----END PGP SIGNATURE-----
Received on Thursday, 12 March 1998 13:39:18 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 27 October 2010 18:14:23 GMT