Re: Is there a way out?

The history of the Internet's standards simply does not bear that
prediction out.  The industry has chosen, time and time again, to adopt
freely-implementable, freely-licensed, open infrastructure standards in
preference to offerings that are encumbered with restrictions.

[I think this is one area where the "stockbroker's caveat" applies: "Past
performance is no guide to the future" i.e. the world has changed, not
necessarily for the better, since the mid-1990s. Many of the things you
refer to below have their origins in work in the 1970s and 1980s - different
times, different customs.]

What happened to the IBM MCA architecture?  The industry chose the
existing, open PC architecture.

[That one is easy: IBM was a sole player trying to dominate in a market
where a perfectly viable alternative existed. These two conditions will
apply less and less as time goes by: the big hardware (and software players)
are learning to move effectively in packs (i.e. standards development
consortia), at least when it suits them, and such viable alternatives simply
do not always exist.]

What happened to the proprietary
local networking architectures?  The industry chose the open,
freely-implementable Ethernet.

[Ethernet's history is far from open in the sense you mean - see US Patent
4,063,220 for the start of the hardware side of things.]

What happened to the dozens of
proprietary mail transport protocols?  The industry chose the open SMTP
protocol.  What happened to any of hundreds of other proprietary
also-ran protocols?  The TCP/IP suite of protocols, openly developed and
documented by the IETF, was and continues to be chosen by the industry.

[I refer generally back to my first comment]

The W3C's existing track record of promoting freely implementable
standards does indeed deserve better than to be bypassed.  This track
record is currently at risk, though, as evidenced by the ongoing debate
on this list.

[For the reasons I gave in my earlier post, RAND does not automatically mean
not freely implementable - otherwise how do ANSI, ITU-T, ISO, IEC and others
still function as technical standards bodies? ANSI has allowed a form of
RAND licensing since 1974 - it still lives. Interestingly, one of the
lesser-known things about most IPR policies adopted by technical standards
bodied like ANSI, ISO, etc. is that they still have a built-in preference
for non-patented technologies such that they will only accept them in
proposals/standards where deemed "essential".]

The W3's own stated charter <http://www.w3.org/Consortium/Points/> says
that their purposes include:

   "enabling new forms of human communication and opportunities to share
   knowledge. One of W3C's primary goals is to make these benefits
   available to all people, whatever their hardware, software, network
   infrastructure [...]"

   "[...] promotes interoperability by designing and promoting open
   (non-proprietary) computer languages and protocols [...]"

If the W3C does not make a clear commitment to standards that are freely
implementable by all-comers, without discriminatory barriers like
licensing restrictions or royalty fees, then their position as a
standards organisation for the Web is forfeit, and they will deservedly
wither on the vine as other organisations, which recognise the necessity
of royalty-free standards for infrastructure, take up that role.

[One of the points I often make is that, outside the sphere of regular TSOs
like ANSI, standards bodies have a natural tendency to multiply: technical
standards, as distinct from technical regulations, are voluntary after all
... ]

Received on Thursday, 25 July 2002 17:51:08 UTC