Patent-free standards

         I was a member of HTML-WG and HTTP-WG in the mid-late 1990s, when 
the W3C formed in the shadow of the IETF. We were all concerned that the 
W3C, as a "registered corporate member" org, would serve corporate 
interests primarily, and people's interests secondarily: when in conflict, 
corporate interests would prevail. In the years of W3C's work to date, that 
concern has largely been allayed. Although I represented my corporation's 
interests on the WGs to which I contributed, I recognized that the most 
accessible open standards formed the platform on which we could profit, and 
apparently the extremely competitive member corporations of W3C realized 
that, too.

         The current issue, allowing patented technology to be certified 
"standard" by W3C, lies at the core of that concern. A patent creates 
*less* access to technology, however valuable the technology. This 
restriction defeats the core value of a standard: everyone can use it. So 
W3C becomes a marketing organization for the patentholding organization.

         I quote a message from John Nagle, with whom I agree. I especially 
emphasize his remarks on the GIF format, which is a case study in the 
limits of a patented standard, and underscores the requirement that 
property restrictions be disclosed and avoided before a standard is codified:

" Almost the only patent now that affects most standard web content is 
#4,558,302, the "GIF patent". And that expires on December 10, 2002.

  I own and make money from software patents. But not from the Web, even 
though my unpatented "Nagle algorithm" is in every TCP implementation in 
the world. Patented technology should not be in standardized interfaces. It 
inhibits their adoption and causes the technology base to fork. That makes 
the "standard" non-standard."

- John Nagle
- Menlo Park, CA




--

(C) Matthew Rubenstein

Received on Tuesday, 2 October 2001 13:15:41 UTC