patent policy comments

I found the call for comments only today, and I wanted to contribute.

==> Please note that I am not representing my employer. <==
==> These are my personal comments and observations. <==

The rules on requiring disclosure of standards-related patents 
seem good on quick reading, and something like that should probably
be adopted.

However, I believe that allowing patents that are available under a
RAND license into W3C standards would be a grave mistake.  The W3C has
a number of very powerful, large companies with comprehensive patent
portfolios.  Furthermore, adoption of a patented invention under RAND
terms into a W3C standard could be very lucrative.  The proposed change
would create a strong incentive for companies to cooperate to put each
other's patented inventions into W3C standards, creating a steady
revenue stream to commercial members to the detriment of Web users.
The proposed rules do not guard against that, and I find it difficult
to imagine rules that would.

The worst part of the RAND license is that it may impose payments;
many Internet standards have been pioneered in free, open source
implementations, often created by academics or research labs.

I also believe that there is no need for such a change: there are very
few patents that cannot be worked around fairly easily if they are known
at the time of standards adoption.  Maybe the cost is a few percent in
efficiency or slightly less functionality, but most users and implementors
prefer a simpler standard anyway.  Furthermore, if this kind of enhanced
functionality is worth it to the market, the patent holders can still
build a business around it; I suspect in most cases, they would still
choose to make the invention available for free and gain commercial
advantage from their expertise and additional patents in the space.

The policy might be workable if the RAND license exempts software
distributed under "free" software licenses, like BSD, GPL, or LGPL;
i.e., if payment is only required from individuals or companies that
incorporate the invention into software or hardware that is commercially
sold or licensed.  I would still view this as a suboptimal compromise,

What would happen if the W3C adopted this change?  You can expect that
every standard that includes a patented invention licensed under RAND
would likely soon face an open source, free equivalent that works around
the patent and does not infringe on its claims; the free equivalent may
end up being somewhat less functional.  We have seen the result of this
kind of competition time and again: the free equivalent becomes nearly
universal, even if it is slightly less functional.

Thomas M. Breuel

Received on Sunday, 30 September 2001 19:11:35 UTC