Patent policy comment

If there is one thing to be learned from the success of TCP/IP, the Web,
Internet mail and so on over the past couple of decades, it is that no
standard is likely to succeed unless it has a free implementation.  In
fact I'd go so far as to say that free implementations are more
important than the standards themselves.  It is vital that any standard
can be implemented as widely as possible with the minimum of hassle.

'Non-discriminatory' licensing is a complete misnomer in the case of
software patents.  Such schemes are discriminatory since they make it
impossible for free (open source) software to implement the
functionality which is encumbered.  The idea of RAND licensing might
make sense for physical goods, where there is an inherit marginal cost
in each unit, but not for software.

If the W3C starts producing standards which cannot be implemented
without paying a tax to certain people who hold patents, then these
standards and ultimately the W3C itself will become irrelvant.

The Web's success and the W3C's success are built on standards which are
freely implementable by all and which have reasonable free-software
implementations.  Please do not throw that away by getting caught up in
the current fad for dubious, land-grab software patenting.  A sensible
requirement would be that any W3C standard should be implementable
worldwide in freely distributable and modifiable code.  In other words,
please keep the status quo.

(About me: I am a computer science undergraduate, I recently finished a
six month industrial placement developing Web applications at ArsDigita
so you could count me as a 'developer' also.)

Ed Avis <>
Finger for PGP key

Received on Sunday, 30 September 2001 13:30:20 UTC