W3C home > Mailing lists > Public > www-patentpolicy-comment@w3.org > October 2001

Patent policy change

From: John Arras <jra@cs.umd.edu>
Date: Tue, 2 Oct 2001 18:08:28 -0400 (EDT)
To: <www-patentpolicy-comment@w3.org>
cc: <jra@cs.umd.edu>, <jra@math.umd.edu>
Message-ID: <Pine.SOL.4.33.0110021731120.7291-100000@wonka.cs.umd.edu>


I am a mathematician with a strong interest in programming and
computers. I oppose patenting algorithms, no matter how difficult
they are to come up with, no matter if they are wrapped in words
to relate the math to the real world (thereby making them a word
problem), and no matter whether the solution is eventually implemented
using pen and paper or a thousand machines.

I understand that my view is the minority view, and that there are
pressures to allow patented mathematics to enter the standards that
you set for the Web. The point of standards is to allow everyone
to have a common framework that anyone can use to create
things that can interoperate. Even though the government lets
entities own pieces of mathematics, the owners should not be allowed
to press their private mathematics onto a standards body so that
they can then collect a tax from everyone who wants to use that kind of
math as part of the standard.

A standards body that wishes to be the standard for everyone should
not allow its standards to exclude people. There are  many people who do
math for fun or for research, but not for profit.

One change under consideration is standards with RAND licenses.
If a "standard" comes with a pricetag, then anyone who cannot or
will not pay the price for access to that mathematics will be
prevented from using it. This is the case even with a RAND-type
license. Such a license could state that there is an initial
fee to be paid for using this mathematics, and even if the fee
is low ($100US) or so, eventually there may be so many fees that,
at some point, even small-medium businesses may get priced out of
implementing standards. The same applies if there is a per-unit
license, as many people give away things (such as Apache which runs
a large percentage of webservers). If one million copies of Apache
are given away each year, and the total licensing for these
standards comes to $10 per copy (assuming many licenses that will
be released with licensing fees if this proposal goes through),
then how will Apache come up with $10 million for something they
give away free for anyone to use? The only other choice for
them is to stop following the standards, and the Web splits
into many incompatible standards based on whether or not the
implementors can afford the current pricetag for the "latest"

The other option is to move at a slower pace, but only accept
ideas that can be used freely. This would be a true standard,
because anyone can then use these mathematical ideas without
having to pay an entrance fee. In this case, the w3c would
probably be much slower in adopting standards, but it would
be assured that those standards are available for all to use.
That's what a standard should be. It should be something that
anyone can use, so that everyone will use it. If entities want
to collect a fee for using some sort of mathematics, then let
them, but there is no reason to let them press their math onto
everyone else to make everyone else pay a tax for using that
math. If that happens, they get to pick and choose who can use the
standard based on the price of admission, and the standard isn't
a standard anymore, since not everyone can or will use it.

That should be the choice they have to make: charge a price for
their ideas OR get their ideas used by everyone as part of a
standard, but as they say, they can't have their cake and eat it,

And those are your options: keep the standards open so that they
are true standards usable by anyone, thereby (possibly) slowing down
progress of the standards, or allow companies to start charging
admission, and thereby (possibly) speeding up the process of standards
adoption, but then the standards become the standards only for those
who can pay.

John Arras
Received on Tuesday, 2 October 2001 18:08:46 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:06:44 UTC