Comment on proposed patent policy change

I wish to oppose the adoption of any policy which might cause W3C
recommendations to require technology for whose use royalties must be
paid. I oppose such a policy regardless of the "reasonableness" or "lack
of discrimination" of the licensing process. Indeed, I would be
concerned about any system which might require an implementor even to
register its use of a technology. My reasons are as follows--

1. The administrative burden of seeking and maintaining licenses
   would inhibit implementors. Such clerical effort can be
   prohibitive for a small enterprise. Even in large enterprises,
   it tends to slow adoption by discouraging potential project
   advocates, and to damage innovation by adding to the burden
   on projects which might wish to extend the standard or combine
   it with other technology. Likewise, development for in-house
   use becomes impractical when patents must be licensed.

   The licensing burden grows over time; the more standards a
   product or suite of products supports, the more patents
   have to be licensed.

2. The monetary costs of licensing have the same bad effects,
   but perhaps even more skewed against small, poorly funded
   implementors.

3. The need to seek licenses would, of course, completely
   prevent many open source implementations. Since open
   source has been a major source of innovation in the Internet
   and the Worldwide Web, and it would seem unwise to kill
   that particular golden goose. This is especially true because,
   in the past, open source innovations have perhaps been
   better for standards than closed innovations... they have
   been more likely both to maintain backward compatibility
   with existing practice, and to be offered for future
   standardization.
   
4. Many, many patents in the W3C's field of interest cover subject
   matter which is obvious, and which should not justify a patent
   at all. Unfortunately, both because of a lack of qualified
   examiners and because of regulatory capture effects, patent
   offices have been unduly free in issuing such patents. In the
   US, some unfortunate court decisions have reinforced this state
   of affairs. The result is a patent system badly in need of
   legislative and administrative reform, and it is the duty of
   all responsible technologists and technology organizations to
   support such reform. The policy proposed for the W3C does the
   opposite, serving to legitimize patents that should never have
   been issued, and to encourage people to apply for weak patents
   in the hope of being able to "tax" future standards-based
   products.

5. Even "reasonable" licensing has frequently been used as a barrier
   to entry, helping to perpetuate industrial oligarchies. In
   addition, even under RAND rules, licensors have been known
   to act in bad faith, or to practice "hidden" discrimination.
   Even if these abuses are the exception, rather than the rule,
   they are enormously destructive, and it is unwise to invite
   them.

6. The need to license patents may make it very difficult to
   conduct a development project in secret. Consider the company
   which is developing a new product on a tight schedule, but
   needs to license a patent from a competitor before the product
   can be released.

It is true that some standards bodies operate successfully under RAND
rules, and that some standards requiring licensing have been adopted
without apparently serious damage. However, this has happened mostly in
cases, such as consumer electronics or semiconductor manufacture, where
a few large companies with enormous capital investments make essentially
all of the products. In such a situation, patent licensing does not
greatly increase the already large barriers to entry. This does not
describe the environment in which W3C recommendations are used; in
software, patent licensing costs (including administrative costs) may
frequently exceed all other costs involved in developing a product.

It is extremely rare that a patent covers every possible way to
implement really important functionality. It is therefore usually
unnecessary to standardize around a patented method. In the very
uncommon cases where a patent effectively covers all ways of doing
something extremely useful, the patent holder is in a position to
dictate the terms under which her technology is to be used, and has no
incentive to agree to RAND terms, or indeed to work through a standards
body at all. In such a case, the W3C's position becomes moot.

Please don't saddle us with the need to worry about patents when
deciding whether to implement a W3C standard.

                                -- J. Bashinski

Received on Saturday, 29 September 2001 13:55:43 UTC