Comment Submission for the W3C's Patent Policy Group

(Please accept this submission if you are using a UTC deadline; your submission address in your FAQ online at http://www.w3.org/2001/08/16-PP-FAQ.html includes an extra "s" at the end of your list name in Section 0-4.)


   Dear Sirs and Madams:

   I am an Electrical Engineer who has worked for the commercial market, academia, and government.  I am a member of ICANN's At-Large group as well as the Institute of Electrical and Electronics Engineers.  The patent proposal in question, while possibly beneficial, poses more questions than answers.

   As the patent policy working group's FAQ document notes, there are a variety of patent standards used around the world.  Items may be patented in one country but not another.  The patent policy group's mission (and it is a reasonable one) is to determine the cases when patented material is allowed to be used in W3C standards.  But allowing patent holders to make demands of standards groups for their property is an unreasonable proposition.

   The working group's policy as currently described could require payments to patent holders when no royalties are legally required.  A case in point would be the RSA encryption algorithm, patented in the United States until recently.  While U.S. developers were unable to incorporate RSA support into Internet browsers without payment to RSA Inc., foreign implementations of RSA not requiring royalties were plentiful and in some cases free.  The RAND license does not seem to take into account patent's expiration; royalties collected under it (if allowed) must stop once no longer required.

   The main patent area of the W3C's concern, algorithm patents, are only recognized in a few countries.  Again the RSA algorithm is cited as an example of this.  The vast majority of non-U.S. developers and software users are not held accountable to these patents.  As an international group the W3C should not feel the need to cater to the few countries that do allow these.

   Allowing patented algorithms to be used in standards only encourages speculative patenting of potential next moves.  Rambus and their memory patents should be on top of everyone's mind when it comes to standards groups and patents.  One can find alternatives to patented material if one chooses to look.  Even GIF image developers (once dependent on the U.S. patented LZW compression algorithm) have found patent-free ways of encoding their works.

   While I am by no means an open-source software fanatic, the RAND and RF licenses run contrary to attempts by these projects to synchronize their licenses and interchange formats.  For years the W3C has been attempted to do *exactly* that by synchronizing how academic, commercial, government, and personally designed programs and devices communicated.  Mind you, I have no problems with patents being licensed free of charge and free of restrictions into W3C standards; W3C just should not put any usage limitations on their implementers.

   Standards groups work best when anyone and everyone can implement a standard regardless of race, background, ability to pay, or third-world nation status.  Requiring payments for patents or other patent-based restrictions defined by the vague term "reasonable" runs contrary to these goals.

   While I admire the W3C's attempt to accommodate legal concepts such as choice of venue on behalf of patent holders, the W3C must be aware that there are countries which simply would toss those provisions out.  Given such a case, would the W3C be willing to ban something like HTML 6.0 from usage in parts of the world?

   Sincerely,
   Samuel Greenfeld


   NOTE: These views are my own, and should not be considered those of any other entity mentioned above.  All trademarks are their owners.  Would putting a visible statement like this in every HTML page and XML document be considered a "reasonable" patent license?

Received on Sunday, 30 September 2001 20:10:32 UTC