Policy is self-contradictory and could sideline W3C

1.  I am wanting to push the notion of a patent cross-licensing agreement
    currently called the Open Patent License (OPL) at www.openpatents.org.
    The agreement would allow participants to create products incorporating 
    submitted patents royalty-free:

    1.  In any code distributable under an Open Source license, or
    2.  In any non-Open Source code if all incorporated patents are
        similarly licensed.

    One of the main goals is to let participants use their own patents
    to solve the problems patents cause, in a way similar to the way
    the GPL turns copyright on itself for free software.
    
    (Granted, I'm biased as to this specific patent license-in-progress.)

    Interestingly, both the definitions for RAND and RF in the proposal at
    http://www.w3.org/TR/2001/WD-patent-policy-20010816/#def-RAND
    preclude the use of such a license.  I consider the OPL a solution
    to the some problems of software patents--a company could conceivably
    want to license its patents under the terms of one of the Options
    of the OPL, effectively allowing what would normally be considered
    RF use.  In fact, if all members decided to adopt something like
    the OPL, not only would that allow their patents to generally be
    used in Open Source works, but it would tend to protect against
    slow hijacking of "additions" to standards much more than would
    complete RF patent licensing terms.

    However, even this more-friendly-than-RAND option is not allowed
    by the "no further conditions" clause.

    (Although RF may seem better in individual cases, a defensive patent
    cross-licensing scheme in which use is royalty-free only when no
    non-submitted patents are incorporated may be a better long-term
    solution for making sure patent encumbrances don't become added
    as common-additions-to-standards one-by-one by various companies,
    slowly slowing down progress.  In a defensive patent cross-licensing
    scheme, the addition of any non-submitted patent invalidates the
    ability to use the submitted patents in a royalty-free way, thus
    causing the defector to have to pay "full price", as opposed to
    being able to take advantage of the the royalty-free agreements
    everyone "playing nice" has decided to do.)

    (As a side note, and at the risk of flame wars, a OPL-type
    versus RF-type patent license differences are vaguely similar
    to GPL-type versus BSD-type software license differences.
    IMHO the arguments for each mostly transfer to the patent
    domain.  The problem here is that the RAND license precludes
    some things that would be needed in an OPL-type license.)

2.  According to http://www.w3.org/2001/10/patent-response ,
    One of the "goals of the proposed policy is to ensure that:

         The Web community is not surprised by "submarine" patents
	 whereby unsuspecting participants are forced to pay
	 license fees after their participation in the creation
	 of a Recommendation that they thought was unencumbered."

    Yet, section 8 of the proposed policy, "Member Patent Licensing
    Commitments", merely requires RAND licensing of all non-disclosed
    patent.  It does not require royalty-free licensing.

    The proposal at http://www.w3.org/TR/2001/WD-patent-policy-20010816/
    does not meet the stated goal above. 

    Proposal for modifying Section 8:
    ---------------------------------

    Section 8 should require mandatory RF licensing of these non-disclosed
    patents for the different conditions under which it currently suggests
    merely requiring RAND-type licensing.

    Alternate proposal for modifying Section 8:
    -------------------------------------------

    With the current W3C proposal, a member organization whose patents are
    required for a given standard that had not disclosed the existence
    of those patents WILL STILL BENEFIT FROM THAT NONDISCLOSURE TO THE
    DETRIMENT OF ALL OTHER IMPLEMENTORS AND USERS OF THE STANDARD.

    If the existence of mandatory RF licensing for these submarine cases
    is not acceptable to some parties, then I would suggest an option
    of some type of royalty-free defensive patent cross-licensing agreement
    as I described above that the patent holder be permitted to chose
    instead.

    Note that I am not after "punishing" the member organization that
    possesses the non-disclosed patents.  I am quite happy for them to
    benefit from standards they helped create--that's one of the main
    reasons to participate in the creation of standards after all.
    However, they should simply not gain unfair advantage from having
    a non-disclosed monopoly become part of the standard.  (The question
    of whether the nondisclosure was intentional or not is irrelevant
    in my mind.)

    A mandatory RF patent-licensing requirement would erase the member's
    own monopoly benefit, but a royalty-free patent cross-licensing
    agreement of the style above would not only erase that particular
    unfair monopoly benefit, it would also create a poison pill for
    3rd parties.  3rd parties not agreeing to this cross-licensing
    agreement would then not benefit from the royalty-free licensing
    the agreement provides.  This defecting 3rd-party would then be
    required to pay royalties to all of the patent holders who are
    "playing nice".

    3rd parties who decide to "play nice" would find it cheap
    to do so, while 3rd parties who've decided to encumbering others
    restrictive, (even RAND-type) patent licenses would find it more
    expensive.

    It would be pleasant to think that W3C members not disclosing
    their patents would be allowed to license them in such a way
    as to in effect require 3rd parties to "play nice".

3.  As to the acceptability of RAND terms for W3C standards, looking at
    "About the World Wide Web Consortium (W3C)" document at
    http://www.w3.org/Consortium/ shows how requiring patents incorporated
    in its standards to be merely available under RAND terms are
    incompatible with the W3C as an organization:

    3a.  W3C creation:
    ------------------
         "The World Wide Web Consortium was created in October 1994
	 to lead the World Wide Web to its full potential by developing
	 common protocols that promote its evolution and ensure its
	 interoperability."

    Potential interoperability is severely limited if the creation of
    inter operable code is limited by requiring royalties payments
    for patents.

    3b.  W3C Goals--Universal Access:
    ---------------------------------

         "W3C's Goals

	 W3C's long term goals for the Web are:

	      1.  Universal Access: To make the Web accessible to all
		  by promoting technologies that take into account the
		  vast differences in culture, education, ability,
		  material resources, and physical limitations
		  of users on all continents."

    The principle of Universal access is incompatible with RAND,
    royalty-requiring patent-licensing terms.

    3c.  W3C Design Principle--Decentralization:
    --------------------------------------------

         Design Principles of the Web
     
              3.  Decentralization: Decentralization is without a doubt
	          the newest principle and most difficult to apply.
	          To allow the Web to "scale" to worldwide proportions
	          while resisting errors and breakdowns, the architecture
	          (like the Internet) must limit or eliminate dependencies
	          on central registries. 

    Patents offices themselves, being central registries of local
    restrictions, are just as limiting to the web as the types
    of registries normally associated with such a decentralization
    principle.  Patents can create localized legal choke-points
    of innovation that causing scaling problems in technical
    solutions that are just as badly as scaling problems that
    have technological causes.

    3d.  W3C Role--IP Licensing differences between industries:
    -----------------------------------------------------------

         "W3C's Role

         As with many other information technologies, in particular
	 those that owe their success to the rise of the Internet,
	 the Web must evolve at a pace unrivaled in other industries.
	 Almost no time is required to turn a bright idea into a new
	 product or service and make it available on the Web to the
	 entire world;"

    There is a large consensus in the other responses to this proposal
    that one of the main reasons for the "unrivaled pace" that the web
    has evolved at is precisely the absence of RAND-type terms for 
    patent licensing.  In fact the proposal itself, at 
    http://www.w3.org/TR/2001/WD-patent-policy-20010816/ , mentions
    the inherent compatibility of an unrivaled pace of evolution with
    RF (not RAND) licensing terms in the introduction to section 2:
    
         Both the competitive forces which have lead to innovative
	 technology, and the cooperative spirit which has produced
	 global interoperability standards at an extremely rapid pace
	 have occurred, until very recently, in a market environment
	 without any significant intellectual property licensing
	 requirements. In contrast to other network industries
	 such as telecommunications or transportation, innovation
	 has occurred without recourse to patent rights to protect
	 investment in research and development.

    But then in section 2.1:

         as the Web comes into contact with the telecommunications,
         broadcast media and consumer electronics industries, the
	 tradition of patenting technology from those industries
	 will likely be carried over to the Web. 

    That begs the question.  Given that:
    
    1.  The web has evolved and improved much faster than other
        industries because of the virtual absence of patent
	encumbrances.

    2.  the web will be coming into contact with industries
        in which patent encumbrances abound, and that improve
	and evolve much more slowly because of this,

    why should we throw up our hands and decide as a group
    to limit ourselves to the pace of innovation that these
    other industries have subjected themselves to?  Wouldn't
    it make more sense to retain our principles while opening
    up these other industries to the possibilities of the type
    of faster evolution that the web and the Internet as a whole
    has seen?

    We should view the increasing contact of the web with other
    industries as an opportunity to help them become more
    innovative, not as an opportunity to give ourselves virtual
    lobotomies via non-RF IP licensing.

It's a great idea for the W3C to clarify its patent policy.
However, I would prefer the clarified policy to require
patents to be made available on a RF basis, perhaps with
something like the Open Patent License I'm wanting to
promote being acceptable, but in any event allowing
royalties to be required is not only at odds with the
W3Cs goal's, it will slow down future web development.

(Rather it would, if the community continued to follow
W3C recommendations.  In all likelihood, this policy would
probably result in the W3C being side lined.)

Section 4(e)5 should be amended to disallow royalty
fees, and section 4(e)6 should possibly be amended to allow
for something like the Open Patent License, (these would be
small changes).

In no event should an implementor of a W3C standard be
required to pay royalties.  In no event should a writer
of Open Source code be required to pay for permission
to write an implementation of a standard.

 -Mark Shewmaker
  mark@openpatents.org

Received on Friday, 12 October 2001 00:01:48 UTC