Comment upon PPWG Royalty-Free Patent Policy (W3C Working Draft 26 February 2002)

Dear Daniel Weitzner and W3C Patent Policy Working Group participants,

Thank you for the significant changes made to the proposed W3C Patent
Policy in the 26 February 2002 working draft:

http://www.w3.org/TR/2002/WD-patent-policy-20020226

I am pleased that the W3C is emphasising the importance of Royalty Free
(RF) standards and has eliminated the Reasonable and Non-Discriminatory
(RAND) track.

While the elimination of this track is welcome I notice that the working
draft continues to provide a mechanism for future RAND based standards
at the W3C. To do so requires a Patent Advisory Group to conclude that
development should proceed at the W3C upon RAND terms. The decisions of
the Advisory Committee and the Director will then determine whether a
RAND based recommendation is accepted:

   5.3 PAG Conclusion

   ...

   4. The specification under development should be produced on RAND
      (reasonable and non-discriminatory) terms, either at W3C or some
      other body. Note that there is not yet any process for developing
      or issuing RAND specifications. Therefore if a PAG makes a
      recommendation to proceed on RAND terms, Advisory Committee
      review and Director's decision will be required. It is also
      possible that a PAG could recommend that the work be taken to
      another organization.


I note from the "Summary of 3 June 2002 Patent Policy Working Group
Teleconference" (sent 11 June 2002) that the W3C now intends to give new
names to "its proposed licensing options":

http://lists.w3.org/Archives/Public/www-patentpolicy-comment/2002Jun/0002.html

   One person suggested that the PPWG give new W3C-specific names to
   its proposed licensing options, as the commonly understood meanings
   of RAND and RF are not what is described, and hence introduce
   confusion. The group agreed.


Since there is to be no RAND track it doesn't appear necessary to give a
new name to such a "licensing option". The implications of RAND
licensing for the World Wide Web are now widely understood (and their
introduction overwhelmingly rejected). If a similar license option for
W3C recommendations reemerges in a different guise then a new education
campaign may be desirable.

                              -----------

Please note that the current RF licensing requirements (clause 3) "may
include reasonable, customary terms relating to operation or maintenance
of the license relationship such as the following: choice of law and
dispute resolution."

Inclusion of such a clause could potentially make a RF license grant
incompatible with the GNU General Public License. I raised this issue in
October last year:

http://lists.w3.org/Archives/Public/www-patentpolicy-comment/2001Oct/0033.html

There is at least one W3C Member who may be delighted by the opportunity
to provide a royalty free license that is acceptable according to W3C
policy but just happens to be incompatible with the GNU GPL.

                              -----------

The phrasing of the disclosure requirements (clause 4) could be
improved. The first sentence tends to indicate that only those who
_received_ the disclosure request have a disclosure requirement:

   Disclosure is required when an Advisory Committee representative
   (AC rep), or any other party in a Member organization who received
   the disclosure request, has actual knowledge of likely essential
   claims with respect to a specification.

But the following sentence makes it clear that in fact any member of the
organisation has a disclosure obligation to the AC rep:

   Anyone in a Member organization who has such knowledge must inform
   that AC rep. Where disclosure is required, the AC rep will do so.

The intent of the passage seem to be: (a) to set out when a disclosure
request is deemed to be received (i.e. when any party in a Member
organization receives the disclosure request); and (b) to set out who
must disclose. This may be a clearer formulation:

   Anyone in a Member organization who has actual knowledge of likely
   essential claims with respect to a specification must inform their
   organization's AC rep who will make a disclosure when required.
   The requirement to disclose commences after anyone within the
   organization receives a disclosure request.


4.3 imposes no obligation to perform a patent search or any additional
analysis. While continuing to impose no obligation to perform a patent
search or any additional analysis it would not be onerous to require a
member organisation to at least inform people within that organisation
of the disclosure request (by their standard communications medium such
as email).

And even this minimal requirement would only impact upon those
organisations unwilling to make a RF licensing commitment (when an
organisation makes a RF commitment there are no disclosure
requirements--refer the paragraph entitled "Exemption for those making a
Royalty-Free licensing commitment").


In 4.6 invited experts have to disclose to the extent of their own
"personal knowledge." Could the PPWG please clarify why the term
"personal knowledge" has been employed in place of the earlier term
"actual knowledge" and whether this imposes a greater or lesser
obligation upon the extent of disclosure by invited experts.

                              -----------

I fail to understand why invited experts who clearly have intimate
knowledge of their field are not entitled to participate in a Patent
Advisory Group (5.1):

   Invited experts are not entitled to participate in the PAG, though
   the PAG may cho[o]se to invite any qualified experts who would be
   able to assist the PAG in its determinations.

[Note the grammatical error]

An invited expert may have knowledge about prior art that would make it
more difficult for an organisation attempting to assert non-RF licensing
from inducing the PAG that a recommendation should proceed upon RAND
terms.

If as I suggest invited experts do become entitled to participate, then
a similar provision that applies to any other interested Member who is
the holder of the Essential Claims (clause 5) should also apply to
invited experts:

   In this case the PAG will be open to any interested Member, though
   the PAG may choose to meet without the holder of the Essential 
   Claims in question.

                              -----------

Finally, the definition of Essential Claims sets out a number of
expressed exclusions (that are never deemed to be Essential Claims and
thus fall outside of any royalty free license grant). One of these
expressed exclusions is "enabling technologies" and includes by way of
example compiler technology, object-oriented technology and basic
operating system technology.

If a recommendation results in it being necessary to license one of
these particular enabling SOFTWARE technologies on non-RF terms then it
may not be possible to implementation the RF recommendation using free
software.

It appears that the expressed exclusions from Essential Claims are too
wide and should only include enabling HARDWARE technologies (like the
semiconductor manufacturing technology example).

A recommendation that necessitated the use of a proprietary compiler
employing a patented technique that may only be implemented on fee based
terms does not appear to be a common sense example of a royalty free W3C
recommendation. Yet it appears it could be classified as a RF
recommendation according to the current definition of Essential Claims.


Thank you for considering my comment.

Regards,
Adam Warner

Received on Wednesday, 12 June 2002 10:27:01 UTC