Zolera patent policy comment

First, Zolera Systems appreciates the extension to the review period and the
publication of the clarifying response to public comments released by the
W3C. We also appreciate the effort of the patent policy group to address
conflicting patent concerns in a clear and open manner.

The patent policy frameworks requirement approach of defining a licensing
mode at the start of an activity and using a patent action group (PAG) to
resolve issues is good. Rechartering with Advisory Committee (AC) and
Director approval makes sense for unresolved issues or licensing mode
changes. We propose that changing from reasonable and non-discriminatory
(RAND) to royalty-free (RF) would not require a recharter in the event of no
known patents. This is unlikely, however.

We believe all workgroups should default to RF and that there has to be a
very good reason to charter a RAND group, such as  a known patent in
conjunction with a compelling activity. We agree with the public comment
that RF is preferred and essential for core standards. Making the
distinction between layers will not help as higher layers become core (for
example application protocols such as SOAP are "higher" than IP, but
apparently will be essential to web services). We would like to see RF be
the norm, by requiring all work groups to default to RF. To make this more
effective than a desire, we recommend that in order to start a RAND work
group, it require a 3/4 favorable AC vote and Director approval. (The point
is to make it hard to create a RAND group, but not impossible) We do not
want the W3C to encourage patent land grabs - the incentive should be to
make necessary IP RF to create a standard.

We are concerned that the administrative cost and overhead of managing RAND
licenses will be burdensome for small companies, especially since the terms
are not made known as part of the W3C process. How are terms made
non-discriminatory if not made widely known? We suggest that RAND-mode
recommendations include the detailed RAND terms and payment instructions as
part of the last call recommendation, to minimize the costs of compliance.

We note that the RF terms include the ability to require a grant of a
reciprocal RAND license to all essential claims (#4). Shouldn't the RF
license modify this to be a reciprocal RF license? We think so (Sections 4e,
f)

We support the second formal objection, since late opt-out can derail a
recommendation's value. We would like to modify the objection to not include
"unlimited defensive use of Essential Claims", since this is very broad.

We do NOT support the first formal objection. Reviewing a last call document
to see if one has patents which impact it does not seem onerous, as long as
members are not expected to track the details of all working groups. It is
beneficial to the community to surface and address patent issues of
non-working group members.

We note that the patent policy document will not solve all problems. The
success of the W3C still requires good faith on the part of members for
disclosure and good faith on PAG members not to derail the process.

---
Frederick Hirsch
Zolera Systems, http://www.zolera.com/
Information Integrity, XML Security