Concerns over aspects of PPF

Tim, 
Patent Policy Public Comments list:

I have concerns over aspects of the Patent Policy Framework (PPF), and
would like them recorded. I have implmented an earlier Working Draft of
the PPF, while progressing SMIL Animation (where I was staff contact)
and SVG (where I was chair) towards Recommendation. 

In both cases there was a Patent Advisory Group (PAG) formed and in both
cases, the PAG called for patents and licensing terms from all members
of the relevant working group who were in good standing. These named
patents and teh licensing terms for each W3XC Member were made
publically available, in accordance with the PPF. This was good, and
helped clarify what claims might be attached to these specifications.

One aspect of the PPF is early disclosure of patent claims. This is
good. In the case of the SVG working group, Apple disclosed a patent
which might have applied, early in the work of the working group. Since
this was well before the first draft of the PPF was released, Apple is
to be commended on their foresight in doing this. The SVG WG examined
the patent, which relates to a system of multiple alpha channels (for
example an RGB image would have three alpha channels) and took steps to
ensure that the SVG specification would not infringe this patent
(without taking a position on the validity or otherwise of the patent).

My first concern is this

1) There is no provision in the PPF for a WG to make an official
statement that they have takens steps to avoid a disclosed patent. This
should be part of the Proposed Recommendation materials and, unless the
specific applicability of such a patent is challenged in Member review,
the Directors decision should clearly state that the WG took steps to
avoid that patent and that no-one from the Member review has shown that
they failed. In other words, the best determination from review is that
the specification is believed to have avoided this patent.

Second concern

2) There is no expectation that a Member, having disclosed a patent
early on and the WG having taken steps to avoid it, will at the time of
Proposed Recommendation either 

a) agree that the patent has been avoided
b) disagree but provide a RF license for that patent only
c) disagree and specify the exact fee structure for that patent

Failure to do address these concerns removes any benefit from early
disclosure. It merely spreads fear, uncertainty and doubt in the user
and developer communities, and is directly harmful to W3C's core
mission. 

Concern three:

3) Publication of patent claims is good, but publication without comment
is not; it  provides implicit, W3C-logo-backed endorsement of the patent
claims, which are all presented as if they were valid. There should be a
substantial disclaimer on all patent disclosure pages saying that. The
increased exposure of the patent claim on a very visible and respected
website should not merely be used as free advertizing. It must have some
compensatory drawback, otherwise all companies will simply list any
patent they want to advertize (there being no weeding out of 'clearly
off topic' patents). So for each patent that is listed, the member
should be required to make, at the time of Proposed Recommendation, a
determination of the applicability (or not, see below) and choose one of

a) withdraw the patent as not being relevant to the eventual, finished
specification
b) state that the patent is not essential technology to implement the
specification
c) make no determination of applicability but give a RF license for that
patent used to implement this specification
d) state that is is essential, and that a RF license is given for that
patent used to implement this specification
e) state that it is essential, and 
  i)  describe the exact part of the specification that makes it
essential
  ii) provide the exact terms of the RAND license including fees and
other conditions.

The reason this has to happen at Proposed Recommendation is that the
Membership can then have the option, in case e), say in their review to
return the Proposed Recommendation to Working Draft to remove the part
of the specification that requires the patent. 

In the case of e), the specification should in any case not move to
Recommendation but instead a PAG should be formed. One of the outcomes
of the PAG might be to decide that the specification is still freely
implementable and can proceed to Recommendation unchanged (ie it might
disagree with applicability) in which case the Directors decision should
clearly state this.

I request that the PPF specification be updated to incorporate these
changes, and once done, a mail be sent so that I can review the changes
and either state that 

a) I am satisfied, or that 
b) I do not believe the comments have been addressed.

I oppose the PPF moving to Proposed Recommendation until these issues
have been addressed.

With these changes, the PPF might be a very good tool to help W3C in its
core mission to bring the Web to its full potential, provide universal
accessibility, and so forth by encouraging Royalty Free working groups
and Royalty Free patent licensing. Without them, there is grave risk
that it achieves the exact opposite effect.

-- 
Chris

Received on Monday, 1 October 2001 06:41:05 UTC