W3C home > Mailing lists > Public > www-patentpolicy-comment@w3.org > October 2001

Re: SVG ad inf.

From: Chris Lilley <chris@w3.org>
Date: Sat, 06 Oct 2001 16:49:06 +0200
Message-ID: <3BBF19E2.B9E7AB14@w3.org>
To: Daniel Phillips <phillips@bonn-fries.net>
CC: F J Franklin <F.J.Franklin@sheffield.ac.uk>, www-patentpolicy-comment@w3.org


Daniel Phillips wrote:
> 
> On October 6, 2001 03:38 pm, F J Franklin wrote:
> > On Sat, 6 Oct 2001, Daniel Phillips wrote:
> > > My interpretation is that, according to this proposal, no Recommendation
> > > would be finalized if it was considered to be subject to claims of a
> > > member's patent for which an RF license has not been offered.
> >
> > If RAND licensing is inevitable, then at a minimum I want to know exactly
> > which parts of the specification are covered, and exactly what the
> > licensing terms are.
> 
> As has been pointed out repeatedly on this list, RAND licensing for web
> standards is not feasible because of the large component of free software and
> volunteer authors in implementing those standards.  This sets the web apart
> from other environments in which RAND has been used successfully.[1]

You miss my point. Firstly, I am calling for this at the Proposed
Recommendation stage. Secondly, just as the PPF makes it explicit at the
start of the work whether it is RF or RAND, my suggested changes would
make it explicit near the endo of the work not just tha there was a
vague "claim' that a patent might apply - that merely spreads FUD to
developers - but an explicit requirement that the area of a spec that is
affected be identified and specific license fees (or none) be
identified.

In other words, instread of the current

Company X declares US patent 123456789 might possibly apply to the FooML
spec

I want to see (one of several things, as noted earlier, but if a company
insists on RAND and does not withdraw the patent as being unrelated and
does not give RF license for that specific patent thus making its
applicability moot)

Company X declares US patent 123456789 applies to section 3.19.2 of
FooML and will license this for 20,000 US dollars per software product
and 10 US dollars per copy of software distributed

Sounds scary? Good. I want it explicit which part of the spec is
affected and exactly what the financial cost is. That will allow

a) public pressure to be applied to company X and a bad public relations
mess for them
b) section 3.19.2 of the spec to be removed, and company X agrees that
the patent no longer applies.

> > However, I am deeply unhappy with the idea that RAND licensing may be
> > inevitable, and I believe that the W3C has the power (though perhaps not
> > the legal authority?) to insist on RF licensing in all open standards.
> >
> > I suspect that the PPF will be ratified regardless of its furious
> > rejection by the open source / free software community. If it is, then I
> > hope that, at *least*, Chris Lilley's amendments are incorporated.
> 
> The way I read it, Chris Lilley's proposal does not permit a recommendation
> to be finalized if it incorporates RAND licensing.
> Perhaps Chris would be willing to clarify that point.

Oh, it does, as long as there is a strikingly good reason for doing so
for that individual case. If for example the specification relates to
hardware (in Europe, patents are allowed on hardware but not on
software) then, since hardware is sold and there is no public domain or
freely-given hardware, if the implementors of the specification agreed
that the fees were acceptable to them then it might still be a good idea
to go ahead. 

Or they might decide that the fees were too high and not to go head and
to pick some non-W3C technology instead, which would waste the time of
everyone on the W3C working group for a couple of years who had helped
develop the spec. Peer pressure can be a wonderful force, sometimes ;-)

It would however force explicit licensing fees to be declared up front
and in advance
so everyone would know exactly the situation. I believe that this is in
full agreement with the policy of upfront and explicit statements of
licensing mode for a specification and for contributors to a
specification.

The basic difference between RF and RAND as defined in the PPF is
license fees. But zero license fees are currently allowed as RAND, which
is confusing.

I think it would be a lot clearer if RAND always meant fees and the
patent and licensing page said exactly what those fees were.

It would also, thinking about further ramifications of this, force a
dilemma on the Open source community because what about this one:

Company Y declares US patent 123456789 applies to section 3.19.2 of
FooML and will license this for 20,000 US dollars per software product
and 10 US dollars per copy of software distributed, unless the software
has an Open Source license in which case all fees are waived.

Great for Open source but a total deterrent for Shareware and for small
businesses.

-- 
Chris
Received on Saturday, 6 October 2001 10:49:10 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 April 2010 00:13:41 GMT