Bruce Perens's Journal: Answers to your questions about the W3C patent policy

From: Pierre Machard
> Do you have any reason to think that the position you defend will
> satisfy Patent holders ?

Some of them have threatened to walk off of W3C in response to even so
mild a position as the draft policy. Nobody thinks they'll really do it.
But your question shows the problem: we can't control their behavior
through the standards organization. We need their cooperation to make
this work. Thus, we can't take a draconian position. The only way around
this is to get legislation, which is a worthy but uphill battle.

From: Nick Phillips
> I presume you've seen Rik van Riel's suggestion as posted
> to one of the SPI lists earlier. In fact it wouldn't surprise me if the same
> things triggered his message as yours, either directly or indirectly.

A number of people have suggested defensive patent pools. I think
I remember doing so in a 1997 article in LinuxWorld.com . The three
problems are:

        1. Getting inventions.
        2. Getting money to file for patents. This is both legal fees to
              format the patent claim (which has to be right if you want to
              be enforcible), and filing fees.
        3. Getting money to file lawsuits. If you can't sue, nobody's going
              to be very interested in your claims.

I think that #1 could be handled by the community, #2 could at least be
started with pro-bono assistance from legal and engineering students, etc.
#3 doesn't have to come until later. If you want to run with the project,
please do so.

From: Wouter Verhelst
> It may not be a bad idea to have patent holders turn to a different
> standards body than people that object to software patents. If there are
> expensive 'standards' from one standards body and free standards from
> another, I feel that people would use the free standards, so that the
> patent holders would lose. Even if they have their own standard.

Well, there are about 100 existing organizations they could turn to,
including IETF (which has a joke of a patent policy IMO) and OASIS. It
would be very easy to do. I don't think making them do that would win
us anything.

From: James Antill
From Bruce:
> > The code that makes use of
> > the patented principle must be under the MIT license, which allows a
> > scope-limited patent license. That may be linked into GPL code and
> > distributed.
>
> How does this work?
> Say I have "xmms", which is GPL code that I didn't write ... and I
> want to implement some w3c std. that contains one of these patents. So
> I do the code as an MIT license, but I'm going to have to link it to
> the GPL'd code ... it's going to be a _derivative work_ ... so the
> code is basically GPL, no matter what I put at the top of the file.

The GPL terms on linking are that all parts of the derivative work must be
under a license with _no_additional_restrictions_ on top of those in the
GPL. The GPL does not prevent you from _removing_restrictions_, as long
as you are the copyright holder on the portion of the code in question.

From: Andre Lehovich
> I've been trying to comment on the draft patent policy.
> The link below -- to approve inclusion of my comments in the
> official archive -- doesn't work for me, returing a 404.

It's breaking for everyone, I think. I notified Danny Weitzner, the Patent
Policy Working Group chair.

Attention anyone whose message doesn't appear here: thanks for writing!
As usual, I am buried in mail and stuff to do, so although I read them
all, I can't answer every message individually.

        Thanks

        Bruce Perens