Subject: RE: Patent-based dual-licensing open source business model
From: <stephen@xemacs.org>
Date: Tue, 26 Sep 2006 15:13:00 +0900

Brian Behlendorf writes:

 > Larry's proposed business model is that I am compelled to release
 > as Open Source the source code to my software that implements the
 > patent.

[Aside: there is no "compulsion" involved.]

IUI differently.  AIUI, Larry's proposed business model is that you
may, if you like, (a) restrict your commercial practice of the patent
to distribution of free software that implements it (including not
using it at all as an extreme case), or, again if you like, (b) pay
mutually agreeable royalties for a mutually agreed license, or (c)
unilaterally practice the patent in some other way at the risk of the
usual legal penalities for such practice.

Under (a), you may use that free software *by itself*, but this
presumably doesn't help if it interacts with any undistributed or
proprietary systems that are part of your business.  If your practice
interacts with any such system that is not pure free software, then
you must pay royalties.

Larry (and Prof. Cameron) has made few statements I'm aware of that
clarify the meaning of "by itself" and/or "interacts with systems that
are not pure free software."  The ones I've seen are not encouraging
for FS*B* purposes.  For example, distributing it preloaded for
convenient use on a hardware platform ("mere aggregation" in GPL
terms) costs you your exemption AIUI (at least, Larry hasn't
explicitly denied it), because that is a value-added service not part
of "distributing free software".  Distributing it with a proprietary
operating system is the same by the obvious isomorphism.  I would
assume that use of a proprietary plug-in with software that otherwise
qualifies for the Convenant's exemption also disqualifies you.

My understanding of the intent of the covenant is that it (1)
encourages the propagation of the *knowledge* encapsulated in the
patent, (2) permits usage by entities whose customerhood is more
potential than real because of the transactions costs involved, while
(3) giving up almost none of the real profit potential.

Again AIUI, the point of asking on FSB is to find out if (1) and (2)
are actually totally ineffective, and whether there are (4)
significant potential commercial uses by FSBs under choice (a) above.

Am I close, Larry?

Regards,
Steve