Subject: RE: License Discussion for the Broad Institute Public License (BIPL)
From: "Rory Pheiffer" <rpp@MIT.EDU>
Date: Thu, 11 May 2006 08:58:18 -0400

Hi Brendan,

Thank you for your inquiry.  I understand your concern and hope that I can
clear this issue up some for you.

MIT owns patents for the sole purpose of encouraging commercial development
of the technology through licensing.  The issue for MIT is one of control
over enforcement of the patents.  MIT owns over 2000 patents in the US alone
and many more worldwide.  Most of those patents are licensed to companies or
are intended to be licensed to companies.

The issue for MIT is in the instance where MIT has exclusively licensed
patents that may cover software that we release under the Broad Institute
Public License.  MIT would not knowingly choose to patent an algorithm or
method implemented by open source software and then knowingly withhold
rights to that patent.  The issue is where claims in patents not obviously
or easily identified as covering the software do, in fact, cover the
software.  MIT may or may not be able to control enforcement of those patent
claims if they have been exclusively licensed to a company.  Because of the
size of MIT's patent portfolio, it is impossible for MIT to review all of
its patents, even in a cursory fashion to determine which likely cover any
particular software.  Thus, we can not offer a broad-based license to any
patent that is owned by MIT that covers the use of the software licensed
under the BIPL.  Whether or not granting the right to use the software
creates an implied patent license is still an open question of law and not
one that we choose to address.

In the event, MIT files a patent to specifically cover a particular piece of
software, MIT does offer licenses of that known patent with its software
licenses.  However, MIT is not willing to offer any representations or
warranties relating to these facts or situations because of potential
liability to MIT, which is a non-profit research and educational institution
that provides significant research and educational resources to the world
for free.

I hope this helps you understand why MIT found it necessary to create the
BIPL and make sure that MIT was not granting rights to its patent portfolio


-----Original Message-----
From: Brendan Scott [] 
Sent: Thursday, May 11, 2006 7:37 AM
To: Rory Pheiffer
Cc:; 'Karin Rivard'; 'Jill P. Mesirov';
'Michael Reich'
Subject: Re: License Discussion for the Broad Institute Public License

Rory Pheiffer wrote:
> Hello,
> I am writing to begin a discussion about the license we (the Massachusetts
> Institute of Technology or MIT for short) recently submitted to the Open
> Source Initiative for approval as an OSI Certified license.  The title of
> the license is the Broad (pronounced Brr-ode as opposed to Brr-awed)
> Institute Public License, or BIPL for short.  Below please find the three
> sections which your group has asked me to address:

> The most significant reason why we could not use the MPL is because MIT
> cannot agree to grant a license to any MIT-owned patent that may end up
> being associated with the software.  This is because it is virtually
> impossible for MIT to know if any of its many professors, researchers,
> investigators, or students have, are, or will do work that gives rise to a
> patent claim that covers use of its software under the BIPL.  MIT cannot
> grant rights to these patent claims because the professors, researchers,
> investigators, or students involved with the patent claim would not have
> opportunity to make decisions regarding the licensing of these patent
> if a blind patent grant was part of a license that we distribute under.
> patent grant in the MPL is an encumbrance that MIT cannot agree to give.

Maybe I am not understanding what you are concerned about, but...

The wording of the MPL aside, if MIT owns the patents, on what basis can it
deny a licence over them?  That is, if use of the software necessarily
requires an exercise of the patent, and MIT grants a licence to use the
software, how can you maintain that there is no licence over the patent?  

(and conversely, if MIT does not own the patents, why would the licence be
of concern?)