Subject: RE: For Approval: Broad Institute Public License (BIPL)
From: Karin Rivard <rivard@MIT.EDU>
Date: Fri, 14 Jul 2006 12:27:23 -0400
Fri, 14 Jul 2006 12:27:23 -0400
Larry,
Thank you for this.  A few clarifications:

At 11:46 AM 7/14/2006, Lawrence Rosen wrote:
>You have announced to everyone here that MIT (probably) has patents 
>that you won't license with your software, and that the university 
>or others may come after users if we use your software. If you leave 
>out a patent grant entirely, that will make not only your license, 
>but your software, very risky. I am aware of no modern, large open 
>source project that would accept your contributions under such terms.

MIT has not announced that, on its own, it would pursue users of the 
open source code in order to enforce patents against users of the 
open source code.  MIT's position is that we have not knowingly filed 
any patents that cover the use of the software to be implemented 
under the BIPL.  However, MIT has over 2000 patents in its portfolio, 
many of which have been licensed to third parties exclusively.  Our 
exclusive license enables the exclusive licensee to name MIT as a 
party to an infringement lawsuit as the patent owner--this may only 
be done if legally necessary in pursuing the infringer and only after 
discussion (but not permission) from MIT.  Because of wonderfully 
creative and appropriate patent claim interpretations, it is possible 
that patents we never anticipated covering use of the open source 
code could, in fact, have claims that cover use of the open source 
code.  Although MIT would try to discourage infringement suits 
against open source users, ultimately, we are bound by our 
contracts.  MIT is unable to grant the broad based background patent 
license because it may very well cause MIT to, unknowingly, be in 
conflict with existing license grants.
>
>I also want to point out that MIT's situation as a large, diverse 
>institution where no one person knows everything that others are 
>inventing, is not that unique. IBM, Sun, HP, Apple, etc., are also 
>large institutions, and they all have dealt with this problem in 
>some fashion. I also note that W3C, hosted at MIT and with close 
>ties to the inventive community at MIT, has also solved this problem 
>in its patent policy, not by ignoring patents, but by committing its 
>members to license patents when necessary to practice industry standards.

A university licensing program is significantly different than IBM, 
Sun, HP etc.  Corporations do not relinquish control over pursuing 
infringers as part of their patent license agreements.  MIT does 
relinquish such control.  You may choose to disagree with this MIT 
position, but that is the position and that is the term that is in 
our exclusive licenses.  Thus, our hands are tied--we do not control 
our exclusively licensed patents nor are we in a position to procure 
extensive infringement or non-infringement opinions on our unlicensed 
patents prior to licensing them.  It simply is not practical nor 
cost-effective for our Institution.  Corporations retain control over 
infringement suits and thus do not need to perform infringement 
analyses on all of their patents.

I recognize that other universities are willing to offer patent 
grants in open source licenses despite granting to their exclusive 
licensees the rights that MIT grants.  This is an independent risk 
assessment and policy decision made by each institution.  MIT has 
made its assessment and will not make the broad based patent grant.

As to W3C, that is a very tricky issue.  MIT has not subjected its 
patent portfolio to the W3C patent policy.  You may contact W3C 
directly, perhaps Daniel Weitzner, for more information on how MIT 
operates in light of the W3C patent policy.

It would be unfortunate if MIT could never receive an OSI approved 
open source license because of this issue, but I understand that OSI 
will apply its standards as it sees fit.  MIT can not change its 
Policy for this one license.
>
>As Russ Nelson suggested, maybe MIT can make a more valiant attempt 
>to solve this patent problem, rather than passing it on to users to 
>deal with in the absence of patent information that only MIT has at 
>its disposal.
>Other universities and their technology licensing departments are 
>trying to address this problem in a more comprehensive way. Perhaps 
>MIT can help us create a better solution than a license that places 
>risky software into the stream of commerce.

Agreed.  But solutions take time and policy changes.  MIT's current 
policy is the result of some past occurrences that were a problem for 
us.  To change it would mean the possibility of the past problem 
reoccurring.  The process for change is being looked into, but it takes time.

Karin
>
>/Larry Rosen
>
>Lawrence Rosen
>Rosenlaw & Einschlag, a technology law firm 
>(<http://www.rosenlaw.com>www.rosenlaw.com)
>Stanford University, Lecturer in Law
>3001 King Ranch Road, Ukiah, CA 95482
>707-485-1242  *  fax: 707-485-1243
>Author of "Open Source Licensing: Software Freedom and
>                 Intellectual Property Law" (Prentice Hall 2004)
>
>----------
>From: Karin Rivard [mailto:rivard@MIT.EDU]
>Sent: Friday, July 14, 2006 7:43 AM
>To: license-discuss@opensource.org
>Cc: Rory Pheiffer
>Subject: Re: For Approval: Broad Institute Public License (BIPL)
>
>Dear BIPL Discussion Group:
>
>I am writing on behalf of MIT.  It's not clear to me if this is how 
>the process works, but the group has raised a few issues on which I 
>would like to comment.
>
>It appears from discussion that there are three concerns raised 
>about the BIPL license:
>
>1.  MIT does not explicitly license MIT-owned patent rights that 
>might cover the open source software.
>2.  The license isn't fair because the BIPL requires "contributors" 
>to license their patent rights that cover their contributions, while 
>MIT does not do the same.
>3.  The license is unlikely to be "used."
>
>Here are my comments:
>
>1.  The requirements for OSI certification do not include a 
>requirement that the originator of the software offer a license to 
>originator owned patents.  As has been pointed out in the discussion 
>group, MIT's position  on not offering a patent license in the BIPL 
>is consistent with the GPL, the BSD license, the MIT license, the 
>Educational Community License, and others.
>
>2.  There is a lack of parity in treatment of the Originator of the 
>code and future contributors to the code.  This is true.  MIT will 
>not offer the patent license; however, the requirement on 
>contributors was an attempt to procure for users as many "freedom to 
>use" rights as possible.  If this disparity in treatment is so 
>abhorrent to OSI, it is easily remedied.  MIT will delete from the 
>BIPL all references to any patent grants from contributors.  Thus 
>the BIPL will simply be another open source license that is silent 
>on patent rights.
>
>3.  I do not understand the last comment from the list.  The 
>software is what is used.  The license is the mechanism by which the 
>software is used.  If no one contributes to the development of the 
>software because they do not like the license terms, that is 
>ok.  The fact remains that the software remains freely and openly 
>available for use by the public, which I thought was the 
>goal.  Further, "use" or "usability" is not one of factors that is 
>stated as a requirement for OSI approval.
>
>General comment:  MIT's BIPL license, as submitted, complies with 
>each and every factor listed on the OSI site for achieving 
>approval.  Nevertheless, if the approval committee demands parity of 
>treatment among MIT and the contributors, MIT will delete all 
>references to patent licenses in the BIPL.  If this remedy is 
>acceptable to OSI in order to achieve approval, please let me know 
>and the change will be made.
>
>Thank you.
>Karin Rivard
>
>__________________________________________________
>Karin K. Rivard, Asst. Director and Counsel
>MIT Technology Licensing Office, Room NE25-230
>Five Cambridge Center, Kendall Square
>Cambridge, MA 02142
>Phone:  (617) 253-6966; Fax: (617) 258-6790
>Email:  rivard@mit.edu

__________________________________________________
Karin K. Rivard, Asst. Director and Counsel
MIT Technology Licensing Office, Room NE25-230
Five Cambridge Center, Kendall Square
Cambridge, MA 02142
Phone:  (617) 253-6966; Fax: (617) 258-6790
Email:  rivard@mit.edu


Larry,
Thank you for this.  A few clarifications:

At 11:46 AM 7/14/2006, Lawrence Rosen wrote:
You have announced to everyone here that MIT (probably) has patents that you won't license with your software, and that the university or others may come after users if we use your software. If you leave out a patent grant entirely, that will make not only your license, but your software, very risky. I am aware of no modern, large open source project that would accept your contributions under such terms.

MIT has not announced that, on its own, it would pursue users of the open source code in order to enforce patents against users of the open source code.  MIT's position is that we have not knowingly filed any patents that cover the use of the software to be implemented under the BIPL.  However, MIT has over 2000 patents in its portfolio, many of which have been licensed to third parties exclusively.  Our exclusive license enables the exclusive licensee to name MIT as a party to an infringement lawsuit as the patent owner--this may only be done if legally necessary in pursuing the infringer and only after discussion (but not permission) from MIT.  Because of wonderfully creative and appropriate patent claim interpretations, it is possible that patents we never anticipated covering use of the open source code could, in fact, have claims that cover use of the open source code.  Although MIT would try to discourage infringement suits against open source users, ultimately, we are bound by our contracts.  MIT is unable to grant the broad based background patent license because it may very well cause MIT to, unknowingly, be in conflict with existing license grants.
 
I also want to point out that MIT's situation as a large, diverse institution where no one person knows everything that others are inventing, is not that unique. IBM, Sun, HP, Apple, etc., are also large institutions, and they all have dealt with this problem in some fashion. I also note that W3C, hosted at MIT and with close ties to the inventive community at MIT, has also solved this problem in its patent policy, not by ignoring patents, but by committing its members to license patents when necessary to practice industry standards.

A university licensing program is significantly different than IBM, Sun, HP etc.  Corporations do not relinquish control over pursuing infringers as part of their patent license agreements.  MIT does relinquish such control.  You may choose to disagree with this MIT position, but that is the position and that is the term that is in our exclusive licenses.  Thus, our hands are tied--we do not control our exclusively licensed patents nor are we in a position to procure extensive infringement or non-infringement opinions on our unlicensed patents prior to licensing them.  It simply is not practical nor cost-effective for our Institution.  Corporations retain control over infringement suits and thus do not need to perform infringement analyses on all of their patents.

I recognize that other universities are willing to offer patent grants in open source licenses despite granting to their exclusive licensees the rights that MIT grants.  This is an independent risk assessment and policy decision made by each institution.  MIT has made its assessment and will not make the broad based patent grant.

As to W3C, that is a very tricky issue.  MIT has not subjected its patent portfolio to the W3C patent policy.  You may contact W3C directly, perhaps Daniel Weitzner, for more information on how MIT operates in light of the W3C patent policy.

It would be unfortunate if MIT could never receive an OSI approved open source license because of this issue, but I understand that OSI will apply its standards as it sees fit.  MIT can not change its Policy for this one license.
 
As Russ Nelson suggested, maybe MIT can make a more valiant attempt to solve this patent problem, rather than passing it on to users to deal with in the absence of patent information that only MIT has at its disposal.

Other universities and their technology licensing departments are trying to address this problem in a more comprehensive way. Perhaps MIT can help us create a better solution than a license that places risky software into the stream of commerce.

Agreed.  But solutions take time and policy changes.  MIT's current policy is the result of some past occurrences that were a problem for us.  To change it would mean the possibility of the past problem reoccurring.  The process for change is being looked into, but it takes time.

Karin
 
/Larry Rosen
 
Lawrence Rosen
Rosenlaw & Einschlag, a technology law firm (www.rosenlaw.com)
Stanford University, Lecturer in Law
3001 King Ranch Road, Ukiah, CA 95482
707-485-1242  *  fax: 707-485-1243
Author of "Open Source Licensing: Software Freedom and
                Intellectual Property Law" (Prentice Hall 2004)

From: Karin Rivard [ mailto:rivard@MIT.EDU]
Sent: Friday, July 14, 2006 7:43 AM
To: license-discuss@opensource.org
Cc: Rory Pheiffer
Subject: Re: For Approval: Broad Institute Public License (BIPL)
 
Dear BIPL Discussion Group:

I am writing on behalf of MIT.  It's not clear to me if this is how the process works, but the group has raised a few issues on which I would like to comment.

It appears from discussion that there are three concerns raised about the BIPL license:

1.  MIT does not explicitly license MIT-owned patent rights that might cover the open source software.
2.  The license isn't fair because the BIPL requires "contributors" to license their patent rights that cover their contributions, while MIT does not do the same.
3.  The license is unlikely to be "used."

Here are my comments:

1.  The requirements for OSI certification do not include a requirement that the originator of the software offer a license to originator owned patents.  As has been pointed out in the discussion group, MIT's position  on not offering a patent license in the BIPL is consistent with the GPL, the BSD license, the MIT license, the Educational Community License, and others.

2.  There is a lack of parity in treatment of the Originator of the code and future contributors to the code.  This is true.  MIT will not offer the patent license; however, the requirement on contributors was an attempt to procure for users as many "freedom to use" rights as possible.  If this disparity in treatment is so abhorrent to OSI, it is easily remedied.  MIT will delete from the BIPL all references to any patent grants from contributors.  Thus the BIPL will simply be another open source license that is silent on patent rights.

3.  I do not understand the last comment from the list.  The software is what is used.  The license is the mechanism by which the software is used.  If no one contributes to the development of the software because they do not like the license terms, that is ok.  The fact remains that the software remains freely and openly available for use by the public, which I thought was the goal.  Further, "use" or "usability" is not one of factors that is stated as a requirement for OSI approval. 

General comment:  MIT's BIPL license, as submitted, complies with each and every factor listed on the OSI site for achieving approval.  Nevertheless, if the approval committee demands parity of treatment among MIT and the contributors, MIT will delete all references to patent licenses in the BIPL.  If this remedy is acceptable to OSI in order to achieve approval, please let me know and the change will be made.

Thank you.
Karin Rivard


Karin K. Rivard, Asst. Director and Counsel
MIT Technology Licensing Office, Room NE25-230
Five Cambridge Center, Kendall Square
Cambridge, MA 02142
Phone:  (617) 253-6966; Fax: (617) 258-6790
Email:  rivard@mit.edu


Karin K. Rivard, Asst. Director and Counsel
MIT Technology Licensing Office, Room NE25-230
Five Cambridge Center, Kendall Square
Cambridge, MA 02142
Phone:  (617) 253-6966; Fax: (617) 258-6790
Email:  rivard@mit.edu