This post is a slightly-edited trip report I wrote following my participation in the Licensing and Policy Summit for Software Sharing in Higher Education conference in Indianapolis in mid-October. I’ll speak more about it in future posts.
Dan Greenstein, the co-chair, from the University of Caliifornia was unable to attend, so Bradley Wheeler, CIO of the University of Indiana (all campuses) led the conference.
I confirmed with Brad that all the conference discussions are open. I took almost 20 pages of notes, and will post them to the conference portal soon. Brad volunteered the services of Tina Howard, a technical writer on the staff of IU, to take notes and help organize the conference discussions. I offered my help to Tina in answering any questions she might have about open-source, as she is new to this area. I also took along a digital camera, and handed it to Brad Thursday after lunch, suggesting he might want to take some pictures. He did; I will post them to flickr shortly and then post the URL’s to the portal.
There are two key open-source projects in this area currently underway. Both are enterprise-level applications: Sakai for course management, and Kuali for managing a university’s financial operations.
There were only two commercial firms represented: IBM and rsmart.com, a firm that provides services and support for Sakai and Kuali, and which is also is playing a key role in development. Key figures here are Chris Coppola from rsmart (Kuali) and Joseph Hardin from University of Michigan (Sakai).
There was representation from the following universities: Indiana University (all campuses), University of California (all campuses), MIT, University of Michigan (Paul Courant, former provost was present along with Joe Hardin), Cornell and Penn State. Over a third of the attendees were attorneys. There were also many folks involved in technology transfer and university operations.
Cliff Schmidt of the Apache Software Foundation was present Wednesday night and all day Thursday. His input was greatly appreciated.
There was significant representation from outside the U.S.:
Randy Smith from the Tech Transfer Dept. of the University of British Columbia;
Naomi Korn and Ralph Weeden (counsel) from JISC, http://www.jisc.ac.uk/ , from a group in the U.K. that in their words, “provide world-class leadership in the innovative use of Information and Communications Technology to support education and research;”;
Malcolm Bain from Legistics, a law firm in Barcelona that is currently providing guidance to a consortium of universities in Catalonia that is developing a system that will support 10,000+ concurrent users;
John Norman from Cambridge University;
Robin Stanton, CIO of Australian National University (and former department chair of Andrew Tridgell, author of Samba and a current member of IBM’s Linux Technology Center);
Randy Metcalfe from OSS-Watch, http://www.oss-watch.ac.uk/ . In their words, “OSS Watch promotes awareness and understanding of the legal, social, technical and economic issues that arise when educational institutions engage with free and open source software. It does this by providing unbiased advice and guidance to UK higher and further education.”
I individually spoke during the conference with all the folks named above, as well as most of the other attendees.
Present also were folks from the Mellon Foundation and Ithaka.org. Mellon has provided key funding for the projects as well as funding for this conference (including paying for all the meals). Mellon is very sophisticated in their operations. They don’t just give way money but try to do so strategically, providing initial funding so projects can become self-sustaining. Ithaka specializes in services in this area. For example, Matthew Rascoff and Paul Courant wrote much of a very well-done report on open-source and academia available from ithaka; Barnaby Gibson, counsel to ithaka, was a very active participant in the legal discussions. He also serves as pro bono counsel to the Sakai Foundation. Mattthew, Barnaby and I were of the same flight home to New York, and I spoke with them while waiting for our delayed-flight to take off (Tip to all for the future; avoid Northwest Airlines if at all possible. My original flight to Indiana was canceled, and my returning flight was delayed. A flight to Boston that was supposed to take off about the time of the flight back to LGA was also canceled.)
Someone from a major university approached me and said they were in interested in using IBM’s own “dogear” technology. I mentioned that while Dogear is very good, it provides a tagging function very similar to that offered by http://del.icio.us . He said he knew that. They have an a university portal and he had suggested folks use del.icio.us to record their bookmarks, but many were hesitant to do since the bookmarks were public, while Dogear could be used to keep the bookmarks private inside their firewall. I said I would put him in touch with someone in IBM who could help
Most of the conference was about very specific issues related to Contributor License Agreements (CLA’s) and providing some statements about patents in the open-source license for Sakai and Kuali.
The CLA used to date has been that used by Apache It has been accepted by about 6 universities, including Indiana and Stanford, but both UC (University of California, all campuses) and MIT requested that additional language be included before they would approve. These changes have been accepted. The language is similar, and relates to the following problem.
Existing CLA’s (in my view) are directed towards two groups: the open-source developers who write and maintain the code, and the corporations who contribute open-source code or use open-source code in their products (IBM does both). The developers just want some assurance they won’t be sued. The companies want to be comfortable using open-source code in their products. The open-source folks recognize the need for a more rigorous review process; that is a price they must pay to get more corporations to use their code and also to get those corporations to help contribute their own developers to repair and refine the code.
Corporations own the IP of their employees. But universities are in a different situation. For example, in some (most?) universities the situation is as follows:
a) Undergraduates own their work;
b) Graduate students who are paying for their own education own their work;
c) Graduate research assistants don’t own their work; some universities require they sign an agreement to this effect, as does IBM with each of its employees;
d) Faculty have a more complex relationship. Indeed this is the main concern.
Faculty by tradition and also in most cases “by policy” have their own rights to their work. For example, an English professor who writes a book gets the proceeds of its sales. EE and Computer Science professors have asked for similar rights for the code they write; and in some cases have gotten the same rights. “Policy” is especially tricky in that university policy is “black letter policy,” based on explicit documents that in some cases are over a hundred years old. Revising these documents is a very non-trivial process, one not to be undertaken lightly.
All this is made more complicated by the special value associated with patents in the area of genetics, pharmaceuticals and bioengineering. And it gets even more complicated because universities are more and more focusing on the money to be made via technology transfer, and so some are less willing to make their work available in as open a form as has been the norm in the past.
When a professor gets a patent, a university typically has the first right to license that patent, sharing the proceeds with the inventor. However, if the university passes, then all the patent rights remain with the faculty member in the form of an exclusive license. However, the Apache CLA says that a contributor grants rights to all of its patents that related to the code being contributed. However, both MIT and UC are unwilling to accept this language, as they say that can’t determine at the time of contribution if a member of their faculty may have a patent that reads on the patent. So if they sign the CLA they are potentially reducing the potential income to the holder of the patent, and if this indeed proves to be the case then they can be — and have been — sued by the faculty member; As one attorney put it simply, “What is the language that says we don’t know what we don’t know?”
The net of the added language — as I can best recall — from UC and MIT is to say the contributions are not from the institution but from the contributors as individuals.
There was some discussion of revising policy to define the notion of a “enterprise software box,” by which is meant that faculty would have to agree that the university reserved the right to license their patnets for use in open-source code being developed by the university and jointly with other universities for their own purposes, as is the case with Sakai and Kuali. Both UC and MIT agreed to attempt some policy revision, though this is by no means easy. For example, UC may have to deal with over 165,000 signed agreements in effect with various faculty members in the past and present.
Cliff Schmidt, the “legal advisor/member” of the Apache Software Foundation (ASF) was present all day Thursday. He is not an attorney but specializes in legal issues on behalf of ASF. He mentioned they were in the process of revising their CLA, and might be amenable to adding language suitable for academic contributions. That is, ASF might add language that would apply only to academic contributions, so the changes wouldn’t affect the existing language for corporate contributions. However, he did say it was unlikely such changes would be put into effect before mid-2007.
The universites had already agreed before the meeting to try to move to the Apache License 2 if possible. Some of their current work is licensed under the Educational Community License, the ECL: http://www.opensource.org/licenses/ecl1.php. As an aside, I pointed out ECL that in my personal view ECL is incompatible with Apache V2, BSD, or MIT because of the requirement to give, “Notice of any changes or modifications to the Original Work, including the date the changes were made.” (The PERL Artistic license has similar language.) Many seemed surprised by this observation.
Brad Wheeler, the chair, sent the attorneys out of the room Friday morning to see if they could come up with some actual language, as all agreed it was hard to discuss how to best proceed without some actual language at hand. They labored for just over an hour (I timed them) and returned to say they had come to an agreement, but were unwilling to share the language until they had had a chance to review it with their own organizations.
By the way, attorneys Charles Drucker of UC and Karin Rivard were very impressive and quite eloquent in expressing their concerns. As an aside, before dinner on Thursday, I said to Karin that I found it ironic that the creators of the BSD and MIT licenses were now moving to Apache 2. Karin said she preferred BSD in its current form to MIT’s license (another irony.)
There were a few votes and such taken during the discussion. I made a point of abstaining from any vote as I didn’t think it proper for me on behalf of IBM to participate in this way. I did participate in the discussion, particularly in some aspects of open-source and project management where I sensed that the folks involved had limited experience, but I made clear when the conference started that I would only be giving my own views based on my experience, and should not be perceived as offering any official views from IBM.
There was a consensus reached near the end of the meeting that it was important to have some patent-related language in the outbound license. The ECL license doesn’t have any. When I pointed out that they were providing a very high level of Due Diligence — one that went beyond that which might be needed for commercial entitiies to use their code, as the code was primarily meant for use just by their own and other universities — they said it was important to have a very rigorous process so they call sell their work to the university CFO’s and Chancellors.
The final consensus was:
1) The community would attempt to incorporate the UC and MIT changes into the existing CLA as soon as possible;
2) Patent-related language would be added to the outbound license, which would be Apache 2 (preferred), modified Apache 2 (less preferred), or modified ECL (least preferred);
3) The optimal solution would be for ASF to revise its CLA to recognize the special nature of academic contributions, and perhaps even revising the outbound license, but it was understood that any changes on the outbound side by ASF could not weaken the existing language, that is it would still have to be acceptable to the corporate community;
4) That, since ASF changes might not be available until mid-2007, the community might proceed by revising ECL and seeking approval from the Open Source Initiatve (OSI) for approval of the revised ECL. The revised ECL would be a substantial revision; indeed, it would be basically just Apache 2 with language added to make it acceptable to the academic community, at least that part of the community developing enterprise-level applications for its own use.
I told the group as a whole that many of the folks involved in managing IBM’s Open Source activities were located in Westchester and I would be happy to arrange informal meetngs, both with Research or to meet on an informal basis with some of our attorneys. Barnaby Gibson from ithaka is based in NYC, and while we were chatting when waiting for our flight to board said he might take me up on the offer.
Friday morning I sat between Randy Metcalfe of OSS-Watch, an open-source information provided based in the U.K., and Jacqueline Ewenstein, an attorney from Mellon. I will try to build a relationship with Randy; I can already tell by the incoming searches to my blog that he has started reading it. I congratulated Jacqueline, Ira Fuchs, and Don Waters from Mellon for their support of the conference.
A highlight of the trip was meeting Paul Courant from University of Michigan. He is a wonderful and very wise man, and I was especially pleased to learn, when I asked him if he was any relation to Richard Courant, that he was his grandson. I asked Brad to take a picture of the two of us and plan to post it on my blog soon. I spent almost twenty years at the Courant Institute of Mathematical Sciences (CIMS), the world’s foremost school of applied mathematics, a reputation due principally to Richard Courant and the many of the colleagues from Germany he enlisted to join him when he fled Nazi Germany to find a new life. I almost crossed paths with Paul in that he worked he summer of ’66 as a computer operator and I came to CIMS in early September of that year.
While waiting for my return flight I learned that Matthew Raskoff of Ithaca is from New Rochelle, which I knew was the home of Richard Courant any many of his CIMS colleagues, including Kurt Friedrichs. Matthew said he was a good friend of one of Kurt’s sons. I mentioned I had had Kurt as one of my professors. My most vivid memory of him is that one morning back in the 60’s when we was into his 80’s I saw him waiting impatiently for the elevator to arrive. He just couldn’t bear to wait an extra minute or two to get to his office so he could start work on mathematics!
On my flight to Indianapolis I fell into a conversation with a fellow passenger and he said I should make every effort to visit the Indianapolis Speedway, home of the “Indy 500.” When the conference ended early Friday afternoon, Julie Dreesen, one of Brad’s people who did a great job manging the logistics of the conference, helped me confirm the Speedway was open. I visited the Speedway on my way to the airport and, being unfamiliar with my rental car controls, managed to leave the lights on when I went to visit the Indy 500 Museum. The net of this was that I had to ask the help of the kind folks who run the Speedway (it is a private company), so they could “start my engine.” This is the topic of my most recent blog post, https://daveshields.wordpress.com/2006/10/22/gentlemen-start-my-engine/
My thanks to Brad Wheeler of Indiana University who sought IBM participation, and to Mike King of IBM for inviting me to participate in the conference last week in Indiana. It was a great honor and privilege to be able to attend.