I have just finished attending the Fifth Annual Open Source Think Tank, hosted by Andrew Aitken and I at Meritage in Napa Valley. Andrew and his team did a great job of organizing the event. The Think Tank is a great forum for discussing the important questions facing the industry, but equally important, we have structured the Think Tank to provide plenty of time to meet and get to know other attendees (more on that later!). Colin Bodell, VP Web Platforms for Amazon, said it best: he always leaves with a thick sheaf of new cards and many new relationships. I provided my annual summary of Open Source Legal Developments, including both 2009 and 2010 (you can see the powerpoint at http://www.docstoc.com/docs/34875054/Open-Source-Think-Tank-2010-Legal-Issues)
This year, we focused the discussions on commercial problems through the use of three formal “business cases” with detailed facts and questions:
1. Public Sector: How should the State of California adopt and manage open source?
2. Mobile Sector: Selecting a Mobile Platform for Application Development
3. Cloud Computing and Open Source
Much of the discussion was focused on the cloud and how it will effect open source. The opinions ranged across the spectrum: many participants saw the cloud as a great opportunity for open source, but a smaller (but vocal) group noted that the cloud could be a major problem for the open source model. In particular, the concern is that with cloud vendors taking responsibility for the “software stack” customers will be less concerned about open source advantages, leaving such issues to the cloud vendor. As more companies move to the cloud, the customers of open source companies will shift from end users to cloud vendors: consequently, open source companies could have fewer and more sophisticated customers.
Another challenge for open source vendors is the more sophisticated response of commercial software vendors: commercial software vendors are building entire “software stacks” and providing such stacks as a comprehensive solution to end users, thus reducing the need for open source applications. The commercial software vendors are also using “cross subsidies” between the different parts of the stacks to compete with open source vendors selling a single application.
The cloud also poses new legal challenges. The interpretation of open source licenses (and commercial licenses) in this new environment is just beginning and lawyers in the industry have very different opinions on the issues. Some of the issues under discussion at Think Tank include: (1) is provisioning software in the cloud a “distribution” under open source licenses and, thus, triggering obligations under GPLv2, GPLv3, MPL, EPL and other “copyleft” licenses and (2) are software stacks in the cloud an “aggregate” as defined the GPL family of licenses and, thus, not a derivative work (a derivative work might require making source code of all of the programs available if the programs are distributed under the GPL type of licenses).
We also had plenty of opportunity to get to know the other attendees: we had a winetasting on Thursday night in the “cave” at Meritage (including tasting Port brought directly from Portugal by one of the attendees). After the winetasting sponsored by Scott Collison of Geeknet, we moved on to the bar at the Meritage where we shared our experience over stronger potations until well past midnight. On Friday, after the morning discussion and a wine tasting in the afternoon (or golf), we had a dinner at Artesa http://www.artesawinery.com/index1.html hosted by Microsoft. We tasted the Artesa champagne, chardonnay and cabernet sauvignon. The site was magnificent with great views from a hilltop across the green hills of Carnaros and the wines were very good. After dinner, we enjoyed using a telescope to scan the night sky (I saw the rings of Saturn for the first time!).
The Think Tank was a great mix of fun and work. Olliance Group will be summarizing the conclusions, so stay tuned!
The SCO litigation with Novell is finally over: a jury recently decided that SCO does not own the copyright in Unix. The decision was not a surprise, given the district court’s earlier decision granting a summary judgement reaching the same conclusion. However, a decision in favor of SCO would have reopened the issue of whether Linux includes code from UNIX. This decision does not deal with the suits against IBM and other UNIX licensees (which are based on contract), but it prevents SCO from suing third parties who use Linux but are not UNIX licensees.
This decision provides the opportunity to review the SCO litigation and its mistakes: the SCO litigation is likely to become an example of how not to enforce intellectual property rights. Ironically, the principals at SCO had purchased rights in another software program and had successfully obtained a significant settlement in litigation. SCO made the following errors:
1. SCO did not have ownership of the necessary intellectual property rights: the agreement between SCO and Novell was limited to contractual rights. These contractual rights were very limited and Novell prohibited SCO from “amend[ing], waiv[ing] or modify[ing]” any rights under the Asset Purchase Agreement (”APA”). Novell also retained the right to instruct SCO to amend, supplement, modify or waive the rights in the APA and if SCO does not comply, Novell can exercise such rights directly.
2. SCO had not done sufficient due diligence on its rights and was requesting Novell to transfer the rights even as SCO was suing IBM. In fact, the agreement providing SCO with the potential right to own the copyrights in UNIX software (the famous Amendment 2) was not discovered until months after the litigation by SCO had commenced.
3. SCO sued the wrong party: IBM. IBM has one of the highest investments in Linux and has millions of dollars to spend on defending Linux and thousands of patents to supplement this defense. This strategy was very odd.
As I have mentioned in the past, the SCO debacle will be remembered for many years for poor planning and poor execution.