The Shuttleworth Foundation is committed to open philanthropy: using community, transparency and other open source principles to create a better world. How We Work is a series of occasional articles that take a critical look at one aspect of our open philanthropy practice. Our aim is to reflect and improve upon our efforts while also sharing what we've learned with others.
The Shuttleworth Foundation believes in open innovation. It is core to the society we want to build. Early on, we made a decision that what we do and fund should be under an open license. Our goal was to make it easy for people to use, adapt and improve whatever our staff and partners created. We wanted maximum viral impact, and we saw open licensing as the first step in this direction.
As it turns out, making open licensing work isn't easy, and going viral is even tougher. In the three years since embracing open licensing, we've bumped up against confusion over IP [this is the first time this is being used so rather "intellectual property (IP)"] ownership, partners who are not willing to share, and lawyers who don't 'get' open. Also, in many cases, we've simply lost track of materials our partners have created. They may be open, but no one can find them. Not even us.
The good news is we're pushing past all of this, putting in place more systematic open licensing and archiving policies. As we do this, we thought we should write down how things have gone so far and explain where we are headed in the future. Hopefully, this will help others move into open licensing more quickly and successfully in the future.
Insisting on open licenses is partly about walking our talk. We are committed to openness in principle, and believe that open innovation can facilitate economic and social progress. It's important for us to have policies in place that are built on this principle. We also need mechanisms that model how these principles and policies can work in action.
However, the big opportunity with open licensing is far more practical: it has the potential to dramatically increase the impact, reach and scale of the ideas we invest in as a foundation. Good content, technology and research released under an open license is far more likely to be picked up, used and recirculated than materials that require permission to use. Simply having to ask for permission either deters or slows people down when they are looking for materials to use. Open licensing takes this inefficiency away and increases the likelihood that your materials will be used. This is especially important for small bits of research, blog postings, podcasts, photos and the like which, over time can help build up a global pool of open knowledge that others can use to build great things.
There is also a hope that open licensing can increase quality and speed up development for some of our bigger projects. For example, open licensing - combined with networking and community building - will make it possible for teachers all across South Africa to contribute to our Siyavula free textbook project. Each teacher can build and improve on the work of those that came before. Over time, other groups of teachers from other countries may take the materials for their own use, improving them further and sharing them again (our licenses require this). While it's far from certain that this collaborative Wikipedia-effect will take hold with our educational materials, we don't even have a chance unless everything is under an open license.
With this combination of principles and pragmatism in mind, the Shuttleworth Foundation implemented its first open licensing policy in 2004.
Initially, this was simply a tacit agreement within the team that all materials created by staff, consultants or partners (aka grantees) would be under an 'open license'. While there was no written policy, openly licensing was regularly discussed during contract negotiations with partners and consultants. In most cases, contracts stated that the partner and the Foundation jointly owned whatever was created through the project in question, and that both would release resulting material under an open license. Members of the internal team released materials they had created under open licenses as well.
While this looked good on paper - clear, simple, aligned with our goals - making it work in practice was another matter
One challenge was getting partners to actually open license their materials. In some cases, partners who were unfamiliar with open licenses would balk at the open idea up front in contract negotiation process, leading to long, drawn out negotiations. In others, they would say 'I didn't know that's what you meant by open' towards the end of the project, and then refuse to apply an open license. And, in one case, the partner agreed to use an open license, but then used restrictively licensed artwork in their materials, making it impossible to actually apply the open license at the end of the day. All of this was exacerbated by the fact that the Foundation and partners were 50/50 owners of most materials. This joint ownership approach created a deadlock whenever there was a disagreement on open licensing.
Another major challenge was simply keeping track of materials and software that were created. From the beginning, a good number of partners were happy to see their materials openly licensed. However, most partners didn't have a place to post these materials on the Internet. The Foundation only had a basic web site and did not have a specific policy to track and archive materials created by partners. The result: much of the research and content created using Foundation resources in the early days is impossible to find today, even if it is under an open license. It turns out that a good archiving and knowledge management policy is just as important as the license itself.
The bottom line is that we didn't have the kind of innovation or viral impact that we were hoping for. Not even close. In fact, we basically had an 'open licensing principle' (this is what we believe) and not a functional open licensing policy (this is how it will work). In 2007, we set out to fix this.
An overhaul of the Foundation's open licensing policy began in 2007 and went into place in early 2008. This written policy aims for a more cut and dry open licensing approach, making it easier for partners to understand and discuss our approach up front. There is standard contract language. Expectations on which licenses can be used are clear. At the same time, efforts to address archiving and access issues are also underway.
As in the past, the new policy requires that all resources created using our staff or funds must be 'open'. However, the details are much more concrete: software, reports, manuals, research results and other materials must be released under an open license and posted using an open format on a publicly accessible web site. While it is possible to motivate for other options, the preferred licenses are the GNU General Public License for software and Creative Commons Attribution Share Alike Generic 3.0 for everything else. These obligations are clearly stated in standard consulting and partnership contract templates, and are often a point of discussion as contracts are developed.
We've also tried to resolve the problems created previously by co-ownership, making sure that every piece of new intellectual property created has a single owner. While who this owner is needs to be decided on a case by case [rather "case-by-case"?] basis, it should always be a person or entity who is likely to champion and promote the idea in question. In the many cases, this is the author or originating organization. They have the the most vested interest in making sure that people know about, use and improve their materials. However, this isn't necessarily the case with massively collective works, like a wiki or collectively written piece of software. In these cases, a single owner like the Foundation or a network organization who will steward the idea on behalf of the whole group may be a better option. This stewardship approach is more and more common in the open source world and is standard policy at companies like Canonical (the company behind Ubuntu Linux).
It's important to emphasize that ownership here is not about access or use: the open licensing policy ensures that everyone has access. Ownership is about stewardship; making sure there is someone to run with and protect the openness of an idea for the long haul.
The final issue we needed to address was the accessibility and archiving of materials. The recently re-launched Shuttleworth Foundation web site will help with this, offering a more flexible and easy to use place to post materials. However, this is just a small piece of the puzzle. There also need to be policies in place and people checking to make sure that everything produced gets posted somewhere. This needs to be built around a simple workflow that will ensure it becomes part of the natural processes that staff and partners use when they are producing and publishing materials.
One possible approach would be to require that partners post everything they create somewhere public (their own site, Wikipedia, an existing open access research or software repository, etc.). The Foundation would in turn commit to pointing to this research and keeping a copy for archival purposes. All of this would be included in consulting and partnership contracts. The Foundation will develop and begin to execute a plan along these lines in the second half of 2008.
It's worth noting that there has been some early traction with these ideas, even if they aren't fully implemented. Simply putting open licensing in our standard contract language has led to many useful conversations with partners, deepening their understanding of the whole open innovation cycle. In turn, this has had a bit of a viral effect, with partners using open licenses even in unrelated projects and agreements with other parties. Also, on an operational level, discussions about licensing are much faster as our stance is clear cut. We are no longer reinventing the wheel with every contract.
See also: Shuttleworth Foundation Open Resources Policy
While there is a long way to go with all of this, it feels like we're getting closer to the right track. Requiring open licensing and rigorously archiving will ensure that all the materials created with Foundation resources will always be open and accessible. Picking the right owner or steward will increase the likelihood that the ideas we back are promoted, that communities form around them and that improvements are integrated and re-released. Hopefully, time and tweaking will result in an approach that really works.
We certainly believe that there is enough potential here that others - foundations, governments, research institutes, universities - should be looking at open licensing very seriously. The practical reasons are clear: increased likelihood of impact and scaling for ideas they fund, in ways that could never even be imagined by design. There are also less tangible but equally important benefits that come from the faster feedback loops and the promotion of open, collaborative ways of working. It's worth taking the time to ask: what are my reasons for keeping this or that idea closed? Unless there is a real bottom line reason, set your ideas free.
Of course, choosing to adopt an open licensing policy doesn't mean you won't meet obstacles along the way. The most common source of push back [rather "push-back"?] is foundation lawyers. The vast majority of these lawyers don't understand, or don't even know about open licenses. As a result, they often flag vague, unspecified risks related to going open. This has happened to us on a number of occasions. In the end, our legal experts have backed our interest in open licensing, underscoring our belief that the real risk is that resources created with passion, ingenuity and money, are lost, wasted or do not achieve the impact which they should. There is no real legal reason not to adopt an open licensing policy.
Another possible roadblock is foundation staff who are obsessed with 'tracking impact'. We've heard people like this say: if it's open and people can take it, I can't track the impact of what I've funded. We would ask in response: is it more valuable to loosely observe that you've had a huge impact because people are using and sharing what you've funded, or to rigorously track the fact that you've had almost no impact at all? Is it better to know all your possibilities for impact in advance, or to open things up so that people can evolve what you've funded into ideas and impact that you could never have imagined?
If you do decide to 'go open', it's important to take the time to be thoughtful about how it happens. Our experience suggests that there are three issues to pay particular attention to: license choice (choose a license like CC BY SA that has maximum viral impact); ownership (think about who has a stake in making ideas travel and keeping them open); and accessibility (make a clear plan for access and archiving). These are three areas we tripped up on, and that we're now working to improve. Watch this space. We'll be back with an update as we learn more along the way.