So a couple weeks ago I had the pleasure of participating in the formation of the new Open Source Geospatial Foundation. I’m quite excited by the possibilities, but as I’m trying to keep this blog more focused on a few ideas instead of a news source or a soapbox for me to stand on, I have left it to others to talk about specifics.
One of the main topics of this writing for me is to play with some ideas on the future of business and production, augmenting capitalism towards something even more just and democratic than our current system. I still have a post bouncing in my brain about the potential of sustainable innovation performed by legal entities that are non-profit by nature, which is what I’m going for with The Open Planning Project. But another potentially interesting hybrid model may be emerging with this new Open Source GeoSpatial Foundation (aka OSGeo).
I believe OSGeo has the potential to be a very unique entity in the world. The closest model that we have is the Apache Foundation, but it soon became obvious that there are several key differences. One is that we already have a number of open source geospatial communities that function very well, and more importantly, function differently from one another. The new foundation must be able to take all in under one umbrella, without forcing any individual project to give up the architectures of participation that they’ve found success with. There is a new infrastructure, that some projects can gain a lot from joining and embracing, but for others it could be a step backwards. There are certainly advantages to having every project on the exact same infrastructure, from a ‘marketing’ perspective, but those may not outweigh sacrifices a project must make in its method of getting things done, the tools that participants use to collaborate. Thankfully all the foundation members involved see the importance of finding the best balance between the two, hopefully utilizing things like a common look and feel, but functioning on different infrastructures.
The other main difference I see from the Apache Foundation is that we’re looking to do ‘more’. Apache is basically a legal structure so that big corporations who put money and energy in to the code can’t sue each other. It operates on very low overhead, and the majority of the energy of the foundation itself is focused on working out the legalities of their licenses. And it maintains the infrastructure to support one of the most successful architectures of participation, with many incredibly high quality projects. But it does very little marketing, it does not seek to legitimize ‘open source’, that’s much more the function of the Open Source Institute. What we’re looking to do is more cross sectional, but focused on a single domain – GIS, instead of just software in general. And there’s a lot more work to be done than just set up technical infrastructure (indeed few of our projects are really lacking in that), and form a legal entity to hold the code (which is important, for sure, and something none of the os geo projects really have right now, but honestly shouldn’t be too difficult).
The ‘more’ is yet to be fully defined, but there seems to be energy towards many things, including marketing type activities (legitimizing open source in the GIS arena, for example), curriculum development, ushering in academics to open source, supporting local user groups, advocating for more availability of geographic data, cross collaborations between projects on standards and reducing code duplication, and even bringing architectures of participation to geographic data.
This is a lot, and I must admit a part of me fears a foundation with a huge overhead that has to be supported through corporate sponsers year after year (I’ll try to explicate the reasons for this in another post). But if we don’t default in to other models of organization, I think this new foundation has a great potential to form a new sort of entity, which I will call a ‘foundation of participation’. We are blessed with some great start up funding, and I must say I was incredibly impressed by the people from Autodesk, and I came in with a lot of skepticism. The large capital expense on the collabnet software scares me, but Gary Lang had the excellent idea of creating an ‘insurance fund’, where we estimate how much time and money it would take to migrate to a less expensive solution. If we get that in place, then we are free to explore just how far we can take things, making the larger version sustainable, without having to worry about all being lost if things don’t work out as we hope. We can just pull back to a more modest version.
As for creating this foundation of participation, what I think it could be is a hub of all sorts of ideas and energy around open source GIS that are not directly related to code. Many of us are thinking about the same types of things, such as opening geographic data, curriculum development, conferences, local user groups, ect. But there’s no obvious way to start up these things in the way that there is for an open source software project. One can instantly bootstrap with sourceforge, getting all the tools needed, expanding to other infrastructures as needed. And there are many models of how to organize people in software development, from a benevolent dictator model to consensus based decision making, and many hybrids in between. The good code naturally filters to the top, and often software that is better than anything out there is produced.
What I’d like to see is the OSGeo foundation be able to gather a lot of the energy around the ‘other’ tasks. To allow GIS ‘users’ to join and make a difference, to have their ideas filter up, even if they don’t write code. Tyler Mitchell has been raring to go on such things, and I suspect he is not alone. The problem right now is that the barrier to entry is too high. Tyler basically had to write a book, and get a correlated blog on O’Reilly, doing it all himself just because he really wanted to promote the great tools he had come to learn about. I hope the foundation can come up with some ‘best practices’, of how others can write articles, can speak at conferences and the like. Right now we’re replicating lots of work on basic presentations about open source GIS, and many people are relearning the same things about how the tools fit in with one another. My hope is the foundation of participation can reduce that replication of work. Allow anyone to upload the slides that they used, gather good quotes and stats, collaborate on comparisons between the various packages. Good ideas about how to sell an Open Source based system in Brazil should be posted so others can replicate their success. The arguments that work for certain situations, and fail in others. Curriculum development falls in a similar category, as several disparate efforts are making courses based on OS GIS software, but few others know about them. There are many more tasks like this, I’ll leave it to Tyler to gather them all up.
My point is that these activities should be enabled in the same way the open source software is enabled. Create architectures of participation, which will naturally be different from each other in specifics, but do share a common essence. We should think hard about what tools enable this, and what social structures can contribute to and grow the architecture. And indeed we need a holarchy of participation where the efforts play into and out of one another, building something greater than any constituent parts. Thankfully we have a solid base of open source software projects to build upon – the people in this geospatial arena that most naturally think in a collaboratively way.
If we look at the foundation as a potential architecture of participation, if that’s what we’re trying to do, to allow many more people to collaborate in building a ‘movement’ of sorts around open source geospatial software, then it puts into clear contrast some of the potential worries I’ve had about it. In my previous post I outlined three points as things that enable architectures of participation. One is clearly met, as few would disagree that the things that people want to do with the foundation are useful. Indeed to some extent we are talking about something which is meta-useful, a space of safety and collaboration from which other efforts can grow. The other two I have worries about. One is to lower the barriers to entry as much as possible. The board is working on this, and has made some good strides, but in some ways I think the technical infrastructure of collabnet that makes our core is not actually oriented around architectures of participation. It is great infrastructure for open source software projects themselves, provides all the base tools, but isn’t yet hip to supporting the activities around the software that can be more collaborative. The other is to treat your users as your co-developers. The big worry for me on this front is that we will have a divide between ‘members’ of the foundation, the 45 we decided to start with, and those who didn’t get nominated or elected. We need to think about how to make the foundation inclusive, to have the governance structure really be bottom up. Indeed I’d almost like to see the board make itself obsolete, to have power truly reside in the committees – for software projects and indeed other projects – to have the board as their steward.
As we’re already quite long, I’ll address each of these in its own post. But overall I think the energy of the foundation is really in the right direction towards a very cool inclusive organization, that could accomplish a lot of good. I do believe we need to not default in to other models of how different software foundations work, which is more difficult as it’s easier to just follow what’s come before. We have the potential to do something much bigger than just a place so corporations can’t sue one another. Indeed the software projects already run themselves quite well, and we don’t have much to add there. It’s everything around the software where the foundation can flourish. I think we could make Open Source the standard in the GeoSpatial arena faster than anyone has done in other domains. But that involves much more than just making good software. We just have to figure out how to actually capture a majority of the energy around the software, and get it to just snowball to something far bigger than any of us. If this foundation can accomplish that, then I think we have a bright future indeed.