Although I’ve previously published the slides for the talk I gave at LinuxWorld 2008 in San Francisco, I thought it might be useful to add some additional comments in the blog about each of the eight predictions I made. This is not the full text of what I said nor a full discussion of the slide, but just some ideas that flesh out what I meant. The full one hour video of the keynote talk is now available at the conference website.
Some people have interpreted this slide as being all about ending the proliferation of new open source licenses. I do think we probably have enough licenses to keep us busy for a long time, but I was trying to make other points.
First, even if new licenses arrive, almost everyone will ignore them. That is, the Open Source Initiative might approve new licenses, but it will be the core set of licenses we have now that will dominate the next decade, just as they are doing today. To me, those licenses are:
- GPL v2 and LGPL v2
- GPL v3 and LGPL v3
- Apache License 2.0
- Eclipse Public License
- Mozilla Public License
- Artistic License
- BSD licenses
- MIT license
Feel free to throw in one or two more if it makes you happy or if you think I inadvertently left one out.
According to the list of most widely used FOSS licenses at Black Duck Software, these licenses account for almost 93.5% of all free and open source projects.
All the rest of the free and open source licenses will be fighting over less than 10% of the projects out there. One license not on the list is the Affero flavor of GPL v3 which could strongly come into play if free software gets used more in cloud computing and SaaS. If the license does get widely used, I’m going to cover myself and just say that it is a version of GPL!
So it’s not so much that we won’t see more licenses come on the scene and thus proliferation must be kept in check, it’s more that no one will care. Go ahead, create that vanity open source license, but think twice if you believe you are making the world a better place. Probably you’re just annoying people instead of getting them interested in the software you’ve created.
I should note that it’s not simply how many projects use a given license, it’s the importance of the projects. The Apache License is only #5 on Black Duck’s list with 2.81%, but the IT and general world would be in a very sorry state without the work being done at the Apache Software Foundation. (And yes, people outside the ASF also use the Apache license.
Second, of the licenses that are in wide use by significant projects, I don’t think we’re going to see significant legal changes. There might be some tinkering with them, but I doubt that there will be a GPL v4 in the next 10 years.
Third, as FOSS gets used more and more by big organizations with significant legal infrastructure, there will be a lot of pressure to keep using the existing, well understood licenses. If you made the leap to open source, you probably don’t want to spend a lot of additional attorney dollars (or Euros or Yen or …) figuring out the latest license du jour, as I termed it in the talk.
Indeed, part of a strategy to get open source more widely used in the enterprise could very well be to freeze license changes for the next few years.
Finally, though we might not see more licenses appearing and becoming popular, I do think we’ll be dealing more with how to assemble multiple pieces of software that come under separate licenses. Sample request to your IP attorney: “I want to release a product that uses custom code, LGPL v3 code, Apache code, and Eclipse code. Tell me how to do it.” The formulas for doing this and other combinations will become better understood in the next few years.
Previous: LinuxWorld 2008 Prediction #5: SMB