Since my last post, I’ve gotten some good feedback from folks who are knowledgeable about the issues. I hope to write about this topic in more detail after I’ve gotten back from EDUCAUSE and had some time to do a little more research, but in the meantime, I want to at least mention two of the points of criticism that I’ve received about the first post and that deserve some attention.
First, one reader pointed out that the protection in the Apache clause that is weakened in the ECL is actually untested legally. Apache, as I understand it, puts the burden on the contributor to vet the IP with the institution they represent. If an employee of BigResearch U contributes code that that is covered by a patent held by BigResearch U, then there is an implied granting of a license covering that patent. In theory, this is good because the benefit of the doubt goes to the adoptee. It removes uncertainty about possible infringement.
Except that maybe it doesn’t. I’m told that this implied license has not been tested and might not hold up in court. So, my reader argues, the additional ECL clause that I complained about simply makes explicit a legal wrinkle that may exist with the Apache license anyway. And by making it explicit, the license forces the community to look to other, better tested mechanisms.
Which brings me to another major point brought up by a different reader. In my first post, he argues, I only talked about half the equation. I didn’t talk about the contributor agreement that goes along with the ECL and adds protections at the time of code contribution. That’s a fair criticism, particularly given that not all open source projects have this protection. (For example, I don’t think Moodle uses contributor agreements, although I’d appreciate input from anyone who knows for sure.) You can only evaluate the total patent protection provided by the ECL if you look at both the license and the contributor agreement together.
As I wrote at the top of this post, I’ll be going to school on these topics over the coming weeks so that I can post more about them. Ironically, with all the coverage of edupatents I’ve done over the past year, the intersection with copyright licensing is just not something I’ve gotten into very much. I want to thank Chris Coppola, John Lewis, and all the folks who have been helping me privately to get up to speed on these issues. With a little luck, we’ll get some more public conversation going among the various knowledgeable folks in the community so that the larger community (including me) can broaden its understanding of this facet of the edupatent problem.
Michael,
The whole business of orchestrating intellectual property practices for the good of our open source communities is certainly a complex beast. I’ve learned that the practices we employ to educate our community and do our due diligence on the inbound side of the IP equation have as much to do with real protection on the outbound side as the language of the license. Both are important certainly, but it’s easy to make claims in an OS license without doing the work on the inbound side to make sure you can make those claims. Sakai and Kuali are doing a lot to make sure that the claims we’re making are reliable ones. I can see in this post that you’re becoming aware of this. We can and will do more to document what we’re doing and why… and I encourage your readers to participate in the dialog as we continue to improve our practices…