Two Thank-Yous

Created
Sat, 11/09/2010 - 23:15
Updated
Sat, 11/09/2010 - 23:15

I'm well known for being critical when necessary about what happens in the software freedom community, but occasionally, there's nothing to do but thank someone, particularly when they've done something I asked for. :)

First, I'd like to thank Matthew Garrett for engaging in some GPL enforcement (as covered on lwn.net). He's taking an interesting tack of filing a complaint with US Customs. I've thought about this method in the past, but never really felt I wanted to go that route (mainly because I'm more familiar with the traditional GPL enforcement processes). However, it's really important that we try lots of different strategies for GPL enforcement; the path to success is often many methods in parallel. It looks like Matthew already got the attention of the violator. In the end, every GPL enforcement strategy is primarily to get the violator's attention so they take the issue seriously and come into compliance with the license.

I've written before about how GPL enforcement can be a lonely place, and when I see someone get serious about doing some — as Matthew has in the last year or so — it makes GPL enforcement a lot less lonely. I still think I can count on my hands all the people active regularly in GPL enforcement efforts, but I am glad to see that's changing. The license stands for a principle, and we should defend it, despite the great length the corporate powers in the software freedom world go to in trying to stop GPL enforcement.

Secondly, I need to thank my colleague Chris DiBona. Two years ago, I gave him quite a hard time that Google prohibited hosting of AGPLv3'd projects on its FLOSS Project Hosting site. The interesting part of our debate was that Chris argued that license proliferation was the reason to prohibit AGPLv3. I argued at the time that Google simply opposed AGPLv3 because many parts of Google's business model rely on the fact that the GPL behaves in practice somewhat like permissive licenses when deployed in a web services environment.

Honestly, I never had definitive proof at Google's “real reasons” for holding the policy it did for two years, but it doesn't matter now, because yesterday Chris announced that Google Code Hosting now accepts AGPLv3'd projects0. I really appreciate Chris' friendly words on AGPLv3, noting that he didn't like turning away projects under licenses that serve a truly new function, like the AGPL.

Google will now accept projects under any license that is on OSI's approved list. I think this is a reasonable outcome. I firmly believe that acceptable license lists must be the purview of not-for-profit organizations, not for-profit ones. Personally, I tend to avoid and distrust any license that fails to appear on both OSI's list and the FSF Free Software License List. While I obviously favor the FSF list myself (having helped originate it), I generally want to see a license on both lists before I'm ready to say for sure there are no worries about it.

There are two other entities that maintain license lists, namely the Debian Project and Red Hat's Fedora Project. I wouldn't say that I find Debian's list definitive, mainly because, despite Debian's generally democratic slant, the ftp-masters hold a bit too much power in interpreting the DFSG.

As for Fedora, that's ultimately a project controlled by a for-profit corporation (Red Hat), and therefore I have some trepidation about trusting their list, just as I had concerns that Google attempted to set licensing policy by defining an acceptable license list. As it stands at the moment, I trust Fedora's list because I know that Spot and Fontana currently have the ultimate say on what does or does not go onto Fedora's list. Nevertheless, Red Hat is ultimately in control of Fedora, so I think its license list can't be relied on indefinitely (e.g., in case Spot and/or Fontana ever leave Red Hat at some point.)

Anyway, I think the best outcome for the community is for the logical conjunction of the OSI's list and the FSF's list to be considered the accepted list of licenses. While I often disagree with the OSI, I think it's in the best interest of the community to require that two distinct non-profits with different missions both approve a license before it's considered acceptable. (I suppose I'd have a different view if OSI had not accepted the AGPLv3, though. ;)


0I must point out that Chris has an error in his blog post: namely, FSF's Code hosting site, Savannah accepts not just GPL'd projects, but any project that is listed as “GPL-Compatible” on FSF's Free Software License List.