Google’s open source programs manager, Chris DiBona has asked the Open Source Initiative to delay consideration of Google’s WebM license, and in doing so has called on the OSI to be more open. Specifically, DiBona said Google “will want a couple of changes to how OSI does licenses” and that he thinks “that OSI needs to be more open about its workings to retain credibility in the space”.
DiBona statement came in response to Bruce Perens’ request for the OSI’s License Discuss mailing list to consider the WebM license introduced by Google for its VP8 video codec, and follows the declaration by Open Source Initiative board member, Simon Phipps, that WebM is “not open source”.
DiBona’s request reads as follows:
“Please hold off on submitting this while we determine certain compatibility issues internally at google. We’ll engage with osi in a couple of weeks, likely as not. I would also point out that we’re uncomfortable with make licesne proliferation worse and in the event we do submit it, we will want a couple of changes to how OSI does licenses.
1) We will want a label explicitly deterring the use of the license.
2) We will want the bod list archives open for any discussions of webm. We are not comfortable with OSI being closed.
3) We need to know OSI’s current corporate status. I heard that osi was a california corporation again, but I would like to know, from the group, that this is true for 2010 and that there aren’t any issues there.This might sound stridant, but I think that OSI needs to be more open about its workings to retain credibility in the space.”
.
The request led to a predictably, and some would say justifiably, forthright response from OSI board members, including Simon Phipps and Russ Nelson, while Larry Rosen expressed his support for DiBona’s comments. Peace appears to have been quickly restored – for now – thanks to a level-headed comment from Josh Berkus.
451 CAOS Theory opinion:
Earlier this week we argued that the OSI’s license review process needs to be overhauled but while we agree that there is a need for debate over the OSI’s processes, the License Discuss mailing list is not the place for that debate to occur. Google has shown poor judgment and a lack of respect for the OSI in directly linking debate over the license approval process to discussion of a single license. It is hard not to read DiBona’s request as a list of demands (even if that was not the intention).
Google could, and arguably should, have used an existing OSI-approved license for WebM. Google could also have discussed the license with the OSI prior to its publication, although we would note that Google did not itself submit the license to License Discuss and the WebM website makes no claim that WebM is “open source”, that we are aware of. At the very least Google should have worked through its “compatibility issues internally” before publishing the license, although clearly time was of the essence given the current politics surrounding video codecs.
However, we would also note that much of what Google is asking for is already available from the OSI’s processes. What the Open Source Initiative needs is not more openness, but more initiative. This episode further highlights that the current open source license approval process is unable to keep up with the pace of software development and licensing trends and needs to be reviewed if the OSI is to continue to be held in the respect it deserves.
Subscribe via RSS
4 comments ↓
[…] – Most definitely it will be later – Google has asked the Open Source Initiative to delay consideration of Google’s WebM license, and in doing so has […]
Matt, if an overhaul of the OSI approval process results in the approval of more licenses, that is _not_ good news! If we consider the 66 licenses that are currently on the approved list, there are 66 * 65 == 4290 combinations of just two licenses, each of which potentially needs a lawyer’s analysis. And it is not unusual for a project to combine more than two. There are also 11 superceded or retired licenses in addition to the 66, and many of them are still used on active software, giving us 5852 combinations of two licenses in total.
Obviously, any overhaul of the OSI process should reduce the number of licenses, and thus the combinatorial problem, so that we don’t have as much legal confusion. That this has been under consideration for years, and hasn’t happened, is because of the obvious problem: which licenses do you eliminate, and thus which project teams do you alienate by killing their license.
So, although the OSI board has its problems (and FYI: I am not on that board), what we really need them to do is hold a line: don’t let the definition of Open Source be diluted until highly-restricted software can be called Open Source, and don’t approve more licenses unless there is a _really_ good reason. And so far that’s what they’ve done.
Thanks
Bruce
Hi Bruce,
Thanks for the insight. I’m talking about making the process of approval (or disapproval) quicker, not easier. I would completely agree that the definition of open source should not be diluted.
It is worth noting that figures from Black Duck (and others) suggest that the ten most used licenses account for over 93% of open source software in use, so while the theoretical number of combinations is alarming, in practical terms the issue is not as bad as the statistics suggest.
Matt
I disagree, the OSI needs both more openness and more initiative and the two problems are related.
The “combinations” thing is overblown anyhow. When you install a copy of Windows you’re bound by countless license agreements and terms you don’t even likely read. Several embedded packages and technologies again have other terms. I’d hate to even speculate what the interactions were. Oddly a side effect of openness is that you now know some risks so you fear more. The risks you’re blissfully ignorant of you don’t concern yourself with. (this is not you yourself but you in the generic)