On Mon, Feb 18, 2013 at 2:21 PM, Richard Fontana fontana@sharpeleven.org wrote:
On Mon, Feb 18, 2013 at 01:21:16PM -0500, Pamela Chestek wrote:
As a related aside, whether commercial vendor (or user, especially in the case of AGPL-like licenses) sentiment towards FLOSS licenses has any material impact in their uptake is an unanswered question. A key assumption made by the FSF in the way it handled its quasi-public drafting process for GPLv3 (2006-2007) was that in some way this was so (even though the official view of the FSF remains that 'popularity' of their licenses is a non-goal). I see no post-2007 evidence to justify this view one way or another. I could elaborate on that though it might get a bit off-topic.
An interesting question and I will continue to pursue it, as off-topic as it might be. I always thought of Red Hat an accidental advocate of GPL. That is, it built a business around this hot new thing Linux, and Linux happened to be GPL, so it had no choice but to cope with it. My decidedly anecdotal observation is that it's generally really hard for a company (and people, for that matter) to truly cede control if they can find any way to avoid it. Who are the successful companies who use a true open source, copyleft model (not open core, not dual-licensed) and also own most of the code? So your observation sounds right to me.
That said, I suspect that anyone who isn't sophisticated about copyleft (which is most everyone) believes that this is already a feature of the GPL licenses. In other words, the great fear of GPL is that it will "infect" your proprietary code and you will be forced to release the source code. (Correct me if I've misunderstood; I think this provision will have that effect.)
It is not intended to have that effect, because the license in general is not intended to have that effect. (If anything I would like to somehow make copyleft-next strong copyleft while making clearer that this does not somehow equate to mass 'infection' of a larger proprietary code base, though I am not quite sure how to do this.) It is, however, intended to increase the (perceived) likelihood of enforcement.
I misspoke a bit - the perception is about what the penalty is if there has been an inadvertent "infection," not that the license itself causes the infection per se. But from my experience in the proprietary world (which was Progress Software, which may be part of the reason too), it's the accidental infection they're worried about, not the deliberate creation of a derivative work. And the reason it is accidental is because it's so bloody unclear when you've created a derivative work.
So I guess what this means is, if you want to avoid the "infection" concern, but want strong copyleft, then you need to be very crisp and perhaps less expansive about what a "Derived Work" is. But then we're back to trying to appease an audience that probably isn't going to adopt copyleft in the first place anyway.
Pam