I’ve recently seen a couple of attempts to present non-open source licensing as almost open source. What I’m talking about is Commons Clause and fair source. This sounds a lot like the failed shared source model promoted by companies such as Microsoft in the early 2000. Yes, I’m looking at you redis and neo4j.
What I find particularly disturbing is the way that both, but especially Commons Clause, attempts to piggyback on the good name of open source. The naming is very close to Creative Commons, and the way that the clause is intended to be introduced is very confusing for the user as the user would see a familiar open source license, and then just a “small” exception.
I would argue that the way to go about this is to combine a strong copyleft license such as GPLv3, LGPLv3 or AGPLv3, with a CLA. That way a dual licensing model could be used to create a business model. If that is not possible, the [A/L]GPLv3 needs to be improved – not complicated by additional clauses added by a third party.
The consequence that anyone accepting the strong copyleft can do whatever they like with the software is not a bug in the copyleft license, but instead a feature. Actually, this is one of the core four freedoms of free software.
At the end of the day, my cynical self gets the impression that this is something that you add to an existing project once it has gained traction and then hope that it does not render in fork. If something like the Commons Clause would have been present from the start, the projects would never had gained the user bases and communities that they are enjoying today.
It’s not entirely unfair to mirror Creative Commons in the naming since the Non-Commercial CC licence is what “Commons Clause” is trying to do. I guess RMS would say this is inevitable with the term “open source” although obviosuly “free software” has simliar naming problems.