but then AGPL is not the problem, CLA is, right?
If the investors eventually stop paying the bills (or if the set of people that want to contribute is larger than the set of people the investors will pay), then fork it. This seems like a win-win, in that developers win because they get Free tools, and in that developers win because the get paid well to develop open source software.
- I want this project to be a success (and open source) - To make it a success I need to live - To live I need to pay for food
Commercialization in OSS is pretty straight forward as there aren't many options: - Support, managed hosting, paid plugins/features, sponsors.
With the exception of Sponsors, all of these require a number of changes to code and infrastructure that should not be public and this would not be possible without a CLA.
Ergo: CLA -> OSS Developer gets to live.
Yes, it does open the door to license changes and that is a whole other story. But what it boils down to is that funding an OSS project without a CLA is basically impossible.
Again, I agree with the reasoning for libraries, devTools, infrastructure etc given that those libraries end up in commercial products. However in the case of full fleshed enduser applications AGPL is a great protector against the amazons of the world.
As an end user, I'd rather not use GPL3 software that has a carve out that enables the software to be used by privacy hostile service providers that routinely force-upgrade their user customers to worse versions of their systems, but that prevents people from just sticking a customized version of the software on a gizmo and selling it to me.
After all, I can least air-gap and not upgrade most gizmos.
Anyway, I wish they'd rename AGPL to GPLv4 so the infamous "or later" clause applied to it.
Depending on the side your choose, could you explain a bit more on how and why Affero helps or does not help in the 1990s purpose of original 1990 license?