The author (who also responded) isn't referring to small libraries or utilities that are written by a single person and don't have much public contribution.
You have to ask yourself though, if those trying to assume control can't get together and sell the idea of a new fork to the userbase why would they be the best people to successfully run the original project?
Assume control is a loaded phrase. And many people believe a fork should be a last resort.
> These simply can't be forked without broad consensus around which fork to follow.
Thus:
> if those trying to assume control can't get together and sell the idea of a new fork to the userbase why would they be the best people to successfully run the original project?
That's exactly what I mean by "broad consensus."
They can and the user will decide.
Designing your governance and succession structure to avoid this is better for the long term health of your project.
I don't think it is true. Certainly it takes more work. Broad consensus may be part of that work.
If you cannot reach consensus to produce an equal product as a unilateral decision maker then the benefits of dictatorship are still outweighing the disadvantages.
If another unilateral decision maker runs a fork, people may move to it if it is better. That's them voting, it's not dictatorship, it's representative democracy.
Think of a different scenario: Linus does something that pisses a lot of people off. (Even more than usual) The biggest contributors decide they don't want him in charge. The users don't want anything to do with him. A few of the most influential contributors announce a fork, some of the fans publicize it, and then the major distros base themselves off of the new fork.
At this point none of the forks that you're referring to take upstream changes from Linux, instead they take them from the new fork.