Just idly...
> Deciding which is dominant may well be a matter for your ideological biases rather than an objective absolute.
No, it's really not. That's what objective means; that you count the numbers and acknowledge that one number if bigger than another, by a significant margin.
> If I can pick my own metric, I can prove anything!
Sure, but there are some pretty solid metrics that no one is going to argue you're 'fudging the numbers to suit your bias on'. For example:
- Total number of projects with a given license.
- Total number of dependencies on projects with a given license.
- Total number of new projects with a given license per year/month, normalized.
- Total number of new dependencies on a given license per year/month, normalized.
What is this rubbish about deciding which license you like the most because you think it has some kind of pretty ethical features, or calling it 'copy left' vs 'reciprocal behaviour'.
You're just muddying the waters and making people who support GPL look stupid.
Statistics can be jimmied to show a specific tale, its true... but you know what?
A bunch of graphs, a raw data set I can check myself and a repo with the software that generated the graphs; I'll believe what I see in that any day over some arbitrary talk about how you can't really ever believe anything because heck, everyone has a bias.
Hand waving is a lot easier than actually digging up data and checking the numbers; but it's largely meaningless he-said-she-said-they-said.
A little while ago I met some folk who were absolutely convinced that hg was winning the hearts and minds of people around the world (over git). They were and are wrong. So is this post. ...yet people continue to argue the point.
I think it's easy to end up trapped in a bubble where all you see is people who agree with you, and you can't see what's really going on.
Everyone now and then you actually do need to dig through some hard data yourself and reevaluate your own assumptions I think...