The problem is that open source is given away for free, and companies are exploiting that. It's estimated that open source was worth at least $143M of Instagram's $1B acquisition. But the open source developers got nothing.

The solution as I see it is Supported Source (https://supportedsource.org/)

The estimate on Linux was at least a billion. That's what IBM alone was willing to commit to defending it against Microsoft/SCO suits if I interpreted the reporting correctly. I'd love to see how much money people saved on IIS, etc. thanks to Apache deployments. We'd first need to know what percentage would've paid for a server in the first place as many will only set up if it's free. Then, convert that to even half IIS + Windows Server licensing I bet it's a fortune.

I bet many of them would just go back to pirating stuff.

This is what FOSS actually made in the last decade, moved many that would never pay for software from pirating it to just downloading whatever they can find in open source repositories.

But the developers still don't get anything back.

I didnt think about that. Good point. Counter would be BSA. They tip employees to reveal stuff like that. They could visit each Global 2000 to pay employees for info on what software was used and how often. Plus network fingerprinting. Just collecting first years fees as payment would cover cost of BSA and payouts.

That would only work in countries where organizations like BSA do exist and are actually effective.

Which isn't the case in many countries.

Another good point. I might have to think further on it.

The problem is that a payment system for hundreds of extremely tiny pieces does not work. See my car example. You have to make a decision "do I pay those people" on a level that is too small.

Similar reason why companies like Netflix exist. Imagine each movie would try to sell itself, they would get their own webshop, one per movie... and similar to why subscriptions exist and many other bundling arrangements. Making payment decisions comes at a psychological cost, but just because it's "psychological" and your brain doesn't pay some other part of itself with coin or credit card doesn't mean it's not a cost economics cares about. Cost in economics is not just money.

Another example is Twitch. In addition to giving mostly gamers (but they also have programmers streaming and other non-gaming activities) a technical platform to stream content from home they also give them a monetization opportunity. While you can subscribe to streams individually, which is of course not what I suggested, they now have a program where you can subscribe to a handful of channels for free if you have an Amazon Prime subscription (http://www.theverge.com/2016/9/30/13125824/twitch-prime-amaz...). So this is a partial example of this "bundling" of small individual items to increase the overall spending of money compared to having to make individual purchases.

I see this as an opportunity for achieving something similar for people who contribute code, there could be a business opportunity to bundle a large number of those offerings and in turn offer the programmers a way to monetization, to get paid out of the subscription pool. It's much more difficult though because the audience on bot sides, those offering as well as those buying, is much more discriminating, and the situation already is far more fragmented. For things like gaming streams it never was (very fragmented) to begin with. Imagine Github had a subscription and would pay out to projects according to use (stared, downloads, whatever).