You are right. I am not trying to rewrite history, but I also wonder if, had the planners thought the internet would become as big as it is, would they allow it to be as unrestrained as it was at the beginning?

<< We inherited the current mess that we call the internet because several layers of it were centralized to satisfy corporate interests. They are responsible for our current predicament in the first place.

Separately, it does open an interesting question. Right now the push is to centralize, BUT lets speculate if they would push for decentralization if it meant it became useful for a different purpose ( solar system internet -- assuming private space exploration takes off). I wonder if they would try to cooperate vs force 'their' satellite ( I am assuming a lot now ) communication standard.

> had the planners thought the internet would become as big as it is, would they allow it to be as unrestrained as it was at the beginning?

Interesting question. I think that the arpanet took that design because it started as a research project. The corporations today are unlikely to have ever adopted such a design. I don't know how the corporations back in the day were. And as for the actual planners, the relevant question is if they had any reason to believe that it wouldn't grow so big so fast. We know so many examples where research labs and academia came up with products that are revolutionary. Perhaps they did imagine the possibility and were generous enough?

> Right now the push is to centralize, BUT lets speculate if they would push for decentralization if it meant it became useful for a different purpose. I wonder if they would try to cooperate vs force 'their' satellite communication standard.

That's a very tricky question too. Here's what I think. They would probably cooperate and create an open standard - but only because they want to compete with the dominant player with the first-mover advantage. And that standard would also be so complex that it defeats the purpose of being open, and only they can practically setup anything with it. This is trend that we see widely today - the web standards, kubernetes, bios (or equivalent) firmware, many parts of the Linux software ecosystem, etc. They don't go for the simplest, most logical, orthogonal and easy-to-implement designs, ever.