[car_analogy]

> Genuinely open computing would make customisation and sharing available to everyone, not just tinkerers who know what a command line is.

..you're blaming FOSS for not solving the "programming is hard" problem?

> To most of the population there is zero difference between the two.

Yes, if you ignore all the secondary and ecosystem effects, such as getting to use the resulting free software. I have not once looked at the source of the linux kernel, GNOME, KDE, or Firefox, yet I benefit enormously from the development method and spirit that gave birth to them. That is not zero difference.

[jcranmer]

As a programmer, the number of times a project's closed-source nature has prevented me from fixing a problem I had as a user is... one. The number of times I've availed myself of the open-source code of a project to fix a problem I had as a user is... zero--and that includes the projects I was a maintainer of! Both of these numbers are dwarfed by the number of times being closed-source hasn't prevented me from fixing a problem (as it turns out that reverse-engineering a file format is often an effective solution that rarely needs source code).

So the direct utility of open-source in being able to fix your own itches is in fact extremely rare, even for people for whom programming isn't hard.

There is a better point about FLOSS creating an ecosystem of usable utilities for getting stuff done, one I have availed myself of on innumerable occasions. However, I will also point out that the Stallman stance on software has impeded this goal on several occasions, since GPL or even LGPL [1] licensing can prevent reuse of code. This leaves me wondering how necessary the FSF/GNU stuff was in bootstrapping the open-source ecosystem.

[1] LGPL requires you to link software in a particular way to avoid the license spreading to your code.