> Instead they didn't, and the world is better for it. Why should they change now...?
So this is where we fundamentally disagree: I do not think the world is better off with an uncompromising FSF. I think the world and state of Free Software would have been significantly more advanced if they had adopted a more pragmatic attitude. I don't think that's "compromising" per se, but rather accepting that things don't change in a day.
Also consider an alternative universe where the FSF and/or GNU had invested in to a good code hosting platform and a system to setup payments for people developing software instead of the ecosystem rolling from one proprietary platform to another over the last 25 years.
I mean, no hard FSF/GNU, no GPL, no Linux - with only the BSD/MIT license around, any good kernel feature would have been simply coopted and repackaged like the TCP stack was, keeping commercial software always ahead. From that perspective, the world is undoubtedly better off like this. Even the projects that compromised just a little bit with LGPL often got swallowed: see KHTML, for example.
> consider an alternative universe where the FSF and/or GNU had invested in to a good code hosting platform
FSF/GNU were created by nerdy hippies, expecting them to ever be great at running such a service is unrealistic. Besides, they did provide platforms, they just were never as good as commercial alternatives.
So this is where we fundamentally disagree: I do not think the world is better off with an uncompromising FSF. I think the world and state of Free Software would have been significantly more advanced if they had adopted a more pragmatic attitude. I don't think that's "compromising" per se, but rather accepting that things don't change in a day.
Also consider an alternative universe where the FSF and/or GNU had invested in to a good code hosting platform and a system to setup payments for people developing software instead of the ecosystem rolling from one proprietary platform to another over the last 25 years.