> This is a common misconception about Free software. It's not about freedom of its users or authors, it's freedom of the actual code to be available for others.

That‘s wrong. Free software is all about freedom for users. Stallman wrote whole essays explaining that.

Of course, back in his time, pretty much every user was also a programmer. But he is still very clear about that point, never writing about programmers when he means users.

It‘s also why people still fight their silly argument whether BSD or GPL is "freer". The BSD camp optimizes for developer freedom, the GPL for user freedom. Both camps think everybody sees the world their way (philosophically), so the other camp must be wrong and stupid.

Firstly, GPLed programs with proprietary modifications can easily be trapped inside single organizations.

The original MIT-licensed program continues to exist and be available regardless of the proprietary versions and embeddings.

A copyleft license just asserts, using copyright power, that such things should not exist.

MIT and BSD licenses just allow certain works to legally exist that copyleft advocates think should not exist. The copyleft is concerned with squelching the existence of things, whereas the MIT and BSD licenses mostly promote existence.

(They are concerned with squelching the existence of plagiarism: an author's name and copyright notice removed from the source code work, and possibly replaced with another. Also, squelching the situation where the compiled work is accompanied by documentation which neglect to give attribution to the presence of that BSD-licensed code.)

> The original MIT-licensed program continues to exist and be available regardless of the proprietary versions and embeddings.

> A copyleft license just asserts, using copyright power, that such things should not exist.

Well, a copyleft license asserts that proprietary versions and embeddings of the licensed software should not exist. GPL software doesn't assert anything about the existence of BSD-licensed software. Or at least, not unless you link them together into one thing.

It is true that rms would prefer that such works not exist. (Or that's how he felt originally, at least.) But the GPL is silent on the issue, as it has to be. There's no clause saying "the author promises not to release any software with a non-copyleft license, nor to name a child Dorothy, nor to eat the flesh of an aquatic mammal."

It's my understanding that the MIT license has no such restriction on publishing modifications. The only restriction is that existing code must stay MIT licensed.

Code lacks agency. It's not a person.