Do you seriously think that’s how IP law works? If you weren’t able to write GPL Rust code, Rust would not be free software. That would require the Rust project to issue software licences to programmers that stipulate that you must not create GPL-licensed software using Rust.
Rust is free and open-source, like most programming languages. That means you are allowed to make whatever software you want with it, including GPL software. There’s nothing stipulating that you can’t…
You can write GPL rust code, you can also write non-gpl rust code. Gnu coreutils are gpl, if I choose to write functionality compatible with coreutils from scratch in rust I can relicense that as I please (ie not GPL) OP is postulating the driver behind rust rewrites is not for the language features but to allow coreutils functionality to be relicensed as closed source software.
I mean, you can create a GPL fork of the Rust coreutils if you so please. Or you could do a rewrite in any programming language of your choice and license permissively.
In any case, I profoundly cannot bring myself to care about the fact that you can legally create a proprietary fork of permissively licensed FOSS. I don’t think it’s right to impose any restrictions on what people can do with software/code, which of course conflicts with the fact that other people can take your code and restrict what other people can do with it. So choosing between copyleft and permissive licensing is a balancing act of that contradiction. I don’t think it’s wrong to end up on the side of permissive licensing.
OP is postulating the driver behind rust rewrites is not for the language features but to allow coreutils functionality to be relicensed as closed source software.
And that postulation is extra laughable because non-GPL coreutils implementations always existed. And by always, I mean they actually predate the GNU implementation itself (which was a originally a “rewrite” btw 😉).
And yes, they don’t target GNU compatibility, but some of them are perfectly serviceable as is (e.g. the freebsd ones), and adding GNU compat to them would have been infinitely easier than starting a Rust implementation from scratch anyway.
Another laughable aspect regarding the coreutils/uutils case in particular is that uutils didn’t even start as a corpo-driven or corpo-backed project. It was literally a Mozilla employee having fun in his free time, especially during Covid, and community contributions in the same vein. Rewrites of non-GPL projects (e.g. sudo) were ironically much better backed.
From my experience here and elsewhere, the people who make such stipulations don’t even know which licenses the core packages in their own systems adopt. Many of them don’t even know what these licenses’ provisions precisely entail. They just perpetuate some retarded circlejerk probably started by some clueless+malicious e-celeb that goes like:
linux distro -> C -> gpl -> not corpo -> good
rust -> not gpl -> corpo -> bad
Do you seriously think that’s how IP law works? If you weren’t able to write GPL Rust code, Rust would not be free software. That would require the Rust project to issue software licences to programmers that stipulate that you must not create GPL-licensed software using Rust.
Rust is free and open-source, like most programming languages. That means you are allowed to make whatever software you want with it, including GPL software. There’s nothing stipulating that you can’t…
You can write GPL rust code, you can also write non-gpl rust code. Gnu coreutils are gpl, if I choose to write functionality compatible with coreutils from scratch in rust I can relicense that as I please (ie not GPL) OP is postulating the driver behind rust rewrites is not for the language features but to allow coreutils functionality to be relicensed as closed source software.
I mean, you can create a GPL fork of the Rust coreutils if you so please. Or you could do a rewrite in any programming language of your choice and license permissively.
In any case, I profoundly cannot bring myself to care about the fact that you can legally create a proprietary fork of permissively licensed FOSS. I don’t think it’s right to impose any restrictions on what people can do with software/code, which of course conflicts with the fact that other people can take your code and restrict what other people can do with it. So choosing between copyleft and permissive licensing is a balancing act of that contradiction. I don’t think it’s wrong to end up on the side of permissive licensing.
And that postulation is extra laughable because non-GPL coreutils implementations always existed. And by always, I mean they actually predate the GNU implementation itself (which was a originally a “rewrite” btw 😉).
And yes, they don’t target GNU compatibility, but some of them are perfectly serviceable as is (e.g. the freebsd ones), and adding GNU compat to them would have been infinitely easier than starting a Rust implementation from scratch anyway.
Another laughable aspect regarding the coreutils/uutils case in particular is that uutils didn’t even start as a corpo-driven or corpo-backed project. It was literally a Mozilla employee having fun in his free time, especially during Covid, and community contributions in the same vein. Rewrites of non-GPL projects (e.g. sudo) were ironically much better backed.
From my experience here and elsewhere, the people who make such stipulations don’t even know which licenses the core packages in their own systems adopt. Many of them don’t even know what these licenses’ provisions precisely entail. They just perpetuate some retarded circlejerk probably started by some clueless+malicious e-celeb that goes like:
linux distro -> C -> gpl -> not corpo -> good
rust -> not gpl -> corpo -> bad