Re: Versioning rust (firefox91 again)

[pin <voidpin%protonmail.com@localhost>]

> The extra maintenance burden you talk about comes from engineering
> decisions by various rust-using upstreams to depend on feature only
> added in recent versions

Exactly, Rust itself doesn't force projects to adopt the latest features.
Although, sometimes is tempting. Like when going from 1.69 to 1.70, where
once_cell was incorporated into rust itself and projects could drop an external
crate from their dependency list.

> Things will get better once the language stabilizes, or even gets
> standartizized. But as we have seen with the recent C and C++ std
> versions discussion, non-dead languages always are able to introduce
> some kind of versioning pain in the real world.

True, I just don't see an EOL web-browser as the reason for versioning.

As mentioned already, I understand keeping some applications for users.
I wouldn't think it's bad to keep a minor application that has a dead upstream
or, that has been EOL. But, a web-browser is a different story, it's a security
risk.

I'm not saying Rust shouldn't be versioned, I just don't see the point, yet.