Re: rust - volunteers sought...

[Havard Eidnes <he%NetBSD.org@localhost>]

> It's unfortunate that the rust compiler has so many problems.

Yes, it is.

However, so far (knock on wood!) there have been no major
problems on amd64 and i386, at least.

> I wonder if we are arriving at having versioned rust packages, so that
>
>   - you can add the new one when it works for amd64
>
>   - we can have basically a switch statement based on os/version/cpu

Maybe -- you're thinking similar to what we do in wip with
rust177, rust178 etc.?  Plus add logic for what's tested and
supposed to work?

>   - we need to agree on rules about how old rust needs to be usable when
>     updating things that use rust (if rust were as stable as it claimed
>     this would be less of an issue and it might be better than it used
>     to be)
>
> This would let us decouple fixing everything from updates, and it seems
> like it might generally reduce the stress level.

The problem would be that "required rust version" might exceed
the rust version certain platforms are able to produce, and "then
what?"  There is a tendency in the rust environment to insist on
the "latest and greatest" with minimal to nonexistent backwards
compatibility, so this will then create a new set of problems.

Regards,

- Håvard