On Sat, Jun 27, 2026 at 3:51 PM Danilo Krummrich <[email protected]> wrote:
>
> The only way this could silently break would be to use ForLt!(), but then
> don't
> use the explicitly provided safe cast_ref() method, and instead open-code it
> with an unsafe transmute() assuming covariance from ForLt!().
Yeah, that is the not-compile-error case I was considering.
i.e. I was trying to think if there could be an issue with the name
reuse, since it can always be a pitfall. For in-tree I agree it isn't
a problem. For out-of-tree users, given it is very new and it is
unlikely they had that case, I think it is also fine.
(But e.g. if it were something that had been there for a long time,
e.g. in an LTS already, then I would say it would have been better to
pick another name).
> I can add them on apply, or we leave them as good first issue, as you prefer.
Up to you! I didn't want to add noise/nit work, but if it is fixed
already, then that is great.
[ I also saw some non-Markdown comments in the other commits, e.g.
// Invariant ForLt types (e.g. containing a Mutex)
in the sample. ]
Thanks!
Cheers,
Miguel
