@WG-prioritization/alerts issue #74215 has been requested for prioritization.
It looks like it works as intended Oo
I personally also think this is expected breakage/unspecified behavior
I am biased as I wrote the PR implementing this change though
"but we could probably change the Hash impl to always hash the ~same number of bytes"
This might be a way to fix this, I don't really have an opinion on if this is worth it though
esp considering that there are other harder to fix reasons why the layout of
mem::Discriminant may change between architectures,
e.g. conditional compilation + some interesting (i don't think currently used) optimizations.
So I think it is better for this to be "always" broken than for it to work in most cases and suddently blowing up for some edge case.
this is more of a "notice of this isn't covered under stability guarantees of rust releases"
Wait, I just spend some more time looking into this... there is something going wrong here, I would expect a call
write_isize in both cases
but that's somehow not the case
oh god :face_palm: this change is still in beta, so I obviously won't get the expected behavior on stable...
@lcnr what should we do with the prioritization request here? it doesn't seem that we can do much for now?
one option is just leave things as they are until we make a decision
I think this is a decision for the lang team for now, so I don't think it's that useful to prioritize this beforehand.
I-prioritize and ping
Issue #74215's prioritization request has been removed.