Luca Cappelletti
Luca Cappelletti
I am not sure I understood, what result would you expect from that? I have found some solutions out there, but they involve compiling against Postgres, which seems overkill.
I can try that!
> Did someone actually manage to fix this? I did, in PR https://github.com/pgcentralfoundation/pgrx/pull/2068
This issue should now be closed as PR #2068 has been merged.
Basically, in most cases the derive is enough to cover everything, and honestly when I opened the PR I had just done that thinking it should be it. Afterwards, as...
Small update: fixed errors estimating size on your crate side, but also identified an error on the mem dbg side. Working on that now.
Now if you rerun the same script as above, you will find that all estimates are matching.
Would it be a adequate solution to use, instead of a wrapper type, plainly [`chrono`](https://docs.rs/chrono/latest/chrono/index.html) structs? That's what I am wrapping in my own use case.
Which problems did you encounter? Is there an issue relative to it? I wasn't using [`pgrx::datum::TimestampWithTimeZone`](https://docs.rs/pgrx/latest/pgrx/datum/struct.TimestampWithTimeZone.html) because I wasn't aware there was another struct other than `pgrx::pg_sys::TimestampTZ` honestly. It may...
The gist of my use is that I am comparing several libraries that use different hash functions (stuff for set cardinality estimation) and testing how performance changes with different hashes....