On 2025-Mar-03, John Naylor wrote:
> I imagine this kind of integration work is messy and difficult, which
> is why I would be reluctant to write something that way if I have a
> choice.
I've marked this commitfest item[1] as Returned with Feedback. Please
feel free to reopen it once you have an updated patch to share. I think
John's reservations on whether to absorb another requisite shared
library is something to strongly consider -- I notice that x86-simd-sort
is not yet packaged by Debian, for instance.
Maybe your needs *are* served better by introducing simd-sort in
pgvector after all? Alternatively: would it be too bad to reimplement
simdsort in C within Postgres rather than using a third party library?
[1] https://commitfest.postgresql.org/patch/5607/
--
Álvaro Herrera Breisgau, Deutschland — https://www.EnterpriseDB.com/
"If you have nothing to say, maybe you need just the right tool to help you
not say it." (New York Times, about Microsoft PowerPoint)
