> 21 апр. 2021 г., в 21:01, Stefan Keller <sfkel...@gmail.com> написал(а):
>
> What's needed - and what many here as well as the "ML-In-Databases"
> paper from Kraska et al. (2021) are saying - is, that a new index
> (like a learned index) should be implemented as a PostgreSQL
> extension.
BTW, you don't have to implement index from scratch. You can take one of many
existing and modify towards Learned Index.
Index contains many parts: scans, builds, vacuums, wal, opclasses etc. And
simply adhering to API is a lot of work.
I'll be happy to help you to with formulating key differences and, probably,
code of an extension.
Best regards, Andrey Borodin.