On Thu, Sep 29, 2022 at 1:16 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
> The intention seems to have been to change the
> behavior only for the conn == NULL case. So maybe we need to
> fix it as attached.
Yeah, that makes sense. Sorry for the oversight.
> This'd still be broken for the
> multiple-libraries scenario, but I admit that that's pretty
> hypothetical.
Since the goal is to let clients decide which connection options to
hardcode based on the SSL implementation, I think it stays
forward-compatible with multiple libraries, as long as this API
returns the "default" library that you get if you're an older,
clueless client. We would need a new API of some sort to let newer
clients figure out their choices.
Thanks,
--Jacob