Michal Kleczek

> The recent problem I'm seemingly stuck with is `hs-opentelemetry` is using `UnliftIO`, which seems not quite composable with our (implicit, correct?) monad stack. So the deeper into the call...

Until this is implemented my solution is to have two instances of PostgREST behind an NGINX routing requests based on HTTP method. The only problem with this is that PostgREST...

I am not really sure if ordering by distance specifically is not too limiting. In SQL it is more general ORDER BY [expr] DESC/ASC. It is just that distance operator...

What I have prototyped is something similar: Let hasql-pool implement a kind of a multi-pool where you can create multiple connection pools where all would share a single global pool...

I have played with the idea some time ago and have a preliminary implementation. The main goal was to trace SQL statements so it required Hasql adjustments. I also wanted...

It was working last time I tried :) As mentioned in https://github.com/PostgREST/postgrest/pull/3105#issuecomment-1863025619 - static binary builds not tested.

I can submit a PR with https://github.com/mkleczek/postgrest/blob/temp-opentelemetry/flake.nix BUT I did not spend time on making sure all build types work correctly - especially static binary build.

Maybe https://datatracker.ietf.org/doc/rfc7807/ is relevant?

I've marked is as bug as in my opinion this is actually a security vulnerability (and something that _can_ be considered as violation of RFC spec).

> Chiming in, I would be one of those people submitting a problem if this were to change. > The actual solution proposed is to make `jwt-aud` a regular expression....