[throwawaymaths]

I wonder if having that sort of "safe concurrency" causes developers to overuse concurrency and introduce coordination costs.

Do we know that tokio's concurrency strategy is optimal for database access?

[fsociety]

Let’s assume that this is true. Then is the solution to go back to C where it is hard to get concurrency running correctly? Seems backwards thinking.

[throwawaymaths]

No, the solution is to keep the thing that's running quite well in c unless you can prove that it's better?

[maximilianburke]

Starting with a claimed 20x speedup seems like proving it's better? https://modelz.ai/blog/pgvecto-rs