Reqwest default retry handling

[untold spoke]

Hi @neon juniper , after reading the PR again, I think it is best to always enable NACK retry handling and make it opt-out via some function on the custom retry logic. My reasoning is that this behavior has been the default previously. In addition, reqwest actively aims to be a high-level HTTP client, so being opinionated on low level concerns should be totally in scope of the crate. We can always add some documentation explaining this default behavior for people that need low level control (as an aside, perhaps reqwest is no good fit then?). What do you think about that?

If you want to stick to the current status quo, I would advocate making the nack handling a pub function, so we can include the default behavior in our custom retry logic. I can open a PR on that

[untold spoke]

Citing the first lines of the crate docs: "The reqwest crate provides a convenient, higher-level HTTP Client.

It handles many of the things that most people just expect an HTTP client to do for them.". Doing some low level retry logic really fits in that catagory imo

Reqwest default retry handling