Daniel McCarney

> cpu added this pull request to the merge queue 51 minutes ago This seems to be jammed up. The [last job](https://github.com/rustls/rustls/actions/runs/9175714619/job/25229290743) (Build+test (stable, ubuntu-latest)) is completed, but spinning. I'm...

> I'm going to pull it out and requeue and see if that does the trick. Looks like there's a [broader outage](https://www.githubstatus.com/incidents/nj47vccwm2zj).

:wave: Hi there, > In our application we would like to make a decision based on contents of SNI in ClientHelloPayload Have you looked at using the [`Acceptor`](https://docs.rs/rustls/latest/rustls/server/struct.Acceptor.html) API? The...

I left a comment with [a WIP branch](https://github.com/rustls/rustls/compare/main...cpu:rustls:cpu-acceptor-plaintext?expand=1) on https://github.com/rustls/rustls/pull/1916 but I think it makes more sense to discuss here. Using the acceptor API and the fns on an accepted...

> Could it just be a boxed Error? Yup! That's a better idea, not sure why I didn't think of that initially. I will change it out. > I think...

> Yeah that's the sort of thing I had in mind. OK, I won't pursue turning the WIP branch into a PR in that case :+1: Should this issue be...

> There seem to be two radically different approaches here and we can't get a consensus From my read I think Option 3, the status quo, is where maintainer consensus...

I think this has been addressed. There are now ffi bindings for `rustls_acceptor` and AFAICT `mod_tls` has been adjusted accordingly.

> I'll write a PR to fix it so make doesn't build client and server. I went ahead and made this change: https://github.com/rustls/rustls-ffi/pull/473 I'm going to close this issue for...

Is the intent that we would match the most significant component only, but would release/version independently in the less significant components? E.g. a rustls-ffi 0.23.x release would be implied to...