Nice work! What do you think of adding a test to tests/h2-tests/src/client.rs?

Used the file that most closely matched what you said. Basically a carbon copy of client_builder_max_concurrent_streams.

Thanks for including the test! I think this is probably good to go... Though, I'd want to ask you, since you're thinking about this topic, you might have certain expected behavior in mind. Should we be testing for that behavior?

What I need is covered by the current test (and small fix). That is, the setting is included in the SETTINGS frame with the value set by the user (and the client successfully handles requests with that setting enabled).

Might be worth to also test that it will throw a protocol error in the server requested size > header_table_size case? (Though I'm not sure if this should be another function / part of the same test / part of the codec read ones)

@seanmonstar Is there anything I can do to help this get merged?

I don't remember what test I was asking for... What's here seems good. Just needs conflicts resolved and we can merge.

Rebased onto master and squashed the commits together. Did compile and pass the tests locally.

Just making sure this is not waiting on CI? Doesn't seem like the error is related to code touched by this PR.

Sorry, I fixed CI in #722, restarting it here.