Benchmarks!

[Nemo157]

Should be pretty easy to throw together some benchmarks comparing the different async variants against their synchronous counterparts to see how much overhead the async state machines introduce.

[dekellum]

Hi, have you considered the implications of doing compress/decompress on the reactor thread vs. some other coordination strategy or dispatching to non-reactor thread?

Good benchmarks should be able to inform any "default" strategy atleast?

See: http://gravitext.com/2020/01/13/blocking-permit.html

Does the newer async-compression worry about blocking a reactor thread? Should it?

[dekellum]

If you go so far as to benchmark compression of HTTP bodies to/from the network with some amount of concurrency (a full stack kind of simulation), then you might want to try something like this to see if size of Stream<Item> buffers (as it relates to time blocking reactor threads) matters:

dekellum/blocking-permit#3