Joshua Scott

As long as we are going to a new major version, I'd be in favor of dropping those, especially if it will make it easier to get the protocol implementation...

@habutre I don't see the need for benchmarking; iolists are almost certainly going to be faster. I think I looked at this before and didn't find any binary concatenations, so...

@dantswain you've been looking at the protocol stuff more, do you know the answer to this?

I think refactoring to use the child specs would be good, and I think will be required to support Elixir 1.10 without generating compile warnings. KafkaEx.GenConsumer does additional work besides...

in the latest release 0.11 we have fixed the stream API to support the global `sync_timeout` setting. Perhaps you need the wait_time to be less than the sync_timeout?

@shamilpd My apologies, I thought I had already responded to this 😞 The fixes you're proposing seem reasonable. As far as the code you mention in server.ex, that code is...

@technomage lz4 isn't yet supported, but a PR would be welcome: https://cwiki.apache.org/confluence/display/KAFKA/Compression lz4 is type `3` - it would be added to https://github.com/kafkaex/kafka_ex/blob/76c752b82c48d04281b8fcf4ad4f14c4fbed335b/lib/kafka_ex/compression.ex to basically wrap the lz4 compress/decompress operations.

Typically the most annoying part was going through and creating release notes. I believe Github does this automatically for you now (on the releases page). I can push up to...

This seems like a good idea to me. I think perhaps we could go with the second version where we can extract the name from `opts`, since normally a GenServer...

@dantswain As far as I can tell this is done?