Drop async support

[methane]

Currently, meinheld maintenance is inactive and maintenance cost is very high.

Major maintenance cost comes from async features (e.g. spawn, msocket, patch) For example, Travis fails recently while test_logger.

I think most meinheld users just want stable and high performance WSGI server. And there is asyncio and aiohttp for async requirements.

How about remove all async features and keep meinheld simple and stable?

[methane]

https://twitter.com/mopemope/status/939826388848087042

@mopemope agreed to remove async support.

So, my plan is:

• 0.7: Deprecate greenlet related features. All tests are running with normal thread.
• 1.0: Remove greenlet related features completely.

[kajuberdut]

I would like to challenge this thought (4 years later): "I think most meinheld users just want stable and high performance WSGI server."

As of now, Meinheld is the premier async server for performance in the world of python: "One thing the two benchmarks have in common is that Meinheld tests did very well in both. Can you guess why? Meinheld is written in C, while every other async server is written in Python. That matters."

https://blog.miguelgrinberg.com/post/ignore-all-web-performance-benchmarks-including-this-one

I would argue based on this that Meinheld's async features are the most distinguishing and valuable feature of it, and what sets it apart from every other option.

Since there doesn't appear to be ongoing maintenance, this point is moot. Nevertheless, I think the argument is pretty strong.