Fix examples/docs for websocket send commands being a coroutine

[balloob]

What do these changes do?

Version 1.3 made websocket send commands awaitable: b3c80ee. However that change did not update tests/examples. This PR finishes that work.

It went unnoticed because when a send method is called, the bytes are actually written to the socket and a coroutine is returned for the drain command.

Noticed it because setting PYTHONASYNCIODEBUG to 1 started printing errors about unawaited coroutines.

Are there changes in behavior for the user?

No

Related issue number

No

Checklist

• I think the code is well written
• Unit tests for the changes exist
• Documentation reflects the changes
• If you provide code modification, please add yourself to CONTRIBUTORS.txt
  • The format is <Name> <Surname>.
  • Please keep alphabetical order, the file is sorted by names.
• Add a new entry to CHANGES.rst
  • Choose any open position to avoid merge conflicts with other PRs.
  • Add a link to the issue you are fixing (if any) using #issue_number format at the end of changelog message. Use Pull Request number if there are no issues for PR or PR covers the issue only partially.

[codecov-io]

Codecov Report

Merging #1886 into master will not change coverage.
The diff coverage is 100%.

@@           Coverage Diff           @@
##           master    #1886   +/-   ##
=======================================
  Coverage   97.06%   97.06%           
=======================================
  Files          37       37           
  Lines        7556     7556           
  Branches     1314     1314           
=======================================
  Hits         7334     7334           
  Misses        100      100           
  Partials      122      122

Impacted Files	Coverage Δ
tests/autobahn/client.py	97.09% <100%> (ø)	⬆️
examples/web_ws.py	95.18% <100%> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update beb1262...7aa27a9. Read the comment docs.

[asvetlov]

Thanks!

[balloob]

When I started yielding from sending messages I quickly ran into Python bug 29930: 2 drains in parallel will raise an AssertionError error.

Do you think that we should add this to the aiohttp docs?

[AraHaan]

Yeah, this is exactly why I use the latest commit of 3.6.1+ by building python from source every day for 32 and 64 bit windows. Also all the more reason to use 'async def' and 'await' as well. I also can't wait for the day when python 3.4.x reaches no further patches by EOL so that way aiohttp can drop python 3.4 and go full 'async def' and 'await'.

[asvetlov]

I think the main showstopper for dropping Python 3.4 support is absence of 3.5 in Debian stable.

[AraHaan]

Yeah, if I was part of Debian development I would change that to latest commit of 3.6.1+.

After all I do not think 3.6.1+ from github would not break most things for 3.4. But it would allow people to use 'async def' and 'await' on the fly without needing for them to manually compile 3.5 or 3.6 to begin with.

[thehesiod]

@asvetlov stretch shipped, IIRC it's 3.5

[asvetlov]

Yes, I know.
We should give people several months for migrating on Debian Stretch.
Maybe Autumn is good time for shipping aiohttp 3.0 with dropped Python 3.4 support.

@balloob do you know plans for Raspberry PI?

[balloob]

Raspberry PI is expecting a Debian Stretch based release over the summer. Last time it was ~5 months after the Debian stable release.

(source - last paragraph "One final thing")

[lock]

This thread has been automatically locked since there has not been
any recent activity after it was closed. Please open a new issue for
related bugs.

If you feel like there's important points made in this discussion,
please include those exceprts into that new issue.