Fix wait_on_start type-instance condition checking to use fail_time w…

[jlnav]

Addresses #1468.

Basically True and False got interpreted as 0 and 1, valid integers for wait_on_start,
meaning the Executor.fail_time attribute never got used.

Now we use that attribute if we actually got True. Any other integer >1 means we wait for that time. False being 0 should simply mean we don't wait.

Thanks @cchall!

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 90.93%. Comparing base (09c6f52) to head (065134f).
Report is 111 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #1474      +/-   ##
===========================================
- Coverage    90.93%   90.93%   -0.01%     
===========================================
  Files           76       76              
  Lines         7679     7678       -1     
  Branches      1139     1139              
===========================================
- Hits          6983     6982       -1     
  Misses         534      534              
  Partials       162      162              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[shuds13]

This used to be a boolean, what is the use case for waiting N seconds?

[jlnav]

I don't exactly remember; I introduced the change in #1226.

Based on grepping through all of github and known frameworks that use libE, I can probably change the option back to boolean only.