Make hatch-rate parameter deprecated instead of killing it right away.

[cyberw]

The renamed events are still incompatible, but that should impact much fewer users.

Should we still call it 2.0? I guess we should.

ping @lhupfeldt

[codecov]

Codecov Report

Merging #1526 into master will decrease coverage by 0.03%.
The diff coverage is 0.00%.

@@            Coverage Diff             @@
##           master    #1526      +/-   ##
==========================================
- Coverage   82.10%   82.06%   -0.04%     
==========================================
  Files          28       28              
  Lines        2509     2509              
  Branches      381      381              
==========================================
- Hits         2060     2059       -1     
- Misses        354      355       +1     
  Partials       95       95              

Impacted Files	Coverage Δ
locust/argument_parser.py	77.19% <ø> (ø)
locust/main.py	20.44% <0.00%> (ø)
locust/runners.py	83.23% <0.00%> (-0.20%)	⬇️
locust/user/task.py	97.29% <0.00%> (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 1bc5c5e...d00b847. Read the comment docs.

[lhupfeldt]

How does the renamed events affect users? I'm too new to Locust to determine whether it is necessary to call it 2.0.
If you follow strict semver then any breaking change should get a new major version - but in my opinion that can also be confusing for users that will generally expect major new functionality in a new major version.

[cyberw]

I think there is enough new functionality to call it 2.0. The breaking change is that if a user listens to the hatch_complete event, they will have to change it to use spawning_complete instead (example: https://github.com/SvenskaSpel/locust-plugins/blob/8a84100bff1ecffe6164be2a9bb631c631ef11b4/locust_plugins/listeners.py#L109)

Whether we call it 1.2 or 2.0 is mostly a matter of taste I guess. Parts of the general OS community (and browser developers etc) seem to have gone the way of releasing lots of major versions (chrome is now 84.x and firefox something like 78.x :)

[lhupfeldt]

2.0 seems fine to me.