Upgrading number of PyLint jobs from 1 to 10.

[dhermes]

Marked "don't merge" for now because I want to actually time the difference on the entire codebase. Working on it now.

[tseaver]

LGTM

[dhermes]

@tseaver May not be worth it.

At the current HEAD in master (or it was current until #2400 was merged)

$ git log -1 --pretty=%H
64353b5639a03ce9d2538eaf5fc112ea9019662e
$ rm -fr .tox/lint
$ time tox -e lint
...
real 3m17.283s
user 3m7.844s
sys 0m8.716s

at the HEAD in this PR

$ git log -1 --pretty=%H
dad810adf293f06dea50a459095a0e60924eb26f
$ rm -fr .tox/lint
$ tox -e lint --recreate --notest
$ time tox -e lint
...
real 3m26.578s
user 3m15.860s
sys 0m18.952s
$ time tox -e lint
...
real 3m21.792s
user 3m11.124s
sys 0m18.400s

So the extra jobs either

• Add overhead without adding speed
• Don't actually get set unless I do something else

[dhermes]

pylint-dev/pylint#479

[tseaver]

That PyCQA issue is supposed to be long-ago fixed.

[dhermes]

@tseaver I'm pretty sure we don't get the speed-up because of the changes introduced in #1974 (I was on paternity leave at the time so didn't fully understand the change). In particular, the fact that we run Pylint on each file separately.

[dhermes]

Will revisit this some other time.