Delete stray comment

[asottile]

No description provided.

[codecov]

Codecov Report

Merging #5496 into master will increase coverage by 0.2%.
The diff coverage is n/a.

@@            Coverage Diff            @@
##           master    #5496     +/-   ##
=========================================
+ Coverage   96.08%   96.28%   +0.2%     
=========================================
  Files         117      117             
  Lines       25564    27265   +1701     
  Branches     2475     3009    +534     
=========================================
+ Hits        24562    26253   +1691     
- Misses        698      713     +15     
+ Partials      304      299      -5

Impacted Files	Coverage Δ
src/pytest.py	100% <ø> (ø)	⬆️
testing/test_pathlib.py	98.24% <0%> (-1.76%)	⬇️
testing/test_parseopt.py	97.72% <0%> (-0.21%)	⬇️
src/_pytest/deprecated.py	100% <0%> (ø)	⬆️
testing/deprecated_test.py	100% <0%> (ø)	⬆️
testing/acceptance_test.py	98.06% <0%> (+0.05%)	⬆️
src/_pytest/pathlib.py	89.2% <0%> (+0.12%)	⬆️
src/_pytest/config/argparsing.py	88.5% <0%> (+0.29%)	⬆️
src/_pytest/outcomes.py	95.83% <0%> (+0.44%)	⬆️
testing/test_assertion.py	98.11% <0%> (+0.54%)	⬆️
... and 8 more

---

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 3f3f3e7...bfba33e. Read the comment docs.

[Zac-HD]

Unrelated to this PR, I just want to say that I'm not enjoying working with codecov... there is simply no way that this PR added 1701 lines of code, and if it made any difference to what code was covered then we have a flakiness problem.

I have no solutions short of Hypothesis-style "require 100% coverage of everything and fail CI otherwise" (IMO reasonable but obviously unpopular), but I'm finding it pretty annoying even when it isn't hanging forever.

[asottile]

Unrelated to this PR, I just want to say that I'm not enjoying working with codecov... there is simply no way that this PR added 1701 lines of code, and if it made any difference to what code was covered then we have a flakiness problem.

I have no solutions short of Hypothesis-style "require 100% coverage of everything and fail CI otherwise" (IMO reasonable but obviously unpopular), but I'm finding it pretty annoying even when it isn't hanging forever.

Totally agree, I mostly just ignore codecov

[nicoddemus]

Unrelated to this PR, I just want to say that I'm not enjoying working with codecov... there is simply no way that this PR added 1701 lines of code,

To be fair, those 1701 lines of code are regarding the "coverage diff". Probably it got confused from some other run with less coverage... I'm not sure