chore(tests): Increase time bias limit to 0.02

[Xeonacid]

On my RISC-V board, it would be a little more than 0.01, such as 0.012354135513305664.

=================================== FAILURES ===================================
___________________________ test_function_times_out ____________________________

    def test_function_times_out() -> None:
        """Test time outs with retry_until()."""
        ini = time()
    
        def never_true() -> bool:
            return False
    
        with pytest.raises(WaitTimeout):
            retry_until(never_true, 1)
    
        end = time()
    
>       assert abs((end - ini) - 1.0) < 0.01
E       assert 0.01162266731262207 < 0.01
E        +  where 0.01162266731262207 = abs(((1708230824.4211185 - 1708230823.4094958) - 1.0))

tests/test_test.py:44: AssertionError
___________________ test_function_times_out_no_raise_assert ____________________

    def test_function_times_out_no_raise_assert() -> None:
        """Tests retry_until() with exception raising disabled, returning False."""
        ini = time()
    
        def never_true() -> bool:
            return False
    
        assert not retry_until(never_true, 1, raises=False)
    
        end = time()
    
>       assert abs((end - ini) - 1.0) < 0.01
E       assert 0.011825084686279297 < 0.01
E        +  where 0.011825084686279297 = abs(((1708230826.5377028 - 1708230825.5258777) - 1.0))

tests/test_test.py:72: AssertionError
_______________________ test_function_times_out_no_raise _______________________

    def test_function_times_out_no_raise() -> None:
        """Tests retry_until() with exception raising disabled."""
        ini = time()
    
        def never_true() -> bool:
            return False
    
        retry_until(never_true, 1, raises=False)
    
        end = time()
    
>       assert abs((end - ini) - 1.0) < 0.01
E       assert 0.012354135513305664 < 0.01
E        +  where 0.012354135513305664 = abs(((1708230865.1922984 - 1708230864.1799443) - 1.0))

tests/legacy_api/test_test.py:58: AssertionError
___________________ test_function_times_out_no_raise_assert ____________________

    def test_function_times_out_no_raise_assert() -> None:
        """Tests retry_until() with exception raising disabled, returning False."""
        ini = time()
    
        def never_true() -> bool:
            return False
    
        assert not retry_until(never_true, 1, raises=False)
    
        end = time()
    
>       assert abs((end - ini) - 1.0) < 0.01
E       assert 0.010137319564819336 < 0.01
E        +  where 0.010137319564819336 = abs(((1708230866.2246602 - 1708230865.2145228) - 1.0))

tests/legacy_api/test_test.py:72: AssertionError
=========================== short test summary info ============================
FAILED tests/test_test.py::test_function_times_out - assert 0.011622667312622...
FAILED tests/test_test.py::test_function_times_out_no_raise_assert - assert 0...
FAILED tests/legacy_api/test_test.py::test_function_times_out_no_raise - asse...
FAILED tests/legacy_api/test_test.py::test_function_times_out_no_raise_assert

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (48ea20b) 88.76% compared to head (5432630) 88.38%.
Report is 74 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #529      +/-   ##
==========================================
- Coverage   88.76%   88.38%   -0.38%     
==========================================
  Files          37       37              
  Lines        3685     3797     +112     
  Branches      526      549      +23     
==========================================
+ Hits         3271     3356      +85     
- Misses        281      304      +23     
- Partials      133      137       +4     

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

[tony]

While I don't see a CLA message showing you signed it, I see you on https://cla-assistant.io/

[tony]

Potentially fixes #481, cc @johanneskastl

[tony]

Also related to #477 and #478 (comment).

[tony]

@Xeonacid FYI: After merge, I may extract the 0.02 into a constant that can be higher.

[Xeonacid]

@Xeonacid FYI: After merge, I may extract the 0.02 into a constant that can be higher.

Glad to see that! Or you may tell me how much you want so I can just fix in this PR.

[tony]

No.

I take that back, I don't want this.

[Xeonacid]

No.

I take that back, I don't want this.

Could you please explain why? As the issues you mentioned, it seems flaky and reproduces on many architectures.