Accompanying tox issue

example.cmd

REM With single envs it works as expected:

REM expecting: nothing printed, no FAIL
tox -q run -e py310-ok & if ERRORLEVEL 1 echo FAIL
REM (OK) expectation met.

REM expecting FAIL:
tox -q run -e py310-fail & if ERRORLEVEL 1 echo FAIL
REM (OK) expectation met.

REM With multiple envs it does not work as expected!

REM expecting no fail, as ALL environments were ok
tox -q run -e py310-ok,py310-ok2 & if ERRORLEVEL 1 echo FAIL
REM (OK) expectation met.

REM expecting to FAIL as at least one environment has failed:
tox -q run -e py310-ok,py310-fail & if ERRORLEVEL 1 echo FAIL
REM (NOT_OK) expectation *not* met

REM expecting to FAIL as all environments have failed:
tox -q run -e py310-fail,py310-fail2 & if ERRORLEVEL 1 echo FAIL
REM (NOT_OK) expectation *not* met

main.py

if __name__ == '__main__':
    print("Running main")

setup.cfg

[options]
packages = find:

[options.packages.find]
where = .
include = main.py

test_fail.py

def test_fail():
    assert False

test_ok.py

def test_ok():
    assert True

tox.ini

[tox]
env_list =
    py310-{ok,ok2,fail,fail2}
minversion = 4.4.6

[testenv]
py_modules =
    main.py
description = run the tests with pytest
package = wheel
wheel_build_env = .pkg
deps =
    pytest

[testenv:py310-ok]
commands =
    pytest {tty:--color=yes} test_ok.py {posargs}

[testenv:py310-ok2]
commands =
    pytest {tty:--color=yes} test_ok.py {posargs}

[testenv:py310-fail]
commands =
    pytest {tty:--color=yes} test_fail.py {posargs}

[testenv:py310-fail2]
commands =
    pytest {tty:--color=yes} test_fail.py {posargs}