BUG: pytest 9.3 (dev) rejects conflicting CLI options (`-R` is defined twice)

[neutrinoceros]

breaking commit: pytest-dev/pytest@532b201
corresponding PR: pytest-dev/pytest#14661

Because this plugin re-defines -R, it conflicts with pytest-remotedata, making it entirely impossible to run pytest.
I don't know that this re-definition serves any purpose now, so I propose the simplest fix possible: drop it here, and with it the entire plugin becomes empty and unneeded.

[neutrinoceros]

Maybe the rabbit hole actually goes deeper than that. I don't see -R being defined in pytest-remotedata. For completion, this is what I see by running pytest --help with pytest dev + pytest-astropy + pytest-remotedata

Traceback (most recent call last):
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/bin/pytest", line 10, in <module>
    sys.exit(_console_main())
             ~~~~~~~~~~~~~^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/src/_pytest/config/__init__.py", line 253, in _console_main
    code = _main(prog=_get_prog_name(sys.argv))
  File "/Users/clm/dev/gh/pytest-dev/pytest/src/_pytest/config/__init__.py", line 223, in _main
    config = _prepareconfig(new_args, plugins, prog=prog)
  File "/Users/clm/dev/gh/pytest-dev/pytest/src/_pytest/config/__init__.py", line 410, in _prepareconfig
    config: Config = pluginmanager.hook.pytest_cmdline_parse(
                     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^
        pluginmanager=pluginmanager, args=args
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    )
    ^
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_hooks.py", line 512, in __call__
    return self._hookexec(self.name, self._hookimpls.copy(), kwargs, firstresult)
           ~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_manager.py", line 120, in _hookexec
    return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
           ~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_callers.py", line 167, in _multicall
    raise exception
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_callers.py", line 139, in _multicall
    teardown.throw(exception)
    ~~~~~~~~~~~~~~^^^^^^^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/src/_pytest/helpconfig.py", line 124, in pytest_cmdline_parse
    config = yield
             ^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_callers.py", line 121, in _multicall
    res = hook_impl.function(*args)
  File "/Users/clm/dev/gh/pytest-dev/pytest/src/_pytest/config/__init__.py", line 1266, in pytest_cmdline_parse
    self.parse(args)
    ~~~~~~~~~~^^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/src/_pytest/config/__init__.py", line 1617, in parse
    self.pluginmanager.load_setuptools_entrypoints("pytest11")
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_manager.py", line 417, in load_setuptools_entrypoints
    self.register(plugin, name=ep.name)
    ~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/src/_pytest/config/__init__.py", line 571, in register
    plugin_name = super().register(plugin, name)
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_manager.py", line 169, in register
    hook._maybe_apply_history(hookimpl)
    ~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_hooks.py", line 580, in _maybe_apply_history
    res = self._hookexec(self.name, [method], kwargs, False)
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_manager.py", line 120, in _hookexec
    return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
           ~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_callers.py", line 167, in _multicall
    raise exception
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pluggy/_callers.py", line 121, in _multicall
    res = hook_impl.function(*args)
  File "/Users/clm/dev/gh/pytest-dev/pytest/.venv/lib/python3.14/site-packages/pytest_remotedata/plugin.py", line 16, in pytest_addoption
    parser.addoption(
    ~~~~~~~~~~~~~~~~^
        "--remote-data", nargs="?", const='any', default='none',
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        help="run tests with online data",
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        choices=['astropy', 'any', 'github', 'none'])
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/src/_pytest/config/argparsing.py", line 125, in addoption
    self._anonymous.addoption(*opts, **attrs)
    ~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^
  File "/Users/clm/dev/gh/pytest-dev/pytest/src/_pytest/config/argparsing.py", line 370, in addoption
    raise ValueError(
    ...<4 lines>...
    )
ValueError: option dest 'remote_data' already used by ['-R'] (this is the option that maps to dest 'remote_data'); pass dest='remote_data' explicitly to share the destination

[neutrinoceros]

apparently the bug already goes away when combining both pytest-remotedata and pytest-astropy's dev versions. So the fix might be to simply release one or two of them. Still digging

[neutrinoceros]

This is pretty confusing, I don't understand the exact conditions needed to reproduce the crash (but I do see it consistenly in a private CI repo). I suspect it might have to do with how pytest visits plugins in the order they were installed, though I'm not finding clear evidence supporting this hypothesis at the moment.

[pllim]

Do you use yet another plugin that also have -R ? I think Brigitta ran into that before but I don't remember which plugin it was.

[neutrinoceros]

I'm not installing any plugin not pulled from astropy's extras

[pllim]

OK, so you say new releases would fix the problem?

[neutrinoceros]

I can't say for certain, because I did not fully understand the failure mode. It couldn't hurt though.

[bsipocz]

The -R has been cleaned up years ago, so I suppose the installer picks up an old version somewhere? (The option has been moved upstream as a clash with another plugin was reported (pydata-leaks), and there was only a very short version period when this was in pytest-remotedata). So I fail to see how a release, that practically only contains GHA version updates would solve this issue; if there is a problem somewhere it may still be present afterwards.

[pllim]

Hmm...

[neutrinoceros]

The -R has been cleaned up years ago

could you clarify ? It's still defined here. Where else was it cleaned up ?

I suppose the installer picks up an old version somewhere?

I checked the logs, and my runs had

• pytest-remotedata==0.4.1 (latest)
• pytest-astropy==0.11.0 (latest)
• pytest==9.2.0.dev101+g850036218

I also got logs showing success with pytest==9.2.0.dev117+g36704f9b5, so it really looks like an already fixed regression in pytest dev now, so we can close this.

So I fail to see how a release, that practically only contains GHA version updates would solve this issue; if there is a problem somewhere it may still be present afterwards.

for posterity, I was suspecting #79 could be the important change.

[neutrinoceros]

This is now showing up on astropy's release branch 7.2.x

https://github.com/astropy/astropy/actions/runs/30742027272/job/91481014362

Re-opening as a solution might still be needed there.