Yann Kaiser

That's odd, I thought it wasn't released yet? On a modified version of `examples/multicommand.py` I get this when running it on `4.1.1`, `4.1.2`, `4.2.0`, and `4.2.1`: ``` $ python3 -m...

As in epsy/sigtools#26 few other incompatibilities affect supporting 3.10 across clize/sigtools just yet. I'll make sure `repeated_test` is also updated in the process. Thanks for bringing it up!

It doesn't look like there is a way to have aliases for an 'alt=' function yet, sorry. I'll look into it.

I can't really do that, users might use `-v` for ``--verbose`` or anything else, whether or not they use ``--version``

It remains that `clize.runner.Clize._process_alt` (https://github.com/epsy/clize/blob/master/clize/runner.py#L203) would need to be updated to handle a dict with tuple-like keys like clize.runner.cli_commands and I still haven't gotten around to it, sorry.

You're calling `run(clize_log_level, ...)` once in each module, therefore arguments get interpreted twice. You can't notice it normally because your function does not product output of its own, but when...

Oh no. don't worry, I just had a busy week. --- It looks like this would work correctly, good job on restructuring things into functions. Is there still an issue?

Your second method would indeed be how it should work. As it is currently, argument decorators flat out ignore anything indicated on their first parameter, because it is entirely determined...

In other argument parsers: - [argparse's subparsers](https://docs.python.org/2/library/argparse.html#sub-commands) - [click's groups](http://click.pocoo.org/3/quickstart/#nesting-commands)

Fixed with `repeated_test 2.3.1`, `clize 5.0.0`