Jérome Perrin

Ok, then it's maybe better I close this and let you handle it ?

Thank you @vince-fugnitto I did not notice your edit, I'm sending a pull request to drop the default keybinding in that case.

If my bisect is correct, the regression comes from 9569493453a39d63064ed7c20653987ba15c99e5

> If my bisect is correct, the regression comes from [9569493](https://github.com/pypa/virtualenv/commit/9569493453a39d63064ed7c20653987ba15c99e5) The reproduction steps were about using python3 to create a python2 virtualenv. Before 9569493, the created `1.venv/bin/python` is a...

Thanks for the feedback, I did not know that of that "discouraged interface". I assume this is the "deprecated method" from the examples in https://stackoverflow.com/a/16775731 , ie: ```python def application(environ,...

In my understanding, the docstrings matters only for publishing (when an object or method is accessed by URL in an HTTP request) and not for "restricted python security" (when accessing...

To prevent confusion, this is about cancellation, but a completely different one from the `Cancel notification for unknown message id ...` issue that was discussed in https://github.com/openlawlibrary/pygls/issues/240#issuecomment-1174672020 .

Thanks @tombh > Do you think it warrants anding to the changelog? yes, we should probably add a line in *Added* section, with something like that: Support cancel notification for...

@tombh thanks, I changed API to expose futures directly, as discussed. While doing this I realised that `progress.create` and `progress.create_async` methods were not tested, so I added test for them...

> Oh right, that's a bit more significant than I thought then. I suppose one idea at least is to patch something like sys.stdout.write with a simple regex to exclude...