Open Stack

> From: Jeremy Stanley [fungi at yuggoth.org]
> Sent: Tuesday, April 23, 2019 2:45 PM
> To: openstack-discuss at lists.openstack.org
> Subject: Re: The python black project.

*SNIP*

> Now it's starting to sound like you're suggesting auto-proposing
> follow-up changes to each user-proposed change while it's still
> under review.

That was what I was trying to explain and failing. That is what I am proposing. :)

> I think it's probably technically possible, though
l> ikely to ~double our overall change volume insofar as review and
> testing go (especially when folks inevitably ignore them and so they
> get rebased and updated every time the parent change is updated).
> And before you say "it's okay to skip running CI jobs on those" it's
> really not, since we're not likely to blindly trust autopep8 not to
> propose a subtly broken modification.

Yeah, I wouldn't recommend blindly trusting it either... it could short circuit all the other tests though, like:
do the linting tests before any other tests. if it fails, end review of the whole PS and submit back a cleanup patch. Then start over with the testing. So test wise, it wouldn't be too expensive. But it could easily double the number of reviews like you point out, if that is an issue.

> Jeremy Stanley

Thanks,
Kevin