Yup, I have a Draft PR already, once the Scheduler HA and the Task Instance Endpoints PRs are merged, I will rebase that PR.
On Wed, Oct 7, 2020 at 8:00 PM Tomasz Urbaszek <turbas...@apache.org> wrote: > As we are close to releasing 2.0 alpha it seems feasible to resurrect this > thread. > > I think we should run black on 2.0 as the last commit before releasing it. > I would even argue that it should be done for alpha release. > > What do others think? > > Bests, > Tomek > > On 2020/06/27 22:51:04, Ash Berlin-Taylor <a...@apache.org> wrote: > > Yes, to Black in principle. But (and this is a big but) no, not yet. Or > not without testing how it affects our ability to cherry pick back to the > release branch. > > > > (My default would be to assume it makes them harder/almost impossible > and this should be the almost last thing we do before we release 2.0) > > > > -ash > > > > On 27 June 2020 22:02:41 BST, Philippe Gagnon <philgagn...@gmail.com> > wrote: > > >It's a good idea. > > > > > >It will make reading the codebase easier, and besides the whole Python > > >ecosystem is slowly moving towards adopting this code style. I > > >personally > > >have been a fan ever since the project launched. > > > > > >With regards to open PRs requiring a rebase, it's an annoyance for sure > > >but > > >if we do decide to standardize on any code style (which we should, > > >black or > > >not), we'll have to pull the bandaid eventually. > > > > > >Just my two cents. ;-) > > > > > >Philippe > > > > > >On Sat, Jun 27, 2020 at 4:13 PM Kaxil Naik <kaxiln...@gmail.com> wrote: > > > > > >> Hi all, > > >> > > >> I would like to open the discussion to enable Black ( > > >> https://github.com/psf/black) - *The Uncompromising Code Formatter* > > >for > > >> automatic formatting of Airflow's entire codebase. > > >> > > >> I have created a WIP PR at > > >https://github.com/apache/airflow/pull/9550 > > >> > > >> Some of the caveats: > > >> > > >> - All the currently open PRs would have some kind of conflict > > >errors > > >> - It *might *make backporting harder (but should be ok'ish if we > > >enable > > >> black on v1-10-test but not 100%) > > >> - There are known issues with "line-lengths" not being honoured by > > >black > > >> (https://github.com/psf/black/issues/1161) and the workaround is > > >to use > > >> "#fmt: off". > > >> > > >> I would love to hear what the community thinks about it. > > >> > > >> Regards, > > >> Kaxil > > >> > > >
