The lazy consensus passed. I will proceed with some initial PRs to pave the way
On Tue, Nov 25, 2025 at 9:38 AM Jarek Potiuk <[email protected]> wrote: > I think it's more of an idea and precise timing of moving each of them is > to be determined - but yeah, what you described is my thinking too. > > On Tue, Nov 25, 2025 at 7:52 AM Amogh Desai <[email protected]> wrote: > >> Jarek - do we want to clarify which distributions we want to do it for >> right now once the consensus is reached? >> Or is the consensus for the general idea of "we will follow this method >> going forward"? >> >> I assume only task-sdk and airflowctl, with docker_tests, k8s_tests, and >> helm_tests to come later? >> >> Thanks & Regards, >> Amogh Desai >> >> >> On Mon, Nov 24, 2025 at 1:56 PM Jarek Potiuk <[email protected]> wrote: >> >> > And link to discussion here of course: >> > https://lists.apache.org/thread/7fc1lj4hoyxv4tnms666bn6y287t9c0n >> > >> > On Mon, Nov 24, 2025 at 9:10 AM Jarek Potiuk <[email protected]> wrote: >> > >> > > Hello here, >> > > >> > > Following the discussion in I would like to ask for a lazy consensus >> > on: >> > > >> > > * using `e2e` for end-to-end kind of tests (docker-compose driven with >> > > python-on-whales, and a common approach on mounting the local sources >> for >> > > fast iteration >> > > >> > > * placing them under `tests` folder in relevant distributions >> alongside >> > > `unit`, `integration`, `system` >> > > >> > > Example task-sdk structure after moving: >> > > >> > > task-sdk >> > > src >> > > tests >> > > unit >> > > task_sdk >> > > some_unit_test.py >> > > e2e >> > > dags >> > > logs >> > > task_sdk >> > > some_end_2_end_test.py >> > > >> > > No final decisions yet on kubernetes tests that are using helm-chart >> - we >> > > will discuss it later. >> > > >> > > The lazy consensus will end Thursday, 27th of November, 10 am CET : >> > > >> > >> https://www.timeanddate.com/countdown/generic?iso=20251127T10&p0=262&font=cursive >> > > >> > > No need to answer unless you do not agree with the consensus. >> > > >> > > J. >> > > >> > > >> > >> >
