As pointed out on Zulip, we recently moved to nextest and are lacking documentation for it (e.g. linking to the official nextest docs, outlining some of the most useful options, etc.).
One trick @jdm mentioned on Zulip that might be worth documenting:
Also, I helpful thing I just figured out: use --stress-count infinite with a test filter (like test_show_and_hide_ime) to run a test repeatedly and increase confidence that a particular change makes intermittent timeouts go away.
As pointed out on Zulip, we recently moved to nextest and are lacking documentation for it (e.g. linking to the official nextest docs, outlining some of the most useful options, etc.).
One trick @jdm mentioned on Zulip that might be worth documenting: