On Mon, 18 Aug 2025 at 01:58, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Perhaps the release notes could be useful as a sanity check though.
> Say, if a person appears in the last three release-note lists and
> is not in the curated list, maybe that is a name to take another
> look at.
I think now that committers are using a more consistent format in the
commit messages that includes email addresses, it should be easier to
script up that sort of thing. e.g. HAVING MAX(commit_timestamp) -
MIN(commit_timestamp) > INTERVAL 'N years' AND count(*) > M. Perhaps
it'd still need a manual review process, but something like that might
stop people falling through the cracks and being forgotten.
David
