That script should really be moved to the tools repo.
Also I think there are a large number of PRs which the script isn't
pinging at the moment, but which are completely stale and haven't been
touched (for years in some cases). Perhaps we could have a "no activity"
ping...and after so long of continued pinging and still no activity it
gets auto-closed. Activity being counted as a comment or push from
anyone (not just a label/milestone change).
Matt
On 14/03/2022 10:37, Mark J Cox wrote:
Unfortunately the autocloses happened due to the bug now fixed[1].
But they can always be reopened again.
[1]
https://github.com/iamamoose/openssl-metrics/commit/49927d122e39d0a534e82f4a611fc9a06e84a95b
Mark
On Mon, 14 Mar 2022 at 10:35, Tomas Mraz <to...@openssl.org> wrote:
On Mon, 2022-03-14 at 10:29 +0000, Mark J Cox wrote:
We have a script that runs daily and makes sure things needing action
for OTC/OMC are pinged if they get old. It also autocloses issues
where it was waiting for the reporter with no action, or waiting for
a
NDA for a significant amount of time.
Because 3.0 wasn't out, it ignored everything with the "post 3.0.0"
milestone. It's time to turn off that exception. However looking at
the PR list this will cause a large number of PR's to change state:
It will ping OMC about 4 stale issues waiting for OMC
It will ping OTC about 14 stale issues waiting for OTC
It will ping committers about 12 stale issues waiting for committers
It will close 5 issues waiting for the creator to make changes >90
days
It will close 4 issues waiting for a CLA for >180 days
Any objections to this, or a preferred time in the future to make the
change?
I would not autoclose the issues to be autoclosed - IMO the script
should just ping at least once and autoclose only after a week or so,
if there is no update.
--
Tomáš Mráz, OpenSSL
