Rotation of the release manager role

[Julian Foad <ju...@apache.org>]

Stefan Sperling wrote (in private; quoted with permission):
[...]
> I have managed the last couple of releases, and I don't think that
> relying on a single person for this is sustainable. In the past,
> we've had release managers like Ben and Julian who basically grabbed
> the RM hat and did not consistently let go of it until they moved on from
> the SVN project.

Stefan you make a good point. When I did RM for that period it didn't consciously occur to me to ask for someone else to take over after a while as a matter of project health, it just seemed like a job that still needed doing. And that's despite my subconscious memory having heard you write about this in the past.

> I am still around and willing to help get releases out
> the door and fix bugs. But my position has always been that I do not want
> to develop an exclusive responsibility for release management.
> Having too few people with RM experience makes the project unhealthy.
> The only way the previously acting RM can avoid that is to step aside.

Agreed.

Can we do something to make this more explicit, visible, memorable? Such as adding a step zero in the release procedures:

Procedures:
Step 0: Were you RM for the last 1.x release or last two patch releases or so? Then it's time to step aside. (See rationale elsewhere.) Please email dev@ for a volunteer...

Thoughts?

-- 
- Julian

Re: Rotation of the release manager role

[Daniel Sahlberg <da...@gmail.com>]

Den tis 11 jan. 2022 kl 16:44 skrev Julian Foad <ju...@apache.org>:

> Stefan Sperling wrote (in private; quoted with permission):
> [...]
> > I have managed the last couple of releases, and I don't think that
> > relying on a single person for this is sustainable. In the past,
> > we've had release managers like Ben and Julian who basically grabbed
> > the RM hat and did not consistently let go of it until they moved on from
> > the SVN project.
>
> Stefan you make a good point. When I did RM for that period it didn't
> consciously occur to me to ask for someone else to take over after a while
> as a matter of project health, it just seemed like a job that still needed
> doing. And that's despite my subconscious memory having heard you write
> about this in the past.
>
> > I am still around and willing to help get releases out
> > the door and fix bugs. But my position has always been that I do not want
> > to develop an exclusive responsibility for release management.
> > Having too few people with RM experience makes the project unhealthy.
> > The only way the previously acting RM can avoid that is to step aside.
>
> Agreed.
>
> Can we do something to make this more explicit, visible, memorable? Such
> as adding a step zero in the release procedures:
>
> Procedures:
> Step 0: Were you RM for the last 1.x release or last two patch releases or
> so? Then it's time to step aside. (See rationale elsewhere.) Please email
> dev@ for a volunteer...
>

Yes, why not add this as the first headline under
https://subversion.apache.org/docs/community-guide/releasing.html#release-creating
?

There is also a kind nudge under
https://subversion.apache.org/docs/community-guide/roles.html#release-manager
("the RM role is currently rotating among various volunteers.")

Kind regards,
Daniel