On Wed, Feb 01, 2012 at 08:56:01AM -0500, Benson Margulies wrote:
> I'm also trying, more pointedly, to head off the 'AWOL mentor' problem
> by asking mentors to think about, and state, the commitment they are
> making.

A lot of harsh words have been directed towards "AWOL Mentors" lately.  The
more I've thought about it, though, the harder it gets to muster indignation
towards these well-meaning volunteers.

The root cause of Mentor attrition lies in the fundamental assymetry between
the investment of a podling's core contributors and the investment of its
Mentors.  People come and go in open source, and because Mentors have
proportionally lower personal investment in their podlings, it is predictable
that they will drop away at a greater rate.  Heck, we probably have enough
data at this point to build a nice model and calculate how many fractions of a
Mentor a podling can expect to lose month-by-month.

Advising Mentor candidates up front that they are making a long term
commitment doesn't change the investment assymetry.  I'm skeptical that it
will accomplish much beyond scaring off some people who could otherwise have
done some good.  Even Mentors who "go AWOL" will often have made valuable
contributions before they drift away.  Since the bulk of Mentoring work
happens at the start of Incubation, is it so bad to have a lot of help around
when a podling needs it most?

For what it's worth, in my assessment, the problem is not that podlings shed a
few Mentors along their journey -- it's that unlike Apache PMCs, podling
oversight structures are fragile and do not easily withstand personnel churn
or self-replenish.

It seems that we only look for new Mentors at times of crisis, when the
official count drops below three.  I humbly suggest that we might seek to
expand podling IPMC representation during times of plenty, and that instead of
looking to our beloved usual suspects, we look to standout podling
contributors who have demonstrated a thorough understanding of the Apache Way.

Marvin Humphrey


---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org
For additional commands, e-mail: general-h...@incubator.apache.org

Reply via email to