[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

TC voting and governance process (was: systemd bad press? score card?)

Joey Hess <joeyh@debian.org> writes:
> Russ Allbery wrote:

>> I think we're still in the middle of our process, which I understand
>> that a lot of people outside the project find baffling and protracted.

> Well, not only outside the project.

> The tech ctte has always operated in the past by coming to a consensus
> and then voting to satisfy the constitution's procedural requirements.
> Since the constitutional procedure as applied to the small scope of the
> ctte is now demonstratably broken in cases where a consensus cannot be
> reached, why are we waiting around for the full horribleness of it to
> manifest?

> I'd be in favor of a GR to fix the constitution to not mandate that the
> ctte use untested and nonfunctional procedures.

This message sparks the somewhat unbaked thoughts that I had when Guillem
posted his draft GR.

I think it's appropriate and important for the project to periodically
examine its governance structures and double-check that they are working
in the way that the project as a whole desires.  Events that test those
governance structures provide a good opportunity for doing so with some
concrete data attached.

I can only speak for myself, but I suspect this sentiment is common among
the TC members: my work on the TC is intended to serve the project as a
whole.  If at any time that work is not helpful, not well-directed, not
handled in a way that's appropriate for the project, or otherwise at odds
with the needs of the project, that's feedback that I want to hear.  If
the mechanisms we have in place for doing that work aren't appropriate, we
should fix them.  It's always hard to know how much weight to put on the
concerns of the people who speak up on mailing lists, so it would be very
helpful to get that feedback in some formal form, after appropriate debate
and involvement of people who may have opinions but don't want to get into
the mailing list discussions.

Over the course of this discussion, we've already identified one issue
with which many of us on the TC are uncomfortable, namely the 2:1 majority
requirement for the project to overrule the TC by GR.  There was already
some discussion of a constitutional amendment to remove that provision.
Anthony Towns also raised some important points about the way that we
treat FD in our votes, and I must say that the degree to which our voting
system fails later-no-harm came as quite an unpleasant surprise to me.
That was always there in the abstract, but it sometimes takes something
like this to make the problem concrete.

At the moment, I find the idea of other people discussing these issues and
any others that have arisen and coming up with a good plan of action while
I sit back and watch horribly appealing for some reason.  :)  I'm not sure
if now is the best time, or if we should wait until we've gotten to some
steady-state position on the init system debate and then start the
procedural discussions.  I'll leave that to other people's discretion.

The only thing that I would ask for as an outcome here is some clear
guidance on how we should handle contentious issues.  Right now, we have a
resolution method that's biased towards arriving at a decision, even if
it's a contested decision, in the form of a provision for a casting vote.
If the project wants that changed, I want to see it changed as well.  But
we need to know what should replace it.  Should we not make a decision?
Should it become a GR?  Should we have a higher bar before we even
consider a question?  I think that's a hard problem, and may require some
discussion.  The TC needs guidance there so that we know what's expected
of us in controversial matters.

Russ Allbery (rra@debian.org)               <http://www.eyrie.org/~eagle/>

Reply to: