[Elections] Checking in / next steps

[Peter Crowther]

[Not a great email to read.  Long and political, and I apologise for
that.  However, I think there are the seeds for several relevant
discussions in here, as well as my usual musings and general dross.]

> From: [...] Daniel Vainsencher
> Many are stating it, but do you (or someone else) mind actually 
> *arguing* why the elections are an urgent priority for this team?

Let's try the following as straw men.  What do folks think?

1) The current leaders said they'd allow elections on a specified date.
Those with integrity will at least wish to do what they said they'd do.
Any without such integrity should be removed from office before their
lack of integrity can do further damage.

2) We have a community that is fragmented and fragmenting further.  I
conjecture that: one of the reasons for this is the way in which the
current leaders came into power; a second is the way that some subset of
them have chosen to use that power without much consultation (yes, I
will state event(s) and name(s) that I consider to fall into this
category if you wish to enquire further); and still another is that the
current leaders appear to be unworried about the fragmentation, which
will tend to accelerate that fragmentation.  Every day this situation
remains unaddressed is another day of the community slowly fragmenting.
We don't have the person-power for that!

3) Some (at least two) leaders wish to stand down.  They signed up for a
tour of duty of a specific length; we should not prolong that.

> As I see it, Squeak has managed reasonably well with unelected 
> semi-officials.

Speaking personally, I disagree - see 2 above.  However, I'm aware that
my opinions on this are contentious.

> there are several reasons to invest time (a couple 
> of months, 
> not a year) in having smooth tools for voting (including on 
> elections):
> 1. If its smoother, more people will participate.

Relevant if the opinions of the lazier participants would change the
outcome; irrelevant if they wouldn't.  Do we have any feel for how
smooth something has to be in order to get an acceptable level of
participation?  For that matter, what *is* an acceptable level of
participation?  100% is unrealistic, unfortunately.  We see somewhat
under 50% in local elections in the UK...

> 2. A good system (especially in the sense of using good tallying 
> algorithm) will be more trusted by people, therefore making any 
> elections done be more meaningful.

Can we firm up what 'good' would mean, in the sense of both system and
algorithm?  To me, a good algorithm:

- discards the fewest votes in reaching its conclusions;

- has known and preferably well-described properties to which we can
refer;

- May prefer a set of candidates for which most people vote (Condorcet
does this) or may prefer candidates with a wide constituency (STV tends
to do this).  We need to decide which property we'd prefer.

A good system:

- is automated, not manual;

- is demonstrably independent of any candidate in the election.

> 3. A fully automated system will allow any voter in the community to 
> raise issues, rather than just the officials that can do the required 
> manual system in a non-automated system.

It will also have to allow automated definition of who forms the
community, otherwise it isn't fully automated.  This is, I think, the
source of Lex's and my differing points of view; we have different
opinions about whether, and how, this can be automated.

> 4. Whatever system we come up with should be tested in production on 
> non-critical decisions (web site layout, whatever) before it 
> is used to 
> decide elections that give people official positions that we are not 
> used to having in the community.

Yes.  To me, this indicates that an off-the-shelf system should be
preferred, as such a system would give more time for such tests and
would also come with certain assuances that others have tested it in
production.

> 5. An election is a part in a system. What are we electing 
> people for? 
> what decisions can the elected make beyond others? for how 
> long are they 
> elected? holding an election without considering these questions is 
> meaningless. We will need discussions of this on squeak-dev (even if 
> guided by some proposals from this team), and we will then need to 
> decide on one of them.
> 
> This is the most important point - we need to make the whole 
> system one 
> that the community is happy with, not just the list of 
> people. This will 
> not happen on a deadline just because that's what fits some 
> preconceived plan.

Indeed.  However, if this is part of our [your - I have recently been
reminded that I am not a member of the team] remit then this team is at
best poorly named and badly under-resourced.

I'm going for an initial guess of 'the current board structure, but
candidates may run with the express intention of changing that
structure', 'as now' and 'I'll walk out of the community if it's longer
than a year, and things are moving so fast that I'd much prefer to see
an interim government of 6 months' respectively as my own answers to the
above questions.  I strongly suspect that we'll see candidates coming
forward whose sole aim in standing is to try to change the way Squeak is
controlled; I would argue that any system we put in place should allow
change in this way until we get the arrow pointing in roughly the right
direction.  Once we know roughly what's wanted, we can start to firm up
the structure*.

> You're right - our goal is not building tools. And in my opinion, our 
> goal is also not "to have an election". Our goal should be 
> social, and about involving the community.

... in what?  Tea and cakes and polite agreement that The World Isn't
What It Should Be and that Something Should Be Done, and then everyone
toddle off home with the vague unease that maybe something more should
have happened just then?

We can stop this and go home tomorrow by saying "we have decided how to
bootstrap the election process.  Here you are, folks.  We are
bootstrapping it by mailing squeak-dev and stating that we have decided
that the wider community should come up with the details of Squeak's
governance.  There you go, folks, you can take it from here, we've
finished."  We'd be laughed out of court.

I agree that any system we construct will require ratification by the
community.  How does one do that, though?  We have no defined process
for it; we need to bootstrap the process.  We could use the system to
ratify itself, but that has dangers - at one extreme, we pick (say) Alan
and say "You have command.  All you need to do is ratify it by saying
so, because you have command."  At another extreme, we stagger around
forever trying to work out a process for ratifying the pocess for
ratifying the process for ... for ratifying the process for deciding the
decision-making system.

I suspect our initial choice of system, within a wide range of options,
will be a non-event.  The community will get a system that works at
somewhere between 80% and 90% of what it could do; and that will be good
enough for the community to use that system to decide to move to a
system that is more closely aligned to its wishes.  People will simply
accept it as an initial system, as it's a step forward from where we are
now.  If we choose to set the system in tablets of stone and hand them
down from on high... that's a different matter.  Squeak attracts a very
divergent community by its nature; imposing a mechanism and asserting
that it will be fixed for all time will give us the grandmother of all
arguments on the main list.  Let's do something that will work well
enough for a period.

[Aside: have you ever played Nomic
(http://www.earlham.edu/~peters/nomic.htm)?  If not, you may wish to dig
out a set of the rules
(http://www.earlham.edu/~peters/writing/nomic.htm#initial%20set) and
read them.  Most instructional for a debate of this kind.]

		- Peter

* But then, I'm somewhere between anarchist and individualist at heart -
I maintain that societies are aggregates of individuals and that it is
the individuals that define the way in which society reacts rather than
the other way round.  The Squeak community appears to exhibit this
behaviour more strongly than many other communities, I suspect because
it attracts intelligent and individualistic people to its exquisite
personal computing environment.