[Linux-aus] AGM, Election, Name Change & Rusty Wrench Award

[James Polley]

On 16/12/2012, at 4:45 AM, David Newall <davidn at davidnewall.com> wrote:

> On 15/12/12 19:23, Chris Neugebauer wrote:
>> http://en.wikipedia.org/wiki/File:IRV_counting_flowchart.svg
>>
>> In the case of multi-candidate elections, such as the election of
>> Council members, or the name change vote, the terminating case is not
>> that a candidate has a majority, but that the number of candidates
>> remaining is equal to the number of seats remaining to be filled.
>
> I'm not sure this flowchart is exactly appropriate in this instance, but
> it does cover most of the process.  For example, to change the name at
> least 75% of votes cast must support the change, not a simple majority.

But since this election is only being used to get a sense of what the
members want to do, and will not result in the name being changed (it
just gives guidance to the incoming council, if they choose to proceed
with changing the name) this is irrelevant.

> From this, it follows that if over 25% of votes cast support keeping the
> same name, no distribution of preferences can provide the required
> support needed to change the name.

I'd be fascinated to read the "logic" that lead you to this
conclusion. Not that it's in any way relevant, since this election
doesn't need a 75% majority for anything.

> The flowchart referred to is
> deficient in this regard, as is the PHP code to which Anthony Towns
> directed me, at
> http://bazaar.launchpad.net/~memberdb-owners/memberdb/trunk/view/head:/authenticated/election-result.inc.
>
> Another deficiency in the flowchart, also relating to keeping the
> current name, is that preferences following "no change of name" must not
> be distributed, nor may it be eliminated.  Allow me to explain in
> greater detail: For sake of example, suppose only three choices are
> available, "name 1", "name 2" and "no change"; and that first
> preferences fall in ratio 40 : 34 : 26.  According to the flowchart, "no
> change" should be eliminated as the minority candidate; however no new
> name can garner 75% support if 26% of voters support no change and
> eliminating that candidate with 26% of votes would pervert that truth.
> Eliminating "no change" as "last place candidate" would be improper.
> Since "no change" must not be eliminated, it follows that preferences
> cannot flow from that candidate.
>

I could explain why this is complete nonsense, but I'm not going to
bother since it's irrelevant - the 75% supermajority applies on a vote
to change the name; this election is not that vote.

FWIW, MemberDB does report on the votes cast at each stage. As an
example, take https://www.linux.org.au/membership/index.php?page=election-result&id=14
- scroll down to the "Ordinary Committee Members" section.

You can see that MemberDB reports, first, on first-preference votes;
then it shows the results of redistributing preferences for the loser,
and so on.

If 26% of people vote "no change" as their first preference in this
election, that fact will be reported. The information will not be
lost, even though their votes will be distributed amongst their
second, third, fourth etc preferences.

Even if we don't get 26% of people listing "no change" as their first
preference, we'll be able to see how many preferences it does get. For
instance, if a lot of people like one or two names, but would prefer
"no change" over the rest of the names, I expect that we'll see "no
change" get a huge swell of support after a few rounds of runoff.

I look forward to your nonsensical numerological analysis of those
results as soon as they come to hand.

> Of course, it must be admitted that this election cannot directly result
> in change of name.

Ah, you admit that all your previous waffle was irrelevant. So why did
you waste your time and ours?

Given that you have such passion and energy to devote to this subject,
and apparently *oodles* of free time to devote to it, I would love to
see you running for election.

In fact, I just attempted to visit memberdb and nominate you  as an
OCM - but memberdb won't allow me to nominate davidn at davidnewall.com,
because "The email address must be that of a current member of the
organisation. You may want to check with the person that this email
address is the one they registered with.."

If you could let me know what email address you registered with when
you joined LA, I'd be happy to nominate you.

>  That may only be done by special resolution, which
> requires the text of the proposed resolution to be circulated 21-days
> prior to the meeting.  What we currently are voting for is to choose the
> name that will form the basis of that special resolution.  The Council
> probably can, should they so choose, eliminate "no change" if it is the
> minority candidate, but I argue this would be deceit, as well as
> pointless.  The point is, members who want to keep the name unchanged
> don't need to vote in this election; they need only vote against the
> special resolution that necessarily must be proposed.  I urge members
> who want to keep the current name to vote in this election as a courtesy
> to the Council; just as I urge the Council never to treat "no change" as
> a minority candidate.

This paragraphs is just as much irrelevant nonsense as the rest of your email.

If members want to keep the current name, voting "no change" as their
first preference is not just a "courtesy", it sends a clear message to
the council. As you said, regardless of the outcome of this process,
if 26% of people want to keep the current name, a special resolution
to change it will fail at an SGM should the council ignore this
message and proceed with the name change anyway.

But by the time it gets to an SGM, the council members will have
wasted a lot of time and effort reserving new names and preparing for
the SGM. Letting the council know *now* that you prefer no change
isn't just a courtesy to the council, it's a way to make sure the
council doesn't waste its time and money.

>
> Finally, so as not to disadvantage members who are unable to attend the
> AGM, I ask that a postal ballot be conducted, for which a (non-committee
> member) returning officer must be appointed.

The name change vote will not happen at the AGM, so this is just as
irrelevant as the rest of your screed.

Unless this one paragraph is meant to be read in isolation, and does
actually refer to the AGM and not the hypothetical SGM where a vote on
the name change would be held. In that case, the process we have in
place ensure that members who are unable to attend the AGM are not
disadvantaged is proxy votes; please explain why you think appointment
of a proxy still leaves you at a disadvantage.

>
> It is a matter of great disappointment to me that the previous council
> mindfully and explicitly chose to discard the old constitution without
> including electronic voting at elections in the new one; that provision
> had been approved by members (even if the Director General had not yet
> been notified.)  It cannot be said that this was an oversight as the
> point was raised during extended debate.  I regret that a postal ballot
> will be a significant expense, however the only alternative (under the
> new constitution) would disenfranchise a great portion of our members.

You're correct in saying that the current constitution does not
explicitly list electronic voting as a voting mechanism. It also
doesn't explicitly mention voting by show of hands; voting by voice;
voting by division; secret votes on ballot papers etc. If the voting
method needs to be explicitly mentioned in the constitution, we're up
a certain creek without a paddle.

>
> _______________________________________________
> linux-aus mailing list
> linux-aus at lists.linux.org.au
> http://lists.linux.org.au/listinfo/linux-aus