[maemo-community] [Council] Voting system progress?

From: Andrew Flegg andrew at bleb.org
Date: Sun Jan 18 19:20:50 EET 2009
Dave wrote:
> Andrew Flegg wrote:
> >
> > Some kind of range voting seemed to get a lot of traction on ITT from
> > some election experts; but was dependent on being written up in a way
> > which makes sense in plain English.
> Honestly, I'm not a fan of range voting. What's wrong with preferential
> voting, which is easier for voters? Single Transerrable Vote, Schultze,
> and Condorcet are all available as counting methods for these - I prefer
> the easy-to-understand STV over the others, and we have software that
> does STV counts.

Agreed. I was clumsy with my terms. s/range/better/. Sorry for the confusion.

Ryan had a task to raise the issue, and there seemed to be consensus on ITT. His conclusions are here:


There seem to be two proposals there. Ryan, are you imagining putting both to the vote, was there a consensus on -community & ITT? Having two "change" options: does that mean people will complain about splitting votes if "none of these" wins; or similar? (What we *do* if none of these wins, I'm not sure. Probably, continue the next election under the current rules and hope the next council take it as a sign to pick up the task again?)

Let's consider the practicalities: what is the s/w we've got capable of? The email/token stuff worked well last time, and so proven software is a plus :-)

Presumably, the existing s/w can handle the one-member, one-vote referendum; so we'll want to keep that installation no matter what (should referenda have STVs?)

> Now that I read this, indeed, we say one vote per person, the 5
> candidates with the most votes get elected.

Yes. The rules are codified, is that so strange?

> I don't like premature red tape, but if you want to run a referendum to
> say that voting should be using a preferential voting system, with a
> counting system to be decided by the council before the start of the
> election, and that the 5 candidates who win the election get elected,
> I'm fine with that.

I'm not sure I'm happy with the current council choosing how the next election is run. That just seems open to abuse.

However, if it outlines the approach, the of the s/w used are implementation details, of course. However, I think the outline should be there to make it clear, ahead of time, how the winners of the election are chosen. With one-member, one-vote it's straightforward; a plain english description of STV or RRV should be achievable - or it fails on one of the requirements of a voting system: understandability without a statistics background.

> I don't think you should get more specific than that on saying exactly
> what system will be used for exactly this reason - we'll end up having
> annual referenda to change what should be implementation details.

I would hope that a better set of rules would mean that changes wouldn't be quite so frequent.

As for the karma requirements, I'd suggest that if we want to change them, that's a separate referendum (avoiding the issues Debian's just faced with the lenny-release voting). It could be, that with ITT now counting towards it, the numbers aren't so contentious.

How many ITT posts alone would someone require to meet the 25 point limit? (Such a member would hopefully be atypical)



Andrew Flegg -- mailto:andrew at bleb.org  |  http://www.bleb.org/
Maemo Community Council member

More information about the maemo-community mailing list