I actually don’t like the system “N votes for N seats” – it seems much better to have M votes for N seats, where M < N – perhaps much smaller.

Think in terms of preferences – say there are 5 candidates and 3 seats. You may have 2 candidates you like a lot, and no clear preference for the other 3 – your 3rd preference gets as much weight as your first two. If you give everyone fewer choices, then you only get people voting for the candidates they’re passionate about – imagine how the election might have turned out if everyone had only 3 votes for 7 seats. All totals would have been a lot lower, but we would not have people getting “meh” votes.

STV would work well, IMHO, if we could automate counting properly. And if we Aussies and Irish could explain it to all the list-voters and first past the posters But in the absence of STV, it seems better to limit the number of ballots people have to ensure that we get people voting for the candidates they really like.

Instead, we could:

a) Make sure that there is an interest leaving behing the current system

b) ask for alternatives i.e.
http://en.wikipedia.org/wiki/Instant-runoff_voting

c) in case we get more than one alternative, discuss and choose the right one, based not only on theory but also in practical terms i.e. free software available, functional and stable.

When Luis resigned, I effectively lost one seventh of my vote. If I had less votes to cast, then I’d have lost even more of my vote.

Quim: you are right. I should get off my arse and follow through about this.

Regarding availability of software, you might consider looking into Devotee, the Debian Vote Engine. It supports electronic voting based on GPG-signed emails, and excellent generation of reports afterward. I don’t know if it supports multiple-winner elections, but if not, I suspect support for such would not prove incredibly difficult to add; the well-thought-out design makes development quite straightforward.