Elections January 2006

elections@lists.squeakfoundation.org

11 participants
36 discussions

Checking in / next steps
by Brent Vukmer 12 Jan '06

12 Jan '06

Hello folks. I apologize for being silent so long. I think I overdid my traditional Christmas celebration/hibernation... :) Thanks Lex, Daniel, Peter, Ken and Cees for your discussion. I'd like to hear from Benjamin and Masashi, too. I will post a longer email (probably sometime tonight EST) with my thoughts about the upcoming election. (In preparation for that, I'm re-reading the elections posts so far.) Here are some initial thoughts about our short-term priorities, in which I realize I'm stating the obvious :). The immediate priority for this team is the upcoming election in February. We need to come up with a list of candidates and a means for choosing the candidates, and then we need to actually run the election. I am very strongly biased towards using a low-tech solution to I would prefer that we spend a minimum of time in the short-term on tool-building and reputation-system algorithm analysis. I've skimmed the various suggestions (SqP, Condorcet, Agora, wiki) In the short term, I prefer that we err on the side of simplicity, and focus on "social engineering" rather than building tools.

4 5

RE: [Elections] Re: Checking in / next steps
by Peter Crowther 12 Jan '06

12 Jan '06

> From: [...] Lex Spoon > Is there a website people can be pointed to for information Not yet, to my knowledge, but I agree with you that there needs to be at least a Swiki page soonest. - Peter

1 0

RE: [Elections] A simple SqF board elections solution using SqP
by Peter Crowther 12 Jan '06

12 Jan '06

> From: [...] Cees De Groot > Please. please please please. KEEP IT SIMPLE. Technologically and organisationally? :-) > You haven't been tasked with running the next elections for President > of the United States. We don't need biometric identification, trusted > third parties or a notary public to count vounts, whatever. Just the > Net version of 'please raise hands'. > > We are just a bunch of guys in a *friendly* community. We have our > technical differences, but I really cannot imagine any inside party > seriously attempting to mess up these elections. Nor an outside party. Lex, at least, has repeatedly raised the points that (my interpretation): 1. He regards the current leadership as illegitimate; 2. He does not trust any system for selecting voters if it can be affected by the actions of a small power group; 3. (reading into some of Lex's messages) I *think* he's concerned about (2) in the case of SqP because some of the roots are part of the current leadership. Lex, these are my interpretations and there's a strong chance I'm misrepresenting you here; I'd very much appreciate clarification of your position on these points. For the record, my own positions are: 1. The current leadership made a power grab in a power vacuum. It was not legitimate by any *democratic* standard*; legitimising the position of that leadership would be prudent if it is to be able to resist accusations common to most revolutionary governments. 2. I'm concerned about the selection system in theory, as it clearly *could* be constructed so as to be biased; 3. I am not concerned with a SqP-based selection system in reality, as I've seen no indication that the roots are behaving inappropriately. > A system is needed that is reasonably objective and which stands a > reasonable chance of listing the right people on the voter's registry. Where 'the right people' needs definition. > It's easy to discuss this issue to death, to build in very strong > safety guards - but let us do something *simple* and just go for it. To me, the key question when evaluating a possible system is "how likely is it that the inevitable errors in this system alter the result?" There will be errors in any system we devise. One problem I have with answering that question in the case of Squeak is that we are at the same time trying to work out who "the right people" are and construct some form of voting system for those "right people" to return multiple candidates. I'd much rather hand the second job off to some body of people who have done it before, hence my suggestion of a known system with known properties (a Condorcet variant). Unfortunately, I don't think we *can* hand the first job off to people who have done it before, so we're stuck with having to work it out. > On 1/8/06, Peter Crowther <Peter(a)ozzard.org> wrote: > > Indeed. For recording and counting votes, however, I'd settle for a > > system where the system's owner has no interest in tampering > > with the result. > > > Should I take that as an insinuation? I won't. Just let me put > forward, for the record, that I don't have any interest in tampering > with the result (as owner of the system where SqP runs). I wouldn't > know why... No, you shouldn't, and I should have worded it as '... the system's owner *can be seen to have* no interest in...'. Any election system run by some of the people up for election runs the risk of accusations of bias, either in voter selection or in counting - see, for example, the 2000 and 2004 presidential elections in Florida, where accusations of bias were levelled in both cases. Systems that are clearly run by an external body are at least freer from accusations of bias in counting. The castaways are clearly already sensitive to the perceived legitimacy of their power, hence the long-standing promise of elections in February. We should also be sensitive to the perceived legitimacy of those elections. They must be seen to be unbiased, otherwise the undercurrents that are still present in the community from the original formation of the power structure will remain or get stronger, with an increased tendency for the community to fragment. Assuming we prefer the community to be as large and as cohesive as possible, this would be a poor result. Personally, Cees, I'd be quite happy with a voting system you wrote and/or ran; I can't see any way *in fact* that you'd want to tamper with it. But it's difficult to get rid of the possibility of the system being *accused* of bias in such a circumstance. Has that made my position any clearer, or has it just made things worse? - Peter * Of course, this begs the question re democracy being the appropriate mechanism for governance of the Squeak community, yada yada... but this is called the elections team, so I'm assuming that a democracy with an elected leadership is considered appropriate. The community ran under a very different model for a number of years, and one of the issues appears to be that it's unclear that an elected democracy *is* an appropriate model for Squeak.

6 8

A concrete proposal for registering voters and conducting the elections
by Peter Crowther 12 Jan '06

12 Jan '06

This is derived from the discussion on #Squeak (see log at http://meme.b9.com/~1c0c02fabeddba6877abc034~/cview.html?channel=squeak& date=060111 with start time 21:00). Goals: 1) Use pre-built systems where possible to minimise development time. 2) Allow several communities to register their interest in the upcoming board elections. 3) Use a recognised voting system with known properties rather than an ad-hoc one. Assumptions: a) Other communities' list admins will exclude any emails they consider unsuitable before submitting lists. b) Nominations close at some time N, before time T (see below). Proposed method: i) Voters in the Squeak community must register on SqP and have been ranked as Apprentice or higher by T - 48 hours, where T is the start of voting. [Alternative: A rank of 'voter' is created between Observer and Apprentice, and this or higher must be achieved.] ii) After T - 48 hours, the list of apprentices and higher ranks [option: voter and higher] is retrieved from SqP and matched against email addresses. iii) The stakeholder organisations listed at http://swiki.krampe.se/castaways/9 are contacted and asked to pass over email addresses of any voters they wish to include in the election. Actually, we'll contact them earlier, but ideally we want a list as near to T - 48 hours as possible. iv) The list of voters V is constructed. This is exactly the union of the SqP list and all the stakeholder lists. v) After time N, the list of candidates C is constructed. Candidates who are willing to stand and meet the criteria described at http://people.squeakfoundation.org/article/53.html make the list. v) Before time T, a private election is created at CIVS (http://www.cs.cornell.edu/andru/civs.html) using Condorcet voting with the CIVS Ranked Pairs completion rule [alternative: MAM completion rule]. The list of candidates used is C; the list of voters used is V. vi) At time T, a message is sent to squeak-dev and all stakeholder lists stating that voting is open, giving the URL and the duration D for which the vote is open (1 week). [Alternative: another duration.] vii) At time T + D, the vote is closed, the result is declared and feedback is solicited from the community on the process. Note that T + D must be before 17th Feb in order to meet the present requirements for leaders stepping down. viii) At some time greater than T + D, Peter (at least) pours himself a large drink and celebrates that this particular hurdle has been cleared. ix) Continuously after T + D, the elections team monitors feedback and tries to improve the process. Comments? - Peter

1 0

RE: [Elections] I am really worried...
by Peter Crowther 11 Jan '06

11 Jan '06

[Directed to elections and Goran privately. Goran, please feel free to reply privately; I reserve the prerogative to forward any response or portions of it to the list if I feel it would benefit the process] > From: [...] goran(a)krampe.se > And please, please don't start *arguing* with me - > because that would just make things even worse. Punch us and then put your glasses on and say we can't punch back? Goran, that's simply cowardly - and if you claim it's a prerogative of being the boss, I'll point to this as exactly the kind of abuse of position that I'm concerned about. > This team is not getting disolved after the election How do you know? That is up to a subsequent administration, not this one - and definitely not you, as you're stepping down. > Get a simple voting system in place ASAP > and make it transparent. Agree. > I agree with the "worries" about accusations > later on about how legitimate the election was - but come on, > keep it on a practical level here. The coup was perceived as illegitimate by a proportion of the Squeak community, the merger with SqF was murky at best. Are you *surprised* the people who care enough to subscribe to the list have worries about the perceived legitimacy of the election? > Personally I would actually prefer to just > hack up a system that accepts votes from all subscribers to squeak-dev > and add a confirmation email mechanism to ensure one vote per person. Thank you for your input. Speaking personally, I would note that it could have been sent in a less antagonistic manner; you probably haven't encouraged the team to consider your view dispassionately. > 3. You have so far only done one thing: discussing a voting system. You should possibly read rather than skim before coming in with all guns blazing. I believe the candidate selection mechanism - as opposed to the voter selection or the voting mechanisms - is so uncontroversial as to have gone through on the nod in the middle of some other messages. > you have done nothing on the most > important task of finding candidates and building an awareness of what > is coming on squeak-dev. If we don't have candidates it > doesn't *matter* what voting system we have! Agree entirely. The team's communication on squeak-dev has been shocking; it needs someone to act as that. I'd be happy to act as a communicator, but would need to be on the team first. > 4. You are actually IMHO missing the absolutely most > important risk. And > no, it isn't fraud - it is the lack of participation. I suspect the percentage participation will be higher than in the original coup. > And > if the voting system for example only allows voting on one > candidate for > the board - then hey, it would make it even worse. Again: read the mails. I'm not aware of a single system that has been proposed here that is limited to a single candidate. Why raise a spectre that doesn't exist? > PS. Why haven't you tried making this team larger? The team process encourages consensus. Read the archives. Think about the points of view that have been expressed. Do you honestly believe that it would have been easier to reach consensus with more people on-list? 'cos I don't. - Peter

2 1

RE: [Elections] I am really worried...
by Peter Crowther 11 Jan '06

11 Jan '06

> From: Cees De Groot [mailto:cdegroot@gmail.com] > Brent is actually the team lead, Peter. Well aware of it. > And as it was said a bit > earlier in the discussion, candidate selection was a no-brainer > ("passed with a nod in the middle of a sentence" - was that the > wording?), so I think Brent also felt that way. Also well aware of it. In which case I assume it's simply cock-up that the candidate selection article went up without the surrounding wording concerning proposer/seconder in the same email? Or is there a later one I've missed? > The fact that the team lead "went public" with searching for > candidates prompted me to add my candidacy to the article (and to blog > about it in order to give it more visibility). Reasonable. - Peter

2 1

RE: [Elections] I am really worried...
by Peter Crowther 11 Jan '06

11 Jan '06

> From: [...] Peter Crowther [...] > The team's communication on squeak-dev has been > shocking; it needs someone to act as that. I'd be happy to act as a > communicator, but would need to be on the team first. Update (I tend to read election mail before squeak-dev, so missed this): Brent has put up an article on SqP and pointed folks on Squeak-dev to that. Cees has added his name to that as a candidate. Consensus on here has gone out of the window in the process. This may or may not be a good thing, depending on whether it's better to have something done that people disagree with or nothing done that people disagree with. Given that Brent and Cees have moved the debate on without reference to this team, is there actually a reson for the team's existence? Or do we accept a coup over the elections, disband and go home? - Peter

2 1

RE: [Elections] A simple SqF board elections solution using SqP
by Peter Crowther 11 Jan '06

11 Jan '06

> From: [...] Daniel Vainsencher > I hope this message wraps up the current discussions: > 1. There seems to be wide spread consensus on how the voting itself > should occur, which is summarized in the requirements page at > http://minnow.cc.gatech.edu/squeak/5835 As far as I can tell, the Debian algorithm is only suitable for selecting a single winner from a set of alternatives (it selects the Conddorcet winner); it is unsuitable for selecting multiple winners. For elections to a board, we need a somewhat different algorithm, hence my links yesterday. > So as far as I can tell, there is not reason implementation for that > should not proceed. I believe there is at least one third-party Web site running suitable software if we wish to outsource this. > 2. The only objection to using SqP as a basis for the voters list was > Lex, and at least Ken, Peter and I think this is a reasonable > basis. I think it is close enough to consensus. > > So I think that the voting system as described in the link above, is > reasonably well accepted, if someone has objections, make them heard. I am not raising objections other than the practical one above (I hadn't looked sufficiently closely at Debian's Condorcet algorithm). > Peter, Cees, you seemed to be interested in implementing this > system or > parts. If you are, just go ahead and and do it, there are no live > alternative proposals. If you don't want to implement it now, > please let > the list know, and I'll post a request for implementation to > squeak-dev. I'm interested in assisting in implementing it. I've also just been hit with a bunch of work we weren't expecting, which means my time for Squeak will be curtailed for a period. That period may well be sufficient to take us past the elections. I wonder whether we use other appropriate software for this period? - SqP as it is now to register voters - anyone who has Apprentice per the rather loose definition at http://people.squeakfoundation.org/certs.html by (say) 12 Feb is eligible to vote. If we are going to do this, we should announce it *now* to give people notice that a SqP account is of more value than it was previously, as it's their voting registration! - A mailing list to accept statements from candidates and formal support from those who nominate, second, third... however far we go down that route... them. - A portion of a Wiki for candidates to post their statements. - A third-party Web site such as http://www.cs.cornell.edu/andru/civs.html with either Beatpath or Civs completion rules to receive and tally the election results. I'd much *prefer* to use our own versions of these tools; I think we're going to be up against it to get them stable in the time we have available. - Peter

3 3

I am really worried...
by goran＠krampe.se 11 Jan '06

11 Jan '06

Hi fellas! Ok, let me try to make this as short as possible so that you bother to read it. I have followed the archive and I just can't help commenting. These are my *personal* comments, nothing else. And hell yes, this is a critical posting - but so be it, hopefully it will make my words come through to you, and I don't care if I make anyone upset, this is all too important for me to worry about that. Now - I am *really* worried that this is not going well, so please, please hear me out. And please, please don't start *arguing* with me - because that would just make things even worse. In order to prevent that, I will not reply to any replies to this post (but I will reply to private email) - you may of course spend time discussing this post with each other - fine. 1. Daniel wrote that he didn't see why there is a "rush" for an election. The thing is - we *promised* that we were stepping down at this date "no matter what" and that is what is going to happen. Period. So yes, there is a *rush*. This is *very* important to the current board and even more so to me *personally*. Let me underline the above by saying that *I* (can't tell for the rest of the board) am stepping down even if this team utterly screws it up and leaves us in a vacuum again. Ok? And yes, if this fails I will blame this team, you bet. 2. It seems to me that you are fumbling the ball. Please, please, please - save these endless discussions about voting systems and politics for later. This team is not getting disolved after the election, you will have plenty of time later too. Get a simple voting system in place ASAP and make it transparent. I agree with the "worries" about accusations later on about how legitimate the election was - but come on, keep it on a practical level here. Most of us haven't even met in real life and I could be a 12 year old girl from Uzbekistan for all you know, I could even be 10 people on squeak-dev. Given all emails I write it sure sometimes feels like that. :) Personally I would actually prefer to just hack up a system that accepts votes from all subscribers to squeak-dev and add a confirmation email mechanism to ensure one vote per person. Sure, I could subscribe 100 people to rig the election - but such a thing could pretty easily be detectable. Anyway, get something done - Cees has repeatedly shown that he wants to help - use him for god's sake. Then it is faaaar more important that you focus on 3 and 4 below IMHO. 3. You have so far only done one thing: discussing a voting system. In fact you initially got another immediate task (which as far as I can tell have been largely ignored) and you have done nothing on the most important task of finding candidates and building an awareness of what is coming on squeak-dev. If we don't have candidates it doesn't *matter* what voting system we have! 4. You are actually IMHO missing the absolutely most important risk. And no, it isn't fraud - it is the lack of participation. Just look at how many signed up on this list! Hell, I am not worried about fraud - but I am dead worried about only getting 11 people voting instead of 1000. And if the voting system for example only allows voting on one candidate for the board - then hey, it would make it even worse. So in utterly short words: - Get a *simple* voting system set up NOW. A transparent one. - Find candidates. Start asking for them on squeak-dev. Start emailing people in private to get them to step forward. Create awareness that this is happing NOW. - Worry about the participation. Ok, end of rant. If you are pissed off at me now and are itching to tell me all about it - wait 10 seconds and then instead decide to show me how wrong I am by getting this team to *deliver*. I don't care about ANYTHING else. regards, Göran PS. Why haven't you tried making this team larger? You are so worried about accusations from the community towards the board later on about the election - if I were you I would be more worried about such accusations *towards this team*.

1 0

RE: [Elections] Checking in / next steps
by Peter Crowther 11 Jan '06

11 Jan '06

> From: Cees De Groot [mailto:cdegroot@gmail.com] > These elections are a > first test of the water. Whatever the team comes up with, there'll be > criticism. It's version 1.0. At best :-). > We'll test it by running this year's > board elections. Then, the new board, which is more democratic than > the current board, but probably not 100% optimally democratic, can > evaluate with the Elections team and the community what can be done > better, and then there's a year to improve. Which, I think, is al the > more reason to start off with something simple and improvable... Indeed. And, in my opinion, not necessarily in Squeak. - Peter

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Elections January 2006