Skip to Content.
Sympa Menu

sm-discuss - Re: [SM-Discuss] voting process (was: Team Lead Nominations: Project Lead)

sm-discuss AT lists.ibiblio.org

Subject: Public SourceMage Discussion List

List archive

Chronological Thread  
  • From: "Jeremy Blosser (emrys)" <jblosser-smgl AT firinn.org>
  • To: SM Discuss <sm-discuss AT lists.ibiblio.org>
  • Subject: Re: [SM-Discuss] voting process (was: Team Lead Nominations: Project Lead)
  • Date: Thu, 9 Feb 2006 15:20:25 -0600

On Feb 08, Eric Sandall [eric AT sandall.us] wrote:
> While this is a month late, I figure better late than never. ;) Up for
> nominations is the position of Project Lead. Please follow our Team

The point was raised on IRC that if this goes to a vote (as in, assuming
Eric is nominated again and someone else is as well), Eric shouldn't be the
one to receive and count the votes. This prompted some other discussion
about the voting process in general and concerns that have been raised at
various times. The problem some have with the current method is that there
is no verifiability for the results. We mail our votes to a person, who
tells us what the totals are. While we may all trust that person, this is
not a trustworthy process. The previous approach of mailing the votes to
the open ML also has problems, since people may be intimidated or pressured
into not voting what they really think.

A suggestion was made that we could avoid both of these problems as
follows:

1) Voter sends a signed vote to vote counter (same as today).
2) Vote counter replies (same as today) and includes a "key" unique to that
vote/voter (new).
3) At the end of the vote, the vote counter publishes the full list of
"keys" and their respective votes to the ML/web site, eg:

voter voted
===== =====
dasgs y
a2tad n
asgai n
pqpow y
ladsg n

Total: 3 n, 2 y, n wins

Since each voter knows which key is them, they can look at this list and
protest if their vote is misrepresented (the reply from the vote counter
would be signed with the counter's gpg key so that the voter to key
relationship would be verifiable). And since the votes are public,
everyone can agree that the math is correct. But since only the voter (and
vote counter) know which key is which person, anonymity is preserved as
well as it is now.

What are people's thoughts on this? There would be various ways to
generate the keys; we could get complicated and use cryptographic hashes,
or we could stay simple and just have the vote counter use a throwaway
10-character random string for each vote. I tend to prefer the latter.

Note that while this prevents the vote counter from stealing an election,
it still leaves them with knowledge of all the votes. This means even if
we decided to switch to something like this in time for the PL vote, Eric
still shouldn't be the vote counter. dkowis has volunteered to count this
election, assuming he doesn't get nominated himself. If anyone has a
particular problem with that option, please let Eric or me know.

If we do both (publish the votes and have dkowis process them), the vote
can be verified, and only he would know who all voted for whom.

Attachment: pgpje_VbVDcGG.pgp
Description: PGP signature




Archive powered by MHonArc 2.6.24.

Top of Page