On Nov 3, 2018, at 12:38 AM, Donald Stufft <donald@stufft.io> wrote:



On Nov 3, 2018, at 12:20 AM, Tim Peters <tim.peters@gmail.com> wrote:

[Tim]
Nevertheless, I probably won't vote - I object to public ballots on
principle.  That's been raised by others, so I won't repeat the
arguments, and I appear to be very much in a minority here.

[Eric Snow <ericsnowcurrently@gmail.com>]
Would it help if we only published who voted, and kept their votes
private?  Publishing the actual votes probably doesn't make a
big difference here, relative to the broader Python/tech community.

That would probably be enough to convince me to vote, but I don't want
to hold things up either.  If I'm the only one, why bother?  It's not
like my vote will change the result ;-)

BTW, the years I was on the PSF Board, I always wanted everyone to
know how we voted on everything.  But I was elected to that position,
so was voting as a representative of those who elected me.

But nobody has any more business knowing how I vote on a PEP than,
say, how I vote for the local mayor.  That's between me and my
conscience.  Your vote is between you and yours, and I want actively
_not_ to be able to see how others voted.

Although I'm all in favor of making the PEP ballots public, if
stripped of personally identifying info.
_______________________________________________


FWIW I tend to agree with Tim on public vs private ballots, although unlike him I don’t feel strongly enough to abstain from voting on this one particular vote.

On a practical matter, keeping the ballots secret will rely on either having a trusted person to tally the election results or using some software that will do it for us. There is https://civs.cs.cornell.edu which we could use that does offer private ballots and offers making the ballots (with or without a name attached to them) public. It doesn’t support “pure” Condorcet but it should be easy enough to take the public but anonymous ballots and compute to determine if there was a condorcet winner or if one of the methods had to break a cycle, and if there wasn’t a condorcet winner, just re-run the election. Beyond that, I’m not sure what other options there are for anonymous ranked voting.

Oh, unfortunately this also doesn’t allow publishing *Who* voted without attaching them to a ballot, it’s either public, attached to the ballot, or private (if you’re not publishing the names, the system doesn’t even keep them, it just generates unique voter IDs for each).