[OSGeo-Discuss] Secret ballots for voting

[Even Rouault]

Le mercredi 15 août 2012 19:28:12, Michael P. Gerlek a écrit :
> Someone asked the Board (off-list) why we do not publish the results of our
> elections in detail -- such as how many votes there were for each
> candidate, including any candidates that weren't elected.
> 
> I'm not sure if our bylaws require secret ballots, but that has been our
> process since the Foundation began.There are clearly pros and cons both
> ways for this. If we were fully transparent, everyone see who voted for
> whom and there would be no possibility of cheating. By keeping votes
> private, on the other hand, we allow people to candidly express their
> feelings without feeling any social pressures.
> 
> It's not clear to me that the community favors a change from our current
> secret ballot process, but I know it has been asked about before over the
> years and I know at least one person raised a question about the lack of
> transparency in our last round of voting. If you feel strongly about this
> please feel free to start a thread about it and we'll see what the
> consensus is.
> 
> PS- I can say that 98 of the 144 charter members voted in this last
> election.

(I trust the CRO to do their job honestly, so the following is just some idea, 
not a personnal requirement)

I haven't investigated that if there are reliable and independant 
organizations that could manage OSGeo ballots, but for better transparency, 
the result of all votes could be published in a web page, without the name of 
the voter of course, so that every voter can at least verify if their own vote 
has been properly taken into account.

That could be something as simple as the following :

Voter 1:
Candidate 1A
Candidate 1B
Candidate 1C
Candidate 1D
Candidate 1E

Voter 2:
Candidate 2A
Candidate 2B
Candidate 2C
Candidate 2D
Candidate 2E

[...]
Voter N:
Candidate NA
Candidate NB
Candidate NC
Candidate ND
Candidate NE

So that would be still secret ballot, but with public results. Like standard 
ballots.

Hum, with a bit of extra thinking, in an enhanced version of the above,  you 
would publish this like :

Voter 1 [unique_identifier_privately_given_by_the_voter_to_the_CRO]:
Candidate 1A
Candidate 1B
Candidate 1C
Candidate 1D
Candidate 1E

etc...

The unique_identifier_privately_given_by_the_voter_to_the_CRO could be for 
example the SHA1 of any random text that the voter would have selected, like :

$ echo "here is Even's vote and now some random text that nobody can guess 
zeaklrfjkgaoiejrmlkjf415646." | sha1sum
8d486b0387ffc57e65ff6e31ac0be69dba266334  -

(I originally imagined that the identifier could be given by the CRO, but if 
the CRO isn't honest, he could give the same identifier to two persons who have 
made the same candidate choices, and they wouldn't be able to distinguish 
their ballot)

That way, you can clearly identify your vote in the list if there are several 
people who made the same candidate choices.

Well, the above idea has certainly its flaws. AFAIK, all electronic voting have 
their flaws...

> 
> -mpg
> 
> _______________________________________________
> Discuss mailing list
> Discuss at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/discuss