Debian Project Leader Elections 2005
|Nomination period:||February 7th 00:00:01 UTC, 2005||February 28th 00:00:00 UTC, 2005|
|Campaigning period:||February 28th 00:00:01 UTC, 2005||March 21st 00:00:00 UTC, 2005|
|Voting period:||March 21st, 00:00:01 UTC, 2005||April 11th, 00:00:00 UTC, 2005|
Please note that the new term for the project leader shall start on April 17th, 2005.
- Matthew Garrett [firstname.lastname@example.org] [platform]
- Andreas Schuldei [email@example.com] [platform]
- Angus Lees [firstname.lastname@example.org] [platform]
- Anthony Towns [email@example.com] [platform]
- Jonathan Walther [firstname.lastname@example.org] [platform]
- Branden Robinson [email@example.com] [platform]
The ballot, when ready, can be received through email by emailing firstname.lastname@example.org with the subject leader2005.
The debates this year were orchestrated by Helen Faulkner and Martin Krafft. I would like to extend the thanks of the project for their stellar work in producing an exemplary debate, despite a record number of candidates participating this year. A transcript of the debate is available for review
The debate was held on March 16th, 06:00:00 UTC, 2005, on the IRC server irc.debian.org. There were two relevant channels, namely, #debian-dpl-debate, and #debian-dpl-discuss. The #debian-dpl-debate channel was moderated, but open to anybody who wished to view the proceedings, but only the candidates, the moderators, and those participating in the immediate discussion will be able to post messages. The #debian-dpl-discuss channel was unmoderated - the intention was for real-time, unmoderated discussion of the candidates' responses to the debate questions.
Raw logs from the channels, in addition to the transcript mentioned above, including the behind the scenes channels used to conduct the debate courtesy of the moderators:
This year, like always, some statistics shall be gathered about ballots received and acknowledgements sent periodically during the voting period. Additionally, the list of voters shall be recorded. Also, the tally sheet may also be viewed. Please remember that the project leader election has a secret ballot, so the tally sheet is produced with the hash of the alias of the voter rather than the name; the alias having been sent to the corresponding voter when the acknowledgement of the ballot was sent so that people may verify that their votes are correct.
With 965 developers, we have:
Current Developer Count = 965 Q ( sqrt(#devel) / 2 ) = 15.5322245670091 K min(5, Q ) = 5 Quorum (3 x Q ) = 46.5966737010272 Option 1 Reached quorum: 112> 46.5966737010272 Option 2 Reached quorum: 384> 46.5966737010272 Option 3 Reached quorum: 376> 46.5966737010272 Option 4 Reached quorum: 390> 46.5966737010272 Option 5 Reached quorum: 261> 46.5966737010272 Option 6 Reached quorum: 346> 46.5966737010272
All candidates would need a simple majority to be eligible.
Dropping Option 1 because of Majority. 0.332 (112/337) <= 1 Option 2 passes Majority. 5.120 (384/75)> 1 Option 3 passes Majority. 3.514 (376/107)> 1 Option 4 passes Majority. 3.861 (390/101)> 1 Option 5 passes Majority. 1.418 (261/184)> 1 Option 6 passes Majority. 2.883 (346/120)> 1
The winner of the election is Branden Robinson
I would like to thank all the candidates for their service to the project, for standing for the post of project leader, and for offering the developers a strong and viable group of candidates.
Total unique votes cast: 504, which is 52.22797% of all possible votes.
In the graph above, any pink colored nodes imply that the option did not pass majority, the Blue is the winner. The Octagon is used for the options that did not beat the default. In the following table, tally[row x][col y] represents the votes that option x received over option y. A more detailed explanation of the beat matrix may help in understanding the table. For understanding the Condorcet method, the Wikipedia entry is fairly informative.
- Option 1 "Jonathan Walther"
- Option 2 "Matthew Garrett"
- Option 3 "Branden Robinson"
- Option 4 "Anthony Towns"
- Option 5 "Angus Lees"
- Option 6 "Andreas Schuldei"
- Option 7 "None of the Above"
Looking at row 2, column 1, Matthew Garrett received 394 votes over Jonathan Walther
Looking at row 1, column 2, Jonathan Walther received 18 votes over Matthew Garrett.
Option 3 defeats Option 2 by ( 248 - 220) = 28 votes.
Option 4 defeats Option 2 by ( 244 - 221) = 23 votes.
Option 2 defeats Option 5 by ( 350 - 60) = 290 votes.
Option 2 defeats Option 6 by ( 249 - 185) = 64 votes.
Option 2 defeats Option 7 by ( 384 - 75) = 309 votes.
Option 3 defeats Option 4 by ( 245 - 222) = 23 votes.
Option 3 defeats Option 5 by ( 352 - 95) = 257 votes.
Option 3 defeats Option 6 by ( 244 - 184) = 60 votes.
Option 3 defeats Option 7 by ( 376 - 107) = 269 votes.
Option 4 defeats Option 5 by ( 375 - 80) = 295 votes.
Option 4 defeats Option 6 by ( 264 - 196) = 68 votes.
Option 4 defeats Option 7 by ( 390 - 101) = 289 votes.
Option 6 defeats Option 5 by ( 307 - 99) = 208 votes.
Option 5 defeats Option 7 by ( 261 - 184) = 77 votes.
Option 6 defeats Option 7 by ( 346 - 120) = 226 votes.
Debian uses the Condorcet method for project leader
elections. Simplistically, plain Condorcet's method
can be stated like so :
Consider all possible two-way races between candidates.
The Condorcet winner, if there is one, is the one
candidate who can beat each other candidate in a two-way
race with that candidate.
The problem is that in complex elections, there may well
be a circular relations ship in which A beats B, B beats C,
and C beats A. Most of the variations on Condorcet use
various means of resolving the tie. See
Cloneproof Schwartz Sequential Dropping
for details. Debian's variation is spelled out in the
specifically, § A.6.