Debian decides to allow secret votes
Posted Mar 28, 2022 15:47 UTC (Mon)
by nickodell (subscriber, #125165)
[Link] (6 responses)
>Would you please explain why Option 2 defeated NOTA by 124 votes but at the same time defeated Option 3, which was identical to NOTA, by only 35 votes?
I don't think they *are* identical. Here's the text of Option 3: https://lists.debian.org/debian-vote/2022/03/msg00021.html
I read this option as "none of the above, and also secret voting is a bad idea."
Posted Mar 28, 2022 20:08 UTC (Mon)
by mattheww (guest, #108205)
[Link] (5 responses)
As Debian uses a ranked-choice voting system, it isn't obvious how to measure a 3:1 majority. The way they chose to do that is to use the always-present "none of the above" option as an indicator: there need to be 3x as many votes preferring a "change the constitution option" to "none of the above" as vice versa.
So if the "none of the above, and also secret voting is a bad idea" option hadn't been present, and people who voted it over the winner had voted "none of the above" over the winner instead, Option 2 would have been rejected.
It does seem quite likely that the existence of Option 3 meant that people who opposed the change unintentionally allowed it to pass.
I think this is not a good way to manage the 3:1 majority requirement. A simultaneous approval vote would be better (ie, next to each option with a supermajority requirement, you get a checkbox saying "if this option wins, I consent to changing the constitution").
Posted Mar 28, 2022 21:31 UTC (Mon)
by NYKevin (subscriber, #129325)
[Link] (3 responses)
That's already how NOTA works, more or less. If you rank it above a choice, you are approving of that choice. If you rank it below that choice, then you are disapproving of that choice. Most of the time, a choice only needs simple majority approval by this method, but constitutional changes require a 2/3 supermajority.
In this case, it appears that some voters ranked Option 3 over Option 2, but then also approved of Option 2. You may believe* that is an illogical vote to cast, but it is a valid vote, and in a democratic system, we generally presume that valid votes should be counted even if we think they are illogical.
* It might be the case, for example, that some of these voters wanted Option 2 to lose, but did not want it to lose by failing to get a supermajority, because this would be perceived as a "tyranny of the minority" situation, and so they would prefer the lesser evil of Option 2 winning.
Posted Mar 28, 2022 21:31 UTC (Mon)
by NYKevin (subscriber, #129325)
[Link]
Obviously, I meant to write "If you rank a choice above NOTA, you are approving of that choice..." instead of the other way around.
Posted Mar 29, 2022 9:35 UTC (Tue)
by spotter (guest, #12199)
[Link] (1 responses)
It doesn't make sense to have a choice that is simply a value judgement "this is a bad idea" in such a situation.
Posted Mar 31, 2022 6:48 UTC (Thu)
by NYKevin (subscriber, #129325)
[Link]
Posted Mar 28, 2022 21:31 UTC (Mon)
by calumapplepie (guest, #143655)
[Link]
Why didn't they just do that in the first place? Vote 3-NOTA-2. We must assume they had a reason. There is no need to further complicate the ballot with additional requirements when the current ballot suffices to express that anyways.
Posted Mar 29, 2022 2:07 UTC (Tue)
by IanKelling (subscriber, #89418)
[Link]
Posted Mar 29, 2022 2:38 UTC (Tue)
by IanKelling (subscriber, #89418)
[Link] (9 responses)
It seems to me that this very situation is a significant argument against completely secret voting in Debian. When someone who voted then has influence over a contested election, the transparency of their vote could be a small but useful piece of information to help guide the project toward doing the right thing, and when just a few votes matters, it could make an important difference.
Posted Mar 29, 2022 3:25 UTC (Tue)
by IanKelling (subscriber, #89418)
[Link]
And it turns out, this won by a single vote.
Posted Mar 29, 2022 9:14 UTC (Tue)
by kleptog (subscriber, #1183)
[Link] (7 responses)
If "the right thing to do" was so obvious then the result would've been clearer.
Posted Mar 29, 2022 11:53 UTC (Tue)
by donbarry (guest, #10485)
[Link] (3 responses)
And as this very debate illustrates, there is no such consensus.
To "accept" the result with the proviso that one can then simply relitigate it is to evade the fundamental point: the litigation then requires a 3-1 majority *the other way* to return the status quo ante.
The smugness of those defending the sanctity of process here is rather difficult to take.
I am very disappointed in the Debian leadership and frankly for the significant section of its membership who are moving away from the transparency, not only at the aggregate level, but the individual, which I always thought was one of its many core features.
Posted Mar 29, 2022 13:35 UTC (Tue)
by kleptog (subscriber, #1183)
[Link] (2 responses)
185 people voted Option 2 over NOTA.
That's a 3-1 supermajority, the presence of any other options doesn't change this. The only argument would be that people didn't understand what NOTA meant, but it's hardly a new thing.
The part I find most disturbing is the low turnout, by my reckoning about less than 10%? Many places with constitutions would consider that too low for a constitutional change.
Posted Mar 29, 2022 15:50 UTC (Tue)
by IanKelling (subscriber, #89418)
[Link] (1 responses)
Of course it does. Context matters. Here is an example, https://www.theguardian.com/world/2021/sep/06/three-near-...
Posted Mar 29, 2022 20:59 UTC (Tue)
by kleptog (subscriber, #1183)
[Link]
I've read what I can see of the thread and no-one seems to have been confused by the ballot.
Posted Mar 29, 2022 15:54 UTC (Tue)
by IanKelling (subscriber, #89418)
[Link] (2 responses)
No. These rules are for humans, not machines, and it doesn't work that way. There is already lots of ambiguity, loopholes and ways to twist most rules meant for humans, and then you can claim you are following them to the letter. In the same thread about this, someone pointed out a constitutional change that was voted on years before but never implemented. So, there were already rules that were being followed to the letter with the opposite intended result.
Posted Mar 29, 2022 21:17 UTC (Tue)
by kleptog (subscriber, #1183)
[Link] (1 responses)
At another level, I'm not happy when people talk about results of elections along the lines of "there is no way people could have actually voted this way". The majority of people only ever express an opinion by voting and the results should be taken seriously.
> In the same thread about this, someone pointed out a constitutional change that was voted on years before but never implemented.
I went through the thread twice (I see 18 messages on lists.debian.org) but couldn't find this. Do you have a link?
Posted Mar 30, 2022 22:20 UTC (Wed)
by amacater (subscriber, #790)
[Link]
I think the decision was made originally in 2005 but has proved difficult or impossible to implement.
Posted Mar 29, 2022 13:37 UTC (Tue)
by anarcat (subscriber, #66354)
[Link]
You (a seconder of the winning option) now believe that we need to stop and re-open
discussion on a closed matter that the whole project voted on (which I believe you've
been a staunch advocate for in the past - GRs are good, democracy, etc, right?),
single-handedly overturning the results, because you now have changed your views
on your own amendment that won, and you now believe the constitution is weakened
such that you're now threatening to remove the Secretary because you disagree with
their reading of the rules (who's executed the role of reading those rules since 2009
without prior issue) that is their role in the project to read?
Just checking. Do I have that right?
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
61 people voted NOTA over Option 2.
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
Debian decides to allow secret votes
The entire thing is so ludicrous that I think it's important to restate the objection that said DPL candidate expressed, in the words of paultag:
Debian decides to allow secret votes
Let me get this straight --
It's rather unbelievable when you put it like that. Others (rra) have put it more gently:
This is an absurd escalation when you have no procedural basis for what
you're demanding, and it's quite concerning coming from someone who is
currently standing for DPL. It's also pointless; anyone else who replaces
the Project Secretary will have to do the same thing. The discretion
you're asking for simply does not exist in the constitution.
I guess this weird Debianity will affect the DPL vote more than this actual GR, in my opinion.
Oh, and I wonder if someone has noted that Lechner was replying to an automated mail from devotee, the voting software, which explicitly starts with:
This message is an automated, unofficial publication of vote results.
Official results shall follow, sent in by the vote taker, namely
Debian Project Secretary
Another thing that is, in my opinion, quite telling.