Kurt Roeckx writes ("Re: GR: Constitutional Amendment to fix an off-by-one
error and duplicate section numbering"):
> On Wed, Aug 26, 2015 at 10:12:41PM +0200, Andreas Barth wrote:
> > (i) Delete most of A.6(3) (which implemented the supermajority
> > by dropping options at an early stage). Specifically:
> > - Move A.6(3)(1) (the definition of V(A,B)) to a new subparagraph
> > A.6(3)(0) before A.6(3)(1).
> > - Remove the rest of A.6(3) entirely, leaving A.6(2) to be
> > followed by A.6(4).
As the original author of this wording, this is my fault. Sorry.
> Those 2 things conflict. First you create a A.6(3)(0), then you
> say there is no A.6(3) left. Please clarify what you mean.
I think this is a numbering problem:
(i) Delete most of A.6(3) (which implemented the supermajority
by dropping options at an early stage). Specifically:
- Move A.6(3)(1) (the definition of V(A,B)) to a new subparagraph
A.6(3)(0) before A.6(3)(1).
^^^^^^^^^
I think I intended that this should read A.6(4)(0).
^^^
That puts the definition of V(A,B) immediately before its first use in
A.6(4)(1).
> > + the consequences set out alongside the majority
> > + requirement (or, if unspecified, the default option
> > + wins).
>
> Where are those consequences documented? This is about the
> non-binding statements?
Yes. The consequences are documented `alongside the majority
requirement'. That is, when the constitutional text specifies a
majority requirement, it can now also specify what happens if the
supermajority fails.
In para (v) of the GR text, text is added to 6.1(4) and 4.1(4) saying
that supermajority-failing TC decisions, and supermajority-failing
attempts to use TC powers by way of GR, are advisory.
> > (v) In
> > * 6.1(4) (Technical Commitee power to overrule a Developer)
> > * 4.1(4) (Developers' use of TC powers by GR) (if another
> > constitutional amendment has not abolished that
> > supermajority requirement)
> > in each case after the "N:M majority" add
> > + ; failing that, the prospective winning resolution text becomes
> > + a non-binding statement of opinion.
>
> So I understand that if there is no winner, and so maybe not
> even a normal majority (that doesn't even exist anymore?) for it,
> it becomes a statement of opinion?
No. I'm not even sure what that paragraph of yours means. (If there
is no winner, how could "it" become a statement of opinion ? What
would "it" be ?) As for `normal' majorities, that is already dealt
with by Condorcet. Something that most people disagree with cannot
become the Condorcet winner because it will be defeated by FD/SQ.
The intent of this change is that if the Condorcet(CSSD) winner does
not meet the supermajority requirement, it is still the winning
outcome of the whole vote, but only as a non-binding statement of
opinion.
So for example, suppose in a TC vote we have:
A "we overrule the maintainer [6.1(4)]: this patch to comply with
policy must must be applied"
B "we set policy [6.1(1)]: the policy is wrong and must be changed"
and votes are 5x A,B,FD 2x B,FD,A
The overall Condorcet winneer is A but only by a 5:2 majority, so the
TC does not overrule.
With the existing rules A is eliminated early, leaving B the Condorcet
winner. This is a bizarre outcome: the winning option was disfavoured
by 5/7ths of the TC !
With the new scheme, A is the Condorcet winner (the `prospective
winner' in the wording proposed in the GR text). But it fails its
supermajority.
So `prospective winning resolution text becomes a non-binding
statement of opinion'. Ie, the TC is treated as having said:
A' "we advise [6.1(5)]: we disagree with the maintainer; this patch
to comply with policy should be applied"
This makes a lot more sense as an outcome. The maintainer can
continue to diregard the disputed policy, because the TC hasn't
mustered the certainty needed to overrule the maintainer; but, the
policy is not altered.
If you feel that the proposed wording isn't clear I'm sure Andi would
be happy to hear your suggestions for clarification.
> (Or I might be totally confused about the effects of all the
> changes you're doing. Those are all non-obvious changes that seem
> to change more than the things you want to fix.)
The determination of when supermajority failure occurs is necessarily
entangled with the consequences.
Ian.
