>>>>> "Russ" == Russ Allbery <rra@debian.org> writes:
Russ> This analysis is very sensitive to the percentage of people in
Russ> the minority who would be willing to delay the vote. I think
Russ> a more likely number (probably still too high) would be at
Russ> most 10% of the voters (a quarter of those in the minority),
Russ> which would allow 7 delays, or 21 days (3 weeks), for a
Russ> maximum discussion period of six weeks.
I did less math, but my intuition was the same.
My intuition was that you could probably get six weeks of delay for a GR
that some minority really didn't like.
Beyond that, I guess the political back pressure would be strong enough
to git rid of the delay mechanism after the triggering GR was dealt
with.
I don't think adding a maximum matters to me much because I think that
the practical maximum is somewhere between 6-9 weeks.
I definitely prefer Russ's proposal.
One concrete change I'd request would be language added that said what
the delays should be used for.
Something like "Delays should only be used to provide time to develop
additional ballot options and not to delay the vote on a GR that those
seeking a delay find objectionable."
Such language, particularly if phrased with should language, has no
normative effect based on how the constitution defines should.
However, I think it might have behavioral effects in terms of setting
community expectations.
For example I could ask someone what ballot option they were working on
if they proposed a delay.
And if it was clear that they were not, our normal mechanisms for
approaching behavior inconsistent with our norms could be applied.
Attachment:
signature.asc
Description: PGP signature