On 6/27/2021 9:41 PM, Carsten Bormann wrote:
On 2021-06-28, at 03:21, John C Klensin <john-ietf@xxxxxxx> wrote:
how
much bigger we would need to make the pool to overcome the
number of volunteers from that company, i.e., to make their
fraction of the volunteers small enough relative to the total to
significantly reduce the chances of their ending up with two
Nomcom members.
Easy: randomly (*) select 10 companies (clusters) making up the nomcom, and then randomly select one volunteer from each company.
Let’s see how well this process proposal goes :-)
Grüße, Carsten
(*) OK, you get to choose the exact way this randomness works, and that may include weighting by number of volunteers. So we may still have a way to make sure there is enough big-company representation in the nomcom.
As Brian point out in a different message, we probably can't/shouldn't
do this in the current pass. But it would be useful to gather a few
different models and consider them for the next.
Your model has the benefit of simplicity and fixes some structural
issues of having as many as 4 Nomcom participants (2 voting members,
liaisons and chair) from a single company. That said, it means that
people from large companies are substantially less likely to be selected.
Another possibility is to change the size of the Nomcom while keeping
the 2 per company limits. 20 Members for example. That would dilute
the influence of the larger companies while still allowing for a
reasonable amount of participation from them. Given that the Nomcom is
now doing a larger number of selections per year, an increase in size
might be warranted. That would require a bit of cat herding on the part
of the Nomcom chair and might suggest the need for sub chairs for the
various organizations with vacancies. (e.g. IAB, IESG, Other... ).
There's the other model I've suggested a few times but won't repeat
here. And I'm sure we can find a few more...
Mike
