[dns-wg] yet another heave on the WG Chair selection procedure

On 7 Jan 2015, at 20:15, Nick Hilliard <nick at foobar.org> wrote:

> On 06/01/2015 12:41, Niall O'Reilly wrote:
>>  [2] A co-chair will serve a term of N years, where N is the number
>>  of co-chairs. Terms will be staggered so that one term expires every
>>  year.
> 
> This is also semantically non-deterministic in the case where the number of
> co-chairs changes.

It's not Nick. If the number of co-chairs changes, the term of office would be adjusted to reflect this. ie If/when 3 co-chairs drops to 2, what were the 3 year terms of the two remaining incumbents would automagically drop by a year. Likewise if we go from 2 to 3, the 2 year terms of the incumbents get extended by a year and the new stuckee slips into the gap that's just been created in the selection cycle. This is implicit from what was in the proposed text. Or should be.

Now I suppose we could choose to write this up and spell it out in painstaking detail. But that seems to be overkill and/or shed painting. YMMV.

That said your suggested [2] seems to be a cleaner and less verbose solution:

> [2] Every second RIPE meeting, the co-chair who has been in that
> position the longest will resign.  If more than one co-chair qualifies
> for resignation, then only one co-chair will be required to resign, as
> determined by [drawing lots|some other appropriate means].

Though I'd prefer "once a year" to "every second RIPE meeting" in case the number of RIPE meetings per year changes or one of them gets cancelled say because the Kras catches fire. And while the selection process may well be aligned with a RIPE meeting, I would be very uncomfortable if that gave the impression that the meeting took precedence over the mailing list when it came to making the consensus decision.