This thread is demonstrating a key weakness in defining any process with security concerns on the basis of random drawing. The chance of hitting this particular corner case was really small.
It doesn't much matter here but it could well matter in another circumstance.
On Mon, Jul 8, 2019 at 2:23 PM Victor Kuarsingh <victor@xxxxxxxxxx> wrote:
Dear IETF Community,The feedback has been outstanding and I appreciate all the input in this matter (re: selection considerations with respect to Chair and past chair primary affiliation and selection elimination). Although the intention was to help drive the maximum amount of non-bias and fairness with the process, the following seems clear.1). The wording within RFC7437 is such that the primary affiliation elimination should only apply to the 10 selected voting volunteers. This is valid even in light of Appendix B, Point 5.2). Although there may be merit in applying the elimination rule with consideration of the Chair and Past Chair in terms of their primary affiliation (in light of the Chair's interaction within the Nomcom process), such a posture may drive a precedent. With that in mind, such a decision should be driven by future community updates to RFC7437.With this in mind, it would be appropriate to update the list of selected members (based on the published random results on July 7th 2019) without consideration of the primary affiliation of the Chair or Past Chair. I will send out an updated announcement soon.Once again, I think the discussion of this matter was quite valuable, has raised valid points, and I appreciate all the feedback from the community members.regards,Victor KuarsinghNomcom Chair 2019-2020On Mon, Jul 8, 2019 at 1:03 PM Ted Hardie <ted.ietf@xxxxxxxxx> wrote:Hi Scott, Victor,Forgive the top posting. Thanks for these explanations of the logic applied here. Generally, I think your idea is a good one and would help promote diversity. However, the current document is pretty clear on what the exact steps are in the process, and applying this logic resulted in a different set. Since this interpretation of the rules was not announced until after the algorithm was run, I believe the correct thing to do here is likely to take the group according to the current algorithm and document the proposed changes for future NomComs.Just my personal read of the situation,regards,TedOn Mon, Jul 8, 2019 at 7:57 AM Scott Mansfield <scott.mansfield@xxxxxxxxxxxx> wrote:I have to admit that this kerfuffle is my fault. I'm sorry I haven't replied earlier but had a number of issues with traveling back from Geneva (long story - not very interesting).
So. I read the following sections of RFC 7437 (4.17) and Appendix B
Section 4.17 of RFC 7437:
The Chair randomly selects the 10 voting volunteers from the pool of
names of volunteers and announces the members of the nominating
committee.
No more than two volunteers with the same primary affiliation may be
selected for the nominating committee. The Chair reviews the primary
affiliation of each volunteer selected by the method in turn. If the
primary affiliation for a volunteer is the same as two previously
selected volunteers, that volunteer is removed from consideration and
the method is repeated to identify the next eligible volunteer...
However it says in Appendix B:
The number of nominating committee members with the same primary
affiliation is limited in order to avoid the appearance of
improper bias in choosing the leadership of the IETF. Rather
than defining precise rules for how to define "affiliation", the
IETF community depends on the honor and integrity of the
participants to make the process work.
In Appendix B it says the number of nominating committee members. Nominating committee members include both voting and non-voting members. There is this distinction because there are times when only voting members vote (on slates for example) and times when the entire nominating committee needs to vote (on processes etc). So, I interpreted this to mean that the NomCom Chair should be sensitive to the affiliation of all nominating committee members. I check the previous 5 years of nomcoms and (unless I missed something) there was no nomcom that had more that two people from the same affiliation on the nominating committee ( I will check again to be sure).
I understand what the community is saying about how 4.17 is interpreted. But 4.17 states in the first sentence of the second paragraph that it is talking about the nominating committee. The nominating committee is more than just the voting members. However, following 4.17 to the letter would have a second Ericsson person as a voting member. As stated before the nominating committee is more than voting members, so I thought it prudent to err toward diversity. I can understand how section 4.17 can be seen as applying to only voting members, but the words say "nominating committee" not "voting members of the nominating committee".
I'm not the chair this year and I apologize for causing the confusion. I just wanted to explain my rationale for making the recommendation I did. I certainly am not trying to change the rules, but provide for the most fair and diverse committee based on how I interpret the rules.
The final decision is with Victor. I will support whatever decision he makes.
Regards,
-scott.
-----Original Message-----
From: ietf <ietf-bounces@xxxxxxxx> On Behalf Of Russ Housley
Sent: Monday, July 8, 2019 10:30 AM
To: Bob Hinden <bob.hinden@xxxxxxxxx>
Cc: NomCom Chair 2019 <nomcom-chair-2019@xxxxxxxx>; IETF <ietf@xxxxxxxx>
Subject: Re: Nomcom 2019-2020: Result of random selection process
>>>
>>> FWIW I think you are setting precedent without necessity - only voting members choose, non-voting members advise.
>>>
>>
>> I agree. The relevant operative text is Section 4.17 of RFC 7437:
>>
>> The Chair randomly selects the 10 voting volunteers from the pool of
>> names of volunteers and announces the members of the nominating
>> committee.
>>
>> No more than two volunteers with the same primary affiliation may be
>> selected for the nominating committee. The Chair reviews the primary
>> affiliation of each volunteer selected by the method in turn. If the
>> primary affiliation for a volunteer is the same as two previously
>> selected volunteers, that volunteer is removed from consideration and
>> the method is repeated to identify the next eligible volunteer.
>>
>> It’s clear from the procedure that the input is the voting volunteers, the method is the random selection method, and that the chair’s test for primary affiliation occurs at this point in the process... Furthermore, a cornerstone of the process is that the only elements in the process one cannot determine in advance are the PRNG inputs, and by no means the selection policy.
>
> I agree, this is clearly called out in RFC7437 and the NomCom should follow it.
>
> If RFC7437 is revised to include this, then that effort should consider the liaisons affiliations as well, not just the current and past chairs.
Are the liaisons known from all bodies before the voting members are selected. I know that they have not in the past.
Russ
