[rpd] new policy proposal: AFPUB-2019-GEN-003-DRAFT01: "Chairs Elections Process"

[Fernando Frediani]

Hello Anthony

The point specifically about the Co-chairs be or not form the same 
country is a good practice to avoid monopolization from a single country.
You are right that 'in theory' the only thing that should matter are the 
candidate requirements, but in practice it may lead to scenarios where 
there is a major participation of people from a specific country and 
they may vote for someone just because he/she is from the same country 
without knowing anything about the person.
While I don't consider this a very critical thing as for example the 6 
month period to be able to vote, I think it avoids that possibility I 
mentioned which is what is more likely to happen.

Fernando

On 12/11/2019 15:56, Anthony Ubah wrote:

>

> Hello Jordie and Souad,

>

> I’ve just gone through your proposal. Having once contested for a 

> chair as co-chair I quite recognize a need to revisit the election 

> process. However, I don’t agree with the crust of this proposal.

>

> The stated problem being addressed (below) is accurate, nevertheless 

> personally I find that several areas are debatable, t’s as evident in 

> the views on the mail trails so far.

>

> *“The Policy Development Working Group (PDWG) discusses very briefly 

> how the chairs are chosen. However, there is not sufficient detail 

> about the candidate requirements, neither a complete process.”*

>

> Most observations I have made have already been brought previously 

> highlighted and brought forward by other members of the community, and 

> I see that you(Jordi) have painstakingly tried to address a few of 

> them, but I will still want to reiterate a few of mine.

>

> From section 3.3.1

>

> *• Both chairs can’t be from the same country,*

>

> *except in exceptional situations where there are*

>

> *no other acceptable candidates, in which case one*

>

> *of the chairs will cease in their position at the*

>

> *following election process (following year), either*

>

> *because their term has come to an end or by*

>

> *agreement among the two chairs, failing which*

>

> *the chair who has held the position the longest*

>

> *will automatically cease in their position*

>

> Ant: I am an advocate for freedom and transparency, thus I strongly 

> disagree with any form of restrictions due to the country of origin or 

> resident of the co-chairs. Nomination and election should be a matter 

> of substance. Like Daniel said */“Else we may be trading competency 

> for representative”/*. The focus should be on the candidates’ 

> competence and motivation.

>

> From Section 3.3.2

>

> *3.3.2 About the election of the Chairs*

>

> *• Voting will be conducted electronically, using*

>

> *mechanisms to ensure, as much as possible, that*

>

> *each voter can cast only one vote.*

>

> Ant:

>

> 1.Through what electronic means?

>

> 2.Solely remote voting? When can the community meet and interact with 

> the candidates, or do we rely solely on credentials and public 

> opinions? If so, then what is the need for waiting till the next PPM? 

> Run an election, conclude and announce results on online.

>

> 3.How can voters be identified for transparency in eligibility and 

> vote counting?

>

> 4.What mechanism is put in place to detect multiple accounts run by 

> one person? Because we might be an advent of a scenario of account 

> squatting for voting purposes. Where one person will register multiple 

> dummy accounts in anticipation of election time.

>

> 5.How can off-net collusion be detected?

>

> *• Anyone who has been part of the RPD List for at*

>

> *least 6 months prior to the start of the election*

>

> *process may participate.*

>

> Ant: Fair

>

> *• Any use of the list for electoral purposes, even*

>

> *when by persons clearly supportive of a*

>

> *candidate, may result in their disqualification, if*

>

> *there is evidence of collusion.*

>

> Ant: This is controversial. A member can disenfranchise a candidate by 

> running a pseudo campaign in his favour, just to have that candidate 

> or both disqualified in favour of another. Simply put, one less voter 

> for one less Candidate. Very unbalanced.

>

> **

>

> *• AFRINIC will communicate the names of*

>

> *Acceptable candidates to the RPD List,*

>

> *announcing where candidate information will be *

>

> *published.*

>

> The term acceptable candidate reflects in bullet point 3 of 3.3.1 and 

> bullet point 7 of 3.3.2. Are there separate criteria stipulated for 

> candidates to measure who fall within the “acceptable” threshold, or 

> is it the same as voter eligibility as stated in bullet point 3 of 

> 3.3.2? If not, what other criteria do you recommend for eligibility?

>

> **

>

> *• A period of 10 calendar days will then begin*

>

> *during which the community will be able to*

>

> *contribute relevant information on the candidates.*

>

> *This information, if confirmed, may be published*

>

> *simultaneously for all candidates on the first*

>

> *working day following the end of the 10-day*

>

> *period. As a result of that information, the Board*

>

> *could disqualify any candidate.*

>

> Ant: Quite Ambiguous. Kindly rephrase.

>

> *• Voting will begin on the first working Monday*

>

> *after the period specified above and will remain*

>

> *open for 7 calendar days*.

>

> Ant: When can the community meet and interact with the candidates?

>

> *• If any objections are raised by a member of the*

>

> *community, such objections must be*

>

> *communicated to the Board within 7 calendar*

>

> *days of the announcement of the results. The*

>

> *Board will then assess whether such objections*

>

> *are significant and have been proven. If no*

>

> *objections are raised, or if those aren’t*

>

> *considered, will proceed to ratify the winning*

>

> *candidate.*

>

> Ant: As I said before, I am an advocate of a transparent system. I 

> think narrowing a lot of things down to the Board is a way for locking 

> the larger community away from critical decisions.

>

> Drifting off a bit to the second trail on the subject matter. The 

> existing election process might not be perfect, but the influence of 

> the local community should not be frowned at. The community is 

> encouraged to grow, and there is no better opportunity to do that than 

> by driving active participation of the host community as was witnessed 

> in Kampala, where a lot of IT students participated actively throughout.

>

> Curbing the local influence of the host country/”newcomer voter’ in 

> the election process will achieve infinitesimal gain. Since the 

> Bi-annual PPM is shifted across regions it solves itself organically.

>

> In conclusion, I think the intention of initiating this proposal is 

> very noble, but I would suggest a comprehensive review. It is not ready.

>

> Kind regards,

>

> Anthony

>

> *Best Regards,*

>

> *UBAH ANTHONY **IKECHUKWU ***

>

> Goldspine Nigeria

>

> E-mail: anthony.ubah at goldspine.com <mailto:anthony.ubah at gloworld.com>.ng

>

>

>

> On Sun, Nov 10, 2019 at 11:23 PM <rpd-request at afrinic.net 

> <mailto:rpd-request at afrinic.net>> wrote:

>

>     Send RPD mailing list submissions to

>     rpd at afrinic.net <mailto:rpd at afrinic.net>

>

>     To subscribe or unsubscribe via the World Wide Web, visit

>     https://lists.afrinic.net/mailman/listinfo/rpd

>     or, via email, send a message with subject or body 'help' to

>     rpd-request at afrinic.net <mailto:rpd-request at afrinic.net>

>

>     You can reach the person managing the list at

>     rpd-owner at afrinic.net <mailto:rpd-owner at afrinic.net>

>

>     When replying, please edit your Subject line so it is more specific

>     than "Re: Contents of RPD digest..."

>

>

>     Today's Topics:

>

>        1. Re: AFRINIC Number Resources Transfer Policy (Fernando Frediani)

>        2. Re: new policy proposal: AFPUB-2019-GEN-003-DRAFT01: "Chairs

>           Elections Process" (Caleb Olumuyiwa Ogundele)

>

>

>     ----------------------------------------------------------------------

>

>     Message: 1

>     Date: Sun, 10 Nov 2019 19:01:58 -0300

>     From: Fernando Frediani <fhfrediani at gmail.com

>     <mailto:fhfrediani at gmail.com>>

>     To: rpd at afrinic.net <mailto:rpd at afrinic.net>

>     Subject: Re: [rpd] AFRINIC Number Resources Transfer Policy

>     Message-ID: <fa51b0ca-936b-5dbf-3ea7-f57fd5474722 at gmail.com

>     <mailto:fa51b0ca-936b-5dbf-3ea7-f57fd5474722 at gmail.com>>

>     Content-Type: text/plain; charset="utf-8"; Format="flowed"

>

>     In practice this situation you describe is very hard to happen, we

>     cannot have things in place to treat the very unlikely situation and

>     that Phase 2 is about to happen soon. Until there the vast

>     majority or

>     organization (really the vast!) can get addresses from AfriNic fine.

>     I hardly doubt one can justify anything more than a /13 at once at

>     the

>     moment. Even in a remote hypothesis that is possible the organization

>     can receive the /13 and work with that until transfers are allowed as

>     per Jordi's proposal that has been changed to start with Phase 2 is

>     triggered and that organization will be able to transfer whatever

>     else

>     is needed.

>     One rule for all and much simpler.

>

>     Fernando

>

>     On 10/11/2019 18:51, Owen DeLong wrote:

>     >

>     >

>     >> On Nov 10, 2019, at 10:51 , Chevalier du Borg

>     <virtual.borg at gmail.com <mailto:virtual.borg at gmail.com>

>     >> <mailto:virtual.borg at gmail.com

>     <mailto:virtual.borg at gmail.com>>> wrote:

>     >>

>     >>

>     >>

>     >> Le?dim. 10 nov. 2019 ??21:58, Jaco Kroon <jaco at uls.co.za

>     <mailto:jaco at uls.co.za>

>     >> <mailto:jaco at uls.co.za <mailto:jaco at uls.co.za>>> a ?crit?:

>     >>

>     >>     Hi Chevalier.

>     >>

>     >>     Please allow me to be blunt.? That's short sighted.

>     >>

>     >>     We cannot transfer IN from other regions unless we allow OUT.

>     >>

>     >>

>     >> Agree 100%,

>     >> Then you have no problems with wait till all RIRs are equal run

>     out

>     >> before we etablish full in and out transfer policy no?

>     >>

>     >>     All the other RIRs require reciprocal *compatible* policies,

>     >>     which means bi-directional transfers.

>     >>

>     >>

>     >>

>     >> All RIRs don't all have equal amount of free space. Big difference

>     >

>     > Depending on your definition here, 4 out of 5 have exactly equal

>     > amount == 0.

>     >

>     >>     Not allowing this means we can't get resources in either.

>     >>

>     >>

>     >> While AfriNIC have free space, operators don't need it

>     >> When it run out, then we can allow transfer policy

>     >

>     > This isn?t entirely true.

>     >

>     > It?s possible that an operator needs more than they can get via

>     > current AfriNIC policies due to ?soft landing? limitations.

>     >

>     > In such a case, said operator might prefer to transfer a large

>     amount

>     > of space in even if they are paying for it on the market

>     > rather than suffer with the small amount of space they can get from

>     > AfriNIC due to the current restrictions.

>     >

>     > Is there a valid reason to preclude such a transfer which, in

>     reality,

>     > prolongs the AfriNIC free pool to the benefit of other

>     > organizations in Africa?

>     >

>     > Owen

>     >

>     >

>     >

>     > _______________________________________________

>     > RPD mailing list

>     > RPD at afrinic.net <mailto:RPD at afrinic.net>

>     > https://lists.afrinic.net/mailman/listinfo/rpd

>     -------------- next part --------------

>     An HTML attachment was scrubbed...

>     URL:

>     <https://lists.afrinic.net/pipermail/rpd/attachments/20191110/6961beb6/attachment-0001.html>

>

>     ------------------------------

>

>     Message: 2

>     Date: Sun, 10 Nov 2019 22:22:39 +0000

>     From: Caleb Olumuyiwa Ogundele <muyiwacaleb at gmail.com

>     <mailto:muyiwacaleb at gmail.com>>

>     To: Owen DeLong <owen at delong.com <mailto:owen at delong.com>>

>     Cc: rpd List <rpd at afrinic.net <mailto:rpd at afrinic.net>>

>     Subject: Re: [rpd] new policy proposal: AFPUB-2019-GEN-003-DRAFT01:

>             "Chairs Elections Process"

>     Message-ID:

>            

>     <CAL_ZvK5xNb=LJXC5rCR8bDKUv528XVJGn2grRmxHC1ZE9Go-iQ at mail.gmail.com

>     <mailto:LJXC5rCR8bDKUv528XVJGn2grRmxHC1ZE9Go-iQ at mail.gmail.com>>

>     Content-Type: text/plain; charset="utf-8"

>

>     Owen see reply in line below

>

>     On Sun, Nov 10, 2019, 9:03 PM Owen DeLong <owen at delong.com

>     <mailto:owen at delong.com>> wrote:

>

>     >

>     >

>     > On Nov 10, 2019, at 02:15 , JORDI PALET MARTINEZ via RPD

>     <rpd at afrinic.net <mailto:rpd at afrinic.net>>

>     > wrote:

>     >

>     > Hi Andrew,

>     >

>     >

>     > El 9/11/19 6:19, "Andrew Alston"

>     <Andrew.Alston at liquidtelecom.com

>     <mailto:Andrew.Alston at liquidtelecom.com>>

>     > escribi?:

>     >

>     > Coupla comments on this one:

>     >

>     > *In 3.3: **If both Working Group Chairs are unable to attend the

>     PPM, the

>     > Board will on-the-spot designate a nonconflictive Chair for the

>     session,

>     > that will be assisted by the staff*

>     >

>     > I don?t believe the board should designate, it should preferably

>     be an on

>     > the spot nomination and floor election by show of hands or other

>     mechanism

>     > at the meeting.  There has been and should remain separation

>     between the

>     > board and the PDP process ? since it is the boards duty to

>     ratify the

>     > process followed to declare consensus on any policy passed, and

>     hence,

>     > should a designate declare consensus on any policy, you create a

>     conflict

>     > of interest situation.

>     >

>     > ?    I?m trying to avoid wasting time in the meeting. The idea

>     is that

>     > this is done up-front the meeting, not during the policy

>     session. I?m happy

>     > to change that to the nomination committee if you think that

>     will resolve

>     > the issue. As I mention in my previous email, I didn?t want to

>     use the

>     > committees references because those are called by the board, so

>     it should

>     > be the board, if those committees exist, the one that delegate the

>     > functions. An alternative is to explicitly have some text in the

>     policy

>     > that indicates that the board is that ?top responsible, but it must

>     > delegate that functions thru a nomination committee?, without

>     entering in

>     > this policy on the details of that committee. What do you think ?.

>     >

>     >

>     > I think that the situation where no co-chair attends the meeting

>     is rare

>     > and exceptional enough that using some time at the meeting to

>     nominate and

>     > elect a chair pro tem is a perfectly valid solution and that any

>     other

>     > solution presented so far is problematic.

>     >

>     > Also, assuming that we know before the meeting that the

>     co-chairs will be

>     > unable to attend is a perilous assumption to build into the

>     process, IMHO.

>     >

>

>     Caleb : A clear example of what happened in Kampala where we had

>     only a

>     Co-chair supported by staff to chair a meeting. How will a single

>     co-chair

>     measure rough concensus?

>

>     What you can't predict is visa rejections or if there is a travel

>     ban on

>     the country that one of the co-chair is from.

>     Imagine a scenario where the two co-chairs are from the same

>     country as

>     canvassed by some members on this list, what happens at that PDP

>     meeting

>     when any or both of them suffers visa rejection or there is a

>     civil unrest

>     in that country leading to travel ban?

>

>     We should in most case see the call to even ensure that no two

>     Co-chair

>     emanate from the same country to reduce the risk exposure of the

>     PDPWG and

>     uneventful scenarios like that.

>

>     >

>     >

>     > *In 3.3.1 Bullet point 6: PDWG Chairs will each serve staggered

>     two-year

>     > terms. PDWG Chairs may only be re-elected for one consecutive

>     term but are

>     > illegible to run again after a minimum one-year pause. *

>     >

>     > Should that not say they are eligible after a minimum one-year

>     pause?  I

>     > presume that?s a typo?

>     >

>     > Yes, definitively, word autocorrection made a trouble here!

>     >

>     >

>     > Personally, I am opposed to term limits. We have elections.

>     Co-Chairs are

>     > up for re-election no less than every two years. If people want

>     a new

>     > co-chair, they can easily vote for that. Why deprive voters of

>     choice? Do

>     > you not trust the electorate to make good choices?

>     >

>

>     Caleb: I do see a valid point of of argument here and I'm sure the

>     co-authors have taking note of your point.

>

>

>     >

>     > *In 3.3.2 Bullet point 7: AFRINIC will communicate the names of

>     acceptable

>     > candidates to the RPD List, announcing where candidate

>     information will be

>     > published.*

>     >

>     > This is ambiguous in my view point ? you state in bullet point 3

>     of the

>     > same section that anyone who has been part of the RPD list for a

>     minimum of

>     > 6 months may participate, in section 3.3.1 you specify a one

>     year minimum

>     > pause ? but beyond that ? what does acceptable mean? This needs

>     some kind

>     > of definition ? and I?d be quite happy to say explicitly that if

>     those 2

>     > criteria are met, they are eligible and then let the community

>     decide, but

>     > if it does mean more than that, it needs less ambiguity.

>     >

>     > In 3.3.1, we have:

>     > ?In addition to the candidate?s biographical information,

>     nominations must

>     > include specific information that allows assessing their

>     contribution,

>     > participation and experience in the PDWG. The candidates must

>     also provide

>     > information about what they will like to achieve during their term,

>     > possible improvements to the PDWG, etc.?

>     >

>     > Because AFRINIC (board -> nomination committee) will check all the

>     > information, they will verify that the candidate has been there

>     for 6

>     > months, can be re-elected (if is the case), provide their

>     biographical

>     > information and some statement about what they want to achieve.

>     Only if all

>     > that has been provided, is an acceptable candidate. However,

>     keep reading ?

>     >

>     >

>     > I am very hesitant to open the can of worms wherein the board is

>     allowed

>     > to deem RPD co-chair candidates acceptable or not. Am I the only

>     one that

>     > sees this as a serious avenue for abuse of the process to stack

>     the deck on

>     > available candidates?

>     >

>     > My statement here in no way accuses or reflects on the current

>     AfriNIC

>     > board. It is a question of procedural abuse that I would raise

>     in any such

>     > construct regardless of the trust or esteem I held for the board in

>     > question. These rules should be intended to last well beyond the

>     current

>     > board and must, therefore, consider any possible board makeup.

>     >

>     >

>     > *In 3.3.2 Bullet Point 8: A period of 10 calendar days will then

>     begin

>     > during which the community will be able to contribute relevant

>     information

>     > on the candidates. This information, if confirmed, may be published

>     > simultaneously for all candidates on the first working day

>     following the

>     > end of the 10-day period. As a result of that information, the

>     Board could

>     > disqualify any candidate.*

>     >

>     > This represents a big problem to me.  It states as a result of the

>     > information the board may disqualify any candidate. This is wide

>     open and

>     > allows the arbitrary disqualification of candidates. If someone

>     is to be

>     > disqualified, it should be on the grounds of not meeting a

>     defined set of

>     > acceptable criteria ? that are published and known and codified

>     in policy.

>     > Anything else could result in similar conflict of interest to that

>     > mentioned in the first point in this email.

>     >

>     > If the board->nomination committee, receives any information that is

>     > relevant and confirmed (so not a ?rumour?), it could be publish

>     and any

>     > candidate be disqualified. Let me have an example. A candidate

>     has been

>     > divorced 10 times. Is that relevant to the community? Clearly

>     not. However,

>     > a candidate has been elected as the chair of an ISPs association

>     and then

>     > has not followed his duties and this can be publicly confirmed

>     (example, a

>     > press release from the association indicating that they fired

>     its chair).

>     > Do you think the candidate is acceptable? The board can decide if

>     > publishing a link to the (already) public information, so the

>     community

>     > knows when voting, or directly disqualify the candidate.

>     >

>     >

>     > This is a _REALLY_ bad idea, IMHO. The board should not be

>     preemptively

>     > disqualifying candidates on judgment calls in lieu of the

>     judgment of the

>     > community.

>     >

>

>     Caleb:The last i checked, there are election and nomination

>     committees with

>     just a board representative.

>

>     The process I think allows for an open call for Expression of

>     Interest for

>     anyone who wants to participate in that committee which is

>     independent of

>     the board but has a board representation in that committee that

>     provides

>     guidance and also give feedback to the board.

>

>

>     >

>     > *In 3.3.2 Bullet point 9: If any objections are raised by a

>     member of the

>     > community, such objections must be communicated to the Board

>     within 7

>     > calendar days of the announcement of the results. The Board will

>     then

>     > assess whether such objections are significant and have been

>     proven. If no

>     > objections are raised, or if those aren?t considered, will

>     proceed to

>     > ratify the winning candidate.*

>     >

>     > I would prefer this be done in a more open manner. That is to

>     say that if

>     > an objection is raised, the objection and the consideration

>     thereof should

>     > be made public.  The community should be able to see the

>     objections and why

>     > they were adjudicated in a particular manner.

>     >

>     > The statement doesn?t say that the board ?can?t? publish the

>     information.

>     > But I think those objections should only be publish if the

>     candidate is

>     > disqualified because the objections have been proven.

>     >

>     >

>     Caleb : I think the lexicon use of board as against the Nomination &

>     Elections Committee can serve as an alternative.

>

>     >

>     > No? The objections and the adjudication of those objections must

>     be public

>     > regardless of the outcome. To do otherwise creates a significant

>     potential

>     > for abuse of process. The objections should not be anonymous and any

>     > anonymous objections should not be mentioned. Whoever raises an

>     objection

>     > should be identified right along with said objection.

>     >

>     >

>     > *In 3.3.2 final point: The Board is the highest instance of

>     appeal in

>     > matters relating to the election process. The board may delegate

>     some or

>     > all of the required functions into the Election and Nomination

>     Committees.*

>     >

>     >

>     > Caleb : A suggestion here for governance Committee will be

>     appropriate.

>

>     > I would prefer this be handled by an appeal committee appointed

>     outside of

>     > the electoral process, and whose members are ineligible for

>     participation

>     > in the main election.  Again, I do not believe that the board

>     should be

>     > involved in the functioning of the PDP since it is they that

>     have to ratify

>     > policy that comes through the process.  Hence, as per a few of

>     my other

>     > points ? I would prefer clear segregation.  While I acknowledge

>     and fully

>     > agree that a board member, in his personal capacity, has every

>     right to

>     > participate in discussions around a policy ? since board members are

>     > naturally members of the community, I do not believe that they

>     should hold

>     > a position to influence anything in the election of a candidate.

>     >

>     >

>     > I agree here.

>     >

>     >

>     > I?ve not read recently how the nomination committee is elected,

>     and I

>     > think unless somebody raised problems on that, we should trust

>     that is

>     > working. Otherwise, that requires a different policy or

>     procedure, or the

>     > board addressing it. You don?t think so?

>     >

>     >

>     > The nomination committee is not what is being discussed in

>     3.3.2, so I am

>     > not sure how this comment applies to the discussion above.

>     >

>     >

>     > I?m with you that participants on the committees aren?t valid

>     candidates,

>     > but is not that part already of those procedures, or should we

>     mention it

>     > here?

>     >

>     >

>     > You are proposing a new procedures document that will supersede the

>     > existing one. As such, it should be mentioned here.

>     >

>     >

>     > Similarly, I?m with you about the segregation of functions, but

>     I already

>     > responded to this in a previous email, so not repeating it here.

>     I also

>     > think that it is ?easier? for the community if the board members

>     do not

>     > participate in the discussions, **however** they have the full

>     right to

>     > do so, as community members (as well as chairs), and the only

>     requisite (as

>     > we do in IETF and all the other RIRs) is that if they express a

>     personal

>     > opinion, they clearly say ?hat off this is my personal view on

>     this?, and

>     > this personal opinion is not used in a decision for ratifying or

>     not a

>     > policy (ratification should be based on ?has the process been

>     followed? Has

>     > this policy a crazy impact on the membership and endangers this

>     > organization??).

>     >

>     >

>     > Agreed, but the above 3.3.2 language puts the board as the final and

>     > ultimate arbiter of the appeals process and that is not a good idea.

>     >

>     > Owen

>     >

>

>

>     Caleb Ogundele

>

>     >

>     > _______________________________________________

>     > RPD mailing list

>     > RPD at afrinic.net <mailto:RPD at afrinic.net>

>     > https://lists.afrinic.net/mailman/listinfo/rpd

>     >

>     -------------- next part --------------

>     An HTML attachment was scrubbed...

>     URL:

>     <https://lists.afrinic.net/pipermail/rpd/attachments/20191110/68602c38/attachment.html>

>

>     ------------------------------

>

>     Subject: Digest Footer

>

>     _______________________________________________

>     RPD mailing list

>     RPD at afrinic.net <mailto:RPD at afrinic.net>

>     https://lists.afrinic.net/mailman/listinfo/rpd

>

>

>     ------------------------------

>

>     End of RPD Digest, Vol 158, Issue 56

>     ************************************

>

>

> _______________________________________________

> RPD mailing list

> RPD at afrinic.net

> https://lists.afrinic.net/mailman/listinfo/rpd

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.afrinic.net/pipermail/rpd/attachments/20191112/98b5b137/attachment-0001.html>