[ppml] Policy Proposal 2003-15: IPv4 Allocation Policy for the Africa Portion of the ARIN Region

[Owen DeLong]

> I'd like to address your misconception that AfriNIC is vaporware.
> - AfriNIC was formally proposed in 1997 at Kuala Lumpur
> - Numerous AfriNIC meetings were held in the following years
> - AfriNIC is now a recognised organisation with numerous sponsors from
> around Africa (including private sponsors and government)
> - AfriNIC is in the process of setting up offices in four countries
> (Egypt, Ghana, Mauritius and South Africa)
> - AfriNIC has an elected board (in fact it has had for many years)
> - AfriNIC has two hostmasters (who are in training)
>
OK... I stand corrected. AfriNIC is not vaporware.  An AfriNIC RIR is
vaporware.

> It is true, however, that AfriNIC is not yet able to operate as a
> self-sufficient NIC. That will only happen once the offices are all
> operational, additional staff have been hired and trained and policies and
> fees have been formalised.
>
> The reason AfriNIC will operate as a mini-registry within RIPE initially
> is so that the organisation can draw on RIPE's expertise, learn from an
> establised NIC, and keep startup costs to a minimum.
>
Then, at that time, ARIN and APNIC should transfer to RIPE/AfriNIC the
delegations that apply to Africa and AfriNIC should gain control of policies
for those allocations.  Don't get me wrong, I'm all for that.  I wish 
AfriNIC
all the success and independence it deserves.  I like AfriNIC as a concept.
I did not mean to imply that I didn't think highly of the project.  However,
I don't think that the project is at such a state that it makes sense to
regard speculation about what AfriNIC will do as a guideline for setting
policy in ARIN.

> Mury wrote:
>> This brings up a new question for me.  When AfriNIC becomes a reality are
>> they going to implement this policy any way?  If so, we may as well do it
>> now.  I don't agree with it, but if it's a future reality why wait.
>
> When AfriNIC first starts making assignments/allocations (???), it will
> draw strongly on RIPE and ARIN's existing policies, especially where
> there are similarities. Given the almost unanimous support from the
> African community for 2003-15, I think it is obvious that AfriNIC will
> take this into consideration.
>
> I agree with you wholeheartedly - why wait? The reality is that supporting
> 2003-15 is very similar to supporting the formation of AfriNIC. Both
> ackowlege that the region needs to influence its governance.
>
I support the formation of AfriNIC.  However, there are non-African
ramifications to ARIN adopting 2003-15 which I believe override the
desire to provide this solution for Africa.  I believe 2003-15 represents
a good policy with one flaw.  It is Africa specific.  As long as it is
Africa specific, I will vote against it.  If 2003-15 is amended to cover
all of ARIN, I will vote for it.  I will vote for 2002-3.  I will vote
for an amendment to 2002-3 to expand it's scope to include Allocations as
well as assignments.

> Owen, I don't see how 2003-15 would compromise 2002-3. If anything, I
> believe that if 2003-15 is passed, it will create a strong precedent for
> the passing of 2002-3.
>
You and I have differing views of the political realities of ARIN.  How
many years have you been watching what happens with ARIN?  Admittedly, I
am a recent addition to ARIN PPML and Public Policy meetings, and, Chicago
will be my first meeting as an ARIN member.  However, I have been involved
in ARIN politics as a small(er) and/or large provider since it's inception.

> Where the two proposals differ fundamentally, is that one establishes a
> fundamental acceptance that the African region needs to be strongly
> acknowleged in the setting of policies that affect it. This is why I
> support both 2002-3 and 2003-15 and believe that they should be evaluated
> separately.
>
That is only one place they differ.  There are two others.  In 2003-15, it
creates support for micro-allcations in Africa ONLY.  This will be used by
large(ish) ISPs in north america as a justification that mciro-assignments
and micro-allocations aren't necessary in North America, and, see, we made
this exception for Africa, so the problem is solved.  Further, I have no
reason to believe that African ISPs have the resources or desire to actively
work for 2002-3 once they receive 2003-15.  This is not a criticism, just
reality.  They have much bigger problems to address and much more limited
resources.  If we all support an ammended 2002-3, then we all win.  If we
continue to divide support between 2003-15 and 2002-3, then we will all
lose.

The second way in which they are different is that 2003-15 provides for
allocations, whereas 2002-3 provides for assignments.  Both assignments
and allocations are necessary at the /22 boundary, and that is why
I have proposed an amendment to 2002-3 to achieve that.

> I also find it alarming that you think that Africa will not back 2002-3 as
> well. The reality is that we have felt the pain first hand and accordingly
> are immensely sympathetic to our colleagues world-wide. We understand that
> in order to gain the support of the global Internet community, we need to
> support it too. At the same time, I'd strongly discourage alienating the
> African Internet community, because we are your strongest allies with
> proposals like 2002-3.
>
It's not that I don't think you will back it.  It's that I don't think that
you have the resources to keep attention focused on it after you receive
2003-15 if 2002-3 tunrs into a longer battle.  Be realistic here.  Many
people have pointed out the difficulty and scarcity of resources for
African ISPs.  A prolonged political struggle for something that won't
really effect them is not likely something they will choose to do.

> Yes, we have been rather silent in the past, but the reality is that until
> recently most Africans were unaware that ARIN was interested in what they
> had to say. The time devoted to AfriNIC and ARIN presence at the iWeek
> conference in Johannesburg has created a lot more awareness than there
> was. In the future, you will see more of us both on this list and at ARIN
> meetings.
>
That's fantastic.  I'm glad to see you guys here, and, I'm very glad to
see you proposing policy, even if I don't agree with the proposal.  I
welcome you to the process and I look forward to working with you as things
evolve.  I agree with you that in an ideal world, we could simply support
both policies and it would work out well for all.  However, we don't live
in an ideal world, and, the large providers that traditionally dominate
policy in the ARIN community will use 2003-15 as a means to squash 2002-3
if it passes.  I understand that you don't see it that way, and, I wish I
didn't.  However, in North America, economic self interest is the primary
motivating factor in how most providers have been voting from what I have
observed.  It is in the economic interest of large providers to make it
difficult for large businesses and small(er) ISPs to change upstream
providers.  As such, they tend to vote for policies that allow them to
do so.

Owen

> Owen wrote:
>> I don't know if I can do this on the list or not, but, I hereby formally
>> propose that proposal 2002-3 be amended to include both allocations and
>> assignments.
>
> I support this suggestion.
>
> Finally, I want to address what I see as one of the concerns that 2003-15
> may introduce - what happens when a North American ISP asks, "Why am I
> being treated differently?"
>
> My answer (assuming hypothetically that 2003-15 is accepted) would be:
> ARIN members have afforded Africa the right to influence the policy that
> is applied to their region. Similar proposals (eg. 2002-3) are on the
> cards and apply to the North American region; We strongly suggest that
> you voice your support for these proposals.
>
> Regards
> Gregory Massel
> co-chairman: ISPA (South Africa)
> http://www.ispa.org.za/
>