[rpd] AfriNIC policy AFPUB-2014-GEN-002-DRAFT-01 reject

[Owen DeLong]

> On Oct 23, 2014, at 8:51 PM, Borg <virtual.borg at gmail.com> wrote:
> 
> 
> 
> Borg le Chevalier
> ___________________________________
> "Common sense is what tells us the world is flat"
> 
> On 23 October 2014 00:28, Owen DeLong <owen at delong.com <mailto:owen at delong.com>> wrote:
>> 
>> I can concretely demonstrate that the availability of IPv4 addressing can have an impact on business investment.
>> 
>> 
>> 
>> Impact, sure. but not determining factor in any significant investment that can be said to be "developing africa". It the arrogance of the statement I mention.
> 
> Actually, it could well be a determining factor.
> 
> 
> For mature market maybe, but for new market, no really.

I disagree… If I were considering investing in any sort of network infrastructure company and became aware that getting the addresses necessary to make a business viable may be difficult, I’d want to insure access to those addresses or withdraw my investment.

If that doesn’t meet your definition of a determining factor, I’m not sure what does.

I will agree that addresses alone will not create investment, but, I think it is absurd to claim that a lack of addresses cannot prevent investment.

> 
> Nobody in their right mind is launching a new IPv4 ISP in Asia right now. The lack of available addressing makes such a venture foolhardy at best.
> 
> 
> The service is Internet, not IPv4-based Internet. Any big investiment in Internet infrastructure in Africa will target non-consumption. Those people who has never have Internet - anything  (IPv6, NAT) will be better and still make the investiment worth it. NOBODY invests invests in networking infrastrasture in large scale in Africa just because of availability of IPv4.

ROFLMAO — I wish that was true. However, internet service today, as much as I wish it were otherwise, is pretty much defined in terms of IPv4. If you don’t provide IPv4 services to your customers, you won’t get customers, no matter how much IPv6 you deploy. That will change in the next few years, but today, that is the essential reality.

I’m not saying anyone will invest in large scale infrastructure just because of IPv4 availability. I’m saying that lack of IPv4 availability can certainly prevent investment.

I don’t think we are actually disagreeing so much as we are approaching from different aspects of the words “determining factor”.

You seem to define “determining factor” as an item whose presence or absence is, by itself, sufficient to cause a positive decision to invest. I don’t believe any such item exists in any business.

OTOH, I am defining “determining factor” as an item whose presence or absence can, by itself, cause a negative decision and prevent one from making an investment. For example, if I were opening a cell-phone store, knowing that a company had a 75% retail tax rate on handsets and a cap on the retail price of said handsets might be sufficient to prevent me from doing so. I would consider that a determining factor.

If I’m considering building out a $20,000,000 network infrastructure, the fact that I can’t get enough numbers to satisfy my anticipated customer base would certainly give me pause, if not outright eliminate my investment.

> Africa has more IPv4 space available than anywhere else in the world right now. That makes it more reasonable for new market entrants to invest in Africa,
> 
> The opportunity for investment is because lots of africans are still offline, NOT because there is IPv4. And for those people offline, any Internet (IPv6, NAT) will be sufficient.

We should probably agree to disagree here. I doubt either of us will convince the other.

> What investor (except IPv4 brokers) bases an investment decision on a resource that is known to be not sustainable to get in the futur?

Shell
Chevron
BP

to name but a few.

> but it also makes Africa a target for pulling resources out to exploit them in other regions.
> 
> 
> That the biggest concern and i think a legitime one for africans. Policy might just be an ineffective way - better have the resources being used by legitimate african interests (gouvenments, universities, etc)

I’m all for that, but this mailing list is about address policy and that’s the only tool we have to effect that. While policy may prove not to be 100% effective, isn’t it best to codify our intent and desire in that policy and try to make it work rather than simply avoid doing anything on the basis that policy might not solve the entire problem?

It seems to me that your approach here is akin to saying “I’m not going to build an engine because without brakes, a car cannot be safely driven.” I prefer to build the engine and also try to find a supplier or build some brakes too.

> I also believe that a lack of limitations on the exportation of address space by pseudo-African companies[1] and/or non-African companies[2] will lead to exploitation that does not particularly benefit Africa or Africans.
> 
> ++1 and that is the problem these proposal try to address (not sure how effective it would have be)
>  
> 
>  I’m not sure where brokers came into this or what you mean by arrogant IT people who over-value IT in general and IPv4 in particular, so I’m not sure how to answer that.
> 
> 
> Arrogance = attitude that get someone to say "I am a member of afrinic, investing an making africa a better place" --- 
> Arrogance = believe that in grande scheme of developmental challenges, Internet (particularly IPv4-based one) ranks in even the top 3 of priorities. 
> 
> Don't get me wrong, IT is very important .... depending on existing infrastructure and context. It not a silver bullet to 'make africa a better place’ 

I don’t think anyone believes for one second that IT or the Internet is a silver bullet. However, the simple fact is that it is a positive force for change and does make a difference. It is a technology that can be leveraged to facilitate many other improvements.

> You forget
> 
> a) Addresses != IPv4 (thankfully Liquid also do v6)

I’ll leave it to Andrew to clarify if I am wrong, but I am pretty sure that Liquid would not consider it economically viable to build out an IPv6-only service at this time. If you’re not building an IPv6-only service, then, IPv4 addresses are a critical component in your business.

Look up what I did for the last 5 years, and I think you’ll find that your statement above really doesn’t hold water.

> b) Going with your model - that business strategy is going to be dead in 4 - 5 years as v6 run out.

I don’t expect v6 to run out in 4-5 years. I expect v4 to run out, but I also expect that to drive more v6 implementation. The state of the network today is that an investment in an v6-only infrastructure is not economically viable. In 4-5 years, a v6-only infrastructure will probably be much more viable than it is today.

> c) In term of numbers, Liquid may not have more customers than the mobile operators in africa and somehow they still in business and thriving.

Sure, you don’t need as many high-margin customers to make a profit as you do low-margin customers. This is a long standing reality of many businesses. I don’t know an appropriate set of brand comparisons for Africa, but I’m pretty sure you can look up the US brand names…

Morton’s doesn’t sell very many hamburgers, but they make a lot more on every one than Burger King.

However, getting back to Liquid, you actually need more IPv4 addresses per high-margin customer than you do for low-margin customers like mobile phone users. It is viable, for example, to run an IPv6-only mobile phone network using NAT64 to provide what few IPv4 services you offer. It’s not viable to do that for, say, a cable access provider, DSL provider, or GPON provider. It’s certainly not viable for a B2B provider.

Why? Well, first, people take limitations on their internet capabilities on mobile phones for granted. They aren’t so accepting of such limitations in residential or business broadband. Second, mobile is a volume business, so as long as 95% of customers get 80% of what they want, chances are you’re not going to have too many people leaving just on account of the internet limitations.

OTOH, as a residential or business access provider, failing to deliver all the internet capabilities that the customers expect (or even falling short of what your competitors can deliver which includes things like Skype, Amazon.com <http://amazon.com/>, QQ.com <http://qq.com/>, and others which are unfortunately still not at all IPv6 enabled.

You can argue about immature markets all you want, but the reality is that as soon as anyone gets any level of internet access, they start talking to people about it and they quickly learn if their access is less than what is available elsewhere and it does, in fact, lead to dissatisfaction.

>> I do believe that we need to protect the African resources from being pillaged by non-African entities and IP Brokers.  I will strongly support policy that prevents African resources flowing off continent into the hands of those who have absolutely no link to Africa AND where there is no benefit (quantitative) to the continent. 
>> 
>> 
>> 
>> That seem to me to be a goal many people in the community share. but i wander , with such small IPv4 space compared to other part of world, why is afrinic consumption so small? 
> 
> There are a number of historic reasons relating to the following problem areas:
> 
> 	1.	Difficulty qualifying for resources under ARIN policies of the past (pre-AfriNIC)
> 	2.	Perception problems created by 1
> 	3.	Cost and the perceived cost-savings of NAT (artificial, though they are)
> 	4.	Perceived security/control/etc. advantages of NAT (also artificial)
> 	5.	Limited understanding of the process of obtaining addresses and fee structures
>> I just do not believe at this point we have a policy proposal that fulfills these objectives. 
>> 
> Rather than merely state that the current proposal doesn’t meet the objectives, perhaps it would be more useful if you could propose language or modifications that could be made to the current proposal in order to achieve what I think almost everyone agrees is the valid goal.
> 
> that phrase was andrew, not me. :-)
> 
> but with your nice list of reasons - i think
> 
> - policy (one that establish and discourage abuse of afrinic resources for pseudo-african companies) is good start. 

Agreed… So let’s come up with wording for one. If this isn’t it, how does it differ from what you and Andrew want? How can we either make this proposal better, or, what would we need in a new proposal in order to match those requirements better than this proposal?

> - problems (2,3,4,5) can be solve by education and communication and training

And those efforts are ongoing, but they’re out of scope for this list.

> - standardize process so legitime companies can get address quicker and cheaper

Sounds great as a theory, but the reality is quite a bit harder… What changes to current process, exactly, would you propose in order to accomplish that?

Owen


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.afrinic.net/pipermail/rpd/attachments/20141027/44c08c7f/attachment.html>