|
Is it time for a split between allocation and services for Internet number resources as was the case for domain name resources? Consider the following;
Back in 1996, Network Solutions had essentially four different government granted monopolies.
1) Sole Allocation of unallocated domain “name” resources for the gTLD inventory.
This function is known as gTLD Registry. Once the name record has been created, it is then granted to a registrant who requested the resource. There is an immense number of possible domain “names” with up to 65 alphanumeric choices for each gTLD registry.
2) Sole post-allocation service provider to the holder/registrant of the domain “name” resource.
These services are the updating of the records Whois information, associated linking to IP address for servers, etc.
3) Sole allocation authority for IPv4 Number Resources.
Granting from the “free” pool Network Solutions determined who would receive an allocation based upon the criteria established by ... them.
4) Sole post-allocation services for number resource holders.
These services are mainly Whois management, Routing Registry updating, sub-delegations to ISPs customers if needed, transfer of records and the crucial reverse domain name system (rDNS) through the in-addr.arpa platform.
In 1997, Network Solutions “spun” off the 3rd and 4th monopoly into a non-stock corporation known as American Registry for Internet Numbers (ARIN) which has continued the monopoly for its region after spinning off several other Regional Internet Registers (RIR) which are in themselves monopolies.
In 1999, Network Solutions agreed to split the allocation/registry business from the services/registrar business. This split has generated enormous competition on the services for existing allocations of name resources. The price for services dropped significantly. The quality increased dramatically and with competition came market forces which acted as a form of governance.
The five RIRs do not compete with each other. They have divided up the planet based upon arbitrary geographic lines. They do not need to react to market needs. They do not have a reason to invest in development of advanced self-service tools.
Therefore is it time to introduce competition for post-allocations services for number resources?
Sponsored byRadix
Sponsored byIPv4.Global
Sponsored byDNIB.com
Sponsored byVerisign
Sponsored byWhoisXML API
Sponsored byCSC
Sponsored byVerisign
I think that most folk agree that IP address distribution is the shining light of Internet Governance, while the multi-billion dollar Domain Name industry is not. Monetising IP address, ASN and reverse delegation would be a step backwards from the ideal situation we have now in terms of IP address policy making.
When you say “The five RIRs do not compete with each other.” That is coreect, nor should they. What they do is collaborate, cooperate and communicate…that seems much better than competition IMHO.
It is not quite correct to say that “They have divided up the planet based upon arbitrary geographic lines.” as the areas correspond to ICANN geography. Here in Africa, the Internet technical community was very happy to build our own RIR, one that is responsive to our needs and regional market.
I also don’t agree that “They do not have a reason to invest in development of advanced self-service tools.”, as RIRs have all built cool (advanced, self-service)tools as a direct result of their membership asking for them.
I’ll blog about this in more detail soon on CircleID.
P.S. Peter, I see from your profile that 2 companies that you are currently employed by are set up to provide just these sorts of post-allocation services. http://www.depository.net/ and http://www.denuo.com. Looks like you want to speculate on IPv4 address blocks just like folks speculate on domain names?
Mr. McGinnis, I appreciate your point of view. Maybe I was unclear, currently the services offered by the five RIRs are monopolies in regards to both allocation and services. I reconginize that to some the allocation process has been "a shining light" while others may not be so agreeable. What is not debatable is that the post-allocation services are not competitive and market driven best quality. So yes, I have invested and lead a for-profit commercial Internet Registry called Depository. It is no secret. But here is where you, and maybe the community misunderstand our business model. There are approx. 20,000+ entities that were granted over 1.85 billion numbers before the RIR regime of contract based allocation services was instituted in 1997. They do not have contracts for services from thier respective RIRs and were underwhelmed by the Legacy Services Agreement offered by ARIN. They deserve the services portion of the function of the RIRs. They already have allocations. Almost all we spoken with do not have resources from an RIR and do not intend to request them. Yet the self-service tools of the RIRs are lacking. Therefore because there is no competition for these services, the RIRs have not been proactive and built best-of-breed services. We are not expecting to be the sole company to offer these services but we are driven to be the best available and are willing to let the market decide. Lastly, I appreciate your position. I do. We are mindful of the history. But to say that collaboration is better than competition for services means that you accept the will of a cartel.
What services are the RIRs not providing that its membership is asking for?
Frank, Competition can address many issues and types of "services". Regarding the RIRs, post-allocation services can be distilled into several key groups; 1) Technical 2) Customer Support 3) Pricing/Fees 4) Contractual Addressing each separately as follows Technical In reviewing the various RIRs technical service offerings we found numerous opportunities which a market-based company can provide. Competition is the well-spring of innovation and we expect to have to constantly develop better more secure service offerings. Since we are building them and haven't released them yet, it would be unwise to reveal them at this time. We provide certain core services currently and are investing in developing specific technical tools that the market does not have. We expect competition which will drive us harder to remain relevant otherwise we will lose in the marketplace. Customer Support We use history as our guide. Network Solutions in 1996 was the sole registrar for domain names. They provided zero self-help tools for registrants to update their domain name records. Once the marketplace for registrars was opened there was an immense change in self-help tools giving registrants total control of their name resources. ARIN was spun-off whole cloth from Network Solutions in 1997, that was 13 years ago. They are now just coming online with limited self-help tools for number resource holders to manage their own resources. With a competitive marketplace, this would not have taken so long or be one type fits all. Financial/Pricing Competitive markets versus single entity monopoly pricing has been shown to create lower costs for the customer every time. I do not think that needs to debated here. Needless to say, we are very mindful of market pricing and will be very reactive otherwise our customers will walk. Contractual Not sure if you have every read the services agreements of the RIRs. If not, you should. We include a very important portion in ours. The customer has a right to terminate the contract and does not have to give up their number resource. Portability is the cornerstone of the domain name post-allocation services marketplace. The current non-competitive system by the RIR cartel doesn't allow any portability to another RIR. There is far more I can write on each subject so this is not an exhaustive list by any means but isn't it time that this strange accepted self-appointed monopoly ended? I am not advocating the end to the RIRs, just the introduction of competition for the post-allocation services.
is how to make money. RIRs provide services to their members, some of these are services that the members don't even ask for but are useful to the regional and global Internet community. The motivation behind this idea is (IIUC) to appeal to folk who have legacy blocks and to interact with RIR databases on their behalf. These legacy holders DO get free services from the RIRs despite not having contractual agreements with them regarding this legacy space. I just looked at a /16 obtained by an organisation here in Africa many years ago to confirm the above assertion. There are entries in WHOIS that show that multiple RIRs have changed the /16 inetnum object over the last two decades, and the domain object is in the AfriNIC Database, as well as in the DNS (confirmed using dig @ chia.arin.net). I don't understand why folk would pay for services that they already get for free, unless they can monetise IP address blocks via a transfer market.
Tim, You are correct. Why would any number resource holder pay for services they get for free? That very same question has been asked multiple times in business history. In the US, broadcast television offered entertainment for "free" while Cable/Satellite television was expensive. The answer to why people paid for what they received free was because they perceived value. We already have clients. These companies perceive value for getting better services than the free version. Again, we are not requesting the ability to allocate these number resources. Instead we have reviewed the entire allocation history and found that 44.14% of the entire IPv4 pool (1.85 billion) was allocated without a RIR contract and has no relationship except if they chose to enter into a legacy services agreement. These companies have been approached by each of their respective RIRs and chosen not to enter into an agreement. What astounds us is that people have never thought to challenge the system before.
The suggested changes raise flags in my mind. In particular, I do not consider it acceptable for RIRs to compete with non-RIRs for IP space allocation. If an enforced requirement that RIRs must never purchase any IP space nor have any previously purchased under their control, we begin to provide some level of protection against unfair advantage. Essentially, I firmly believe that RIRs must only provide a service to customers in a non-biased neutral way, one where the RIRs never compete against the customers.
The above stated, I echo Frank’s question above, “What services are the RIRs not providing that its membership is asking for?” Understanding the problem(s) that are trying to be solved first, and then architecting a solution that addresses those is what I consider the proper approach. This thread as it exists currently appears to be a proposed solution to unspecified problem(s).
Marc, We are not requesting nor will we ever the ability to be an authorized allocation organization. We have expended a great amount of effort and time considering the various effects to this proposal. It is our position that there should be a split between the allocation authority and the post-allocation services. Your business deals in cybersecurity. Yet the most glaring weakness we have discovered is propagated by a lack of services by the RIRs. There is no shared registry system (SRS) for Whois for number resources. Each RIR maintains their own separate database with its own data dictionary and no common protocol for publishing Whois to a central data structure. This means that law enforcement & governments are constantly chasing data bases for Whois records. Furthermore, since the legacy/historical allocations are outside the RIR system, the former InterNIC data records are inaccurate and out-of-date. This represents more cumulative space in real terms than the total allocations granted by the five RIRs combined. In listening to the legacy number resource holders, we have been asked to create a number of real and unique services. None of these are being offered by any RIR. So we have used the method of looking at problems with the current model of service delivery first, then built a business case to address the problems. It is just that once we began reviewing the marketplace, we realized that none of the RIRs have "portable" contracts as do service providers to domain resources (registrars) so we asked the question, shouldn't they be required to be competitive? In the responses so far no one has answered that question.
I don’t feel that Peter is making a very good case for his point. So far I hear only a lecture about the wonders of competition and no direct answer to Frank’s question about what’s broken today. There may well be a better way to do the work that the RIRs do but I truly do not believe that the way involves wide open competition.
Consider what that did for domain names. Drove down the cost of registration, sped up responsiveness, led to better tools. Oh, and it turned the acquisition of domain names into a free-for-all three ring circus of name speculators that hoard thousands of domains name to themselves, totally idle, on the off chance that they’ll find some sucker willing to pay crazy money to ‘buy’ the domain name off of the hoarders. I’m sorry but that has had the net effect of crippling the proper use of domain names by entities that legitimately ought to be able to use them.
As a simple example there are thousands of domains in the form of city.state.us that are owned by these hoarders. My home is the victim of one of this situation. The whole point of creating a naming convention like city.state.us (for example boston.ma.us or denver.co.us) is to make it easy for people to find web sites of readily identified places in a given country. The purchase of these domains by people that have no affiliation at all with a particular town or city therefore blocks that town or city from publishing a web site at an address where one would intuitively look. The consequence is that the internet becomes a tangle mess of domains with strange names that are made in order to navigate around the ‘potholes’ of the domain naming system.
I see no point in making more potholes. We have enough of a mess already and capitalism has never been good at creating global unifying structures. The internet, like it or not, is a global structure and needs to be managed as such.
So please, let’s come up with a better approach than a competitive land-grab type of approach. It will create worse problems than it ever solves.
It's very likely that Peter's approach will create a larger economy than what that five RIRs have in their budgets today. While from a capitalist view I can see why Peter may pursue that, trying to create value out of a market, as Dan Coutu points out, if it's anything like domain names, end-users will stay away from it with a 10-foot pole. And by end-users, I'm not including hoarders.