Re: [Ianaplan] CWG draft and its impact on the IETF

[Bob Hinden <[e-mail]>]

Hi,

> On May 21, 2015, at 12:11 AM, Christian Huitema <[e-mail]> wrote:
> 
>> However, to the extent that injecting a new organization and management
>> structure into the system (in the form of PTI) increases complexity or
>> uncertainty, it pushes us some distance along that range, with various sorts of
>> control arrangements influencing the distance.  And, to the degree to which the
>> IETF exercising "plan B" would be destabilizing (however temporarily) of the
>> IETF relationship with the protocol registries, those registries themselves, or PTI,
>> that is, itself, a reason to not move in the PTI direction without a clear picture of
>> the problem it is trying to solve and how that solution would lower instability
>> risks overall.
> 
> There is an underlying question: would the system be less stable if protocols numbers, IP addresses, and domain names were managed by three different structures? Or would it be more stable? Personally, I am not sure. Having system concentrated creates a concentration of power that increases risks for abuse. Having the system separate might increase risks that separate functions may be challenged by wannabe authorities. I am somewhat more worried about the abuse part, but what do I know?
> 

Stepping back a bit and speaking for myself.

Independent for how the IANA function is implemented, that is the PTI, the way it is now (a department in ICANN), or something else, there are still three separate functions.  These are of course, the IETF (for the protocol parameters), the RIRs (for addresses), and the ICANN naming communities (CCTLDs and TLDs for DNS names).  These three functions will not be combined.  This discussion is only about how the IANA is structured.

My personal take on the names community proposal that includes the PTI, is that they want a degree of separation of the DNS naming functions in ICANN in an analogous way that the IETF and RIRs have today for the protocol parameters and addresses.  I don’t think the naming community has this now and they would like to see more separation of policy and oversight.  I conclude they think that the PTI will achieve this.  If they are happy with this approach, then I am fine with it and think we can move forward.

I don’t see any reason why at this point in time the IETF and RIRs should not continue to have agreements with ICANN.  The PTI will be under ICANN (as a separate entity, with oversight from the ICANN board).  The ultimate responsibility will remain with ICANN.  If the PTI breaks in some way, it will be ICANNs responsibility to fix it.  As we know more about how the PTI is structured, it’s bylaws, etc., etc., that could change.

Or to say it another way using the IETF as an example, the IESG will interact with an email address like <[e-mail]> and as long as the people behind that are responsive and meet the conditions in the IETF MOU and SLA with ICANN, then all is fine.  If the implementation is in the PTI or an ICANN department, it won’t matter to the IETF.  The interface is the same, so to speak.

Bob