Let it go, let it go: How global DNS could survive in the frozen lands outside US control
Public comments on revised IANA transition plan
The plan to move control of the top-level of the internet from the US government to California non-profit ICANN has received broad approval following a month-long public comment period.
The group responsible for guiding the fate of the "names" aspect of the critical IANA contract – the bit that oversees the global DNS – has recommended that the contract is moved from the US government to an affiliate of ICANN that has its own board and would be legally separable at a future date if ICANN was deemed to be doing a poor job.
The group's report, which is a revised version of an earlier, more complex design, also recommends the creation of two more groups. One would be made up of IANA's customers, those that run internet registries, and would draw up a service level agreement with ICANN that it would also monitor. A second would act as a formal review body with the power to order a separation of the IANA contract from ICANN.
The majority of responses were supportive of the plan, with many mentioning how it was an improvement on the first draft. However there were still important outliers on the key recommendation of an ICANN affiliate holding the contract.
The fact that the ICANN Board continues to push this line even when the majority of the internet community has made it plain that such an approach is unacceptable could be a worrying sign of conflict ahead.
IANA: What's at stake?
The US government contracts non-profit ICANN to run the so-called IANA functions – a body that runs the highest level of the world's DNS, allocates IP addresses, and ensures developers can agree on the same numbers and protocols when writing software that communicates over the 'net. It's what keeps the internet as we know it glued together.
That crucial contract is coming to an end, and because the US wants to step away from ruling the internet like an unelected king, the future of the IANA functions is being explored by a panel of experts called the Community Working Group (CWG). ICANN, of course, would love to run IANA all by itself.
And on the other side are those that want a separate company that holds the IANA contract and awards it to ICANN. They fear that an affiliate model effectively hands ICANN control of the IANA contract using a different name. The Indian government and the organization behind New Zealand's .nz top-level domain fall into this camp.
However, the vast majority of respondents believe that the affiliate model is a good compromise and the focus of attention has now moved onto how exactly the affiliate (currently called Post-Transition IANA, or PTI) will be structured and run.
Details, details, details
A majority of respondents highlight that while the approach looks about right, the devil is in the detail.
And key within the details is how the new PFI Board would be structured, what its role would be, and what the legal construction between it and ICANN is.
The Swedish government summed up many concerns in its submission: "The next version of the CWG-Stewardship proposal would benefit to include a process for designing the PTI-IANA contract, a process to establish community consent before entering the contract, explicit mention of whom the contracting parties are and what their legal responsibilities would be in relation to it."
Broadly, some feel that ICANN should select all the Board members while others feel that the PFI Board members should be chosen either by IANA's customers or representative groups within ICANN.
And there is the question of the degree of autonomy that the PFI should be given. Some feel it is very important that it be able to operate independently, even suggesting that the PFI is incorporated in a different country, to prevent ICANN from slowly subsuming the organization. Others feel that a high degree of autonomy risks creates additional bureaucracy and process for no discernible improvement in actual services.
The majority of commenters are in favor of greater autonomy but, again, the ICANN Board pushes strongly on the idea of it being in effective control of all aspects of the PFI.
One of the bigger issues from a broader perspective is the fact that so many of the respondents complained about the actual way public comment was sought and received.
Numerous organizations, including governments and business organizations, complained that the plan was only put out to public comment for 28 days, which was not long enough for proper review. This was shortened from the normal 40 days following a decision by two unnamed ICANN staff to do so.
The 28-day time period was even shorter for those whose first languages is not English - translated versions of the plan were not released until the comment period was almost over - something that the Chinese community was particularly vocal about.
ICANN seems pleased with the 53 comments received but considering the importance of the decision and its global impact, it represents a low turnout. The respondents are also only those individuals or organizations that are already closely following ICANN and its processes. In other words, the wider world is not engaged.
Worrying, some of the responses also make a point of arguing for another public comment period when more details are available: suggesting that the opportunity for further public input is not a given.
This is far from the first time that the insular ICANN has been caught out making decisions within its own echo chamber. Most famously, it was forced to revisit critical details in its flagship "new gTLD" program when the wider world first became aware of the plans to open up the internet's namespace. A series of "overarching issues" were identified and work through, delaying the program's approval by four years.
Incomplete and inconsistent?
Two other big issues also emerged from the public comment period.
First, aside from the key details not being available, the work of this review group is highly dependent on the work of an entirely separate group working on improving ICANN's accountability.
The broad issue there is that most - including the US government - feel that improvements need to be made to ICANN itself before it is allowed any level of control over the IANA contract. That work is still out for public comment, adding an extra layer of complexity to the process: what results from one will impact the other.
Secondly, it is not clear how the plan from the "naming" community fits with the plans already developed and approved by the two other main functions with the IANA contract: numbering and protocol assignments. Both of these are equally as important in terms of internet functioning but are less complex and less political.
Both the numbering and protocol communities have decided that they want to have a direct contractual relationship with ICANN. And both insist on the ability to end that contract on their own terms (something that ICANN's lawyers are trying hard to pull out). Both communities sent responses warning they will need to look at how their needs fit with the proposed PFI model.
One group noted that they "do not foresee any incompatibility" but the group behind one of the proposals said it will need to "conduct legal analysis" before being "able to arrive at a definitive conclusion on this issue". And that means more time.
Taken overall the response to version two of the IANA transition plan is: so far so good, but we've got a long way to go. ®