Comment Period |
Important Information Links |
Open Date: |
21 November 2011 |
Close Date: |
30 December 2011 |
Time (UTC): |
23:59 |
Originating Organization: |
ICANN Policy Department |
Purpose: |
ICANN Staff is seeking comments on the Preliminary Issue Report on 'Thick' Whois [PDF, 635 KB]. |
Current Status: |
The Preliminary Issue Report explores the issue of 'thick' Whois and makes recommendations on whether the GNSO Council should initiate a policy development process on the requirement of 'thick' Whois for all incumbent gTLDs. This Public Comment solicitation represents an opportunity to share perspectives on 'thick' Whois, the Preliminary Issue Report contents, and whether the GNSO Council should initiate a Policy Development Process on this issue. |
Next Steps: |
The Preliminary Issue Report will be updated to reflect community feedback submitted through this forum. A Final Issue Report will then be presented to the GNSO Council for its consideration as to whether a Policy Development Process should be initiated. |
Staff Contact: |
Marika Konings |
Email: |
Click here to download the ALAC Statement on the Thick Whois Preliminary Report in PDF format.
ALAC Statement on the Thick Whois Preliminary Report
23 December 2011
The ALAC generally supports all gTLD registries using the thick Whois model, but regardless of outcome, the ALAC strongly encourages the GNSO to initiate such a PDP.
We believe that the Preliminary Issue Report reasonably reflect the issues and specifications for a PDP, with one exception. Specifically the following is problematic:
Are there other models that could / should be considered? For example, in the context of registrar deliberations on this issue, it was suggested that ‘depositing the [W]hois at a common third party across all ICANN-contracted TLDs’ could be considered.
We find this problematic for three reasons:
- Looking at the specific example, having a common third party for all gTLDs is, to some extent, counter to the intent of having registries spread around the world. It centralizes where the long-term intent is to decentralize. Even if a less centralized approach is taken where there are multiple Whois providers, presumably certified or accredited by ICANN, we end up with ICANN taking on a new responsibility without any real substantive benefits. Moreover, it takes a level of control out of the hands of the Registry. This level of control has been important when one looks at the tiered access that some Registries provide to meet local/regional privacy concerns.
- In the more general case, opening up this PDP to an overall review of Whois models (and perhaps ultimately the Whois concept itself) will take what appears to be a relatively straight-forward issue and transform it into the global Whois debate, That debate is being addressed in a number of other ways at the moment, and the GNSO should take care to ensure that this limited PDP does not morph into the larger and far more contentious issue.
- Lastly, consideration of alternate Whois models does not make sense in relation to just a few gTLDs, all the more so when there are other activities in ICANN that are at this time considering the more general Whois question, potentially including alternative models.
The ALAC does note that some in At-Large are particularly concerned with issues related to privacy, and feel that any policy development be done with the full consideration of privacy issues and in particular should not take actions that would lessen the availability of legitimate and legal access to privacy. Moreover, there have also been concerns raised that a thick Whois requires the storage of personal information outside of the country of origin and such issues might also be included in any PDP.
In summary, with the one exception noted above, the ALAC supports the Preliminary Issue Report, and supports the initiation of a PDP on the narrow issue of requiring that all Registries move to or continue to use the thick Whois model. Moreover we encourage the completion of all such deliberations within a time-frame comparable to the expected go-live launch of the new gTLDs.
(Draft version and comments follow)
Draft ALAC Statement on the Thick Whois Preliminary Report
Author: Alan Greenberg with support of Carlton Samuels
The ALAC generally supports all gTLD registries using the thick Whois model, but regardless of outcome, the ALAC strongly encourages the GNSO to initiate such a PDP.
We believe that the Preliminary Issue Report reasonably reflect the issues and specifications for a PDP, with one exception. Specifically the following is problematic:
Are there other models that could / should be considered? For example, in the context of registrar deliberations on this issue, it was suggested that ‘depositing the [W]hois at a common third party across all ICANN-contracted TLDs’ could be considered.
We find this problematic for two reasons:
- Looking at the specific example, having a common third party for all gTLDs to some extent is counter to the intent of having registries spread around the world. It centralizes where the long-term intent is to decentralize. If a less centralized approach is taken, where there are multiple Whois providers, perhaps certified by ICANN, we end up with ICANN taking on a new responsibility without any real substantive benefits. Moreover, it takes a level of control out of the hands of the Registry. This level of control has been important when one looks at the tiered access that some Registries provide to meet local/regional privacy concerns.
- In the more general case, opening up this PDP to an overall review of Whois models (and perhaps ultimately the Whois concept itself) will take what appears to be a relatively straight-forward issue and transform it into the global Whois debate, That debate is being addressed in a number of other ways at the moment, and the GNSO should take care to ensure that this limited PDP does not morph into the larger and far more contentious issue.
The ALAC does note that some in At-Large are particularly concerned with issues related to privacy, and feel that any policy development be done with the full consideration of privacy issues and in particular should not take actions that would lessen the availability of legitimate and legal access to privacy. Moreover, there have also been concerns raised that a thick Whois requires the storage of personal information outside of the country of origin and such issues might also be included in any PDP.
In summary, with the one exception noted above, the ALAC supports the Preliminary Issue Report, and supports the initiation of a PDP on the narrow issue of requiring that all Registries move to or continue to use the thick Whois model. Moreover we encourage the completion of all such deliberations within a time-frame comparable to the expected go-live launch of the new gTLDs.
20 Comments
John Levine
The concern about storage of PII outside the country of origin makes little sense unless you believe that there every country where people register domains has a local registrar, which is not true. Other than that it's fine, and I agree that mission creep (or in this case, mission leap) is undesirable.
Alan Greenberg
One further thought. I would suggest adding a third point to the list of reasons to not consider other Whois models during this exercise.
3. Moreover, consideration of alternate Whois models does not make sense in relation to just a few gTLDs, all the more so when there are other activities in ICANN that are at this time considering the more general Whois question, presumably including alternative models.
Anonymous
Agree: Till there is approriate global agreement on privacy on the Internet - bad idea to have it globally hosted. Leaves open doors for all kinds of suits.
A model where regional infomratioon is retained regioonally might be better: Most nations within a region have agreed regional protocols and laws.
Just a few early morning thoughts.
TT
Eduardo Diaz
I suggest to change the first sentence: "The ALAC generally supports all gTLD registries using the thick Whois model, but regardless of outcome, the ALAC strongly encourages the GNSO to initiate such a PDP." to something like this: "The ALAC supports all gTLD registries using the thick Whois model and strongly encourages the GNSO to initiate such PDP."
The paragraph: "1. _Looking at the specific example, having a common third party for all gTLDs to some extent..." _does not read well. It seems to contradicts itself.
First it states that: "... having a common third party for all gTLDs to some extent is counter to the intent of having registries spread around the world. It centralizes where the long-term intent is to decentralize." and then it says: _"If a less centralized approach is taken, where there are multiple Whois providers, perhaps certified by ICANN, we end up with ICANN taking on a new responsibility without any real substantive benefits." _which is contrary to the first statement.
Maybe the last sentence should be change to: "If a more centralized approach is taken, where there are less Whois providers, perhaps certified by ICANN, we end up with ICANN taking on a new responsibility without any real substantive benefits."
Maybe I am reading the English semantics in a different way but if that is the case, I recommend to re-write the whole paragraph to make it more readable to the rest of us.
Other than this, kudos to Alan and Carlton for generating this statement. I will support it.
-ed
Seth Reiss
On page 91 of the Whois Review Teams Draft Final Report available at http://www.icann.org/en/reviews/affirmation/whois-rt-draft-final-report-05dec11-en.pdf the team included in its recommendations a single portal that makes thick Whois information available to Internet users. The Team did not come to a consensus on whether the portal should be available only for the two existing thin registries or for all ICANN regulated registries. The draft report language is appended at the bottom of this note.
The portal model avoids having a single third third party in possession of thick whois data which addresses some (and arguably all) of the privacy issues and trans-border issues that have been raised. It is also addresses some or all of the security, stability and cost issues. The model was the brainchild of our very own ALAC rep to the WRT Lutz!
Data Access – Common Interface
17. To improve access to the Whois data of .COM and .NET gTLDs, the only remaining Thin Registries, ICANN should set up a dedicated, multilingual interface website to provide thick WHOIS data for them.
ALTERNATIVE for public comment:
To make WHOIS data more accessible for consumers, ICANN should set up a dedicated, multilingual interface website to allow "unrestricted and public access to accurate and complete WHOIS information". Such interface should provide thick WHOIS data for all gTLD domain names.
John Levine
There are dozens of WHOIS portal sites already. What would be the practical benefit of ICANN running yet another one?
Also, I can tell you from experience that if the underlying WHOIS is thick, running a portal is trivial, if it's thin, it's impossible. So whatever problem this is supposed to solve would be solved better by moving to all thick.
Alan Greenberg
Thanks Seth. The issue on the table is whether to change the back-end structure for .com/net/jobs/name, so I don't see the need to talk about portals at this time.
Perhaps those who advocate thick or thin will raise the concept of a portal during the substantive discussion within a PDP, but that is not the stage that we are at now.
Am I missing something?
Ganesh Kumar
John's point about PII is very insightful and a valid one. For some of us coming from different industries that are heavily regulated in terms of handling client information, complying to confidentiality regulations is a big effort in daily lives. Another complicating factor is varying levels of regulations- some stringent and some lax- in different jurisdictions. By offering a blanket support to thick whois, I am not sure we have considered the privacy expectations- legally mandated or otherwise.
Adopting a thick whois may also prevent in future collection of any information that could be used to ensure security, but needs to be kept private.
With so many issues open, initiating a PDP is appropriate.
Fatima Cambronero
Most of the gTLD agreements require “thick” Whois output, all except .com, .net, .names and .jobs. In fact, there are no ICANN standard requirements on this issue and Registries satisfy their Whois obligations using different services. For example, VeriSign offers "thin" Whois for dotCom and dotNet. On the other hand, “thick” Whois is also expected to be offered for new gTLDs.
A “thick” Whois model provides a central repository for a given registry while "thin" Whois provides a decentralized one. Therefore it is necessary to clear up an endless chain of responsibilities that could delay in time and could mean the loss of a domain name holder right.
As At-Large members, it is necessary to be in the position of Internet end users and not in registrant’s privacy perspectives. From an end Internet user point of view, it has to be asked if it would be more positive to be interested in the privacy of the information given to register a domain name or the possibility to have access to such information to pursue criminal offences. There is a public interest involved when trying to reach such domain names registrant’s information; it is also pursued for the security of the legal system, when trying to identify the criminals. While this was not the original function of the Whois service, in practice it is used in a number of other legitimate activities.
We consider it appropriate to initiate a Policy Development Process (PDP).
If approved, the future PDP should also consider a legal background analysis regarding the national legislation from the Registry country, in order to prevent from having a cancelation of the contract based on national rules. Specifically related to “thick” Whois, it is importante to identify if all users data informed for the domain name registration are considered personal data under specific law that regulates treatment, publication and transfer of personal data. ALAC could contribute to this analysis since it’s directly related to Internet end users perspectives.
Moreover, the PDP should also consider if its conclusions for “thick” or “thin” WHOIS implementation will be applied for new agreements or if it will be retroactive to all registered domain name holders.
On behalf of Natalia Enciso, Raquel Gatto and Fatima Cambronero.
Alan Greenberg
Thanks Fatima. You raise an interesting point in your last paragraph. Personally, I think it would make no sense to not make it apply to all registrants (over some reasonable period of time). Otherwise the situation would be very difficult to manage, and the benefits significantly reduced.
But that is something that must be considered if the PDP is initiated. Any such PDP would have to consider implementation issues such as convergence and retro-activity.
Avri Doria
I I tend to think that in general it makes sense to look at other models other than the per Registry thick model that is being forced onto all new Registries and considered for existing registries. I do not see this as mission creep and do see the usefulness in a world where only two models are considered viable with no real work having ever been done on proper modeling to meet the requirements of a whois service.
And I think a PDP that picks a model without considering the requirements that are being discussed elsewhere is probably a waste of time until such time as the requirements are agreed upon.
In the case of the current PDP proposal, the only valid issue I see before the requirements of the service are known and have ICANN-consensus is "should all (incumbent, new and future) Registries be governed by the same rules, including in terms of whois service.
John Levine
We've had WHOIS for thirty years. Do you really think that nobody's ever considered other models?
As a way to throw more sand in the wheels of the gTLD juggernaut, I think it's a great idea. Other than that, sheesh.
Avri Doria
I see no reason for it to delay the gTLD program. Not only is there over a year until there are any new gTLDs, all of these new gTLDs will be subject to the ICANN-consensus policy in regards to Whois.
As for studies on other models for Whois other than Thick and Thin, can you please post the URLs on the methods and outcomes those studies. Perhaps you are right and the work has already been done, but I can't recall it.
Fatima Cambronero
I also like to know about other models for Whois, please John, could you share that information? Thank you.
Alan Greenberg
It may make sense (and as I mentioned in the statement, perhaps that is already going on in another forum), but with respect to the particular subject of this Issue Report, it sounds like an enormous leap out of scope. Moreover, it sounds like a way to guarantee that in the short to medium term, the problems that were identified that led to the request for this Issue Report will stay unresolved.
I would have no problem if in the context of this particular possible PDP, the result is that all registries must use the same Whois model, and that currently, that model is thick (but could change in the future).
Regarding other models in general, there are several hundred ccTLDs that are not bound by ICANN Whois rules. Perhaps a survey of those would prove illuminating.
Avri Doria
Since this is a response on a preliminary issues report, I am not sure that the notion of scope is that tightly constraining. What we are commenting on is whether all issues have been discussed in the issues report and whether we agree with the tentative issues as defined. I think that there is probably a lot of latitude in the scope of repsonse.
John Levine
Unless more than one person thinks this is a good idea, perhaps we could give it a rest.
Avri Doria
Well at least two of us have expressed interest in seeing you back up your claim that this has been studied, in the context of gTLDs.
But I am not sure what you mean by 'give it a rest' does this mean you want to suppress debate between those who make claims and those who question them?
Certainly if I end up being in a minority of one, the consensus will pass me by. But in a conversation between just a few people it seems premature to close a topic while there are still issues to be addressed.
John Levine
Re different WHOIS approaches, the IETF is working on son-of-WHOIS right now. It's not hard to find if you're interested.
As Alan keeps pointing out, the question on the table is thick vs. thin, not imagine any possible WHOIS model that hasn't been implemented (which a single centralized server an interesting example, if only as the worst possible alternative.)
Alan Greenberg
I don't know "how tightly constrained" the scope of a Preliminary IR is, but I sort of assume it is limited to the issue specified in the request, which in this case is "requirement of 'thick' WHOIS for all incumbent gTLDs".