[cooperation-wg] DNS4EU - proposal for a community response - draft

> On 7 Feb 2022, at 14:09, Shane Kerr <shane at time-travellers.org> wrote:
> 
> Desiree,
> 
> Thank you for this proposal!

And thank you for the questions, thoughts and comments Shane.
My brief answers, questions bellow.


> On 03/02/2022 20.16, Desiree Miloshevic wrote:
>> We’d like to know if there is some support from members for having a RIPE community response on this proposal?
> 
> I think that a RIPE community response makes sense.
> 
>> We would really appreciate your feedback or any comments you’d wish to make or if you'd like us to work further on this.
> 
> I have some feedback, inline below.
> 
>> Proposed DNS4EU RIPE community statement comments
>> 1.
>>    RIPE community believes that governance of the DNS resolution chain,
>>    which is such an important element of everybody's Internet
>>    connectivity, should involve all stakeholders and can not solely
>>    rely on legislation and regulatory oversight.
> 
> I think we should be careful with this. For me, it's not really important that everyone who could conceivably be considered a stakeholder to be involved in governance of DNS resolution. Do we really think it is important if bodies like GEMA or other intellectual property organizations are part of this (to pick one of many possible groups who could claim to be a stakeholder)?
> 
> If the goal is to provide the best system for EU citizens and residents, then that should be the focus. Certainly access providers and other DNS operators can proxy their users' interests, and have a lot of expertise, and should be involved though.
> 
Multistakeholder means an opportunity for stakeholders to be involved (e.g. technical community, academia, private sector, govts, civil society) and not that everybody always needs to be involved. It ensures that a decision making processes involve more actors and do not depend on a single stakeholder’s decision.  We should not be deciding who should or not be involved - MSM model means it’s broader than one stakeholder, in this case, the legislator.
 
>> 2.
>>    RIPE Community hopes that any winning bidder will adhere to what we
>>    see as a fundamental property of the Internet, with a diverse and
>>    competitive landscape, anchored on the principles of
>>    multistakeholder Internet governance.
> 
> "bidders"? Or do expect that any contracts awarded would be to a single organization?

No, the word bidder is broader than a single entity but we can change/clarify that wording, thank you for spotting!
In my view, it'd be a good idea If diverse organisations come together and apply as a consortium, 
which I believe is also the idea behind the call.

> 
>> 3.
>>    RIPE Community believes that the responsibility of well-functioning
>>    Internet access including the DNS resolution is with the access
>>    providers. We believe it should stay that way.
> 
> Should it? Do access providers want this? Is this really the best way to provide performant, reliable, and secure DNS to users?

It’s one possible model and of course there are other models that can co-exist. 
Since access providers already provide Internet access why wouldn’t they also provide a DNS resolution?
For smaller providers, I agree that it could be perceived as additional work.

> 
> I think the market forces do not line up properly with recurvive DNS at the provider level... I doubt many customers choose a provider based on their high-quality DNS service, and so it becomes something that companies must provide even though it doesn't make them money. It's a cost.

You are right, but it's also possible that the market can align well at the provider level. With this significant financial help it could be less of a cost. 
It's one of the many options. An access provider should always have a choice of which DNS resolver service to use or and/or run its own.

I Agree that majority of customers do not choose a provider based on its high-quality DNS service.
We know that some browsers, applications, operators, device manufacturers decide for themselves and provide no choice, but perhaps this is changing. I’m sure someone can provide more input.

Some companies can align very well along the market forces (and recursive DNS at the provider level) and make it good for the end user 
and other companies can also align well with the market forces but perhaps act less in the best interest of the end user.

> 
> Public DNS resolvers make recursive DNS pay for itself, either by donations (as in the case of Quad9), paid subscriptions (as in the case of OpenDNS), or using the data gained for some other purpose in their operations (presumably the case for Google and Cloudflare). So money spent is improving their actual service; that is, a benefit.

I read this as: Diversity in the Public DNS resolvers exists as well as different market models and they should be self-sufficient.
Financial gains generated by the service should go towards further service improvements?

However, not sure if you’re suggesting a particular new text for to the statement?
> 
>> 4.
>>    We understand that to be able to minimise some risks when the end
>>    user selects a random DNS resolver, a possible and feasible solution
>>    is to have the access provider run their local DNS resolvers and/or
>>    an additional DNS resolver as a back-up.
> 
> I guess this means that when an end users picks a DNS resolver that there is a back-up. Probably that's fine, although more and more this is not the user picking, but rather their browser picking some DNS over HTTP (DoH) server. Also, note that this is not standard practice today... I don't think that I have ever seen a configuration for resolution that asks for backup servers.
> 

The rationale was not for the end users to have a back up, as they can rarely pick their DNS resolver, but for access providers to have a back up and a choice.

>> 5.
>>    We hope that the EU could allocate DNS4EU funds to the local
>>    Internet community and encourage Internet access providers to run
>>    their local DNS resolvers. Additionally, the funds can be also used
>>    towards the development of open source software for better and
>>    affordable DNS resolution services.
> 
> I'm all for this! There are plenty of companies both for-profit and non-profit in the EU that write DNS open source software and provide free DNS services.
> 

Good to hear you can support this.

Thank you


Desiree
—

> Cheers,
> 
> --
> Shane
> <OpenPGP_0x3732979CF967B306.asc>-- 
> 
> To unsubscribe from this mailing list, get a password reminder, or change your subscription options, please visit: https://mailman.ripe.net/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: </ripe/mail/archives/cooperation-wg/attachments/20220209/2eae5f68/attachment.html>