[DNSOP] Re: Call for Adoption for draft-sheth-dns-integration

[Ben Schwartz <[e-mail]>]

I think a document like this could be valuable.  Finding the right place for it is harder.  As the document notes, the subject matter is mostly about the DNS namespace, not the DNS protocol.  I could see it in DNSOP, ARTAREA, ISE, or even a non-IETF venue like ICANN.

I do think the subject matter will need some significant adjustments.  The introductory text is too long and repetitive.  Section 3 seems like a sales pitch for DNS, to an audience who (I imagine) surely already knows whether they want to use domain names.  The essential content of the document in Section 4 is unusually vague.  It offers only a sort of "design guidelines for design guidelines".  There are also some factual/detail issues, like the lack of distinction between "DNS names containing non-ASCII bytes" and IDNs, and scaremongering about resolver compatibility with new record types.

I think this document would be a more valuable, and more appropriate for DNSOP, if it were more opinionated.  Instead of merely noting the existence of various problems, it should provide a decision tree for what to do about them.  For example, it could say

"DNS labels are permitted to carry arbitrary octets.  However, to reduce complexity and avoid certain security risks, application integrations SHOULD restrict identifying names to those containing characters from the Preferred Name Syntax.  If the DNS names will be displayed prominently to ordinary users, applications SHOULD apply IDN processing, with appropriate defenses against homograph attacks."

Similarly for change of control (is that the same as re-registration?), handling transient resolution failures, etc.

--Ben


________________________________
From: Joe Abley <[e-mail]>
Sent: Wednesday, May 28, 2025 9:18 AM
To: Ondřej Surý <[e-mail]>
Cc: dnsop <[e-mail]>
Subject: [DNSOP] Re: Call for Adoption for draft-sheth-dns-integration

Hi Ondrej,

On 28 May 2025, at 14:37, Ondřej Surý <[e-mail]> wrote:

> The draft is available here: https://urldefense.com/v3/__https://datatracker.ietf.org/doc/draft-sheth-dns-integration/__;!!Bt8RZUm9aw!9o45UezG22LAOlZaRp-0pSAUn3Ol8FnEbCmp8Ba4yZnnuFvBJwwM8vouarJRvB1B20erA9aLpIqnY0I$
>
> Please review this draft to see if you think it is suitable for adoption
> by DNSOP, and send any comments to the list, clearly stating your view.
>
> Please also indicate if you are willing to contribute text, review, etc.

This is the first time I have looked at this document, which surely means I have not been paying attention, and which I mention just to give context to the following.

I think this document's audience is not what is described in section 1.1. I think the intended audience in fact is application and service designers who are thinking about incorporating features of the namespace used by the DNS into what they are building.

I wonder whether dnsop is the best place for that audience to provide input and review. Maybe there's somewhere in the application area that would be a better venue. This working group (and other groups like the DNS directorate) could still provide meaningful review of a document that is worked on elsewhere.

This is not particularly an objection to adoption, just a question of whether this is the best place for it.


Joe
_______________________________________________
DNSOP mailing list -- [e-mail]
To unsubscribe send an email to [e-mail]