Re: QUIC re-chartering: include DNS-over-QUIC?

[Eric Rescorla <[e-mail]>]

On Tue, Feb 4, 2020 at 1:05 PM Erik Nygren <[e-mail]> wrote:

> On Wed, Dec 11, 2019 at 4:38 PM Mark Nottingham <[e-mail]> wrote:
>
>> We've just put out Calls for Adoption for extensions to QUICv1, as we
>> believe that the group has some capacity to discuss them as it finishes
>> work on the core protocol.
>>
>
> Is there interest and bandwidth in picking up work on DNS-over-QUIC (eg,
> draft-huitema-quic-dnsoquic-07
> <https://tools.ietf.org/html/draft-huitema-quic-dnsoquic-07>) as well,
> presumably in coordination with the DNS community?
>

I think it would be OK to define DoQ (though I share Tommy's question about
DoH/3, which needs no additional standardization) but I don't think that
should be done in the QUIC WG, any more than we define HTTP in TCPM. We
needed one driving study but now that the QUIC mapping is done, other
mappings should probably be done elsewhere.

>
> Getting a second protocol using QUIC might help make sure we work through
> issues that may arise early, plus DNS-over-QUIC seems quite attractive as a
> technology for resolver-to-authoritative communication if/when we go that
> way.  ie, it seems strictly better than defining a DNS-over-DTLS
>

You might be interested in https://tools.ietf.org/html/rfc8094

-Ekr



> and also seems to have plenty of advantages over DoT.
>
> The current charter also says:
>
> This [HTTP] mapping will accommodate the extension mechanisms defined in
>> the HTTP/2
>> specification. Upon completion of that mapping, additional protocols
>> may be added by updating this charter to include them.
>>
>
> Best,
>      Erik
>
>