Re: [Doh] Request for the DOH WG to adopt draft-hoffman-resolver-associated-doh

[Peter Saint-Andre <[e-mail]>]

On 1/23/19 12:25 AM, Daniel Stenberg wrote:
> On Wed, 16 Jan 2019, Paul Hoffman wrote:
> 
>> So, does this WG want to adopt this as a work item?
> 
> Disclaimer: I used to work for a browser company but I no longer do. I
> never spoke for them anyway.
> 
> For me, one of the key elements and features with DoH is that I as a
> user have picked a DNS provider I decide to trust. 

You are in a small minority. :-) Very few users will ever change their
settings.

> Be it a global CDN
> provider or my own cloud instance. Any other way, with the ISP or my
> local network admins telling me what server to use, is a major setback
> in my view.

The local network admins in your enterprise might want to have more
control, which seems OK to me because, depending on your company's
policies, they might own or have effective control over your device
(however, your user agent might want to override the suggestions it
receives).

> All forms of opportunistic DoH will make it no better than opportunistic
> DoT, which ultimately will fail to protect my privacy.
> 
> At the same time I of course appreciate that there's a problem/challenge
> here with figuring out what DoH server to use and how to even find out
> that there is one offered. If I don't know any servers, getting one or a
> few suggestions might help. For this, this spec offers ways and I'm for
> the working group adopting this.

+1 to adoption.

Peter