Re: [Add] A proposed charter for ABCD

[Mirja Kuehlewind <[e-mail]>]

Hi Glen,

See a few small comments inline 

> On 9. Jan 2020, at 23:56, Deen, Glenn (NBCUniversal) <[e-mail]> wrote:
> 
[snip]
>  
> [GD] As I said in my other notes – Tommy’s version is very similar in its goals which is a very good sign.  The main difference is how much and what you front load the WG with constraints on where they start and how they have to proceed.
>  
> The areas that I would change would be to remove a few things:
>  
> 	• I would remove the following since I think the getting into informing clients of what is appropriate is both a can of worms, and gets into the realm of telling the clients what to choose versus the clients pulling the information they want and then making their choice based on whatever logic they have in the client.
>  
>     - inform clients about how to select appropriate encrypted and unencrypted resolvers for DNS queries in various network environments and use cases.

So this is the only part in the proposed charter that actually talk about selection. If you want to work on discovery AND selection the charter should say that. Maybe this point as currently phrased a bit too broad and we can be more concrete here? 

I would also be okay to say for now that the group will work on discovery mechanism first and advance selection mechanisms need rechartering. However, as I said in a previous mail, I don’t think we can design the discovery mechanisms correctly if we don’t have at least a few simple selection policies in mind that could be used.

>  
> 2.    Remove (below) the Adaptive DNS draft starting point because as it currently is Adaptive DNS goes beyond D&S and gets into interesting areas of resolver behavior including the MUST requirement that DoH servers support ubiquitous DNS.   That’s more than D&S as I say is interesting but it’s far more than the WG should try to take on if it wants to be able to reasonably deliver output.   BTW – I’m not saying don’t discuss such things. I am simply saying such things would better be done in another WG.
>  
>              The starting point for the group's work is Adaptive DNS (draft-pauly-dprive-adaptive-dns-privacy), subject tothe working group's design choices.

See my other mail I’ve just sent.

>  
> 	• I would also do an edit to make sure it is not specific to DoH and DoT and make it about any DNS. The current rev. mentions unencrypted DNS in some places and focuses on DoH in others. Both Encrypted DNS and even traditional DNS/53 since if the group were to produce a better way of discovering and selecting DNS resolvers there isn’t any reason to limit its use to encrypted DNS. Network environments like enterprises that may choose to use DNS/53 could benefit from it.    
>  
> 	• I continue to advocate for changing the name to not be ADD. Too much baggage comes with it.  I continue to propose the simpler DNS Discovery and Selection or DNS-DS

DNS-DS might be confusing with RFC6763 (DNS-Based Service Discovery).

Mirja