Re: [Add] ADD State of Things Observations

[Andrew Campling <[e-mail]>]

Glenn
The approach you’ve suggested seems like a reasonable way forward, especially if the split is along the lines suggested by Chris.  Ideally we’d first flesh out the user requirements a little more and then have the WG adopt the document so that there is an agreed record of the use cases and associated requirements that we’re aiming to solve.

Andrew

From: Deen, Glenn <[e-mail]>
Sent: 15 October 2020 16:16
To: ADD Mailing list <[e-mail]>
Subject: Re: [Add] ADD State of Things Observations

Chris,

Thanks for the comment.

I meant the proposed complexity grouping as starting points, but if it’s possible to merge RFC1918/CPE with a less complex case that’s very nice.  There do remain other complex environments that won’t be as easy.   Ultimately such lines are drawn are drawn by  authors and the WG group so we choose to draw as appropriate.

Regardless of where the lines are drawn, for now do people feel the general observations around complexity and path proposed to be reasonable?

Glenn

________________________________
From: Chris Box (BT) <[e-mail]<mailto:[e-mail]>>
Sent: Thursday, October 15, 2020 2:12 AM
To: Deen, Glenn
Cc: ADD Mailing list
Subject: Re: [Add] ADD State of Things Observations

Glenn,

I'm happy with the general principle of splitting into 2 or 3 areas and working on those in parallel.

But I'm not entirely sure I agree with where you've drawn the distinction.

(1) Low-complexity environments.  – this would include the case that started the “My single use case” thread

(2) High-complexity environments – this would include the RFC1918 situations,  networks with more advanced technical controls, networks/devices with applied policy controls.

I would see RFC1918-addressed forwarders as very much in the scope of "My single use case".

In fact as Martin said:
This might need the full matrix of DoT/DoH, v4/v6, with/without a forwarder, but this is fundamentally just a single use case.

As others have said, such non-upgradeable forwarders are so common that any "tell me how to contact your encrypted version" protocol will need to deal with them.

Likewise, a consequence of selecting the network's recommended encrypted resolver is that network-applied policy controls may be in scope. So they are not solely found in "high-complexity environments".

But I do agree that it is useful to separate out such more complex items as Enterprise, and the provision of useful information about each possible resolver, such that the client can make a more informed decision if it wishes to.

Chris