Re: [dnsext] Design team report on dnssec-bis-updates and CD bit

[Andrew Sullivan <[e-mail]>]

On 2010-07-10, at 17:33, [e-mail] wrote:

> On Sat, Jul 10, 2010 at 11:05:19AM -0400, Olafur Gudmundsson wrote:
>> On 09/07/2010 6:04 PM, [e-mail] wrote:
>>> On Fri, Jul 09, 2010 at 10:21:08AM -0400, Andrew Sullivan wrote:
>>>> Dear colleagues,
>>>> 
>>>> On 2010-07-06, there was a design team meeting in Bethesda, MD, to
>>>> address the issue of setting the CD bit in the dnssec-bis-updates
>>>> draft.  David Blacka, Olafur Gudmundsson, Mike StJohns, and Andrew
>>>> Sullivan attended.
>>>> 
>>>> In the meeting, we concluded by describing three different models for
>>>> how the CD bit is to be treated and used by a validating resolver
>>>> (i.e. one that has at least one trust anchor), and what interaction
>>>> this has with a local cache.  We based these models on RFC 4035, on
>>>> the discussion we saw on the list on this topic, and on what we think
>>>> people have implemented.  We want one, and only one, of these models
>>>> to be described in the draft as the standard model.  These models are
>>>> outlined below.
>>> 
>>> 
>>>    so the We in the para above is the design team of
>>>    yourself, David, Olafur, and Mike.  This does not meet
>>>    the WG criteria of "five" reviewers.
>> 
>> This is orthogonal, design team can be of any size, the output of an
>> design team can be one of:
>>    a proposal of way forward
>>    a list of options
>>    a comment on a particular design.
>>    a statement of failure to reach conclusion
> 
>    hum... this kind of makes the IETF WG process subject to capture,
>    don't you think?  According to you, a design team of one can make
>    a statement about failure to reach conclusion...  Thats a pretty 
>    low bar.

To be clear, nobody is suggesting this is a WG product. If it were, we could rule on consensus rather than discussing. But we brought this input to the WG as our considered opinion. 

>    except Andrew, as spokesperson for the design team cast a 
>    bias - not just the presentation of options with a request
>    to the WG to come to consensus.  

The report says that we want one option, that's true. I personally prefer that the WG coalesce around one because I think a completely indeterminate behavior on the part of intermediate resolvers would be very bad for user experience. I'd be interested in counter arguments. 

> To compound influence, there
>    was an inferance that the WG chairs would close this topic

Actually, I don't think an inference was needed. And I didn't imply anything. I stated that if the WG couldn't reach consensus, we could leave this topic out of the draft on the grounds that we have no consensus. That's not a threat; it's the chair's job to say such things. As it transpires, I believe now that the tables are a significant enough improvement over the existing 4035 text that we should include them all if need be. But I see no reason why we cannot at least settle on a predictable default.

A

-- 
Andrew Sullivan
<[e-mail]>