[ietf-dkim] "interoperability"???

[Jim Fenton]

Hinestly, I struggled a bit with whether to use the term 
interoperability there.

The point I'm trying to make is that the normative text should be 
sufficient for different implementations to return the same result under 
the same test conditions.  If we were to have an ADSP "bake-off", one of 
the test conditions might be to report the result when given a 
non-existent domain.

I think that this is a valid test case, and that implementations should 
produce the same result.

-Jim

Dave Crocker wrote:
>
>
> Jim Fenton wrote:
>> Dave Crocker wrote:
>>> Having non-normative text that describes it serves to promote the 
>>> idea but not couple it with the fate of ADSP.
>>
>> Having the ADSP result depend on non-normative language in this case 
>> does not meet the bar of interoperability that we need to achieve.  
>> Making it non-normative means that two spec-compliant implementations 
>> of ADSP would return completely different results for non-existent 
>> domains.
>
>
> Sorry, but I can't let this one go by without asking:
>
>
> I completely do not understand the claim of non-"interoperability" here.
>
> Since the record(s) in question are not created with ADSP in mind, 
> then the domain owner cannot be said to be participating in ADSP, with 
> respect to this check.
>
> SO how is inter-operation hurt or hindered by this specification's 
> making the check normative vs. non-normative?
>
> d/
>
>