[ietf-dkim] 1193 considered harmful

[Dave Crocker]

Russ,

> My concern was that you only talked about the backward compatibility.
> In the future, please couple the backward compatibility issue with the 
> other dimensions discussed in the charter. 

Let me see if I understand the requirement you are describing:

1. Someone makes a suggestion for a change.

2. The group discusses its benefits and leans towards adoption.

3. Someone else notes that this breaks validation from pre-IETF signers.

(further steps occurred, but your rule appears to apply to this third step)

It appears that you are asserting that any such expression of concern about 
incompatibility requires repeating whatever the trade-offs are, nevermind that 
the benefits have already been expressed.

It further appears that anytime someone expresses a concern for compatibility 
with pre-IETF implementations, but fails to include other text that discusses 
this as a collection of trade-offs, you or a chair feel compelled to treat the 
concern as if it were a scope or process challenge.

Since I am not aware that any other expressions of benefit or detriment are 
automatically expected to carefully list a full set of trade-offs, please 
explain the assertion of this policy in this case.


  For example, a statement
> that a backward compatibility concern can be avoided if we do A instead 
> of B to meet technical objective C is completely appropriate.  I did not 
> interpret you message in this light.

Like Stephen, you appear to have read vastly more in my text than I put there. 
I never stated nor implied my own preference about the resolution of the current 
topic.  (For reference, I was in fact conflicted about it.)  Rather, I was and 
am concerned that incompatibilities be treated carefully.


> Again, the DKIM charter says:

Thanks for including that, yet again.  I hadn't read it carefully enough before.

d/

-- 

Dave Crocker
Brandenburg InternetWorking
<http://bbiw.net>