ietf-dkim
[Top] [All Lists]

Re: [ietf-dkim] draft Errata on RFC 4871

2009-01-26 20:16:28


Jim Fenton wrote:
Dave CROCKER wrote:
For a base spec to say "the value is opaque" and another spec to come along 
and 
say "I'm announcing the particular, and possibly interesting, scheme that I 
follow for creating that value, and I promise to conform to that scheme for 
all 
such values that I create" is just fine.
  

The phrase "the value is opaque" could be interpreted to mean that it
MUST NOT have any easily-discerned relationship to other things.  So any
message with an Author Signature (as defined in ADSP) would be in
violation of the base spec.  If it said "the value MAY be opaque" that's
probably reasonable.


Jim,

There is nothing in the semantics of "opaque" that says the structure can't be 
there.  It means there is no means of declaring and communicating it that is 
part of the standard and hence the presence of any possible structure and 
semantics is outside of the spec.

I don't know what "easily-discerned" means, as a technical term talking about 
an 
interoperable protocol.

What "opaque" means and what the draft Errata text says explicitly is:

    "Their sub-structures and particular semantics are not publicly
    defined and, therefore, cannot be assumed by an Identity Assessor."

Hmmm.  Perhaps that sentence should be added to the definitions of the two 
terms, rather than appear in the 'relationship' section?

Use of language like "MAY be opaque" would completely defeat the point behind 
explaining that, in terms of technical standards specification in the base 
document -- the string has no guaranteed structure or meaning.  That's the 
essence of opacity.  That some encoders might create structure and meaning and 
that some assessors might discern it is nicely heuristic.  And heuristic ain't 
standard.

Anyhow, I think that clarifying text has a meaning that is clear -- if not, 
then 
we need to fix it -- and certainly quite different from the interpretation you 
are suggesting.

Also, can you provide some technical distinction between "easily discerned" and 
"not easily discerned" with respect to technical implementation?

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net
_______________________________________________
NOTE WELL: This list operates according to 
http://mipassoc.org/dkim/ietf-list-rules.html