[Ietf-not43] transaction settlement issue stated in Appendix B.

Leslie Daigle leslie@thinkingcat.com
Thu, 20 Feb 2003 18:12:14 -0500


Andrew Newton wrote:
> Leslie Daigle wrote:
> 
>>
>>  A service MUST accept and process the query with a bag;   refuse the 
>> query with an error message if it does  not recognize the data in the 
>> bag;  or refuse the query
>>  with a different error message if the data in the bag
>>  is of the proper format but is otherwise unacceptable
>>  (e.g., out of date tokens, or general refusal to handle
>>  referred queries at this time).
> 
> 
> I'm not sure this is needed, especially given that service requirements
> are not longer MUST/SHOULD issues.

I respect the WG's consensus that it shouldn't
be defining policy for the service (by making requirements
on server behaviour), and my concern is just that we
shouldn't specify a protocol and then imply that the
spec needn't be followed (in terms of the semantics).

I think that concern would be addressed by including
a sentence or two, where you distinguish between service
description and protocol requirements, saying that <while
individual servers/service operators may not support all
features that the protocol can support, they must
respect the semantics of the protocol queries and
responses>.  Except, written in English, not this tortuous
phrasing :-)

In other words, you shouldn't return referrals if you
don't mean it, etc.

Leslie.

-- 

-------------------------------------------------------------------
"Reality:
     Yours to discover."
                                -- ThinkingCat
Leslie Daigle
leslie@thinkingcat.com
-------------------------------------------------------------------