From: Bruce Lilly (blilly@erols.com)
Date: Sun Feb 16 2003 - 09:29:20 CST
Charles Lindsey wrote:
>>>4.7.3.6. RFC 987 backwards compatibility
>>>
>>> The mapping defined here is different to that used in RFC 987, as the
>>> RFC 987 mapping lead to changed message IDs in many cases. Fixing
>>> the problems is preferable to retaining backwards compatibility. An
>>> implementation of this standard may recognise message IDs generated
>>> by RFC 987. This is not recommended.
>>>
>>>Which is also an example of where "backwards compatibility" (in my sense
>>>of the term)
>
>
>>No, "retaining backwards compatbility" in this instance means
>>that the installed infrastructure would continue to be
>>considered compliant with the new standard, which is exactly
>>the same meaning of "backwards compaibility" used in other
>>contexts.
>
>
> No, he is saying that if an implementation of RFC 987 produces a message
> ID in the now deprecated style, it may not be recognized by a strict
> implementation of RFC 2156. That is a breech of backwards compatibility in
> the sense in which I use the term.
Whether or not that is the case, it is clear that it is a
breach of backwards caompatibility in the same sense as that
term is used in the other instance in the same document, and
in other RFCs, and in the Usefor WG charter, and not in the
way you now wish to redefine it.