Bob Braden wrote:
Julian,
Hi. I believe that the problem your pointing out is real, common, and
not so easy to solve.
First, you might regard the given definition of Updates as a
"best-effort" definition, with all that "best-effort" implies. ;-) I
am sure that serious suggestions on a more precise definition would be
welcome, but it is unclear how much effort one ought to lavish on
word-smithing here. The word "updates" in its common usage seems to
carry the desired meaning.
The RFC Editor and the IESG frequently struggle with the semantics
of the relationships among related RFCs. It is clear that the categories
"Obsoletes" and "Updates" are only the high-order bits of often complex
relationships. One could add "Partially Updates" and "Partially
Obsoletes" I suppose, but it is not clear where to stop or whether this
will really meet the need. We have recently been advocating the
creation of a series of area-specific "road-map" documents to
summarize these relationships; this would get around the limited
number of bits in the present categories.
Finally, we note that there is an alternative way to do business: the
IETF could shape its documents to fit the categories. That is
essentially what more formal standards bodies do. The IETF, for good
reason, chooses instead to let documents develop in the most expedient
fashion, and then notate them with (approximate) notations like
Obsoletes and Updates.
Bob Braden
Bob,
thanks for the clarification. I think that "updates" and "obsoletes" is
really sufficient (if there's more to say the new spec should clearly
state that in the introduction anyway). The only issue seems to be that
RFC2223bis is very restrictive in what "updates" is supposed to mean,
both not matching what I happen to need :-), and what historically is't
been used for. Looking at the original text in RFC2223:
Updates
To be used as a reference from a new item that cannot be used
alone (i.e., one that supplements a previous document), to refer
to the previous document. The newer publication is a part that
will supplement or be added on to the existing document; e.g., an
addendum, or separate, extra information that is to be added to
the original document.
This seems to be less restrictive and explicitly allows for instance
what RFC2396 does (updating the definitions of URI syntax without
updating the descriptions of particular URI schemes).
Regards, Julian
--
<green/>bytes GmbH -- http://www.greenbytes.de -- tel:+492512807760