ietf
[Top] [All Lists]

Re: BCP or RFC references

2008-08-13 12:37:40
I think it would be better to use phrasing like this:

        BCP 32 (currently RFC 2606)

Tony Hansen
tony(_at_)att(_dot_)com

John C Klensin wrote:


--On Wednesday, August 13, 2008 8:13 AM -0500 Eric Gray <eric(_dot_)gray(_at_)ericsson(_dot_)com> wrote:

Isn't it a little too redundant to include the parenthetical
"RFC 2606 or its successors" along with BCP 32?

This is really a separate topic and one that it would be nice if, after all these years, the IESG, RFC Editor, and, if they care, the IAB would make a decision about and then start reflecting that decision in style guidelines (including the Checklist) and in tools.

While I'm going to use "BCP" in the examples below, the question applies equally well to "STD" numbers.

* Is a citation of BCP NN a reference to whatever the current version of the BCP, and all of the documents that make it up? If so, we need citation and referencing formats for such things that are not tied to an RFC number (or, worse, several RFC numbers). We have no such referencing model and some tools, such as xml2rfc and its bibliographic libraries, make faking one really painful.

* Is a citation of "BCP NN (RFC MMMM)" or "BCP NN [RFCMMMM]" a reference to the BCP or a reference to the RFC with a note that it is a BCP? If the latter, should the form be "RFC MMMM (BCP NN)" or perhaps "RFC MMMM (BCP NN) [RFCMMMM]"? Or should this form be prohibited?

* If RFC MMMM is a BCP, does referencing it without the BCP number mean that future revisions or updates don't count?

* If a particular specification is known much more widely by its RFC number than by its BCP one (which is certainly the case for RFC 2026), what is the approved form of citation if one wants to be clear that the BCP and not the RFC is what counts? Choices include:

   -- Use the BCP number and make people try to find out what
   is being talked about by consulting the bibliography or some
   index outside the document.

   -- Use the RFC number with some text like "or its
   successors", perhaps even "or its successors as BCP NN".

   -- Use the BCP number with the RFC number and hope that
   people figure out the BCP is intended and the RFC is
   specific.

   -- Use the BCP number with the RFC number and a note to make
   the intent clear.

I've clearly got some opinions on this, and they favor clarity over ambiguity, even if the clarity involves some possible redundancy, but YMMD. And some editorial guidance in the Checklist, in 2223bis or some other style manual, would, IMO, really be appreciated.

   john

_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf