[rfc-i] Order of numbered and un-numbered sections in RFCs

Martin Duerst duerst at w3.org
Sat Sep 4 01:20:51 PDT 2004

Hello Paul,

Some comments below.

At 08:42 04/09/03 -0700, Paul Hoffman / VPNC wrote:
>At 9:31 AM +0300 9/3/04, Pekka Savola wrote:

>The reason I say this is because sometimes folks put e.g.,
>>Acknowledgements or Security Considerations in appendices (after the
>(From draft-rfc-editor-rfc2223bis-08.txt:)
>    A published RFC may contain the sections in the following list.  Some
>    of these sections are required, as noted.  The order shown is
>    required, except that the order shown for the sub-items 7a-7f within
>    Body of Memo is generally recommended but not required.
>       1.  First-page header           [Required]
>       2.  Status of this Memo         [Required*]
>       3.  Copyright Notice            [Required*]
>       4.  IESG Note                   [As requested by IESG*]
>       5.  Abstract                    [Required]
>       6.  Table of Contents           [Required for large documents]
>       7.  Body of the Memo            [Required]
>        7a.  Contributors
>        7b.  Acknowledgments
>        7c.  Security Considerations   [Required]
>        7d.  IANA Considerations
>        7e.  Appendixes
>        7f.  References
>       8. Author's Address             [Required]
>       9. IPR Boilerplate              [Required*]

I'd avoid numbers here, because this may otherwise be confused
with the numbering of the sections.
Also, 7a should be something like 'real body' or 'actual spec'
or whatever. Also, my understanding is that since very recently,
a IANA section is required.

>>rfc-editor-rfc2223bis-08.txt includes a pretty good order, though by
>>default I'd change the order or references and appendices.
>I have seen documents with the various parts of the "body" in different 
>positions. Should we make hard-and-fast rules for future RFCs, or leave 
>the body order floating?

I use to think about security considerations and IANA considerations
as part of the actual spec, which means that in the above point
7, they should come first. Having references last is helpful because
that's the point one goes to most often during reading the rest
of the doc, but the boilerplate at the end makes this less easy
than if they were really at the end. If there is an index, it
should usually come after the references.

It would also be good to have contributors, acknowledgements,
and author's addresses close together, because they are all related.
But then we already have conflicting constraints :-(.
(it should be "Author's Address or Authors' Addresses", btw).

Regards,    Martin.

More information about the rfc-interest mailing list