[rfc-i] ABNF (RFC2234)

Bruce Lilly blilly at erols.com
Thu Apr 7 13:28:28 PDT 2005

>  Date: 2005-04-07 07:23
>  From: "Hollenbeck, Scott" <shollenbeck at verisign.com>
>  To: rfc-interest at rfc-editor.org

> We (the IESG and the
> authors) had to consider each point individually to confirm agreement
> with the alleged issues, and in the end there was disagreement that led
> to (still unresolved) questions of how they were accepted in the first
> place.  Suffice it to say that they were NOT blindly incorporated into
> the update.

Without attempting to address the issue of fixing the blame, and
returning to the issue of fixing the erratum problem, I note:

1. The RFC Editor errata page states "If there is credible evidence
   that an error claim is incorrect or misleading, the posting will be
   removed from this list."

2. after it also advised "please verify the errors with the authors
   and/or appropriate area directors of the IESG before sending them to
   the RFC Editor for posting"

3. Past correction of an erroneous erratum (for RFC 2047) was handled
   by a second published erratum which corrected the earlier one.

I believe that the earlier discussion here coupled with the results of
the IESG/authors' deliberations constitutes "credible evidence" that
parts of one of the published errata is incorrect.

So it seems that the problem could be resolved by:

a. removing the offending parts of the offending erratum, which
   possibly simply requires an author or AD to direct the RFC Editor
   to do so in clear terms


b. removal of the entire offending erratum, leaving genuine editorial
   errors in 2234 unaddressed unless an erratum addressing only those
   errors is added


c. a new erratum that corrects the erroneous erratum

As a hasn't yet happened AFAICT, here's a stab at c:
In the published erratum to RFC 2234 dated 19 Nov 2001:

    prose-val      =  "<" *(%x20-3D / %x3F-7E / c-wsp) ">"
                           ; bracketed, possibly multiline string
                           ; of SP and VCHAR without angles
                           ; prose description, to be used as
                           ; last resort

should be:

    prose-val      =  "<" *(%x20-3D / %x3F-7E) ">"
                           ; bracketed string of SP and VCHAR
                           ; without angles
                           ; prose description, to be used as
                           ; last resort

Rationale:  the published erratum made incorrect changes to the grammar
and its description
Now, if one of the 2234 authors or a cognizant AD would send that to
rfc-editor at rfc-editor.org with a note of approval, the problem can be

More information about the rfc-interest mailing list