RFC 2231, "MIME Parameter Value and Encoded Word Extensions: Character Sets, Languages, and Continuations", November 1997

Errata ID: 3269
Status: Held for Document Update
Type: Technical
Publication Format(s) : TEXT
Reported By: Chris Newman
Date Reported: 2012-06-28
Held for Document Update by: Barry Leiba

Section 3 says:

    (2)   the mechanism MUST NOT depend on parameter ordering
          since MIME states that parameters are not order
          sensitive.  Note that while MIME does prohibit
          modification of MIME headers during transport, it is
          still possible that parameters will be reordered when
          user agent level processing is done.

It should say:

    (2)   the mechanism MUST NOT depend on parameter ordering
          since MIME states that parameters are not order
          sensitive.  Note that while MIME does prohibit
          modification of MIME headers during transport, it is
          still possible that parameters will be reordered when
          user agent level processing is done.
     (3) the mechanism MUST NOT alter parameter values that are
          critical to existing MIME processors. This specifically includes
          the "boundary" parameter for multipart types and the "charset"
          parameter for text types.
        

Notes:

Earlier text in the section states "Any such mechanism MUST be compatible with existing MIME processors." The addition of a 3rd item clarifies an additional behavior that is necessary to achieve that requirement. It is a flaw in the RFC 2231 standard if it creates a message format that can not be processed by an RFC 2045 processor but can be processed by an RFC-2045-as-extended-by-2231 processor. I have seen a 2231 split boundary marker in the wild so this elaboration on the existing MUST appears to be necessary.

Note that I do not object if this errata is marked "hold for document update".

Report New Errata