Re: [apps-discuss] [Technical Errata Reported] RFC6839 (4367)

[Larry Masinter <[e-mail]>]

The error reported should be accepted as an erratum on RFC6939 which
says:

"When JSON is written in UTF-8, JSON is 8bit compatible ([RFC2045]).
When JSON is written in UTF-16 or UTF-32, JSON is binary ([RFC2045])."


which is false. RFC 2045 defines 8bit data

   "8bit data" refers to data that is all represented as relatively
    short lines with 998 octets or less between CRLF line separation
    sequences [RFC-821]), but octets with decimal values greater than 127
    may be used.  As with "7bit data" CR and LF octets only occur as part
    of CRLF line separation sequences and no NULs are allowed.



but UTF-8 JSON may well contain sequences that include CR or LF
octets not part of CRLF sequences, and may also contain more than
998 octets between CRLF line separation sequences.

Following http://www.ietf.org/iesg/statement/errata-processing.html:

It’s a matter of judgment whether this error COULD cause implementation
or deployment problems (someone assuming they could use 8bit for arbitrary
utf8 JSON?) but it seems possible. 

The change proposed makes a reference to RFC 7159 which is unnecessary,
it could just say “Encoding Considerations: binary”.

It seems like a tautology to have a media type which in general
requires “binary” but to note “If an instance of the media type
meets the requirements for 7bit or 8bit transport then 7bit or
8bit encoding is compatible”, since you could say that for any
binary media type.

 RFC 6838 “Media Type Registration” section 4.8 says
"Possible values of this [Encoding Considerations] field are:”

and lists 7bit, 8bit, binary, framed as the possible values.

Larry
—
http://larry.masinter.net