From: Bruce Lilly (blilly@erols.com)
Date: Fri May 02 2003 - 13:13:34 CDT
Pete Resnick wrote:
> Since the content of the "Subject:"
> field is unstructured (and therefore has no particular semantics), I
> don't think the discussion of "Re: " belongs in the syntax and semantics
> document.
[...]
> I could be persuaded that the "MUST NOT use anything other than 'Re : '"
> belongs in #2, but I agree that #3 is probably sufficient for both issues.
This is one of those cases where what appears in the other documents
affects the syntax. If the discussion goes much beyond observing that
the appearance of "Re: " is frequent (along with John's observation
that that says absolutely nothing about what article, if any, one with
"Re: help!" is related to), then Subject *isn't* really unstructured,
it's structure is really something like:
subject = "Subject:" SP ["Re:" SP FWS] *(utext [FWS]) CRLF ; "Re:" is case-sensitive, "Subject:" is not
In particular any appearance of MUST or MUST NOT w.r.t. what may or
may not be done with Subject field content would seem to warrant a
corresponding effect on the syntax ABNF. That is not to say that
the discussion necessarily belongs in the syntax document, but that
the discussion *affects* what needs to appear in the syntax.
If the discussion of "Re: " entails its use as a back-reference of
any sort, then that absolutely is part of the semantics (and it
means that in that case Subject *does* have particular semantics)
and is appropriate for the syntax and semantics document. If the
appearance of "Re: " near the beginning of the Subject field body
has any _meaning_ at all, then it has *semantics*, for "meaning"
is a synonym for "semantics". Conversely if "Re: " has no meaning,
then there's not much point of discussion that goes significantly
beyond affirming that lack of meaning (and no need for any MUST or
MUST NOT requirements w.r.t. "Re: " and therefore no need to put it
in the formal ABNF).