ietf-822
[Top] [All Lists]

Re: New I-D: draft-koch-subject-tags-considered-00.txt

2004-11-29 07:34:42

On Fri November 19 2004 07:22, Charles Lindsey wrote:

Generally speaking, meaningful tags in Subject fields

By definition, there are no "tags" in the Subject field; it is an
unstructured field.

(specifically any 
which may need to be globally recognized by MUAs)

The only content that *needs* to be recognized are illegal
content (when generating) and RFC 2047 encoded-words,
and only for two specific reasons:
1. to present content to humans in human-readable form
2. to comply with specific requirements when generating
    fields

And yet software
needs to recognize at least the "Re: " form,

There is no such "need", and "Re: " has no defined
special significance in any unstructured field.  "Need"
implies some role in protocol; there is none in this case.
A role in protocol as implied by "need" further implies
that some protocol element (in a properly-designed
protocol) is unambiguous.  That is certainly not the
case with the "Re: " hack (which is why it is a "hack");
see below for some well-known examples.  There is a
defined, long-standing, standardized indication that
a message is a response to another message; that is
the presence of an In-Reply-To or References field.

for purposes of presenting 
lists of messages in a sensible order

Nonsense.  The Subject field has no inherent role in order
of presentation; sure, it can be used for collating messages
at the user's request, but "Re: " should play no specific role
in that process (unless specified by the user).  Otherwise,
Subjects such as:
  Subject: Re: the note between Do and Mi
  Subject: Re: the element with atomic number 75
  Subject: RE: Raoul Edwards strikes again!
  Subject: Re: the prefix signifying back, again, against, anew
  Subject: Re: the Egyptian sun god
etc. may be treated improperly (including by the methods
in the IMAP sorting extensions draft).  Likewise for
localized variants of "Re: ", such as "Ant: ".

The USEFOR WG
has taken the position that the only one worthy of any official
recognition at all is a single "Re: "

No, the USEFOR WG has not taken any such position. There
is a wide variety of opinion, and what is agreed is that the
"Re: " hack should not be considered part of the field syntax
or semantics (per RFC 2822 and predecessors).  Even that
straightforward issue took literally years to resolve, and there
are some (ahem) who still attempt to reverse that consensus-
based decision.