From: Bruce Lilly (blilly@erols.com)
Date: Fri Jun 18 2004 - 08:13:49 CDT
Seth Breidbart wrote:
> Eivind Tagseth <eivindt@multinet.no> wrote:
>
>
>>Yes, these are still not covered satisfactory. I still think that
>>if we're going to use the term «back-reference», we need a clear
>>definition of the term.
>
>
> How about something like:
>
> A back-reference is something added to the Subject header content to
> refer to (the Subject of) another message. The standard
> back-reference is "Re: ". Other back references in use may include
> "Sv: " (deprecated), "Fwd: " (deprecated), or
> "$NewSubject [was: " . . . "]" (when added by the user). Those that
> are added automatically by software, other than "Re: ", are referred
> to as "non-standard bac- references".
Ignoring typos, it still suffers from the defect that given an article, it
is not possible to identify any part of such an article as a so-called
"back-reference", since the definition incorporates intent ("to refer to"),
which cannot be determined other than by the hypothetical ESP-based reader.
Given the fact that the Subject field is unstructured, it is not possible
to produce an objective definition of back-reference (i.e. one that does
not involve intent), so following up on the converse of Eivind's statement,
since we cannot produce a clear definition that is not misleading, we
should not use the term.
The term does not appear anywhere in RFCs 561, 724, 733, 822, 850, 1036,
2822 (etc.; indeed it does not appear in *any* of the on-line RFCs) or
in GNKSA, so it is clearly not necessary.