Date: Tue, 25 Nov 1997 09:11:25 +0000 (GMT)
From: Philip Hazel <ph10@cus.cam.ac.uk>
To: Bart Schaefer <schaefer@brasslantern.com>
cc: Jacob Palme <jpalme@dsv.su.se>, drums@cs.utk.edu,
        mailnews-l@SEGATE.SUNET.SE
Subject: Re: IETF draft on the "Mail-Followup-To" header
In-Reply-To: <971124111146.ZM29118@candle.brasslantern.com>
Message-ID: <Pine.SOL.3.96.971125085417.28644H-100000@taurus.cus.cam.ac.uk>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

On Mon, 24 Nov 1997, Bart Schaefer wrote:

> Now, I realize that colloquial use typically calls the collection "the
> headers" and each name-value pair a "header".  I've used it that way
> myself for years.  But if we're going to write understandable specs, we
> need to avoid this kind of confusion.  Let's pick one set of terminolgy
> and stick to it.

That was approach taken by Algol 68. Nice language. Shame about the 
terminology. RIP.

IMHO, it's far too late to try to stop people talking and writing about
things like "the 'From' header". Why not go with the flow? Otherwise you
are just creating another confusing word like "domain".

Also, from RFC 822:

     The  syntax of several fields of the rigidly-formated
     ("headers") section is defined in  this  specification; 
             ^
             ^
and though it goes on to say
                                                                      
          The syntax  that  distinguishes  between  header  fields  is
     specified  separately  from  the  internal  syntax for particular
     fields.  This separation is intended to allow simple  parsers  to
     operate on the general structure of messages, without concern for
     the detailed structure of individual header fields.
     
it is quite easy to interpret "individual header fields" as "fields of 
individual headers", i.e. "fields of individual header lines", and I 
suspect this is how many people understand it. I certainly do. Later in 
RFC 822 we have

     3.1.  GENERAL DESCRIPTION                                        
                                                                      
          A message consists of header fields and, optionally, a body.
     The  body  is simply a sequence of lines containing ASCII charac-
     ters.  It is separated from the headers by a null line  (i.e.,  a
     line with nothing preceding the CRLF).     

The fact that it talks about "the headers" leads one naturally to
consider a single one of them as "a header". There are also things like

        Certain field-bodies of headers may be  interpreted 

which should have been "Certain field-bodies of the header" under the
proposed usage. 

I think it is far too late to try to change popular usage. I think it is
well enough known that attempts to control popular usage of natural
language in general usually fail, and I don't see why this case should 
be different. How about

     A message consists of a sequence of "headers" and, optionally, a 
     body. The body is simply a sequence of lines ..... Each "header" 
     consists of ....  The headers are separated from the body by a null
     line.

Philip
   
-- 
Philip Hazel                   University Computing Service,
ph10@cus.cam.ac.uk             New Museums Site, Cambridge CB2 3QG,
P.Hazel@ucs.cam.ac.uk          England.  Phone: +44 1223 334714