> >This option is likely not useful  in most common situations, and someone
> >who requests "IETF" headers ought to be aware of the implications.
>
> FYI, every time there has been an attempt to standardize mailing lists on
> the  Internet  (or  just  informal talk  about  this),  the  overwhelming
> majority of the  people involved felt that only  IETF-like headers should
> be offered, and all other types of headers should be banned - even if the
> user specifically  requests them. I would  then tell people that  I would
> never implement such a standard and  get flamed and flamed and flamed. In
> a normal Internet working group context,  I would then be ignored and the
> work would proceed and lead to  a standard banning other forms of header,
> because  a majority  of the  working group  members (who  may or  may not
> represent a majority of the actual users) felt they were very bad. But in
> the case of mailing lists, an Internet  standard would only be as good as
> the  authors of  the main  packages think  it is,  so these  "key" people
> cannot be ignored as easily and we still have non-IETF headers.
>
>   Eric
 
I subscribe to a lot of lists.  Some use LISTSERV style headers and
some use IETF style headers.  I find advantages to both.  For example,
if a message is sent to a list and cced to someone not on the list,
normal LISTSERV headers drop the cc out of the headers, so it is
difficult to cc the extra person when replying to the list.  On the
other hand, with lists that use IETF style headers, it's not uncommon
to reply direct to the originator and the list.  These extra recipients
tend to build up in the headers after several replies.