Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!wubios!phil
From: phil@wubios.wustl.edu (J. Philip Miller)
Newsgroups: comp.mail.elm
Subject: Re: Confirm on open in elm ?
Message-ID: <1990Nov19.195912.9576@wubios.wustl.edu>
Date: 19 Nov 90 19:59:12 GMT
References: <1990Nov19.162037.7631@amd.com> <1990Nov19.182044.14572@DSI.COM>
Organization: Division of Biostatistics, WUMS, St. Louis, MO
Lines: 94

In article <1990Nov19.182044.14572@DSI.COM> syd@DSI.COM writes:
>
>Electronic mail, using most full feature MTA's, such as MMDF, sendmail
>and smail support the idea of the 'Return-Recept-To:' header  which
>will tell the sender that the letter was  delivered, and when it was
>delivered.  This is like the US Postal Services green post card.
>

One of the differences in e-mail is that there is no requirement that the MTA
on the recepient's system honor the Return-Recept-To: header.  Thus unlike the
USPS green card, if you send it out you cannot be guaranteed that it will be
honored.  I still think it is a good idea since frequently you will be using
it with a correspondant that you communiate with frequently so you will know,
up front that her/his system does or doesn't support it.


>However, their is nothing to say I ever opened your piece of mail and
>read it, perhaps it was signed for and I just threw it out.
>I could let it sit for days or weeks and then read it.
>
but you can't say you did not receive it.  It begins to localize the
responsibility for non-response.  This is why it is used for important
commercial transactions.


>With the concept of the MUA automatically telling you when and what I
>did with your mail, you have implimented a 'computer watchdog' to keep
>an eye on me and 'reduce' my privacy rights.  This is 'Big Brother'ism
>and it bothers me.  (Personal opinion)
>
Well, we utilized the facility heavily in PROFS and RICE/MAIL on IBM CMS
systems and I never heard anyone raise that issue.  Perhaps requiring you to
use a MUA that supported it, or requiring you to turn on the option might be
objectionable, but it seems to me it is a lot like running finger on a network
- it is true that there are privacy and security issues involved, but it is
sure handy to have.  At least let those users who want the facility  the
option of having it available.

>One method of reducing this intrusion is to have the MUA (elm) ask
>for permission to send a response, such as in: 'You have just saved
>a message on which the sender has requested acknowledgement of the save
>Should I send an acknowledgement of the save for you?'
>
I don't think 'save' is an important event, I leave lots of stuff in my mail
spool.

>Similar messages could be asked for read/etc.  The exact header to use
>needs to be discussed between all the MUA's and the author really should
>file an RFC about it.  However, using an X- header temporarly is not
>out of the question.
>
then lets do it!

>Technically: Its not that difficult.  What it entails is placing more
>flags on the Status: line to indicate what replies have already been sent.
>Preventing multiple replies is however a more difficult problem.  Their
>are many ways the mailbox could be left without updating the Status: lines
>thus loosing the infomation that a reply was sent.  Also if other MUA's
>rewrite the status lines using only flags they understand, then the
>flag info could be lost if the user uses a combination of MUA's.
>
In my CMS experience, when things went wrong, the end result was duplicate
acknowledgments.  I think the technical issues that get raised really relate
to the next level of complexity.  If you have this facility, then the next
thing the users will want to do is to log the return requests and the receipt
of the acknowledgments so that they can display (and possibly resend) mail
that has not been received.


>However, no one has yet to propose or file the RFC, nor has anyone
>really discussed the privacy issues besides myself and one or two
>people in personal correspondence with me.
>
no RFC is needed.  RFC 822 povides for the registration of new header types -
although it has never been done [or at least had never been when we checked it
several years ago with respect to this very issue], there is no reason why it
can't be done.


>Am I alone?
>
I hope so :-)

>Note, at best, I would make it an elmrc and/or Configurable option,
>such that it could be disabled by a user or site.
>
I find no problems whatsoever with that.

-phil
-- 
     J. Philip Miller, Professor, Division of Biostatistics, Box 8067
	 Washington University Medical School, St. Louis MO 63110
	     phil@wubios.WUstl.edu - Internet  (314) 362-3617
uunet!wuarchive!wubios!phil - UUCP (314)362-2693(FAX)  C90562JM@WUVMD - bitnet