Re: POP3 behavior on broken tcp connection

Steve Dorner <sdorner@qualcomm.com> Fri, 03 June 1994 10:16 UTC

Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa00966; 3 Jun 94 6:16 EDT
Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa00962; 3 Jun 94 6:16 EDT
Received: from ANDREW.CMU.EDU by CNRI.Reston.VA.US id aa02613; 3 Jun 94 6:16 EDT
Received: (from postman@localhost) by andrew.cmu.edu (8.6.7/8.6.6) id GAA00611; Fri, 3 Jun 1994 06:13:54 -0400
Received: via switchmail for ietf-pop3+@andrew.cmu.edu; Fri, 3 Jun 1994 06:13:54 -0400 (EDT)
Received: from po3.andrew.cmu.edu via qmail ID </afs/andrew.cmu.edu/service/mailqs/q001/QF.chvk90600UdbE1uE5W>; Fri, 3 Jun 1994 06:13:20 -0400 (EDT)
Received: from ux1.cso.uiuc.edu (ux1.cso.uiuc.edu [128.174.5.59]) by po3.andrew.cmu.edu (8.6.7/8.6.6) with SMTP id GAA00479 for <ietf-pop3+@andrew.cmu.edu>; Fri, 3 Jun 1994 06:13:15 -0400
Received: from dorner.slip.uiuc.edu by ux1.cso.uiuc.edu with SMTP id AA00262 (5.67b8/IDA-1.5 for <ietf-pop3+@andrew.cmu.edu>); Fri, 3 Jun 1994 05:12:53 -0500
Received: from [192.17.5.3] by dorner.slip.uiuc.edu with SMTP id AA05327 (5.67b/IDA-1.5 for <ietf-pop3+@andrew.cmu.edu>); Fri, 3 Jun 1994 00:07:26 -0500
X-Sender: sdorner@192.17.5.1
Message-Id: <aa14694107021016e71a@[192.17.5.3]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Fri, 03 Jun 1994 00:07:34 -0500
To: POP3 IETF Mailing List <ietf-pop3+@andrew.cmu.edu>
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: Steve Dorner <sdorner@qualcomm.com>
Subject: Re: POP3 behavior on broken tcp connection

>no, No, NO!  If the TCP connection breaks, the server MUST NOT delete the
>messages.  Anyone ever hear of robustness?  Jeesh.

Ever hear of bad phone lines?  Jeesh.  :-)  :-)

Users are not pleased to discover that, because a phone line crapped out,
the deletions they've made are now unmade.  This is absolute DEATH for
(what I consider to be) the canonical log-in/grab-mail/delete-mail POP3
client, if it's on a flaky link.

I get this "robustness" reported to me as a bug on a very regular basis.
(Though not from users who have installed our version of popper >:-}.)

You have to decide which failure mode is more important; the broken
communications medium or the broken client hardware/software.  My
experience has been that the former is very much more important.

--
Steve Dorner, Qualcomm Incorporated
 "There's nothing wrong with you that can't be cured
  with a little Prozac and a polo mallet." - Woody Allen