Mail behaviour that I hate
There are many things in E-mail that irritate me.
Some of them are:
- People losing their temper
- People who quote the whole 200-line message that they respond to, and
then add "I agree" at the bottom
- People who see a spam on a mailing list, and do "reply all",
sending to all the 20 other mailing lists in the To: line
- Mailers who bounce mail seemingly at random, sometimes letting it through,
sometimes not.
But people won't listen to me, and punishing them by hanging out their
names in public like this isn't going to help much. So instead, I've found
two objects of really irritating behaviour that I've chosen as my particular
windmills to tilt at.
Mailing lists that do not change MAIL FROM
RFC 1123 (Host Requirements) says:
5.3.6 Mailing Lists and Aliases
An SMTP-capable host SHOULD support both the alias and the list
form of address expansion for multiple delivery. When a
message is delivered or forwarded to each address of an
expanded list form, the return address in the envelope
("MAIL FROM:") MUST be changed to be the address of a person
who administers the list, but the message header MUST be left
unchanged; in particular, the "From" field of the message is
unaffected.
Some mailing lists are run like giant aliases, NOT changing the MAIL
FROM address, with the result that bounce messages are returned to the
original sender, and not the mailing list administrator.
Some lists are so big and so poorly maintained that whenever I send
a message to them, I get 2 or 3 bounce messages back.
I hate that to the degree that I won't use such lists.
MTAs that bounce to the From address
RFC 1123 says:
5.3.3 Reliable Mail Receipt
........
If there is a delivery failure after acceptance of a message,
the receiver-SMTP MUST formulate and mail a notification
message. This notification MUST be sent using a null ("<>")
reverse path in the envelope; see Section 3.6 of RFC-821. The
recipient of this notification SHOULD be the address from the
envelope return path (or the Return-Path: line). However, if
this address is null ("<>"), the receiver-SMTP MUST NOT send a
notification. If the address is an explicit source route, it
SHOULD be stripped down to its final hop.
Some mailers send to the From: or Sender: line from the headers. Some
of them are on crippled systems that don't have any place to put the Return-Path:
or similar info; others are simply broken.
A variation is mailers that send both bounce and reply messages to the
envelope "from"; these are "beloved" by listadmins,
but don't worry the peons on the list too much.
The Wall Of Shame
When someone breaks these rules, I've often sent them a message. From
now on, I will start hanging out the names of the relevant MTAs and mailing
lists here, for all to see and be warned.
Of course, my hope is that people will take action to correct the problem;
I'll be all too happy to remove MTAs and lists from this list!
Mailing lists that don't change MAIL FROM
- smime-dev@RSA.COM
- tr29-list@taux01.nsc.com
- mfpa@mfpa.com (hosted at crash.cts.com)
- cf@omg.org and/or docman@omg.org
MTAs that bounce to Sender: or From:
The mailers that are structurally unable to do this right include:
- Microsoft Mail at:
- iss-rt.rt.cs.boeing.com (just to scare me, it even includes all recipients
of the message in the CC: field of the bounce!)
Microsoft agrees that this product is broken, and recommends that sites
upgrades to Microsoft Exchange.
- Microsoft Exchange (early versions) at:
- mirabeau.saclay.cea.fr (Microsoft Exchange Server Internet Mail Connector
Version 4.1.611) (which also gives error messages in French only!)
Also, my warning to postmaster timed out. Not a quality site!
- cocms1.calwdc.shl.com (Microsoft Exchange Server Internet Mail Connector
Version 4.0.837.3)
NOTE: Microsoft says that this is a VERY old beta release, and that
newer releases do this right!
- CC:mail at:
- ccmail.crosscomm.com (1.30/SMTPLink)
(somehow it managed to quote the entire message INCLUDING the right address
to send bounces to, but send the bounce to the wrong one....)
- smtpgate.ccmail.com (SMTPLINK V2.11.01)
- smtplink.ccmail.com (ccMail Link to SMTP R6.0)
- smtplink.sra.com (SMTPLINK V2.11 PreRelease 4)
- ccgate.hac.com (no version; this site also UUEncodes the original message,
just to make for more fun!)
- mfi.com (SMTPLINK V2.11) (also includes the original message WITHOUT
headers, making for some confusion
- battelle.org (no version, but includes the message text without headers,
just like the one above; also, the site does not support Postmaster)
- ccmail.airtouch.com (SMTPLINK V2.11 PreRelease 4)
- CSERVE2.CCMAIL.CompuServe.COM, possibly the same as hil-img-2.compuserve.com
(no version info)
- smtpgate.kofax.com (ccMail Link to SMTP R6.0)
- usa.racal.com (SMI-8.6/SMI-SVR4)
- cm.inc.hitachi.co.jp (cc:Mail SMTPLINK 2.1)
- Lotus Notes Mail Gateway for SMTP at:
- harlequin.co.uk
- internet1.lotus.com (V.02 Beta)
- notes.ub.com (V1.1)
(this one gets extra black marks for not supporting Postmaster)
- forrester1.forrester.com (also does not support Postmaster)
According to Dave Newbold <dnewbold@iris.com>, Lotus has a patch
called "patch 30" that is supposed to fix this problem; I haven't
seen a bounce from his system after that message, so it might have been
a good fix.
Other mailers that seem to do this include:
- EMDUPM12.BITNET
- BayNetworks.com (also, the info in their non-delivery reports is close
to zero)
- mcimail.com
- ilbx.mot.com (apparently a WorldTalk 3.0.4/1.64 X.400 gateway)
- moau.mo.br.np.els-gms.att.net (attmail.com?)
- softsw.ssw.com (Soft*Switch Central V4L40P1A) - their bounces are also
singularly unhelpful in locating the real problem.
- euro.kommers.se - sends to REPLY-TO, which almost guarantees loops.
Also a singularly unhelpful bounce format.
- cix.compulink.co.uk
- nortel.ca - also includes few details - just the address and the subject
- and assumes that you're within their system.
- compuserve.com - according to 'Net lore, they have decided as a matter
of policy not to correct this bug, although they have been made aware of
it multiple times.
- ventritex.com (UGate [Ver. 1.73]-) - replies come from "-MaiSer-@ventritex.com"
- mdn.net (IPAD 1.52 and IPAD 2.0)
Cases that I'm not sure about
In some cases, I just get a bounce, and can't figure out who did the
Wrong Thing. Then I'll list both the ML and the end system here until I
figure out what happened.
Harald.T.Alvestrand@uninett.no
Last modified: Fri Jan 3 08:48:46 1997