Should know better
by Andrew Macpherson on Feb.22, 2011, under E-Mail hosting
The base standard for email RFC822 is close to 30 years old. You’ld think by now that people would be capable of adhering to it, and that deviations would be jumped on by every anti-spam tool in existence.
Take for instance the requirement to have a message-Id. A Message-Id has a very specific format: it starts with a ‘<‘ character, then a unique string often based on the date, time and how many messages the system has processed today, then ‘@’, then the fully qualified domain name (FQDN) of the originating system, and finally the closing ‘>’ character.
The message-Id is designed to trace a message through the e-mail infrastructure. It should only be added by the starting MTA, and subsequent systems, in these SPAM aware times, should refuse to handle messages without a valid one, rather than adding one as they did in the past.
The Message-Id is intended for computer use, not human interpretation, so the FQDN should be that — a domain name that can be checked in the DNS, not europe3.mybigcompany, nor missing entirely.
This is even worse when the company in question sells SPAM and Virus detection software, or reliable communications. Congratulations NORTON – Symantec, congratulations Blackberry, come up and take a bow.
