Message In A Box


          Email. One of the biggest changes in the way people communicate at the close of the twentieth century. Sent with the immediacy of a phone call and read at the convenience of the recipient like traditional mail, it combines the best of both worlds in one form. Even computer-phobic people are getting email accounts now because of the convenience and lack of long distance style charges or post office style delays.

          It is becoming such a part of life for some people that new ways of accessing it are appearing daily. Private companies such as Iridium are launching multitudes of satellites to broadcast web accessed cel phone and pager signals on a global basis. Within the next year or two you will be able to access your email (and vice versa) from anywhere on the planet. No location too remote.

          The crazy part of this is that none of this was planned. It just happened.

          As most people are aware, what we now think of as the Internet grew out of the old ARPANET, a linking of advanced computers to exchange files and data that came onstream in 1969. ARPA, the Advanced Research Projects Agency, was an Eisenhower administration initiative that was meant to put American scientific research on the fast track after the successful and prestigious Russian Sputnik space launches. ARPA was given an open ended mandate and the kind of budget only a decade of prosperity for the richest country in the world could provide. Originally meant to get the American rocket program research into high gear, by the early 1960's it had branched off into information technology and computer networking over leased phone lines.

          ARPANET was originally envisioned solely as a method for sharing data between the expensive and cumbersome computers of various military and university research departments. A popular misconception is that the network was designed to survive a nuclear attack because of its ability to form connections between two computers over many different routes. The more prosaic reality is that often the phone lines would be of poor quality or be too busy so alternate routes were necessary. Even before ARPANET was online though, researchers would often use these connections to communicate informally with their counterparts in other places, bypassing the telephone and teletype systems and using the network for their own purposes. There was no email system or protocols for sending mail to other computers and the use of the network for this was unauthorized.

          The users of this network were extremely bright and unorthodox types though, so email traffic began to grow to the point where some way of standardizing the form was becoming necessary. Ray Tomlinson, an unassuming engineer at BBN, the company contracted to run ARPANET, had made the first email program; originally two programs - one to send and one to receive email - and had designed it for use on local networks. He saw the possibilities for putting this program on the ARPANET and combined the two programs into one. The now ubiquitous "@" symbol was chosen by him to separate the user name from the computer system name because no name or address contained the character. With the incorporation of Tomlinson's program into the file transfer protocol developed for the ARPANET in 1972, several thousand people began using the system and the traffic load started to climb. Within a year, three quarters of the ARPANET traffic was mail.

          This early mail program bears little relationship to what we are today familiar with. Network standards had not been agreed to and your mail arrived as an addition to your mailbox file so that you would have to scroll through all your messages to get to the newest ones. Through the 1970's various programmers here and there would tweak the program to make it run easier and RFC or Request For Comment messages were flying thick as blizzard snow as arguments raged among the individualistic programmers about how mail standards should work. The ARPANET was composed of a wide variety of different and incompatible systems so there had to be common standards everyone could agree on, but determining what they would be was no easy matter.

          Take as an example the email header.

          Here's what you are used to seeing:

Subject: Email headers
Date:  Tue, 1 Dec 1998 15:34:40 -0400 (AST)
From: "Andrew D. Wright" <>
To: Millicent Fakename <>

Short and simple, isn't it? Who from, who to, what about, and when sent, expressed as a relationship to GMT, or Greenwich Mean Time (-0400 or minus four hours). Notice how the sender's name is in quotes? This isn't because I'm using a fake name, but because I'm using an initial for my middle name with a period after it and you cannot use a period (or "dot" as it's referred to in this instance) in a personal name because it's only for separating elements in the computer system's address (

          The problem: too easy to fake messages.

          Here's the full header for this message:

Received: from au141@localhost by id <36703-769>; Tue, 1 Dec 1998 15:34:40 -0400
Date: Tue, 1 Dec 1998 15:34:40 -0400 (AST)
From: "Andrew D. Wright" <>
To: Millicent Fakename <>
Subject: Email headers
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Return-Path: <>

Now you can see a unique message ID, information about the computer I sent the message from, what kind of message this is and where to send a response to it.

          The arguments raged about how much information to put into the headers. A minimalist faction argued that you needed to see next to nothing about the mechanics of the message and that this was just information overload. The other faction argued that you needed full authentification of every detail. Eventually the compromise was that there would be an option in a mail program to allow you to set how much of a header you wished to see and that the rest of the header would be present but invisible to you unless you asked for it. In addition, optional header information would be included by putting an "x-" in front of it. For example, this is from my Netscape mail program and would be tacked on the end of the full header:

X-UIDL: 64c1d9f9ae6494903b9f50acaeea99a7
X-Status: A
X-Mailer: Mozilla 3.04 (Win16; I)
X-Orcpt: rfc822;
      Note the RFC here - this reflects the particular Request For Comment that suggested this header line.
X-Mozilla-Status: 0001
Content-Length: 1068

where my Netscape mail program would tell the recipient what version of Netscape (the browser is known as Mozilla within Netscape) I had used to mail the letter and what Operating System I was using on my computer. Many programs (including PINE, the CCN mail program) will allow you to add your own additional header information. One example of this would be Reply-to: which is used by Userhelp volunteers among others to have a response to a message sent to a different address from the sender's, in this instance to instead of to the particular volunteer's mailbox.

          Email began to move out of the realm of the military and academic users and into the private sector in the late 1970's to the point where there existed almost a half million mailboxes on the Net by 1979. Email programs were still prohibitively expensive, costing several thousand dollars, so for the time being only larger corporations could afford to join the ranks of the military and academic users, but with the personal computer revolution in the 1980's, the floodgates were opened and the numbers began to climb skyward with the speed of a nuclear explosion; by 1990, twelve million people had email.

          On today's Internet there is no accurate figure for how many mailboxes there are. Many people have several different accounts for business and home use and web based email is freely available; Microsoft's Hotmail claims 30 million accounts in their system alone. When corporate mail accounts, standard Internet Service Provider mail accounts, Community Net mail accounts, web based mail accounts, academic mail accounts, and government mail accounts world-wide are tallied, the number is easily in the hundreds of millions.

          Such growth has brought problems as well as convenience to people. The explosive growth in the numbers of users has presented a tempting target to unscrupulous con artists taking advantage of users unfamiliar with the wide open and unpoliced "Wild West" nature of the Internet. Senders of unsolicited commercial email or "spam" prey on users who believe claims that they will be removed from mailing lists if they reply with "Remove" or that the email is sent in compliance with (non-existent) standards or regulations. Spammers get around the message authentification in full headers by hijacking the Send Mail (SMTP) servers belonging to other systems unrelated to them so that people trying to trace the origin of a message get the wrong machine address.

          A war is being fought between spammers and users and system administrators trying to reclaim their bandwidth. Battles rage in the Usenet newsgroups and on the Web as spammers get closed down and relocate to hijack new unsecured servers. According to IMC, the Internet Mail Consortium, in January some 52% of the publicly listed SMTP servers were unsecured and by July the number had dropped to 36%. The price for running an unsecured server can be high: a system that chronically relays spam without securing their SMTP server to prevent this will be treated like a pariah, other systems blocking all mail sent from it until it closes down or mends its ways. Public access to the SMTP server here at the Chebucto Community Net had to be blocked last year after spammers had discovered they could relay their spam through it making it appear it was originating here.

          All told though email works remarkably well for all its growing pains and seems poised to become even more of a factor in daily life in the twenty first century.


You may direct comments or suggestions about this feature to:

Andrew D. Wright,


Back To The Beacon Index

Back To The Beacon Back Issues Index