CTSS Mail

Peter H. Salus <peter@mids.org>

From Matrix News, 8(4), April 1998 <mids@mids.org>, http://www.mids.org +1-512-451-7602, fax: +1-512-452-0127

As I noted in Casting the Net (1995, p. 95), CTSS at MIT had electronic mail in 1965. It was written by Tom Van Vleck and Noel Morris. Tom told John Quarterman: ``I wrote CTSS mail. Licklider mentioned inter-computer mail to me, as I remember, when he asked me if I was interested in a project he had in mind, ``to connect all the ARPA-funded machines together, and see what they said to each other.''

Van Vleck recalled that it was written ``implementing a suggestion by Louis Pouzin and Glenda Schroeder in a CTSS Programming Staff Note. They wanted mail so that operators could inform you when your lost file was retrieved. Noel and I saw it as far more, and we were right.

``I visited MIT on business a few summers ago, and dropped in on Roger Roach, who still had the CTSS listings. We looked at the MAD code for the MAIL command, and it was clearly my code. MAIL was a privileged command, that could do things normal user programs could not: it used the ATTACH.(PROB, PROG) call to switch to the recipient's file directory and added the message to the file MAIL BOX. We can date the command fairly precisely, since the ATTACH call was introduced as part of the `new file system' for CTSS which came up on August 9, 1965. So MAIL was added after that.

``You know how the UNIX mail command works two ways, reads mail if invoked without an address, and sends if invoked with an address? My fault. The UNIX interface derives from the Multics mail command (I wrote the first one) which derives from the CTSS command. And the CTSS command worked that way because I could only have one command -- privileged commands' names were assembled into a table in core A, and the six words or so for another command name was too much.

``Note carefully that I am not claiming the CTSS command was the *first* electronic mail command. At the time Noel and I were writing MAIL, we knew of an electronic mail project at BBN called, I think, MERCURY. And there was a military message system called AUTODIN, what are its dates? DTSS might have had some kind of inter-user mail by then.

``We all owe a debt to Ray Tomlinson, for choosing the @ sign (Multics ARPANet mail first used the control argument -at instead, since the @ was a line kill character) and for extending the pre-existing computer mail paradigm to the ARPANet. But that wasn't the first mail message. Probably wasn't the first inter-computer mail message. [It wasn't. Thomas Marill and Larry Roberts had connected the SDC's AN/FSQ-32 in Santa Monica with the TX-2 at Lincoln Labs via a dedicated 1200bps phone line. They presented their results at the Fall 1966 AFIPS conference. -PHS]

``There was a lot of nervousness at the time about ticking off the Post Office. Calling it `mail' was thought by some to be a Bad Idea, because they feared the USPS would demand that you put 10 cents, or whatever a stamp cost then, into a bucket every time you sent a mail message. If you put a personal note in a parcel, the rule than was that you were supposed to attach first class postage.'' A decade later, the French network, CYCLADES, still didn't have email, even though Louis Pouzin (one of the authors of the memo that first proposed CTSS mail) played a major role at CYCLADES. Quarterman has pointed out that the French PTT was even more dictatorial than the USPS.

``And, unfortunately, I have no recollection at all of the first message that Noel and I sent while testing CTSS mail.''