[ih] Ancient mail question about 551 and 251 reply codes

[Jack Haverty]

In the early 1970s, Licklider's work at MIT pursued his vision of a 
"galactic network" in which users' computers would interact with other 
users' computers over a wide-area network to help the users do 
everything they do in life.  Human-human communications was part of that 
vision, initially instantiated in email implementations. Lick's group at 
MIT built such a mail system; I built the server component.

As a model, we used what we knew, which was the US Postal System. Our 
projects called it "messaging" to avoid any conflict with the "snail 
mail" bureaucracy which had a mandated legal monopoly on "mail".

IIRC (been a long time...) the Postal System at the time had several 
means of handling people who had moved.  First class mail was forwarded 
to the new address for a year or so.  Less-important mail was returned 
to the sender with a label indicating that the addressee had moved.

I don't remember, but I was probably lobbying for the 251 and 551 codes 
to be added to the official protocol, as a way to mimic the behavior of 
the snail mail system.  My server also probably sent those messages out 
during SMTP interactions, and may have done something with any such 
errors it received (at least reflecting the error back to the original 
sender).   I do remember that our users would get error reports as email 
from the message server, e.g., to report that a message remains 
undelivered after trying for a while (days) but the remote host was 
consistently down.   I recall spending some time to make such messages 
"user friendly", e.g., something like "Excuse me, but you might want to 
know that your message concerning <subject> that you sent on <date> has 
not yet been delivered."

I don't recall that we ever actually received any such errors from other 
SMTP systems on the Arpanet.   Few other systems seemed interested in 
going beyond the very basic mail functionality that had started as a 
simple addition to FTP.  Our mission in Lick's group was to research 
human-human communications.  Other sites on the network were focused on 
other research topics.

We also lobbied for electronic "stamps" as a way to provide some 
back-pressure as a limiting mechanism for spam.  No one like that either.

If you can find archives of the "HEADER-PEOPLE at MIT-AI" mailing list, it 
would be a good source of such historical artifacts about email.

Jack Haverty
(MIT 1966-1977)

On 2/12/24 11:30, John Levine via Internet-history wrote:
> For a very long time SMTP has had the 251 and 551 reply codes which
> say that the recipient is somewhere else and give the new address. 251
> means the reciving system accepts the message and will presumably
> forward it, while 551 rejects the message so you're supposed to resend
> it yourself. They were introduced in a different form in RFC772 in
> 1980 and in the current form in RFC780 in 1981.
>
> My question is whether anyone has actually done anything with these,
> like resend the message to the new address, or update address books,
> or at least report the reply to the sender somehow. I don't ever
> recall it, but in 1980 I was still on uucp.
>
> R's,
> John,uucp at computer.org
>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature.asc
Type: application/pgp-signature
Size: 665 bytes
Desc: OpenPGP digital signature
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20240212/f8d1768b/attachment.asc>