1993-10-21 - why an envelope

Header Data

From: pmetzger@lehman.com (Perry E. Metzger)
To: cypherpunks@toad.com
Message Hash: 97f422a07e92d51ee5d35e44aad3f9484d0d2e2341703685f67c4acf676b1897
Message ID: <9310211724.AA02906@kublai.lehman.com>
Reply To: N/A
UTC Datetime: 1993-10-21 17:28:00 UTC
Raw Date: Thu, 21 Oct 93 10:28:00 PDT

Raw message

From: pmetzger@lehman.com (Perry E. Metzger)
Date: Thu, 21 Oct 93 10:28:00 PDT
To: cypherpunks@toad.com
Subject: why an envelope
Message-ID: <9310211724.AA02906@kublai.lehman.com>
MIME-Version: 1.0
Content-Type: text/plain

Enough people have asked that I'll explain why envelopes are seperated
from headers in mail.

Lets say I'm sending mail with the following header

To: foo, bar, baz

Mail gets stored and forwarded many times during its delivery process
on a large internet. If at every stage the only hint the mailer had
for delivery was the header address, every mailer along the way would
have to generate three copies of the mail in order to guarantee
delivery. Kind of nutty, eh?

Thats why there is a seperate envelope. When you generate the mail,
three seperate copies will be produced with seperate envelopes
indicating the mail whould be sent to "foo", "bar", and "baz", and the
mailers in between don't look at the headers at all.

This is a slight simplification because envelopes can contain multiple
addresses -- but thats when you are sending one message to many people
via a common path -- the mailer at the last step where the mail can
take a common path is expected to break the envelope up before sending
it further along.