Sometime on Aug 21, Satya assembled some asciibets to say:
believe that the "From" display on the recipient's client is taken from the header. Correct me if I'm wrong.
You're making the same mistake that anyone who doesn't dabble in
Actually, Krishnan was correct in what he asked. He just asked the wrong question (he omitted the existence of the envelope entirely).
Here's how it's done (this is a simulation of an actual telnet session. No mail servers were desecrated in the making of this simulation):
Let me fill in the numbers you left out.
$ telnet mail.localhost 25 \d\d\d Some SMTP stuff
220 -----> connect accepted. It could be an error code if the server is overloaded
helo myself \d\d\d Moo
250 -----> command executed successfully. It could be various error codes depending on the error
mail from: fake@invalid.com \d\d\d Sender accepted # would be rejected if relay control was enabled
250
rcpt to: fake2@invalid.com
250
data
354 -----> partial command, await further data.
From: fake3@invalid.co.uk To: satyap@vsnl.com Subject: SPAM!!
I spam you! . \d\d\d Message accepted for delivery
250
quit
221 ------> Mt. Mary Steps to Khar Station West.
Philip is setting the mail from: header, and leaving his From: header alone.
No, I'm setting both for remotely destined mails, and neither for locally destined mails. I'm working on mails destined for both locations. When that's done, I will write a paper.
Philip