Email:

Transfer, retrieval and storage

E-mail's are transferred from one computer to another by a Mail transfer agent (MTA) (a mail server is a computer acting as a MTA). The 'Simple Mail Transfer Protocol', abbreviated as SMTP, is the most popular protocol that MTAs use. Sendmail was the first MTA to use SMTP. You can learn more about the SMTP protocol at 'RFC 5322' and 'RFC 5321'. The 'Received' field in the e-mail header, is a list of the MTAs which have handled the e-mail.

After the MTAs have transferred the e-mail, the message will usually reside on a remote server. The user will then retrieve the e-mail from MTA's using an e-mail client program (such as Outlook Express) (also called a mail reader or mail user agent (MUA)) or a webmail service (such as Gmail). Clients and webmail use a retrieval protocol such as IMAP or POP, to download the e-mail from an MTA to the clients inbox. (images, a picture of a client login and webmail login)

Once the e-mail has been retrieved, it then needs to be stored. There are many formats which can handle storage, perhaps the most simple and efficient is Maildir. Which operates by creating unique temporary files for each retrieved message. It will depend on the client or webmail, as to which storage format they will use. Maildir, for example, was designed for the qmail program, but is compatible with other clients. Not all storage formats create unique files for each e-mail, another option is to use a collective database format. mBox is one such example, storing messages in one single file. Why are there a variety of different techniques for storing messages? because, unlike message transfer, the Internet Engineering Task Force (IETF) has not developed a standard mechanism for storage.

E-mail process, step by step (sequence applies to most users)

1. A sender composes an e-mail in a client program, such as Outlook Express.
2. He presses send, the client program then formats the message (in MIME format etc.).
3. The client program then uses SMTP to send the message to a local MTA.
4. The MTA which accepts the message depends on the user's ISP or webmail service.
5. The MTA reads the destination address, such as 'joe@example.com'.
6. The MTA will lookup the domain address 'example.com' in the Domain Name System (DNS).
7. The DNS system will give the MTA the 'mail exchange server' for the domain.
8. The MTA will then send the e-mail onto the mail exchange server using SMTP.
9. The destination user will login his client program.
10. His client program will use POP or IMAP to accept the message from his/her MTA.