電子郵件

範例

The diagram above shows a typical sequence of events ^[1] that takes place when Alice composes a message using her mail user agent (MUA). She types in, or selects from an address book, the e-mail address of her correspondent. She hits the "send" button.

1. Her MUA formats the message in Internet e-mail format and uses the Simple Mail Transfer Protocol (SMTP) to send the message to the local mail transfer agent (MTA), in this case smtp.a.org, run by Alice's Internet Service Provider (ISP).
2. The MTA looks at the destination address provided in the SMTP protocol (not from the message header), in this case bob@b.org. An Internet e-mail address is a string of the form localpart@exampledomain.com, which is known as a Fully Qualified Domain Address (FQDA). The part before the @ sign is the local part of the address, often the username of the recipient, and the part after the @ sign is a domain name. The MTA looks up this domain name in the Domain Name System to find the mail exchange servers accepting messages for that domain.
3. The DNS server for the b.org domain, ns.b.org, responds with an MX record listing the mail exchange servers for that domain, in this case mx.b.org, a server run by Bob's ISP.
4. smtp.a.org sends the message to mx.b.org using SMTP, which delivers it to the mailbox of the user bob.
5. Bob presses the "get mail" button in his MUA, which picks up the message using the Post Office Protocol (POP3).

This sequence of events applies to the majority of e-mail users. However, there are many alternative possibilities and complications to the e-mail system:

• Alice or Bob may use a client connected to a corporate e-mail system, such as IBM Lotus Notes or Microsoft Exchange. These systems often have their own internal e-mail format and their clients typically communicate with the e-mail server using a vendor-specific, proprietary protocol. The server sends or receives e-mail via the Internet through the product's Internet mail gateway which also does any necessary reformatting. If Alice and Bob work for the same company, the entire transaction may happen completely within a single corporate e-mail system.
• Alice may not have a MUA on her computer but instead may connect to a webmail service.
• Alice's computer may run its own MTA, so avoiding the transfer at step 1.
• Bob may pick up his e-mail in many ways, for example using the Internet Message Access Protocol, by logging into mx.b.org and reading it directly, or by using a webmail service.
• Domains usually have several mail exchange servers so that they can continue to accept mail when the main mail exchange server is not available.
• E-mail messages are not secure if e-mail encryption is not used correctly.

It used to be the case that many MTAs would accept messages for any recipient on the Internet and do their best to deliver them. Such MTAs are called open mail relays. This was important in the early days of the Internet when network connections were unreliable. If an MTA couldn't reach the destination, it could at least deliver it to a relay that was closer to the destination. The relay would have a better chance of delivering the message at a later time. However, this mechanism proved to be exploitable by people sending unsolicited bulk e-mail and as a consequence very few modern MTAs are open mail relays, and many MTAs will not accept messages from open mail relays because such messages are very likely to be spam.

Note that the people, e-mail addresses and domain names in this explanation are fictional: see Alice and Bob.

Format

The format of Internet e-mail messages is defined in RFC 2822 and a series of RFCs, RFC 2045 through RFC 2049, collectively called Multipurpose Internet Mail Extensions (MIME). Although as of July 13 2005 RFC 2822 is technically a proposed IETF standard and the MIME RFCs are draft IETF standards,^[2] these documents are the de facto standards for the format of Internet e-mail. Prior to the introduction of RFC 2822 in 2001 the format described by RFC 822 was the de facto standard for Internet e-mail for nearly two decades; it is still the official IETF standard. The IETF reserved the numbers 2821 and 2822 for the updated versions of RFC 821 (SMTP) and RFC 822, honoring the extreme importance of these two RFCs. RFC 822 was published in 1982 and based on the earlier RFC 733.

Internet e-mail messages consist of two major sections:

• Header — Structured into fields such as summary, sender, receiver, and other information about the e-mail
• Body — The message itself as unstructured text; sometimes containing a signature block at the end

The header is separated from the body by a blank line.

Header

The message header consists of fields, usually including at least the following:

• From: The e-mail address, and optionally the name of the sender
• To: The e-mail address[es], and optionally name[s] of the message's recipient[s]
• Subject: A brief summary of the contents of the message
• Date: The local time and date when the message was written

Each header field has a name and a value. RFC 2822 specifies the precise syntax. Informally, the field name starts in the first character of a line, followed by a ":", followed by the value which is continued on non-null subsequent lines that have a space or tab as their first character. Field names and values are restricted to 7-bit ASCII characters. Non-ASCII values may be represented using MIME encoded words.

Note that the "To" field in the header is not necessarily related to the addresses to which the message is delivered. The actual delivery list is supplied in the SMTP protocol, not extracted from the header content. The "To" field is similar to the greeting at the top of a conventional letter which is delivered according to the address on the outer envelope. Also note that the "From" field does not have to be the real sender of the e-mail message. It is very easy to fake the "From" field and let a message seem to be from any mail address. It is possible to digitally sign e-mail, which is much harder to fake. Some Internet service providers do not relay e-mail claiming to come from a domain not hosted by them, but very few (if any) check to make sure that the person or even e-mail address named in the "From" field is the one associated with the connection. Some Internet service providers apply e-mail authentication systems to e-mail being sent through their MTA to allow other MTAs to detect forged spam that might apparently appear to be from them.

Other common header fields include (see RFC 4021 or RFC 2076 for more):

• Cc: Courtesy copy
• Bcc: Blind Courtesy Copy
• Received: Tracking information generated by mail servers that have previously handled a message
• Content-Type: Information about how the message has to be displayed, usually a MIME type
• Reply-To: Address that should be used to reply to the sender.
• References: Message-ID of the message that this is a reply to, and the message-id of this message, etc.
• In-Reply-To: Message-ID of the message that this is a reply to.
• X-Face: Small icon.

Many e-mail clients present "Bcc" (Blind carbon copy, recipients not visible in the "To" field) as a header field. Different protocols are used to deal with the "Bcc" field; at times the entire field is removed, whereas other times the field remains but the addresses therein are removed. Addresses added as "Bcc" are only added to the SMTP delivery list, and do not get included in the message data.

IANA maintains a list of standard header fields.

Body

Template:Refimprovesect

Content encoding

E-mail was originally designed for 7-bit ASCII. Much e-mail software is 8-bit clean but must assume it will be communicating with 8-bit servers and mail readers. The MIME standard introduced character set specifiers and two content transfer encodings to enable transmission of non-ASCII data: quoted printable for mostly 7 bit content with a few characters outside that range and base64 for arbitrary binary data. The 8BITMIME extension was introduced to allow transmission of mail without the need for these encodings but many mail transport agents still do not support it fully. In some countries, several encoding schemes coexist; as the result, by defaulf, the message in a non-Latin alphabet language appears in non-readable form (the only exception is coincidence, when the sender and receiver use the same encoding scheme). Therefore, for international character sets, Unicode is growing in popularity.

Plain text and HTML

Both plain text and HTML are used to convey e-mail. While text is certain to be read by all users without problems, there is a perception that HTML-based e-mail has a higher aesthetic value. Advantages of HTML include the ability to include inline links and images, set apart previous messages in block quotes, wrap naturally on any display, use emphasis such as underlines and italics, and change font styles. HTML e-mail messages often include an automatically-generated plain text copy as well, for compatibility reasons. Disadvantages include the increased size of the email, privacy concerns about web bugs and that HTML email can be a vector for phishing attacks and the spread of malicious software.^[3]