Email

7.2 Email

SMTP

Message transfer agents - Move email across the network.

Source machine establishes TCP socket connection to port 25 on destination machine.

Sample SMTP conversation delivering an email from client C to server S.

DCU mail servers

It was easier to see how mail servers worked when DCU had its own mail servers. But these are being phased out.

Now all DCU email is done through a Gmail service.

Incoming server: imap.gmail.com
Outgoing server: smtp.gmail.com

Of course, these "server" addresses have entire server farms behind them.

If any DCU mail servers still running, try these out:

$ telnet mail.dcu.ie 25

$ telnet mailhost.computing.dcu.ie 25
help
214-2.0.0 This is sendmail version 8.12.10
214-2.0.0 Topics:
214-2.0.0       HELO    EHLO    MAIL    RCPT    DATA
214-2.0.0       RSET    NOOP    QUIT    HELP    VRFY
214-2.0.0       EXPN    VERB    ETRN    DSN     AUTH
214-2.0.0       STARTTLS

Email headers

RFC 822 - SMTP message format
Email headers
How to Get Email Headers on different email clients.

Basic email header fields.
Message consists of:

Header fields
(blank line)
Message body

Other email header fields.
Applications can invent their own. Must begin with X-.
e.g. Some Spam filtering software attaches headers like X-Spam-Status:

Email headers in forensic work

I receive an unpleasant email from Hotmail.
But where is it actually from? America?

"View - Headers - All" (or something similar) in your email client and you can see the actual email headers.
These reveal the path it travelled and also the original IP address, which you can then look up.
Lookup here said it was Eircom, who I then complained to.
I could even see what part of Ireland this came from.

X-Originating-IP
You will soon start spotting stories like this everywhere: Man sends email to MP threatening to kill her children. He is convicted April 2012. How is he caught? "Police traced the malicious email back to Zimmerman's computer by tracking the IP address". Sounds like it was somewhere in the email headers.

MIME

MIME encoding.
Invented for binary data to be sent by email (designed for text data).
Encodes it as simple plain text 7-bit ASCII chars.

Media types (or MIME types or Content types)
- Common examples
IANA
- MIME types
Not just for email. The Web (HTTP) uses these too.

Types of MIME encodings

base64 encoding: For binary attachments.
Can encode any binary stream whatsoever as (a really safe subset of) 7-bit ASCII.
"The resultant base64-encoded data has a length that is approximately 33% greater than the original data"

Content-Type: application/msword; name="Postgrad_Prospectus_Research.doc";
 x-mac-type="42494E41"; x-mac-creator="4D535744"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="Postgrad_Prospectus_Research.doc"
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....

Quoted-printable encoding: For basically plaintext with 8-bit ASCII chars.
Encodes everything as 7-bit ASCII. The 8-bit chars are encoded as: "=nn".
"=" itself encoded as "=3D".

I have had several enquiries =
regarding=20
IEEE Xplore (full-text access to IEEE and IEE publications) and would =
like to=20
inform you of some recent developments. As part of the<A=20
href=3D"http://www.library.dcu.ie/news/index.htm#HES"> IReL</A> =
initiative (funded=20
by SFI and HEA) to provide increased access to e-journals, negotiations =
for IEEE=20
Xplore are at an advanced stage.

MIME headers.

MIME content types.

How to repair corrupted email

Sometimes email files can get corrupted or broken.
If a base64 attachment survives somewhere in the corrupted file, you can extract it as follows:
1. Get the raw email file. (View source or view original if you need to.)
2. It is plain text so you can just edit it with a text editor.
3. Remove everything around the base64 data, including all Header lines (Field: Value) and demarcation lines ("--") and blank lines.
4. Then something like:
```
 
base64 --decode file > file.jpg
```
5. It works! Decodes it back into the original 0's and 1's.
6. You have to know (from surviving Header lines) what file type the file is meant to be.

Standalone email clients

Email clients - Read and send email.

Mail server: mail.isp.com receives and stores your email when you are offline.

To read it you might:

Use a web interface the ISP has set up at: www.isp.com
Use a standalone client to communicate with mail.isp.com over the network, using various protocols.

Both types of email client (web clients and standalone clients) are popular.

(a) mail.isp.com collects your email.
(b) You periodically use a client to connect to mail.isp.com

POP3

The original protocol for standalone clients.

POP3 client sets up TCP connection to the mail server at port 110.

POP3
POP3
- list of commands

Again, possibly try:

$ telnet mail.dcu.ie 110
$ telnet mailhost.computing.dcu.ie 110

Could write script to loop forever, alert you when new message comes in from sender X.

Sample POP3 conversation to retrieve 3 email messages.
You can retrieve and then delete from server. Or retrieve and leave on server.

IMAP

Updated protocol for standalone clients.

IMAP - Multiple clients accessing the same mailbox. Each sees the others' changes.
e.g. Access DCU mail from home and from work. Don't want changes stored on client.
With POP, can "leave messages on server". But no folders. With IMAP, can create mailbox subfolders on the server.
Other features:
- Partial fetch. Just fetch header. Speeds things up. Do not fetch attachments. Do not fetch body of emails you will never click on.
- Server-side searches.
Standalone clients can be set up to keep a running mirror on local machine. As well as writing changes to server.

Web-based email

Web-based email like Gmail

Keep everything on server (like IMAP). No standalone client. Just web browser.

Not everyone uses web client only:

Businesses like security of having local copy of email.
For some uses, web client can be less flexible than standalone client.
e.g. I have programs to "grep" local copy of my email, programs to build mail rules, etc.

Note that most big web email services can be accessed through either web client or standalone client.
Gmail supports IMAP access.

Proprietary email protocols

Microsoft Outlook uses a proprietary protocol to talk to Microsoft email servers.
IBM Notes (formerly Lotus Notes) uses a proprietary protocol to talk to IBM email servers.

Surveys

Survey of email reading environments, 2018.
From litmus.com.
Shows the popular ways of reading email: desktop web browser, desktop standalone client, and mobile.

I personally like using desktop standalone client. (Mozilla Thunderbird with IMAP.)

Survey of top email clients, 2018.
From litmus.com.
Shows a mix of mobile and desktop clients.
Huge amount of reading email on Apple devices. (Is there a problem with data gathering for other devices?)
Also shows a mix of webmail and standalone email clients.

emailclientmarketshare.com from litmus.com