Last update: Sat Sep 24 06:47:07 2005 ... Thu Mar 23 14:00:37 2017
What is my electronic mail address?
Suffix your login account name with @math.utah.edu, for example, email@example.com.
Can I have a different e-mail address?
In rare cases, users may request, and be given, alternate, or additional, e-mail usernames or aliases. Ask the systems staff. This service is not available for undergraduate accounts.
Is there an IMAP or POP server for access from e-mail clients on remote systems?
Yes. We support both IMAP and POP to read your e-mail. The required host name for the incoming server is mail.math.utah.edu. For sending mail, we also support authenticated SMTP through the host mail.math.utah.edu, this would be the outgoing server or SMTP server. As of December 2009 all IMAP, POP and SMTP connections are required to use SSL/TLS to function. With most mail clients, this is simply a matter of selecting either SSL or TLS.
The POP port number is 110 (pop3 with StartTLS). The IMAP port is either 143 (imap with StartTLS) or 993 (imaps with SSL/TLS). The SMTP port numbers are either 25 (smtp with StartTLS) or 587 (submission with StartTLS). Use SMTP port 587 if you suspect your internet service provider (ISP) might be blocking port 25.
With IMAP (incoming) and SMTP (outgoing), the option StartTLS may need to be selected.
The username and password for both IMAP and SMTP are the same.
However, you are responsible for figuring out how to configure your remote e-mail client to use either IMAP or POP services. The remote system management or your service provider may be able to assist you, but our local systems staff should be consulted only as a last resort. Our e-mail services have operated successfully for more than a quarter century, and it is very rare for them to be unavailable or misconfigured: faults are almost always at the remote client.
We recommend using IMAP instead of POP, and keeping your mailboxes on our mail servers (maybe with local copies on your remote computer). That way, you can still have access to your e-mail, even if you don't have access to your remote system. Also, you can then be sure that your e-mail archives are properly backed up nightly: backup policies on laptops and remote systems are frequently unreliable, irregular, or even nonexistent.
What e-mail client can/should I use?
Like the choice of a text editor, selection of an e-mail client is often subject to strongly held views, and no mail single client has proved adequate for everyone.
The systems staff strongly prefer text-only e-mail clients like mm, mutt, pine, and Emacs rmail, since they are much more powerful, and don't waste time displaying images, unpacking attachments, and reformatting text in mostly useless fancy fonts and colors.
Some users prefer handling e-mail with a Web-browser: on most of our systems, the older netscape and mozilla are available, though now deprecated because their development has largely ceased. Their descendant, the newer seamonkey Web browser and e-mail client, is available on most systems. The thunderbird client is a companion to our recommended Web browser, firefox, and if you prefer the browser approach, thunderbird is likely to be the best currently supported tool.
If you are very old, and long for the simpler days of early Unix systems, you can try using the primitive Mail program, or the even more limited mail program.
Most modern desktop window systems offer some sort of default e-mail client that may be readily accessible from an icon on the desktop. For example, many students use dtmail in the Common Desktop Environment (CDE).
The critical point to be aware of is that each e-mail client usually has its own unique format for storage of mail messages, and other clients normally cannot read formats other than their own. This means that you cannot easily switch between e-mail clients if you expect to be able to easily access old e-mail messages: pick a client that works for you, and then stick with it.
How can I access my e-mail securely from a remote Web browser?
Please be aware that attacks are still possible between your keyboard and the local client, or from someone looking over your shoulder. In particular, a keyboard sniffer installed on a machine in an Internet café can record everything that you type, including usernames, passwords, bank account numbers, and credit-card numbers. Such environments should be considered completely untrustworthy, and avoided.
How do I automatically forward e-mail between systems?
E-mail systems on Unix operating systems recognize a special hidden file named .forward and stored in your top-level login directory. It contains a single line with a comma-separated list of e-mail addresses to which mail should be forwarded. The list may contain a username prefixed with a backslash, in which case e-mail is also forwarded to that mailbox corresponding to that user. The list may also contain a filepath naming a file to which mail is to be delivered.
Except for the latter kind of delivery, the presence of a .forward file causes immediate redelivery of incoming e-mail, without storing it in a local mailbox.
If Professor Jones moves to another institution, she can cause her future incoming e-mail at Utah to be immediately forwarded to her new address with a .forward file line of the form firstname.lastname@example.org.
When Professor Smith goes on sabbatical, he might be unsure of the guest institution's computer systems reliability and backup policies, so he arranges for a local copy to be kept at Utah, and another to be forwarded, with a .forward file line of the form \smith, email@example.com.
If you find that you want to simply discard all incoming e-mail messages for a time, you can create a .forward file with the line /dev/null. That ask for all incoming mail to be written to the null device, effectively, and permanently, discarding it, without any possibility of backup or recovery.
Mail forwarding and automated reply messages (aka. vacation) can be turned on and off via the "Autoresponder / Mail Forwarding" option in Squirrelmail
How do I recall a mistakenly sent message?
You cannot! Once you hit the send key, e-mail delivery is virtually instantaneous, and your message is then propagated to all of the recipients, which could be a very large number if your message was to a popular mailing list. If you make a serious mistake in such a posting, a brief apology in a subsequent message might sometimes be in order. If you mistakenly sent personal information, such as a credit card number, then you'd better call your card supplier to cancel the card immediately.
Who owns my e-mail?
The legal issues about ownership of e-mail are uncertain, and vary among legal jurisdictions. Many companies assert ownership of all e-mail on computer systems that they own. In general, at our site, systems staff normally avoid looking at user mail files, but they could be forced to do so in the event of security breaches, or at the orders of a legal subpoena, or of higher-level administrators of the University of Utah or of the Government of the State of Utah.
As a rule, it is expected that e-mail traffic through our systems is primarily connected with research and teaching of mathematics and related subjects, or for general educational purposes. Personal e-mail, if present, should be a minor part of your e-mail traffic. As with Web pages, e-mail should be civil in tone, and should not be used to harass or otherwise inconvenience or offend its recipients, or others mentioned in the message text.
Like all other data on our filesystems, e-mail is archived for long periods. Thus, even if you delete e-mail that you believe falls outside the proper use of University computer systems, it may be recoverable, possibly by legal force, for a very long time.
All mail-client software sets suitable permissions on mailbox files to prevent anyone other than the owner from reading the messages. However, systems staff have management privileges that can ignore file permissions, and you yourself can subvert the normal visibility restrictions by careless setting of permissions of your mailbox files. As a rule, the output of the ls -l file-listing command on such files should report a permission string of -rw-------, meaning read and write for owner, and no access for anyone else.
For further details, see the University Information Resources Policy.
How should I respond to lengthy e-mail?
It is a long-standing practice in e-mail discussions to maintain context of the discussion by including brief snippets of the incoming message identified by one or more leading close-angle brackets, followed by your reply. For mailing-list postings, you should also identify the sender and timestamp of the incoming message. Here is an example of a response:
In mail to this list on Mon, 19 Sep 2005 06:59:11 -0600 (MDT), Joe Blow <firstname.lastname@example.org> wrote: >> I recommend that we sell all our stock in Widgets >> Inc., because the CEO has just been indicted for >> fraud. I disagree; the Wall Street Journal reports today that the company just signed a multibillion dollar contract with the US Army. I think their stock will go up substantially, despite the indictment.
Do not insert a copy of the sender's message at the start or end of your message, even if your mail client has a button or key to do so. This just contributes to the already overwhelming e-mail burden that many people have to deal with, and does not help your readers follow the discussion.
Also, be sure to stick to plain text in the ASCII character set, with lines of reasonable length, and neatly formatted, if you want your messages to be easily read by your recipients. Do not send HTML, or even worse, both plain text and HTML, and don't use extended character sets, such as the ISO 8859-n or Unicode sets, unless that is conventional on the list.
Finally, avoid sending file attachments in your e-mail: not only are they avenues for potential attacks, but they seriously increase network traffic and disk space requirements, especially when sent to lists. If you have a file to distribute to others, as long as it is reasonably public, put it on your Web page, and send a one-line URL that points to the file. Then only those recipients who really want it need to get it.
What do I do when I'm on vacation or otherwise away for a period?
For short trips, just let incoming e-mail pile up. That is the whole point of e-mail: you deal with it only at your own convenience, and only if you care to.
However, if you would like remote correspondents to be informed of your absence, then run the vacation program just before you leave, and just after you return. It will prompt for the information it needs to set up vacation responses, or if that service is already in effect, ask if you want it turned off.
vacation keeps a record of your incoming correspondents, and will generally send them only a single message each week about your absence, and will try hard to avoid sending such messages at all to any mailing lists that you subscribe to.
A typical vacation message looks something like this:
I will be away from my office from Monday 19 September 2005 until Sunday 25 September 2005 Your mail regarding "$SUBJECT" will be read when I return.
However, please be aware that public notice of your absence could make your home a target for theft. For many e-mail users, silence and delay in response is the best policy.
Mail forwarding and automated reply messages (aka. vacation) can be turned on and off via the "Autoresponder / Mail Forwarding" option in Squirrelmail
How do I deal with unsolicited e-mail (spam)?
Sadly, unwanted bulk e-mail has seriously abused the world-wide Internet for several years. Because it is a world-wide problem, national, state, and local laws that have been enacted to combat it are largely ineffective.
The best advice is to ignore spam: just delete it as soon as you encounter it. You are warned that such e-mail often contains objectionable material, including drug pushing, financial scams, political messages, pornography, phishing attacks (which attempt to get you to divulge personal information), purported security checks, racist diatribes, virus attacks, and so on. Life is sometimes hard, and you just have to ignore or avoid things that you do not like.
E-mail clients that use window systems (e.g., some Web browsers on Unix, and most clients on Apple and Microsoft desktop operating systems) should be configured to not open attachments, and not display embedded images until you wish to see them. Set your e-mail client preferences accordingly, so that you can avoid unpleasant surprises from nasty pictures that you probably don't want to see.
Never respond to spam, or follow a Web link from spam messages! That just increases network traffic, and more seriously, confirms that your account and/or system exists. That is likely to make your system more vulnerable to future attacks, and the target of yet more spam.
What spam controls are in force?
We do not do any mail-content filtering, but we do greylisting and blacklisting of e-mail from most off-campus addresses (campus e-mail, and e-mail from .edu and .gov domains, is delivered immediately):
When a system connects to our e-mail
servers, the tuple consisting of
is checked to see if it is in the whitelist. If so, then e-mail is accepted immediately. Otherwise, the remote system is told to try back in 14 minutes. If the remote system then tries again with the same tuple after 14 minutes, that tuple is added to the auto-whitelist for the next 24 hours. Once the tuple is in the auto-whitelist, any further communication happens immediately. Thus, it is only the initial message which is delayed by 14 (or more) minutes.
The actual delay may be much greater than 14 minutes, but that depends on when the remote sending server tries to deliver the message again.
If you look at the header lines in the messages that you receive, there is one with the tag X-Greylist. It will tell you how long the message was delayed.
Can spam controls permanently block delivery?
We sometimes see remote mail systems that do not obey the greylisting request to return after 14 minutes. Either they retry more frequently (and each time are told to come back later), or they retry delivery from a different mail host. In either case, mail is effectively blocked permanently. There is nothing that we can do about this, since the fault lies in a remote system that is not under our control.
If this happens to you, it is your responsibility to contact the remote mail administrators and get them to fix their problem. They are then welcome to contact our local systems staff to discuss the technical issues in more detail, if necessary.
It is rare for mail-delivery problems to be at our end. Our mail server runs on a highly reliable system with redundant hardware components and backup power generation, is closely monitored by a mail-software wizard, successfully delivers and receives several thousand messages daily, and has done so for three decades.
If mail is blocked by the blacklist, this is because the remote host has behaved so badly that we refuse to speak to it. Blacklist entries can be removed on request, but only if we have personal assurances from the management of the remote system that they have resolved the serious problems that led to our blacklisting them.
If you really need to correspond by e-mail with someone on a misbehaving host, please contact our systems staff and ask whether a whitelist exception might be added so that mail from a particular account on that host is accepted immediately, even though blocks remain for all other accounts on that host.
If e-mail delivery still fails, then remember that there are still other communications technologies, including telephone, FAX, postal mail, text messaging, and even better, personal visits.
The system spam controls are not sufficient for me. What else can I do?
While the greylisting and blacklisting techniques have eliminated about 90% of incoming spam since they were implemented, some users still require additional controls. Unfortunately, these extra controls are both complex, and fragile: you can easily misdirect, or even lose, incoming e-mail because of a simple typographic error in filter configuration file. Also, these controls place additional load on our already-very-busy e-mail servers.
Please consult systems staff for further information on such filtering. They can provide you with starting templates of configuration files and blacklists that you can then further tune for your own needs.
How do I subscribe to local mailing lists?
With a Web browser, visit http://www.math.utah.edu/mailman/listinfo/, choose a mailing list, and then follow the Subscribing to ... instructions. Use the same procedure to unsubscribe from a list.
All local mailing lists require membership to post to them, so there should never be any spam on them. All are currently low-traffic lists, with only a few posts weekly.
As with any Internet mailing lists, it is strongly recommend that you avoid posting to a list until you have read it for a few weeks, and can judge the level of discussions and expertise. Otherwise, ill-informed postings by novices often run the serious risk of nasty responses (called flames) from veteran list members.
How do I read local mailing-list archives?
With a Web browser, visit http://www.math.utah.edu/mailman/listinfo/, and follow the first link on the page to reach the list archives.
How do I change the default Web browser in the Thunderbird mail client?
This should be simple, but is not because of poor documentation and lack of customizability. To change the Web browser to, say, firefox, shutdown thunderbird, then use a text editor to add this line to all $HOME/.thunderbird/*/prefs.js files:
When you restart thunderbird, it will use the browser that you specified.