2004 E-Mail Evaluation Criteria
Important Note: The criteria for this year's evaluation effort is heavily
based on last year's criteria.
The product evaluation focused on software that meets the following basic characteristics:
- Run on the University's recommended
desktop standards for 2003-2004.
- Support for "trainable" Bayesian spam filtering rules. (I.e. Client's
interface allows end-user to easily mark messages as spam (or not);
client "learns" what is spam over time; easy access to caught spam
- Strong support for Internet standards: Authenticated SMTP, LDAP (v3)
directory service, and especially IMAP (rev 1).
- Note: This group is focusing primarily on evaluating IMAP clients
vs. POP3 clients based on ISC's recommendation, with a strong
preference for IMAP. However, the group acknowledges transition recommendations
will need to be
those schools or centers heavily using POP3 clients
and whose servers may not be ready to bear the additional disk space
required for IMAP client use.
- IMAP vs. POP
- LDAP is now a requirement since the new Penn online directory will
not support Ph. We will most likely pass along this information to the
Browser Evaluation Team (concerning browsers with bundled email clients).
It's generally stated that whatever Browser is chosen, end-users are
going to automatically use the e-mail client that comes with it, so it
would good to uncover any Netscape Messenger or Mozilla Mail client issues
early. From prior experience, these two clients appear to function seamlessly
with LDAP, but we will investigate this as a team.
- In particular for this effort is meeting the University AAA requirements
clear text passwords (2003 team's link not functioning properly) from being
sent over the network.
direction of the University is towards a Kerberos single logon. With this in
(GSSAPI) Kerberos support and/or TLS as an alternative solution
where Kerberos isn't supported, will be given stronger weight, all other criteria
and functionality being equal.
Additional Background and Related Resources
Information Systems and Computing, University of Pennsylvania