There is now an interactive GUI fetchmail configurator, fetchmailconf.
Code is 64-bit clean and Y2K-safe.
Automatically decodes armored 7-bit MIME into 8 bits (this can be
suppressed).
You can specify which SMTP error is recognized as a spam block.
Support for Kerberos V authentication.
Support for IMAP-OTP authentication using Craig Metz's patches
for UW IMAP.
Support for IPv6 and IPSEC (using Craig Metz's inet6-apps library).
Support for IMAP with RFC1731-conformant GSSAPI authentication.
Fixed and verified support for Cyrus IMAP server, M$ Exchange,
and Post Office/NT.
Support for responding with a one-time password when a POP3 server
issues an RFC1938-conforming OTP challenge.
Support for Compuserve's RPA authentication protocol for POP3
(not compiled in by default, but configurable).
Since 3.0:
Support for IMAP RFC 1731 authentication with Kerberos v4.
Support for multiple-folder retrieval in a single session
under IMAP.
Following SMTP 571 response to a From line, fetchmail no longer
downloads the bodies of spam messages.
Support for a `hunt list' of SMTP hosts.
Support for ESMTP 8BITMIME and SIZE options.
Support for ESMTP ETRN command.
The stripcr & forcecr options to explicitly control carriage-return
stripping and LF->CRLF mapping before mail forwarding.
Since 2.0:
Support for secure use with ssh.
Mailserver passwords can be parsed out of your .netrc file.
When forwarding mail via SMTP, fetchmail respects the 571
"spam filter" response and discards any mail that triggers it.
Transaction and error logging may optionally be done via syslog.
(Linux only) Security option to permit fetchmail to poll a host
only when a point-to-point link to a particular IP address is
up.
RPOP support (restored; had been removed in 1.8).
2.0 and earlier versions:
Support POP2, APOP, RPOP, IMAP2, IMAP2bis, IMAP3, IMAP4, IMAP4rev1. .
Support for Kerberos V4 user authentication (either MIT or Cygnus).
Host is auto-probed for a working server if no protocol is
specified for the connection. Thus you don't need to know
what servers are running on your mail host in advance; the
verbose option will tell you which one succeeds.
Delivery via SMTP to the client machine's port 25. This
means the retrieved mail automatically goes to the system
default MDA as if it were normal sender-initiated SMTP mail.
Configurable timeout to detect if server connection is dropped.
Support for retrieving and forwarding from multi-drop mailboxes
that is guaranteed not to cause mail loops.
Large user community -- fetchmail has a large user base (the
author's beta list includes well over two hundred people). This
means feedback is rapid, bugs get found and fixed rapidly.
Carefully written, comprehensive and up-to-date man page describing
not only modes of operation but also how to diagnose the most
common kinds of problems and what to do about deficient servers.
Rugged, simple, and well-tested code -- the author relies on it
every day and it has never lost mail, not even in experimental
versions. (In the project's entire history there has only been
one recorded instance of lost mail, and that was due to a quirk
in some Microsoft code.)
Strict conformance to relevant RFCs and good debugging options.
You could use fetchmail to test and debug server implementatations.
For anybody who cares, fetchmail is Y2K safe.
Features in common with other remote-mail retrieval programs:
The other programs I have checked include fetchpop1.9, PopTart-0.9.3,
get-mail, gwpop, pimp-1.0, pop-perl5-1.2, popc, popmail-1.6 and upop.
Support for POP3.
Easy control via command line or free-format run control file.
Daemon mode -- fetchmail can be run in background to poll
one or more hosts at a specified interval.
From:, To:, Cc:, and Reply-To: headers are rewritten so that
usernames relative to the fetchmail host become fully-qualified
Internet addresses. This enables replies to work correctly.
(Would be unique to fetchmail if I hadn't added it to fetchpop.)