There is another option, especially for those with dialup connections: don't download spam. SpamSieve works only AFTER you've downloaded the email, including attachments. (Therefore, you could end up with porn on your hard drive, not to mention how long it can take to download large spam attachments.) One solution is to use something like POPmonitor.

POPmonitor lets you download the headers (and a user-defined number of lines) of emails in any POP mailbox. (It's easy to set up multiple accounts.) You can review the emails, choose those that are spam (one can review the first part of the content, as I said before) and have POPmonitor delete them. You can also send fake bounce messages.

POPmonitor can do some filtering and can be set to automatically delete those emails it figures are spam, though that can be dangerous, as I've had very legitimate emails marked as spam by every program I know of, including SpamSieve. The user can mark email addresses for a "white" list (or black list!).

Some will complain that this isn't automatic. That's a choice one needs to make. Some of us would rather have some control over spam filtering and don't look kindly upon downloading, then filtering, moving to another folder, etc. For those with dialup accounts (or who use dialup while away from home), something like POPmonitor is much better than SpamSieve. I started using POPmonitor when I used a dialup account and still use it. It also helps me monitor little-used email accounts once in a while. From a time standpoint, it takes me no more time to use POPmonitor than it did to check that SpamSieve hadn't marked some important email as spam.

FWIW, I understand tha Eudora can be set to download headers only, then download only the ones you want.

I've come up with a way that cuts back on the amount of spam that gets through the normal spam filters by simply adding one rule.

First, if you want to track how well this works, add a new mailbox to Apple Mail and name it something like "Received From Unknown".

Go to Preferences/Rules and add a rule (I named mine "FromUnknown"). In the header list, choose "Received". If "Received" doesn't exist, go to the bottom of that popup menu and choose "Edit Header List..." and add it. Set up the rule so that header "Received" CONTAINS "from unknown" (without the quotes), and choose action to be move-message-to-mailbox "Received From Unknown", or to your junk mailbox, or whatever.

I've noticed that about 90% of all spam that gets through to my inbox had "from unknown" in one of the "Received" headers. I think this is when the sender has tried to mask his address, as all emails I receive from normal people never have this phrase (with one exception). The one exception to this is a mailing list I have subscribed to. It also has "from unknown" in a Received header but since I have a filter that directs all mail from that mailing list into a separate mailbox, and that rule occurs before the one here, it still works as it should.

Like I said, this rule is catching about 90% of all junk mail that gets through normal spam filtering, and it has not once redirected a non-spam email to the junk box. I see no reason why it shouldn't work for you too.

..is to use Gmail, which filters 90% of spam pretty efficiently. Using Apple's Mail, it filters out all the rest, bar perhaps one or two a week (I use three busy accounts which attract about 700 spam mails a week. I used to use SpamSieve in conjunction with Entourage and wasn't happy with it.

I've been a very happy customer with SpamSieve.

Part of my job is to be fluent in multiple email applications. By using SpamSieve, I'm able to use one spam corpus for all the apps (Apple Mail, MS Entourage, Qualcomm Eudora, GyazMail, GnuMail) - a big time-saver. I've found that it works extremely well and learns quickly, and for $25 you get one spam filter that works with everything.

How does this compare with Spamnix, or have you tested it?

Have you tested SpamNix?

It's been my view for a long time that spam filtering should be performed server-side. This has the notable advantages of allowing one to use, as the Guest above put it "a single spam corpus" with any mail client. This also helps bandwidth-limited and dialup users by cutting out the spam at the outset, before it ever gets to the client.

My preferred server-side spam solution is SpamAssassin and an IMAP server. IMAP is particularly nice because it stores all messages server-side by default, as opposed to POP, which defaults to deleting messages once downloaded. IMAP also has much better mailbox support on the server side. (The tendency of IMAP servers to store mail on the server is actually probably one reason why mail providers seem to shy away from it - people can be packrats with their email, and all that old archived cruft would take up a lot of space.)

I have been using SpamAssassin for some months now, and even at the outset, it sucked most of the incoming spam into my spambox. Granted, the spam I get is predominantly difficult to classify by feature (not enough sensible text), so the DNS blacklists are doing most of the work for me. However, once I had the requisite number of spam messages (200) to start using the Bayesian content analysis features, I can count the number of spam that have slipped through on one hand. (And one of those made it through just barely, because it was nothing but a link and a bunch of text cribbed from The Hobbit!). No false-positives for me, either (though in previous testing, I did see one that was a club newsletter that had enough spamlike features that it could easily be mistaken for spam. This was also without the Bayesian filter enabled.)

It's really too bad mail providers don't seem to be more interested in combatting spam at the server level. It seems like a much more sensible approach.

-JMV