C-Command Software Forum

Block domains?

I frequently get spam from the same domains, just different random users. In other words the left side of the @ sign changes, the right side does not. I would like to block those entire domains. This seems simple enough, just add an entry to the block list using either contains or ends with the domain name. However, I do that and it never seems to work.

What is the proper way to block an entire domain? Am I missing something?


You can do this by adding a blocklist rule such as “From (address) Ends With @domain.com”. If this is not working, you can look in SpamSieve’s log to see why SpamSieve predicted the message to be good. Or, if there’s no prediction in the log, that means your mail program was configured such that it never sent the message to SpamSieve for analysis.

So, I’ve checked, and that is how I added the domain to the blocklist. I waited til I got another one and looked at the log today, but it only has entries for how items were trained. I must not understand how to check the log properly. Can you elaborate on that some or refer me to documentation I can read to understand that more?

All my log entries look like this. No notes about why something was blocked or not blocked that I can see…

Trained: Spam (Manual)
Subject: Dog Food Coupons: Save on Pedigree, Purina and more
Identifier: F2xcjhM4UkjmphQddBEtUA==
Actions: added rule <From (address) Is Equal to "DogFoodCoupons@simplelgr.com"> to SpamSieve blocklist, added to Spam corpus (4860)
Date: 2013-03-16 05:48:51 +0000

Trained: Spam (Manual)
Subject: The education you need to earn your X-ray technician degree
Identifier: PZcORLwVRtbGuK0i+A1Hqg==
Actions: added rule <From (address) Is Equal to “XRayTechnician@dot-bruise.info”> to SpamSieve blocklist, added rule <From (name) Is Equal to “XRay Technician”> to SpamSieve blocklist, added to Spam corpus (4861)
Date: 2013-03-16 05:48:55 +0000

Trained: Spam (Manual)
Subject: Win a Million Dollars Every Year for Life!
Identifier: 2GOVbioT73d20TpDWwj7nA==
Actions: added to Spam corpus (4862)
Date: 2013-03-16 13:58:01 +0000

Trained: Spam (Manual)
Subject: Your Equifax score has changed.
Identifier: jHQmLn8hFMFSN5yLHs0oiw==
Actions: added rule <From (address) Is Equal to "Lindsey@coloredribbon.com"> to SpamSieve blocklist, added rule <From (name) Is Equal to “Score Program”> to SpamSieve blocklist, added to Spam corpus (4863)
Date: 2013-03-16 14:08:20 +0000

Normally there would be log entries that say “Predicted: Spam” and “Predicted: Good.” The absence of such log entries means that the mail program has not been asking SpamSieve to analyze any of the incoming messages. Therefore, you should check the setup.

Thanks. While following the instructions, I did find that some settings were not as they should have been. Hopefully, this will fix everything. I appreciate the help and the patience. I should have caught that myself, sorry.

Is it possible to have a shortcut to create this rule? Like ctrl+cmd+d will mark the entire domain instead of only the specific sender? I can’t write script so any detailed response is highly appreciated.

It is possible to do this, and perhaps sometime I’ll write a script for it. However, you said that you were using the trial, so you’re probably just getting started. I would really encourage you not to do this by script. There is normally no need for you to manually create or delete blocklist rules because SpamSieve learns to recognize your spam automatically. If you find that you often need to manually block whole domains, there’s probably something that’s not set up or working properly, so please contact technical support.

Spamsieve has actually improveddrastically the past few days and it seems it learns well enough. When trial expires, I will go for full version.

However, I have noticed that with intervals the same spammers create dozens of new domains and send to all my email addresses and when this happens, I now hope Spamsieve will have learnt by then and need not to re-learn because spam comes from new domains. As you understand, i haven’t studied the script or how Spamsieve works. I hope domain-blocking won’t be necessary and I remain happy.

A bit off topic: How do I get rid of the pop-ups (notifications) when spamsieve finds a suspected spam? It tends to be hassle to close those notifications windows.

Yes, that should be fine because SpamSieve learns from the entire contents of the messages, not just the addresses.

Please see Use Growl global notification system.