deep magic spam filtering? [feature questions] H
    deep magic spam filtering? [feature questions] Synametrics Support
        deep magic spam filtering? [feature questions] Synametrics Support

From: H
Date: 5/4/21 12:13 PM
Topic: deep magic spam filtering? [feature questions]
Type: General Discussions
Post a follow up

I have skimmed the filtering sections of the manual.

 

Does xeams implement TMDA for challenge response?

.. Are my recipients [temp] whitelisted to avoid challenge response?

 

Can an ordinary user be enabled to view suspected message queues like the site admin?

 

Are spam assassin rules supported?

 

Is greylisting supported?

 

DNSBL informed rules?

 

When will there be DANE support?  RFC 6698

 

There should be rules to characterize headers as a single entity.  Size [is, less, greater], as char count.  And size filters for any individual header or X-header.

 

Is there DCC support? (Distributed Checksum Clearinghouse)

 

[I motivated cPanel to adopt DCC and many other great antispam ideas but they are no longer as responsive to ideas. BoxTrapper falls short of TMDA]

 

Don't waste money on Arvixe hosting or their infuriatingly substandard support for any class of service.

 

 

 

Top

From: Synametrics Support
Date: 5/5/21 7:19 AM
Topic: deep magic spam filtering? [feature questions]
Type: General Discussions
Post a follow up

I will number my answers according your original questions below.

Q1: Does xeams implement TMDA for challenge response?
A1: TMDA is an open source script in Python and since Xeams is not written in Python, it cannot use it. However, TMDA addresses 4 thing:

  1. whitelists - supported by Xeams
  2. blacklists - support by Xeams
  3. challenge/response - supported by Xeams
  4. tagged addresses - I am not 100% sure what this means. Most links on tdma.net are broken and therefore, does not know what this covers. Having said that, there are several features in Xeams that could fit in the category of "special-purpose email address". These are temporary address (https://www.xeams.com/temp-address.htm), and Auto-Learn senders (https://www.xeams.com/adaptivefilters.htm)

Q2: Can an ordinary user be enabled to view suspected message queues like the site admin?
A2: Yes. Like administrators, users could also log in to Xeams web interface to manage several things, such as:

  • White/black listing
  • Out of office messages
  • Mailing list subscriptions
  • Viewing/restoring quarantined messages
  • Manage passwords for End-to-End encryption

Q3: Are spam assassin rules supported?
A3: No. Xeams' filtering mechanism is not based on SpamAssassin or any other open source project. Therefore, you cannot import rules between the two systems.

Q4: Is greylisting supported?
A4: Yes. Check https://www.xeams.com/greylisting.htm for more information

Q5: DNSBL informed rules?
A5: DNSBL, Yes. I am not sure what you mean by "informed rules". Xeams comes pre-configured with several publicly available RBL servers. You can add new servers if you like using the web interface.

Q6: When will there be DANE support?  RFC 6698
A6: I have added this as a feature request. Our dev. team will evaluate this and will try adding this feature.

Q7: There should be rules to characterize headers as a single entity.  Size [is, less, greater], as char count.  And size filters for any individual header or X-header.
A7: Interesting idea. Could you give me an example of when would you use the size of the header a criteria to assign score, considering most headers can be forged.

 

 

 

 

Top

From: Synametrics Support
Date: 5/5/21 7:39 AM
Topic: deep magic spam filtering? [feature questions]
Type: General Discussions
Post a follow up

I forgot to address your last question in my previous reply.

Q8: Is there DCC support? (Distributed Checksum Clearinghouse)
A8: Xeams has its own mechanism to handle this. There are two rules:

  • Xeams has its own RBL server, which could blacklist misbehaving IP addresses
  • IP reputation filter, which looks at last 7 days of IP addresses
  • Fuzzy Search, which creates a fuzzy checksum of previously received messages and assigns score if similar message comes back
Top