Re: [MlMt] How is span score computed?
On 2024-09-26 at 09:47:21 UTC-0400 (Thu, 26 Sep 2024 09:47:21 -0400) William Allen is rumored to have said: I subscribe to my local newspaper’s daily bulletin. Recently I noticed I wasn’t getting it anymore and after looking a the junk folder saw it had a spam score of 4.0. Just looking at the mailings I can’t see any difference. Is there any way to understand what is triggering a score for a particular piece of mail? Likewise, is there a way to override the filter for a particular sender? MailMate itself does not score messages. The scores it can detect are those determined either by SpamSieve locally or by your email provider's spam filters. Without knowing which is relevant in your case, it isn't possible to say how to adjust it. So if you have installed SpamSieve, consult its documentation for how to adjust its scoring. If you haven't installed SpamSieve, the score is being added by your mailbox provider and you should ask them what adjustments are available. SOME (not all) mailbox providers claim that by removing the $Junk flag and/or adding a $NotJunk flag and/or moving mail from a "Junk" or "Spam" mailbox to the INBOX will be noticed by their filter maintenance systems and lead to future similar messages no0t being marked as spam. SOME also claim that if you add the sender to an address book linked to their mail system (such as Google, iCloud, or Exchange Online/MS365) it will prevent future mail from being labeled Spam. [Puts on Apache SpamAssassin maintainer hat for the following tangent] One of the most common free and open-source toolkits included by mail providers as a part of their spam filters is SpamAssassin, maintained by the Apache Software Foundation. Anyone can use SpamAssassin and modify it however they like. It is sometimes useful to use a SpamAssassin scan to figure out what may be considered spam by systems that use it *and* by other tools that use similar scanning approaches. If you're comfortable working with command line tools and understand how to setup a Perl runtime environment (probably with MacPorts or Homebrew) it can be useful to install SpamAssassin and use it to answer such questions as "why was this marked as spam?" Unfortunately, the most easily findable website offering the general public SpamAssassin scans is miserably misconfigured and misleading. If you do find and use it or any similar tool online, you should understand that any spam filtering requires site-specific information to work well, so public scanners are always going to make mistakes based on their lack of knowledge. I don't link to public scanners because they have that innate flaw, but some people find them helpful. -- Bill Cole b...@scconsult.com or billc...@apache.org (AKA @grumpybozo@toad.social and many *@billmail.scconsult.com addresses) Not Currently Available For Hire ___ mailmate mailing list Unsubscribe: https://lists.freron.com/listinfo/mailmate
[MlMt] How is span score computed?
I subscribe to my local newspaper’s daily bulletin. Recently I noticed I wasn’t getting it anymore and after looking a the junk folder saw it had a spam score of 4.0. Just looking at the mailings I can’t see any difference. Is there any way to understand what is triggering a score for a particular piece of mail? Likewise, is there a way to override the filter for a particular sender? ___ mailmate mailing list Unsubscribe: https://lists.freron.com/listinfo/mailmate
Re: [MlMt] How is span score computed?
Thanks! That was very informative and helpful. Regards, Bill On 26 Sep 2024, at 11:09, Bill Cole wrote: On 2024-09-26 at 09:47:21 UTC-0400 (Thu, 26 Sep 2024 09:47:21 -0400) William Allen is rumored to have said: I subscribe to my local newspaper’s daily bulletin. Recently I noticed I wasn’t getting it anymore and after looking a the junk folder saw it had a spam score of 4.0. Just looking at the mailings I can’t see any difference. Is there any way to understand what is triggering a score for a particular piece of mail? Likewise, is there a way to override the filter for a particular sender? MailMate itself does not score messages. The scores it can detect are those determined either by SpamSieve locally or by your email provider's spam filters. Without knowing which is relevant in your case, it isn't possible to say how to adjust it. So if you have installed SpamSieve, consult its documentation for how to adjust its scoring. If you haven't installed SpamSieve, the score is being added by your mailbox provider and you should ask them what adjustments are available. SOME (not all) mailbox providers claim that by removing the $Junk flag and/or adding a $NotJunk flag and/or moving mail from a "Junk" or "Spam" mailbox to the INBOX will be noticed by their filter maintenance systems and lead to future similar messages no0t being marked as spam. SOME also claim that if you add the sender to an address book linked to their mail system (such as Google, iCloud, or Exchange Online/MS365) it will prevent future mail from being labeled Spam. [Puts on Apache SpamAssassin maintainer hat for the following tangent] One of the most common free and open-source toolkits included by mail providers as a part of their spam filters is SpamAssassin, maintained by the Apache Software Foundation. Anyone can use SpamAssassin and modify it however they like. It is sometimes useful to use a SpamAssassin scan to figure out what may be considered spam by systems that use it *and* by other tools that use similar scanning approaches. If you're comfortable working with command line tools and understand how to setup a Perl runtime environment (probably with MacPorts or Homebrew) it can be useful to install SpamAssassin and use it to answer such questions as "why was this marked as spam?" Unfortunately, the most easily findable website offering the general public SpamAssassin scans is miserably misconfigured and misleading. If you do find and use it or any similar tool online, you should understand that any spam filtering requires site-specific information to work well, so public scanners are always going to make mistakes based on their lack of knowledge. I don't link to public scanners because they have that innate flaw, but some people find them helpful. -- Bill Cole b...@scconsult.com or billc...@apache.org (AKA @grumpybozo@toad.social and many *@billmail.scconsult.com addresses) Not Currently Available For Hire ___ mailmate mailing list Unsubscribe: https://lists.freron.com/listinfo/mailmate___ mailmate mailing list Unsubscribe: https://lists.freron.com/listinfo/mailmate
Re: [MlMt] How is span score computed?
On 26 Sep 2024, at 8:09, Bill Cole wrote: MailMate itself does not score messages. The scores it can detect are those determined either by SpamSieve locally or by your email provider's spam filters. Without knowing which is relevant in your case, it isn't possible to say how to adjust it. So if you have installed SpamSieve, consult its documentation for how to adjust its scoring. If you haven't installed SpamSieve, the score is being added by your mailbox provider and you should ask them what adjustments are available. How does an email provider mark their spam score? Is there a header field that they use that MailMate recognizes? --Randall ___ mailmate mailing list Unsubscribe: https://lists.freron.com/listinfo/mailmate
Re: [MlMt] How is span score computed?
On 26 Sep 2024, at 14:26, Bill Cole wrote: Yes. I believe MM recognizes various headers generated by different tools. I believe X-Spam-Score and X-Spam-Status (both optionally added by SpamAssassin) are both recognized, and possibly the X-less equivalents The X- prefix was deprecated last century. Sad it's still being used for new header fields. Sad that we don't have a single standard header field for conveying server-side spam scores. --Randall ___ mailmate mailing list Unsubscribe: https://lists.freron.com/listinfo/mailmate
Re: [MlMt] How is span score computed?
On 2024-09-26 at 17:13:46 UTC-0400 (Thu, 26 Sep 2024 14:13:46 -0700) Randall Gellens is rumored to have said: On 26 Sep 2024, at 8:09, Bill Cole wrote: MailMate itself does not score messages. The scores it can detect are those determined either by SpamSieve locally or by your email provider's spam filters. Without knowing which is relevant in your case, it isn't possible to say how to adjust it. So if you have installed SpamSieve, consult its documentation for how to adjust its scoring. If you haven't installed SpamSieve, the score is being added by your mailbox provider and you should ask them what adjustments are available. How does an email provider mark their spam score? Is there a header field that they use that MailMate recognizes? Yes. I believe MM recognizes various headers generated by different tools. I believe X-Spam-Score and X-Spam-Status (both optionally added by SpamAssassin) are both recognized, and possibly the X-less equivalents -- Bill Cole b...@scconsult.com or billc...@apache.org (AKA @grumpybozo@toad.social and many *@billmail.scconsult.com addresses) Not Currently Available For Hire ___ mailmate mailing list Unsubscribe: https://lists.freron.com/listinfo/mailmate
