Join the AMA (Ask Me Anything) with the Firefox leadership team to celebrate Firefox 20th anniversary and discuss Firefox’s future on Mozilla Connect. Mark your calendar on Thursday, November 14, 18:00 - 20:00 UTC!

Search Support

Avoid support scams. We will never ask you to call or text a phone number or share personal information. Please report suspicious activity using the “Report Abuse” option.

Learn More

Need hook added to Thunderbird for spam detection

more options

I receive spam that can't be filtered out by adaptive filtering or by the email filters, but would be very simple to filter out using a little script-writing, if the raw email was available to the script. Any language would do, including the standards, PHP and JavaScript.

I get each email only once from a different account/domain/IP address, and the wording is always different, but it is clear that two spammers are behind all these emails, since only two of my many email addresses are used as destinations and since they share a common style or type of content. It would be easy to check for certain characteristics of these emails using a script language that had access to the entire raw email before it gets into my Inbox.

For example, one category may always include a fixed pattern of HTML. Or another may always include characters in a certain part of the UTF-8 encoding space. Such patterns can't be detected by filters but could be detected by short scripts.

I can't add features to Thunderbird due to the complexity of building Thunderbird, and due to the small disk space on my computer, so I would like to request that a developer add a "hook" feature to call a script to determine whether an incoming message should be put in a Junk folder or sent to the Inbox as usual.

I'm okay with the hook being anything convenient, such as a local call to a named script file, or even a remote call to a website REST endpoint.

Such a spam solution would be so incredibly effective that I (or someone else) could set up a service to distribute and update such a script or endpoint code, preventing spam not only for me, but for others. There could be a community of programmers who contribute to such a standard script. Each massive spammer could be given a name, just like each virus or worm is given a name.

But I need the hook added to Thunderbird first, please!

I receive spam that can't be filtered out by adaptive filtering or by the email filters, but would be very simple to filter out using a little script-writing, if the raw email was available to the script. Any language would do, including the standards, PHP and JavaScript. I get each email only once from a different account/domain/IP address, and the wording is always different, but it is clear that two spammers are behind all these emails, since only two of my many email addresses are used as destinations and since they share a common style or type of content. It would be easy to check for certain characteristics of these emails using a script language that had access to the entire raw email before it gets into my Inbox. For example, one category may always include a fixed pattern of HTML. Or another may always include characters in a certain part of the UTF-8 encoding space. Such patterns can't be detected by filters but could be detected by short scripts. I can't add features to Thunderbird due to the complexity of building Thunderbird, and due to the small disk space on my computer, so I would like to request that a developer add a "hook" feature to call a script to determine whether an incoming message should be put in a Junk folder or sent to the Inbox as usual. I'm okay with the hook being anything convenient, such as a local call to a named script file, or even a remote call to a website REST endpoint. Such a spam solution would be so incredibly effective that I (or someone else) could set up a service to distribute and update such a script or endpoint code, preventing spam not only for me, but for others. There could be a community of programmers who contribute to such a standard script. Each massive spammer could be given a name, just like each virus or worm is given a name. But I need the hook added to Thunderbird first, please!

All Replies (4)

more options

There are already a number of quality email spam products available that process email prior to it being processed by an email client. Adding a new function of this magnitude would reduce the resources to support the email function. However, you're free to submit the recommendation at bugzilla dot mozilla dot org.

more options

I appreciate and respect your opinion, although it does seem to make a very negative assumption. Optionally calling a local script or even a remote endpoint is a simple coding change, essentially just a few lines of code. The most complex part of it is adding the option to the options menu. How complicated can that be for someone familiar with Thunderbird coding? I have submitted the request to bugzilla.

Editorial: Why have we put up with spam so long, when we have successfully blocked ads on the Web through shared URL lists and have successfully blocked single-IP spammers using blacklists? No one seems to have thought of the simple solution that I described in this posting. It would obviously work well, allowing us to match whatever the spammers think up, yet would be a simple addition to any email client, particularly for open source software like Thunderbird that anyone is allowed to modify.

more options

All of your email arrives on server first and then Thunderbird just downloads it. Have you tried to get the spam email dealt with by the server spam filters so you never even download it? If you report at that level it would be more convenient to you.

In 'Message Filters' under the first drop down - 'FROM' there is a 'Customise' option not sure what you need but you can use it to detect other headers.

Some people use programs like MailWasher or Spamfighter

more options

Both the server and Thunderbird filters use conventional rules that once worked, quite a long time ago. But spammers make money from their spamming, so they apply their ingenuity and find ways to get around such rules. In other words, access to the headers just doesn't work anymore. It is necessary to look for subtle clues, most of which are in the body of the message, as well as going beyond simple pattern matching of the headers.

For all the spam messages I have received in recent months, I can see simple ways a program (script) could detect one of these spam messages, even though filters and antispam programs don't work. However, there is currently no way to interface such a script-based solution to Thunderbird. All we need is a hook, and we can take it from there.