Rspamd 3.12.0 released with Ollama support, separate read and write servers in fuzzy check, HEIC file support, added proxy features, better error handling and various bug fixes
Rspamd 3.12.0 released with Ollama support, separate read and write servers in fuzzy check, HEIC file support, added proxy features, better error handling and various bug fixes
I've released an SPF filter for OpenSMTPD: https://netsend.nl/opensmtpd-filter-spf/
I've created a spamtrap filter for OpenSMTPD. If you have any dedicated traps sprinkled around or maybe some old e-mail addresses that are only used by spammers these days, you can use them to protect your other mailboxes.
The #MIMEDefang Project has a major update coming in v3.5, with the work mostly done by Giovanni Bechis. Substantial feature additions and bug fixes! MIMEDefang is one of the most mature and versatile tools for mail filtering via the Milter API (available in Sendmail and Postfix.) It has been my personal years-long favorite tool for hooking #SpamAssassin into MTAs, long before I joined the project.
The first release candidate is at https://github.com/The-McGrail-Foundation/MIMEDefang/releases/tag/v3.5-rc1
Playing around with a #milter with #postfix and discovered that there is a quatantine feature that moves affected mails into a seperated hold queue! It even supports to set a reason from the milter!
Only problem? For the last 12 years postfix does abolutely nothing with that information. Nice xD So if I want to use that in my milter I need a seperate quatantine log that records the reason with the queueid of the message...
JUST RELEASED!
#MIMEDefang 3.4.1, the mature #milter that lets you integrate #SpamAssassin, #ClamAV, and other filtering tools with #Sendmail, #Postfix, or any other milter-capable #MTA and to enact any #email policy or practice that you can express in #Perl.
One thing I'd like to do is move the #SRS portion of the mail stuff into #rspamd. rspamd can trivially rewrite the envelope from as a #milter. I just hardcoded bounces@server as a test, and that worked fine; so the question is how much work implementing SRS itself will be in #Lua.
This would avoid the need for #postsrsd and #postforward.