On Sat 16/May/2026 00:24:41 +0200 Douglas Foster wrote:
Ale, I write in the context of a business domain, where acceptable emails must be acceptable based on the combination of organizational policy. An organization is likely to block some lists as not acceptable because they are not business-related. You don't need to use the company network to chat with your quilting club, unless your business is quilting-related.
For example, anyone working at, say, Cisco should be interested in participating in IETF mailing lists.
My point was entirely about authentication. When we know that all messages from a list will arrive unauthenticated, it is foolish to use authentication as a reason to block a subset of those messages. Ignoring authentication does not mean whitelisting from all content checks; that is an independent decision. And should be easy for anyone to detect a mailing list, based on a stream of messages, with multiple From domains and a single Mail From domain, all without authentication. It may be a list or an ESP or an auto-forwarder or a fraudster. It is reasonable to investigate and determine which situation applies, then configure filtering rules based on what is learned.
I don't think it is reasonable, in response to such a clear-cut request as p=reject, to start investigating whether this is the first message of a stream that might later exhibit certain characteristics. Detecting mailing lists reliably is not easy, and attempting to do so veers toward omniscience.
Compared to the simple authorization step that users already undergo when subscribing to a list, a complex classification process based on statistical inference looks totally overkill.
The comment about "personal devices" was not essential to the argument. My organization blocks access to personal webmail sites because we assume that they are inadequately protected against spam, so personal email needs to be retrieved on personal devices. If an organization is blocking a mailing list as an unacceptable risk, then it should take measures to prevent that restriction from being circumvented using personal email accounts on in-house computers.
Hm... not totally clear, but it seems unrelated to SMTP.
I think you are suggesting that organizations should also figure out whether their users have actually subscribed to a list before allowing list messages to penetrate the network. I agree, and I have said as much previously. I have vetted all of the mailing lists that send messages into our network. I have previously observed that GoogleGroups.com and Groups.Outlook.Com are known sources of malicious traffic, and they are blocked by default. Hosting services for organizations could delegate the allow/block rule to their clients. Mailbox Providers for the general public have the hardest problem to solve, but so not need my advice.
Exactly. Luckily, they use a distinct domain name, GoogleGroups, rather than routing everything through Gmail; otherwise, domain-based whitelisting would be even more problematic than it already is.
Best Ale -- _______________________________________________ dmarc mailing list -- [email protected] To unsubscribe send an email to [email protected]
