You get a number of emails from a problem domain, so you want to blacklist it. You leave the MUA, open ISPConfig, go to email > spamfilter > whitelist, select the scope, and save. That's a very cumbersome procedure. I'm looking for something easier. I was thinking of creating an email account like wblist@, where mail forwarded to that account by a local user would trigger a script that does the above function using the ISPConfig API. So, you use an email in front of you in the MUA, set the subject to begin with w or b, set the body to the address or domain to w/b, and send the email to that account -- all from the MUA, avoiding the need to log into a different application. To avoid accidents, like blocking gmail.com, the wblist@ address itself would be blocked against inbound mail from anyone but the administrator. For end-users the process would be to forward to a different address, like please_block@, which would be reviewed by the admin. Note that this is not the same as a mechanism for reporting Junk=Spam, though it's similar in concept. I tend to over-complicate things sometimes. Maybe I'm missing something. What is YOUR process for maintaining the w/b lists? Do you agree that the process to go through the ISPConfig UI is cumbersome? Do you agree that an interface to the ISPConfig API for would be a useful way to automate this process? Is there a better way? Thanks!
I would not automate it just by forwarding it. Can't think of a much better solution right now though, other than improving your spamfilter With a good spamfilter I barely have to block domains.
Thanks @Th0m. Note that I said that would only take place if the mail comes from an administrator who is consciously send the directive to an automation process. I know we agree that automating a domain block without review is plain ol stupid. So to be clear, a user might flag mail items as Junk, and in review I determine the domain actually sending the mail is legitimate, but a persistent source of rubbish. So I make that determination to block the domain. NOW, we have the feature to do blacklisting, and whitelisting as well! But it seems cumbersome for I or anyone reading this to have to open the open the ISPConfig panel to do this specific operation. Yeah, I've been doing a lot with Rspamd lately, and improving spam filters is an ongoing exercise. The thing is, when we learn_spam or learn_ham or use fuzzy processing, we leave it to the internals to figure out why a mail item is ham or spam. w/b-listing is a clear declaration of exactly what we want to do and why. It has its place and should not be over-used, nor under-used. So I'm just brainstorming a way to simplify the mechanics of performing an operation that we must now do manually. Sometimes the answer is "just do it the hard way" ... thinking of better ways to approach a problem is how we lead to innovation and enhancements.
