Allow rejection of incoming mail with a specific local-part in a catch-all domain

Furio Ercolessi shared this idea ago

Proposed

I love using a catch-all domain, for the possibility of giving out a different address to different entities without having to add an alias or do anything else - like i can give <acme@mydomain> to a company named acme i purchase stuff from.

But: if an entity becomes annoying, it would be nice to be able to shut down a specific address within my catch-all domain. In other words, if <acme@mydomain> starts to receiving mail from acme that i do not want to see any more and unsubscribe does not work, or if that address leaked out to spammers, it would be extremely nice to be able to configure the system so that mail to this address gets a "550 User unknown" at the SMTP level because that address is in a deny table which is consulted by the MTA before the normal lookups used to look-up the destination mailbox for the catch-all mydomain.
This is, I think, the only feature in mailbox.org that I miss.

I like this idea 0

Replies (2)

veil ●

I am also an intensive user of the catch-all function. I love catch-all because it gives me a lot of freedom in how I organize my emails. But unfortunately, this comes at the price of receiving a lot of SPAM.

The suggestion would be an extremely helpful feature and would certainly reduce spam in exactly the right way. I am also sure that it would make mailbox.org stand out very positively from other providers. I don't know of any provider where this is possible. Only as a filter rule, but not at the SMTP protocol level.

Reply URL

Early G ●

Is there now way to do this now? I am receiving a ton of spam as a result of a certain stock broker getting hacked and would like to cut the spam down...

Reply URL

veil ●

Unfortunately not. I also think that mailbox.org doesn't pay much attention to the English-language section of this userforum.

Maybe it would be a good idea to post the suggestion again in the German section (https://userforum.mailbox.org/). ¯\_(‘’)_/¯

It would be such a huge help to me if this feature were available.

URL

Vicki ●

You should be able to achieve this with the custom rules or with sieve filters. If you create a rule to catch emails to that specific user & domain, then reject them, and stop rule processing, before proceeding to the normal catch-all rule it should work.

For example, placing a sieve rule like this one above your normal catch-all:

if envelope :is "to" "acme@mydomain.tld" {
    reject "Reject message";
    stop;
}

Docs: https://kb.mailbox.org/en/private/e-mail/can-i-use-sieve-mail-filtering/

I think for actual spam you might want to just have them deleted or filed to spam rather than rejected though, since you don't want to be bouncing spam mail back.

For example, placing a sieve rule like this one above your normal catch-all:

if envelope :is "to" "acme@mydomain.tld" {
    reject "Reject message";
    stop;
}

Docs: https://kb.mailbox.org/en/private/e-mail/can-i-use-sieve-mail-filtering/

I think for actual spam you might want to just have them deleted or filed to spam rather than rejected though, since you don't want to be bouncing spam mail back.

URL

Furio Ercolessi ●

To Vicki: yes and no. In principle, rules seem to allow rejection on recipient. In practice, I tried it and it does not always work, for reasons that I was not able to understand. Furthermore, that system is somewhat complex and not completely supported (they won't examine rules we have created to find problems, and I can understand this could imply significant support costs). Furthermore, it is not obvious that SMTP reject after the DATA stage works under all circumstances (I do not even know if Mailbox.org does that, do they?), and bouncing after the SMTP transaction has been closed is well known to cause a plethora of nasty problems, so rejection via sieve would be likely not to work as smoothly as a "550 User unknown" during the SMTP transaction and before the DATA stage. And yes, I would like the sender to be informed, as they may not necessarily be hard core spammers. For instance, it could be an entity I purchased products from, they subscribed me to a marketing list without my consent, and their optout system does not work. These things happen.

URL