[kent m]

...that - before it displays an incoming message - will send a reply to all incoming messages and then automatically delete the message if a dead email address is found?

I would like this as a spam filter.

I still find that Mail will occasionally set a friend's message to "Junk", so that's not exactly working for me. However, as all spam seems to be sent with fake return addresses, an attempt to contact returns a "no recipient" style message, I see this as a better way to filter them.

Perhaps via AppleScript?

k

[cpac]

that's an awful lot of steps, and wouldn't that send an auto-reply to every single email you get in, even the valid ones?

If your problem with the built-in junk mail filter is that your friend's emails end up in it, you have a couple of options:

First, keep junk in training mode for a while until it stops marking your friends' emails as junk.

OR

Second, modify the junk rule (in Preferences>Junk>Advanced) to exclude emails from your friends' addresses.

either would be more elegant than the solution you're proposing.

[kent m]

hello,

I feel that the more elegant solution is mine...

potential solution...

sure it would send a message to each person that sends a message to me, but it would preferably happen invisibly - I don't want to elicit a response from the individuals, I only want a response from the server level. my potential solution would only allow a message through that returns a true response to whether or not the recipient exists...

if it reports that the email address is a valid one then the message can be let through.

the Mail junk filter is fine, but there's always the chance that a personal message will be culled out. I never want to let that happen.

I have a personal website that has occasionally elicited a lot of messages from strangers, and I don't want to shut that out, so it's not as simple as having an "acceptable" list of addresses that I only accept mail from.

as it's my experience that spam senders NEVER use a verifiable return email address - why is this not an elegant solution? It would happen invisibly.

I'd love to find a way to do this and any help is welcomed.

[Earth Mk. II]

In SMTP you don't know if an address is valid or not until the mail hits the server at the receiving end - it just gets passed along until it hits its destination. Its up to the receiving server to decide what to do with the message.

I don't think your solution is implementable without sending an email, at least not without rewriting the protocol (SMTP is an inherently insecure protocol - it's called "Simple Mail Transfer Protocol" for a reason) or client software. And that could get more annoying than helpful.

Also, not all spam uses fake reply addresses, so I doubt that would even fix the whole problem.

You're probably best off re-training Mail's spam filter. That, or get all your friends to use PGP and filter out anything that isn't signed with a valid key.

[jasong]

Actually, what he is asking for is possible. I use a tool at work to occasionally check the email addresses on our mail list. It basically acts like it is sending a message to the mail server, and it records if the server accepts the message for a given user or not.

That being said, the problem with using this as a general purpose spam filter is the amount of bogus traffic it would generate. It would literally double the number of mail requests going on if everyone used it (every incoming message would trigger an outgoing response), leaving mail servers unable to effectively deliver the mail.

-- Jason

[kent m]

Right, exactly. Every reply I've sent to a spam message has bounced back as 'undeliverable'. So, what I want is a filter to withhold displaying a message until a return address has been verified. If a reply is 'undeliverable' the message doesn't get let through to me.

Ideally this should happen at the server level, before I ever see it, but in lieu of that I wonder if it can be done myself. Perhaps by setting up a local firewall/server machine that would act as a 'gatekeeper' before letting the mail through to my personal machine?

At this point I'm not too concerned about doubling message traffic and don't think it would happen that way, really. They'd be pretty small packets and wouldn't be stored. Also, ideally, once an address was 'verified' as 'legitimate' it would no longer need to be continually rechecked - this would only happen to messages from 'strangers' as a 'one time occurance' - more of a 'first-time check-in' proceedure.

And considering this, it seems that a thing like this may perhaps spur towards a solution at the host level?... if the servers see that traffic amount is doubling for a certain kind of mail - spam, which is the 'real' bogus traffic - they might be more inclined to set up their own filtering system.

jasong, what was the tool that you used to do the checking?

kent

[jasong]

At this point I'm not too concerned about doubling message traffic and don't think it would happen that way, really.

yeah, really it would happen that way. It wouldn't double the amount of storage needed because as you said, no message is being sent. But the majority of mail is only a few k in size, the hard part is the delivery of the message, and this is what would be doubled, the whole process that got the message to you, would have to be done in reverse.

I don't see anyway this could be done in Mail, but I am sure someone could write a email program that did this. Since you asked, the program I use is called Email Validator Advanced but it is Windows only (sorry, what I have to use at work). I don't know of any Mac programs that do this because I don't have a personal need for it.

Good luck and keep us updated on any progress you make on this.

-- Jason

[kent m]

it would double the amount of traffic for new incoming messages, triple maybe, but that would only happen once for new messages from verifiable addresses.

Once an address has been verified as "legitimate" a record of that address and it's status could be stored and it would not be checked again. That would solve that issue...

Thanks for the link to the mail app. Ah, windows... But maybe I can find something similar and coherse a programmer friend of mine into helping me out.

k

[fortepianissimo]

Originally posted by kent m:
it would double the amount of traffic for new incoming messages, triple maybe, but that would only happen once for new messages from verifiable addresses.

Once an address has been verified as "legitimate" a record of that address and it's status could be stored and it would not be checked again. That would solve that issue...

Thanks for the link to the mail app. Ah, windows... But maybe I can find something similar and coherse a programmer friend of mine into helping me out.

k

Is this what you're talking about?

http://www.macupdate.com/info.php/id/11905

[kent m]

Originally posted by fortepianissimo:
Is this what you're talking about?

http://www.macupdate.com/info.php/id/11905

Not quite, I think. This seems to be a tool for verifying addresses contained in a mailing list, not for verifying addresses on random incoming mail. You give it a list of mail addresses and it finds out if the addresses are valid or false.

Still it's got common points. maybe there's something here that could be adapted...

Thanks for the link!

kent

[fortepianissimo]

Originally posted by kent m:
Not quite, I think. This seems to be a tool for verifying addresses contained in a mailing list, not for verifying addresses on random incoming mail. You give it a list of mail addresses and it finds out if the addresses are valid or false.

Still it's got common points. maybe there's something here that could be adapted...

Thanks for the link!

kent

Yes - I'm not knowledgable about what's working behind the scene. If I knew it, I guess it's relatively easy to hook the mechnism up with Mail.app via AppleScript...

(and then I need to learn that as well, but guess probably not too difficult to pick up)