Security and SpamTopCompatibility and issues with SMTPim2000 Concepts

im2000 Concepts

Why does the Email address not contain an account name and a domain.

The tight asociation of a user to a host does not exist anymore, and this is one of the main advantages spammers take to do their harm - forgery of Email addresses, either by inventing them or by pretending to be an account of another person.

Directory Services like LDAP are yet disconecting the concept of the host and the user, we are in the `"era`" of the netcitizen.

im2000 Mail addresses identify an "eternal mail address" asociated with user, institution, charge, etc. in fact - anything which, what or who can be sent a message to. A global distributed database, similar to DNS stores some vital data needed by the mail exchange, amongst them:

What happens when there are multiple recipients to a message?

Each recipient is sent a notification with it's own set of credentials. The mail storage server keeps track of who has retreived the message and who not.

How will this protocol know when to delete the outgoing message file? How long will the message file remain before being automatically deleted and therefore no longer retrievable?

This is determined by the senders will and desire. She/he can determine to give the message an expiration time, urgent message normally are not urgent any more after some time. She/he can determine that the message file be deleted upon successful retrieval of (all) recipient(s), of course with an optional notification to the sender if the message has not been retrieved by a due date. There is also the option to leave a message forever on the server, which will be the case of a public mailing list - the mail spool is the archive in this case. And finally imagine an advertising company allowing a fixed number of anonymous recipients pick up a message - for example a limited offer. I'm sure that somebody will come up with another deletion criteria soon!

The current system leaves a message for a client to read at his leisure. This system will require all clients to retrieve their mail within a given period. What happens when they go on an extended vacation?

You can configure a mail robot at the recipient's side, which collects all messages postet to it during vacation. You can configure a vacation program, which ask's each sender to expire their messages not until after the vacations. This gives the sender a chance to not post messages (which means storage costs) which have no value after this time.

Wouldn't there be a significant increase in non-delivery notifications by people who don't read spam or other messages sent to them?

The system can send notifications with a geometric back off time interval to the recipient. The message can simply expire, or can generate a notification to the sender upon "non-delivery" who decides then what steps have to be taken.

Advertisers (amongst them some spammers) aren't interested in people that don't read their mail, so they will post a message, notify once and again and then forget about it.

Real spammer, the ones that don't care about creating notification storms, will soon be identified and blacklisted.

Not knowing the content of a message will have it's drawbacks and may require more traffic on the Internet.

A business that has to accept email from customers will not know the difference between mail from a spammer and mail from a potential new client. How will this be handled?

When sending a message, some "comittment" about it's contents quality is made beforehand, like: if you are my buddy, you are on my whitelist and so every mail from you is trusted and welcome. If you are a quality advertiser, you comit to send only messages when a customer has explicitly requested it, and it will contain the restriction (products, máximum volume of mails, availability, price range) the customer has asked for. If the sender breaks the comitment, the recipient at least can revoke the permision to send notifications. A previous unknown sender can make no comitment and the recipient can proceed as she/he pleases - not allow notifications at all, allow any unknown sender to notify her/him, trust some "header" information, lookup the identity and corresponding web-site of the sender in the database before retrieving the message, check some black/white/or greylist if the sender is known to them and how she/he is qualified. etc. etc.

When you mean no recipients, do you mean `"undisclosed-recipients`"? Mailing lists frequently use Blind Carbon Copy all the time to hide one subscriber's email address from another.

There is no need to subscribe to a mailing list for reading. You just go there and look what's new. No notification is sent to anybody, no foreign address is transmitted in a post a certain reader fetches. The reader will reads the message when she/he has time to do it, not when it is posted.

However there are several variations to the mailing-list theme. Look up the docs and webpages.

There will always have to be a bounce notice of some sort to acknowledge a mis-spelled or discontinued email address.

If a user stops paying ISP's to receive notification messages for her/him, the global registry of the account will not yield a place where to notify and a "bounce" to the sender will be issued. It will not contain the message, as the message is yet on the senders server.

A user can even register with several ISP's to receive notifications, so one can work in diferent cities (like I do) and read mail "locally".

The ISP can give expiration or validity data of the registry and so a smooth change to a new provider can be pre-programmed.


Jorge.Lehner@gmx.net

Security and SpamTopCompatibility and issues with SMTPim2000 Concepts