I get SPAM from companies which appears to come from different people, but is evidently all coming from a few particular companies. Clearly they are spoofing their IP address and even reviewing the message headers does not reveal a true source point. As an ‘outsider’ I can’t understand why this is such a problem to deal with.
My understanding of the system is that an email gets split up into small packets which can travel by any route to the final recipient ISP. The recipient server reassembles the message from the packets and delivers it to the account holder. If a packet gets lost on the way the recipient server would ask the sender for a re-transmit. Now if the sender’s IP address has been spoofed then this re-transmit cannot get actioned and the message would not get completed. Presumably the message gets deleted after some suitable time-out period.
The most obvious solution is then for the recipient ISP to deliberately throw away a packet from each incoming message in order to force a re-transmit. This gives a low-level automatic verification that the sending address is valid.
The fact that this obvious remedy is not being used suggest that my understanding of the low level TCP/IP system is inadequate. Perhaps somebody would be so kind as to point out what I am missing.
My understanding of the system is that an email gets split up into small packets which can travel by any route to the final recipient ISP. The recipient server reassembles the message from the packets and delivers it to the account holder. If a packet gets lost on the way the recipient server would ask the sender for a re-transmit. Now if the sender’s IP address has been spoofed then this re-transmit cannot get actioned and the message would not get completed. Presumably the message gets deleted after some suitable time-out period.
The most obvious solution is then for the recipient ISP to deliberately throw away a packet from each incoming message in order to force a re-transmit. This gives a low-level automatic verification that the sending address is valid.
The fact that this obvious remedy is not being used suggest that my understanding of the low level TCP/IP system is inadequate. Perhaps somebody would be so kind as to point out what I am missing.
