>I re-read the customer message and I think I understand the issue better. Let me explain to you and Borislav to see if there is a solution:
>
>The customer - for security reasons - restricts the use of SMTP. Therefore, they control access on the Exchange server via a list of authorized IP addresses. Therefore, they assign a bunch of static IP addresses for PCs using my application and enter those IPs in the list of authorized (on the Exchange server). Therefore, in their case, if a PC changes the NIC or PC itself gets replaced, the user has an issue emailing. And we have hit this problem a couple of times already. So they want to know if my application can send email from the server and not from the client PC. I suppose the SMTP will be the same but initiated on the server instead of on the client PC.
>
>Now, does it make sense and is it possible to do what they want?

Perhaps the user-side of your application might save pending e-mail requests to some kind of DBF or other structure, and a separate process on the server regularly checks for unprocessed (unsent) records, i.e., e-mails.