i wonder, why do we have so many of that stupid mail protocols, why just not to get rid of it and create a new mail server with simpler and more logical configurations similar to nginx and do all the mailing via http(s) protocol, that way the servers will have much less software, libs, dependencies, etc. I know - a crazy thought, but i think it is possible, no ?
Generally things are the way they are for history. You can look at the Linux filesystem as an excellent example of “because REASONS!” The largest reason why Mail services are foundationally the same as they were 40 years ago is because an entire ecosystem of enhancements, products, and practices exist around them. Changing the underlying protocol used for Mail would mean that spam and content filters, security features (like DNS enhancements) and client applications would all need to updated to whatever these protocol or server changes would be. That’s alone huge challenge, but you also have to convince all the implementations to adopt these changes, right? Not only every Linux but also folks like Microsoft Exchange, otherwise you run the risk of your protocol only being used by a small subset of the users. Worse, how terrible would it be if your users on your protocol could email each other, but not anyone who uses a different email system?
This is the best!
i wonder, why do we have so many of that stupid mail protocols, why just not to get rid of it and create a new mail server with simpler and more logical configurations similar to nginx and do all the mailing via http(s) protocol, that way the servers will have much less software, libs, dependencies, etc. I know - a crazy thought, but i think it is possible, no ?
Generally things are the way they are for history. You can look at the Linux filesystem as an excellent example of “because REASONS!”
The largest reason why Mail services are foundationally the same as they were 40 years ago is because an entire ecosystem of enhancements, products, and practices exist around them. Changing the underlying protocol used for Mail would mean that spam and content filters, security features (like DNS enhancements) and client applications would all need to updated to whatever these protocol or server changes would be. That’s alone huge challenge, but you also have to convince all the implementations to adopt these changes, right? Not only every Linux but also folks like Microsoft Exchange, otherwise you run the risk of your protocol only being used by a small subset of the users. Worse, how terrible would it be if your users on your protocol could email each other, but not anyone who uses a different email system?