What worries me rather more than people bending the SMTP standards to make them meet their needs is that the idea of open communications is being lost.
Open source and open specifications are fine. But they are much less important than open infrastructures. What matters to me with mail is that it is the one communications network everyone I know is on. I don't have that for messaging (30+ competing closed networks) and I don't have that for voice any more. The telephone system is becoming more and more useless every day because it has no technical controls to block spam.
We faced a similar issue with sharing files over the Internet in the early 90s. FTP worked, kinda but the specification was really not fit for purpose. It is layered on Telnet for a start and it assumes a particular UI model which is a really bad one. But FTP was the way everyone did things so there was absolutely no point in ever thinking about replacing it with something better.
Until some people did - HTTP
The Web was not simply a replacement for FTP though, it was a whole different application that just happened to subsume enough FTP functionality to replace it. And it did that in a very particular way, by introducing an interoperability layer over the old and the new so that you could use an FTP server to serve the Web if you wanted to.
I think we could certainly hive off many of the functions SMTP doesn't do very well and develop a protocol that supports those functions decently because it is designed to support them, they are not afterthoughts. Mailing lists would be one of those functions.
SMTP is still a decent protocol but what people have done with it has created an infrastructure that makes fixing it pretty much impossible.
As far as I am aware, nobody other than myself is working on a mailing list protocol that is end-to-end secure right now. We could try to fit those features into the old infrastructure but why do that when everyone will need new client and server support?
