Hacker News Reader

Comment by pmlnr | Hacker News Reader

Don't try to "replace" email, ever. Build, name it is a X, just not as "to replace email", because by doing that, it's DoA.

View on HN

Yup!

There are many, many protocols for exchanging messages, chats, files, etc. (basically, what e-mail does), they all come, get some traction, last for two years, then die (except maybe irc, but except for a few nerds, it's hard to get anyone on there anymore). Some get killed by google, some just get replaced by "the next new thing"... but e-mail still lives, and works, and does what is needed.

It doesn't do what's needed. It provides a channel for spam, insecure delivery etc. It's too broken. We do really need a new email.

Yes, but the point is, don’t sell it that way.

The fact that it has problems is entirely tangential to the fact that it does what's needed. What's needed is "I send a thing to a person and it soon-ish shows up on their end", with the bonus of knowing that someone would have to expend significant effort to intercept it. This has been true since the invention of writing and very few scenarios needed more. Anything on top of that (spam rejection, steong verification and encryption, etc.) is nice to have, but clearly not critical, because otherwise we still wouldn't all be relying on paper mail, email and sms for even the most critical of things.

So yes, we really do need something better than email (and we have plenty of options), but it doesn't have to (nor will it any time soon) replace email.

None of what you wrote speaks to not doing what is needed.

We do not need new email. We need a way to authorize messages sent and also a way to enburden senders for the content they send. In pre-internet era UUCP was widely used to deliver messages, UUCP providers used to charged their customers for each message they sent which made spam broadcasting quite costly. What I see is a kind of distributed message transportation system that takes currency from sender and shares a fraction of it with recipient. I.e. sender always pays for sending their stuff, recipient gets paid for interruption. SMS messaging is very close to this.

Real transport level security will be possible with MTA-STS (rfc8461). Use DNSSEC, enforce TLS, and verify server certificates. Also I haven't had much trouble with spam recently (spam filter works fine).

The point of MTA-STS is not having to use DNSSEC, which nobody does anyways.

> except maybe irc, but except for a few nerds, it's hard to get anyone on there anymore

Well IRC isn't really meant to be used for persistent communication anyway

This!

People do not want email replaced. They may want a new thing enough to use it in huge numbers.

And then that thing has to endure long enough, be open enough to see even a modest level of trust email has at this point.

Try to stop me :-)

Also addressed here: https://news.ycombinator.com/item?id=25804869#25807379

(I'm the author.)

Reply on news.ycombinator.com