The Basecamp team has published a new application of their creation, Hey.com. It purports to fix many observed problems of email. In my opinion after watching the introduction video of its features, it does quite a good job at that. What especially attracts my attention is the feature where you screen each sender.
The screening feature takes the emails from a sender first seen by the system to a standby area, and you can decide whether those emails and later emails should go to your inbox, your newsletter tab, your paper trail, or be rejected.
This feature immediately reminds me of the very incisive criticisms of email by Jonathan de Boyne Pollard (jdebp), about non-solutions to unsolicited bulk emails, colloquially spam emails. Those non-solutions are still as prominent as when the article was first published.
To summarize, the problem of unsolicited bulk email is: 1) unsolicited; 2) bulk. Yet, the prominent solutions, such as machine learning based filter, trusted email server list, etc., do not target the actual problems. Because those solutions do not target the real problems, they introduce their own problems.
For example, machine learning based filters can put your important emails into the spam folder erroneously. A trusted email server list favors incumbent service providers and cuts off small email operators and personal servers, going against the open spirit of the internet.
The solution Jonathan proposed was a new open protocol, called Internet Mail 2000 (IM2000), first conceived by Daniel J. Bernstein (djb). The key insight is that unsolicited bulk emails are a problem because the bulk burdens the receivers. The idea is to turn the Simple Mail Transfer Protocol (SMTP) around, and burden the sender with storing the outgoing emails. When senders want to send a message, they must first solicit. Only when the receivers are willing will they actually retrieve the messages. But with SMTP already dominant, IM2000 did not receive much attention outside of djb enthusiasts.
Hey.com does not aim to upend SMTP, but only tackles the first aspect of the problem, that spam emails are unsolicited. The screening simulates a solicitation protocol. Only if you agree to receive emails from a sender will you see those emails in the inbox. 20 years after IM2000 was first conceived, something similar has finally made it to the market.
That said, it is not clear whether Hey.com may reject emails from indie operators.
djb also predicted the difficulties of IPv6 adoption, against the committee's optimism. The history has proved his predictions correct.
Incidentally, IM2000 is quite similar to the instant messaging (IM) systems nowadays.
Another similar thing is push notifications.
I actually knew about Hey.com from Basecamp YouTube channel, before I knew about their debacle with App Store.
Oldest comments (0)