4 ms·
And who decides how much that should be? The recipient. Right now I can't stop you from sending me spam if you know my email address. Spam filters work somewha
by byoung2 5y ago
And who decides how much that should be?
The recipient. Right now I can't stop you from sending me spam if you know my email address. Spam filters work somewhat, but I should be able to say: block everyone except these 100 people on my contact list or transactional emails I have opted into. Everyone else needs to pay me $0.50 before the subject line even hits my inbox. And maybe I can even set up tiers for times of the day or days of the week, or charge more on days when I get a lot of email.
- dane-pgp 5y agoSuch a system would indeed solve the economic conundrum of how to set the price, you're right. It's worth pointing out, though, that having per-recipient prices would require a change to the protocols to allow the sender to know in advance how much it will cost for them to have their email delivered (and thus whether it is worth them writing the email in the first place), which breaks the "store and forward" architecture of email. If we're going to break the architecture of email like that, in a backwards incompatible way, we might as well just switch everyone over to Matrix, or try some other protocol changes which are just as invasive but don't have the extra complication of money changing hands. For example, recipients could indicate to a prospective sender whether they are on their spam whitelist or not, and if not, they could indicate what sort of email they would accept, such as "20 UTF-8 characters only", which is enough to receive a password reset code or friend request but not enough to explain why a Nigerian prince needs your help.
- byoung2 5y agoIt's worth pointing out, though, that having per-recipient prices would require a change to the protocols to allow the sender to know in advance how much it will cost for them to have their email delivered (and thus whether it is worth them writing the email in the first place), which breaks the "store and forward" architecture of email. Not necessarily...people are already writing emails on spec because there is no guarantee that I will read it. It's just that they send it to thousands of people hoping that the fraction who read it will make it worth the time spent writing the email. Without changing the architecture of email you could reject emails (e.g. 550: rejected, unsolicited emails require $0.50 payment). Or you could passively ignore emails that don't meet criteria (you could piggyback on the read receipt functionality to provide confirmation). All of this is backwards compatible with existing email