Live data from Hacker News

Sending newsletters should cost something

rmhsilva.com

31–40 of 70 posts

Re: Sending newsletters should cost something

#31
post #6

If you are planning a proof-of-work system for email, you need to address the issues in this paper: https://www.cl.cam.ac.uk/~rnc1/proofwork2.pdf And since blockchains are public, you need to address anonymity and communications tracking (what the spooks call metadata): public keys are pseudonyms and there has been a lot of successful research on decloaking “anonymized” data.

You just need a handshake where the receiver sets the level of work required. Now you can whitelist people. You can also provide modest levels of work for "more reputable" email addresses, and set the default high for everyone else.

Done.

Re: Sending newsletters should cost something

#32
post #23

Earlier quoted context omitted.

No. And I've heard proposals like this going back at least a couple of decades. The theory goes that if you make email expensive (at scale) charging even a trivial fee will make shotgun blasts cost prohibitive. While (very much in theory), a few cents won't deter the casual emailer. Though more likely, as with SMS at one point, people would find ways to route around the expensive pipe because bits are bits at the end…

>> While (very much in theory), a few cents won't deter the casual emailer. Better yet. Require proof of work where the receiver sets the challenge level. Now you can whitelist people by offering trivial challenges for them. If you really want to email me it's going to take 2 minutes of CPU time - this will be done by the new mail clients in the background, so individual personal cost is essentially zero but bulk mai…

Sounds like you're just increasing the demand for botnets.

Re: Sending newsletters should cost something

#33
post #23

Earlier quoted context omitted.

No. And I've heard proposals like this going back at least a couple of decades. The theory goes that if you make email expensive (at scale) charging even a trivial fee will make shotgun blasts cost prohibitive. While (very much in theory), a few cents won't deter the casual emailer. Though more likely, as with SMS at one point, people would find ways to route around the expensive pipe because bits are bits at the end…

>> While (very much in theory), a few cents won't deter the casual emailer. Better yet. Require proof of work where the receiver sets the challenge level. Now you can whitelist people by offering trivial challenges for them. If you really want to email me it's going to take 2 minutes of CPU time - this will be done by the new mail clients in the background, so individual personal cost is essentially zero but bulk mai…

Don't take this the wrong way but... Oh great. Just as the environmental tire fire that is bitcoin and the other crypto-currency bros is winding down, let's replace it with email.

Re: Sending newsletters should cost something

#34
post #23

Earlier quoted context omitted.

No. And I've heard proposals like this going back at least a couple of decades. The theory goes that if you make email expensive (at scale) charging even a trivial fee will make shotgun blasts cost prohibitive. While (very much in theory), a few cents won't deter the casual emailer. Though more likely, as with SMS at one point, people would find ways to route around the expensive pipe because bits are bits at the end…

>> While (very much in theory), a few cents won't deter the casual emailer. Better yet. Require proof of work where the receiver sets the challenge level. Now you can whitelist people by offering trivial challenges for them. If you really want to email me it's going to take 2 minutes of CPU time - this will be done by the new mail clients in the background, so individual personal cost is essentially zero but bulk mai…

That's where we got the idea of proof-of-work in the 90s [1]. The problem is that this makes legitimate use cases like mailing lists very expensive, while providing little protection against spam from botnets.

[1] https://en.wikipedia.org/wiki/Hashcash

Re: Sending newsletters should cost something

#35
post #21

Earlier quoted context omitted.

The dig at capitalism is a bit puzzling since this may be the perfect setting to demonstrate the principles of capitalism. Let them deploy their capital and see how market responds. Their theory is that "sending an e-mail should cost something". Maybe that cost is $2.30 per message, and maybe its $0.0000000023 per message. Either way its their capital at stake.

right and that... sucks? and is bad? A lot of the passion/hobby newsletters I receive would not exist if the person writing them had to pay to send the emails. Adding a cost to distributing information does not increase the quality of distributed information, it just increases the percentage of information distributed that is trying to make the sender money. Maybe I'm misunderstanding the article. Are they talking on…

Any email. But the receiver could refund/credit/whitelist certain senders; there's no reason this mechanism would have outsized impact on hobby lists other than the subscriber would have to somehow mark them as "safe".

In theory, this whole approach should work pretty well. I doubt the practical hurdles are surmountable, however.

The "prove-work-to-deliver" approach seems to have all the same benefits without a complicated, slow, and privacy-impacting ledger.

Re: Sending newsletters should cost something

#36
post #8

It seems like the point of this is to reduce junk mail by forcing a financial cost to sending mail thereby causing the sender to really consider the content of their message before sending it—except there is a real cost associated with sending traditional mail and there is still a huge amount of garbage traditional mail sent out every day. Maybe I’m missing the point, though?

But mail is always priced at a fixed rate. What if your inbox worked on a bonding curve, such that as it filled up it became more expensive to send to (except for whitelisted addresses you approve)? What if your inbox required that the sender owned a particular NFT (proof of membership in a community)? I mean there are infinite rules to play with in that sense. That's what excites me about crypto.

> What if your inbox worked on a bonding curve, such that as it filled up it became more expensive to send to (except for whitelisted addresses you approve)?

When my inbox fills up with spam I can't get legitimate mail because my inbox is full wouldn't essentially the same thing happen? I dunno why not just use the whitelist and be done?

Re: Sending newsletters should cost something

#37

Not sure if its a good idea, Email servers get hacked daily, If you’re auto-sending payment to receiver , now there is an incentive for hackers , to spam themselves with your inbox and you pay out of your wallet for that. Now, yes one could say, that’s no problem just have a multi-wallet approval method , so unless the second one approves, it wont. But now that makes it a bit more complex, especially for newsletters…

Well in the mail you have to pay for the stamp. It's about a dollar. You could charge 10 cents...in fact, that's not a bad idea for a cryptocurrency.

Because the existing crypto currencies support 10c transactions if they grow to any kind of audience!

Re: Sending newsletters should cost something

#38
post #28

Bill Gates proposed sending email costs money. I remember this list from the 00's which included all kind of failed proposals to combat spam. Newsletters are no issue in EU though. They need to provide a clear and easy way to unsubscribe or they break the law.

I still get newsletters I never signed up for tho. They are breaking the law but it's profitable for them to do so.

Re: Sending newsletters should cost something

#39
post #28

Bill Gates proposed sending email costs money. I remember this list from the 00's which included all kind of failed proposals to combat spam. Newsletters are no issue in EU though. They need to provide a clear and easy way to unsubscribe or they break the law.

You don’t get spam from non eu countries?

Re: Sending newsletters should cost something

#40
post #6

If you are planning a proof-of-work system for email, you need to address the issues in this paper: https://www.cl.cam.ac.uk/~rnc1/proofwork2.pdf And since blockchains are public, you need to address anonymity and communications tracking (what the spooks call metadata): public keys are pseudonyms and there has been a lot of successful research on decloaking “anonymized” data.

This is great, thanks for the link. Love how it was written in 2004.
Post reply on HN