Earn.com will ruin email if it succeeds (2018)

Earn wasn't making anyone use their service.

I'm not sure exactly how earn.com worked, but when I implemented this for my own email, all such emails would go to a "quarantine" folder until the sender responded to the autogenerated email.

So yes, the sender's email did get to me. Thankfully, I would be blissfully unaware of it until the sender confirms they're human. The sender needs to do it only once. There's a great chance that if the sender is not willing to go through that hoop, that they are not someone whose email I want to read.

This isn't "killing" the open nature of email. As designed, the email was delivered to me. I just added a workflow to it.

> if you had someone's email you could be sure the message directly reaches him

You don’t get it. I don’t want your message to directly reach me. I want you to kindly f-off and don’t waste my time with endless offers of shit that I don’t need. You think your message is worth it? Pay me for my time to look at it then. I think it is a great idea. I wish someone did it for phones as well. “p2t2p won’t answer your call as you’re not in the list of approved callers. Press hash sign to make a paid call for a small fee of 1 AUD a minute if you really think your phone call is worth it."

Doesn't work at all. Most spam calls are from overseas using throw away voip accounts. Do not call just gets ignored.

Slightly different issue. In my experience, the problem tends to be from robodialers operating on behalf of scammers or fly-by-night marketing agencies. As the US phone system is currently operated, it's trivial for these kinds of actors to spoof originating phone numbers and hit every valid US phone number in sequence.

No. It is not a block list. It only applies to US based organizations who decide to honor it. It does nothing to prevent international callers from using (or forging) a US based number to very urgently talk to you about your lapsed car warranty or your Microsoft Windows error.

No, because most of the spam calls in US are actually criminal fraud, eg: extended warranty. Australia carved out enough exemptions for their DNC registry to make it less useful. Political parties can spam you via phone whenever they like.

It does not work because it is trivial to fake the source number.

A popular form of spam pretends to call from a phone number very similar to yours (they use the same leading 6 digits) to make it look like a neighbor is calling.

Hopefully the STIR/SHAKEN implementation will widen and this will be put to bed.

> Does it not work?

It's ineffective against true spam calls, as it only binds law-abiding callers with a presence in the US.

The UK also has a series of anti-spam and anti-solicitation measures (see this by ofcom, our communications regulator: https://www.ofcom.org.uk/phones-telecoms-and-internet/advice...). They are very effective, many people are educated on reporting so even out of jurisdiction offenders (e.g windows has an issue, or the more recent parcel fee texts) are shut down pretty quickly. We are currently passing (or have passed) some measures to make the service provider liable for out of jurisdiction calls and texts.

The major issue we are not addressing is ‘legal scams’. Charities harassing old people, tradesmen faking issues to the uneducated - that kind of thing.

I haven't gotten a spam call in years that wasn't from a spoofed number.

Pick it up, and it's "We've been trying to reach you about your car's extended warranty"

Miss it and call back later, because it's a local area code? You get some very confused, usually elderly, person who wants to know why you called them.

It might as well be a data source of verified numbers. The FCC is totally useless when it came to enforcement

The recepient configures an auto reply with “No thank you” just keeps receiving money?

Presumably, the sender pays unless you respond.

We could probably invent an IMAP extension for that. Once Dovecot supports it, plenty of open source webmail solutions can then also support it.

> At least this way you know they didn't receive it and if you really did care about them specifically you have recourse.

That was the point I mentioned in another comment: They are notified and if they can't bother doing something tiny for me to get the email, chances are they're not vested in contacting me.

Because:

1. Often takes more than 5 seconds.

2. I often continue to receive newsletters after unsubscribing (not just for a few months, but forever - the unsubscribe feature doesn't work).

3. Why spend 5 seconds for each newsletter when I can use a solution that doesn't require any of my time?

4. This method has the benefit of killing spam pretty much instantly - not just newsletters. Why would I want to hunt for an unsubscribe link when I can kill two categories of problems with less effort than killing a few individual emails?

I don't sign up, usually. My email gets onto lists without my explicit permission when I sign up for services/buy products and then I start getting emails.

It takes a lot of effort to maintain an email these days. Unsubscribing often doesn't work or is temporary.

I can go on and on. Legitimate newsletters that I may sign up for get lost in all of that, and stay around long after I care to read them anymore.

> You are always in control of whether you want to read an email or not.

Demonstrably false. You cannot ignore a category of information without first without interacting with it - as you must first establish that it _is part of that category_. It's impossible to know a spam message that got thru automated filters is spam or not until I read at least the subject line.

Nope. The open protocol email was built on would still work, just some people would now use a service that blocks newsletters and other annoying spam. Only his business model would be effected, the rest of could merilly go on without consequence of Earn.com's success.

That's because postal revenue from spammers helps defray the cost of sending all the regular mail.

As someone who runs a newsletter and is active in the email space (and sees the other side of what happens when a message is sent to thousands of people), I think my concern with this approach is that it comes off as passive-aggressive against everyone, good-faith or not, especially in cases where someone signed up for an email in a double opt-out format where they had to verify that they wanted the messages in the first place. And during this period, if I remember right, a lot of people were tweeting their email accounts in an effort to encourage Earn.com as a revenue stream. To my eyes, it felt like it didn’t solve a problem, but instead brought gatekeeping to email and was not being used for its intended purpose.

I think the problem I have with what I wrote three years ago is that my emotion around this was way over the top, and as a result my legitimate concerns around the model (and how it could damage honest use cases for sending messages, such as newsletters sent in a one-to-many format, or journalists emailing potential sources), got overshadowed by hyperbole. I could imagine hundreds of people doing this on a single list, overwhelming a single user or organization. I would prefer to see an approach like this managed at the standards level rather than from a startup, because I could too easily see it misused.

I have a lot of complex thoughts around this model, because it’s gatekeepy and unworkable at scale, but this clearly whipped-off post didn’t properly convey them.

Not to you, maybe. Good thing you're free to keep your own email. Clearly getting your emails scanned isn't a issue most people care about or Gmail wouldn't be the defacto standard.

Nothing new under the sun, I guess :P

That's interesting, though. The attack would require the collusion of many parties. That's very feasible if you consider how easy it is to create email addresses today.

One could imagine the system requires a one-time registration fee orders of magnitude greater than the spam penalty. That wouldn't be a solution on it's own, but pair that with delays in applying penalties and an arbitrage system and I think we're looking at an increasingly stable system but with ballooning complexity.

Don't know if it's practical, and as you've mentioned I'm sure many people have gone down that road, but I'll be daydreaming about it for the next little bit :)

not necessarily. Even in my prototype sender pays email environment (camram/2 penny blue), the cost of the stamp was pushed on the message originator i.e. the email client. The mail server delivering the mail to another server on the Internet doesn't need to take that message if it doesn't have postage.

There are a few other ideas the project explored including using captive zombie nets generating postage for a company's email using spare cycles on desktops. Truth be told, whole distributed postage stamp generating model looked a hell of a lot like bitcoin mining.

Email and instant messaging aren't in competition they fulfill different communication roles and while each can be used as the other (there isn't a deep fundamental difference) each is habitually used in a very different manner. Asynchronous notifications about large tasks shouldn't be done in a medium where that notification is possibly going to scroll off of a screen in a few hours[1] (maybe I'm busy and want to review that long document sometime next week) - while low friction direct communication becomes sloggy through emails - nobody I've ever met would ever email AFK before going off to lunch. And, specifically, I wouldn't want them to mass-email the whole company with ever status change - being able to see availability at a glance makes communication smoother.

Companies should use both.

1. Message reminders in most clients can sorta do this but I still prefer emails.