Hacker News Reader

Comment by shrubble | Hacker News Reader

This method basically has not yet been used much, but it is in place.

The dangerous issue is that if a spam operator has 33% legit traffic, do you kill the spam operator and the good traffic with it, or what? Kill the traffic... innocent people harmed, or leave the traffic, spam continues.

View on HN

> if a spam operator has 33% legit traffic, do you kill the spam operator and the good traffic with it, or what?

Yes, we should with these long-time, serial offenders. Having legit traffic is just a fig leaf cover for them anyway. Any legit reseller clueless or negligent enough to accidentally stumble into business with these guys will switch providers as soon as their customer's calls stop connecting.

Also, no real telecom providers are routing meaningful amounts of traffic through these shady operations. Any legit traffic on their networks is mostly coming from fly-by-night, bottom-feeding telecom resellers in the same countries the spam calls originate from. Any retail customers of those resellers are probably paying ripoff prices for unreliable per-day or per-call service anyway. It's not people with normal pre-paid monthly service from any legit telco you've ever heard of.

Kill the spam operator.

Also end-users should get information from every hop, so we can block whatever we want with full information, client-side, uBlock-origin style.

You tell the operator to cut off the spammers, and if they don't do it in a reasonable time you cut off the operator.

Kill the traffic. The operator has to do a better job to stay in business.

The operator's business is not the concern here.

If a bank is comprised of 66% of their customers being narcos, the bank gets shut down.

You don’t get to facilitate in illegal shit and hide behind your legitimate customers. Likewise, if you are a customer of theirs and know they heavily transact with illegal services, it’s on you for getting blocked.

Naw they just get a slap on the wrists from the Feds and move on. No jail time even. Now if you and I laundered money to the Mexican drug cartels...

"too big to jail because they are too big to fail"

https://www.investopedia.com/stock-analysis/2013/investing-n...

Their legitimate customers are. If that means getting the operator to clean up their act, so much the better. If they won't, then at some point their legitimate customers will suffer.

And will probably have to jump to a more expensive provider who isn't subsidizing them with spammer revenue.

> The dangerous issue is that if a spam operator has 33% legit traffic, do you kill the spam operator and the good traffic with it, or what?

You are implying that "you" means the telco or FCC decides on behalf of "everyone." That is not the correct viewpoint. If the telcos are the common carrier, they dont get to decide. I am the customer; I get to decide. Problem solved. No additional regulation or debate is needed. This isn't hard.

If the spam calls have spoofed source numbers, the provider should be within their rights to refuse the traffic regardless of common carrier status.

I am the customer and I would love to see the data of which providers _originated_ each call that I'm about to answer. That would make it trivial to set rules about which ones don't even ring. But until I can have that data, I wish they'd just drop the obvious junk.

>I am the customer and I would love to see the data of which providers _originated_ each call that I'm about to answer.

The telcos have this information, but they only usually relay the CallerID (which is user-specified, ie "spoofable") to the end user. ANI, RPID, and now SHAKEN/STIR information which does identify the origin and origin carrier are simply not passed to end users to do anything with, or at least I have not been able to get them to do it despite having capable interfaces.

> I am the customer; I get to decide. Problem solved.

If only we actually had that ability. The best mechanism available to me is what I do: if I get a call from a number that isn't in my phone book, I don't answer it.

> If only we actually had that ability.

"We" do, if people would ask properly for it instead of trying instead to break the customer/common carrier contract.

The carriers withhold information from customers that is useful to determine the nature of traffic and whether it should be accepted or rejected. The amount of metadata that accompanies a modern phone call is substantial, and the carrier typically relays only ONE FIELD to the customer.

Customers should demand access to all call/circuit/packet metadata that is necessary or useful to implement their own traffic policy. To the average person, I can see why it might seem that the carriers appear to be ideally suited to police this problem, but the correct way for them to do it without violating their obligations as a common carrier is to empower their customers with the information and the tools to do it on their own.

Thus describes the incentive, and the onus, for the provider to prevent spam on their network.

If a vpn has 33% legit traffic do you ban every IP range of their service?

Yes

The Spamhaus approach (and one that is increasingly popular in telecom) is to say "f*** it, kill the provider." A lot of people disagree with this obviously.

Spamhaus doesn't block anything.

Spamhaus publishes information on network space which originates large volumes of spam.

Mail system operators use that information as they see fit.

<https://www.spamhaus.org/sbl/>

This is technically true but practically irrelevant. A Spamhaus listing is the Internet equivalent of a sanctions designation and the process to get removed is equally as Kafkaesque as OFAC's. Not to mention the fact that addresses in the [DROP list](https://www.spamhaus.org/drop/drop.txt) effectively can't even connect to the broader Internet because every important tier1/tier2 uses these lists and drops all their packets.

Reply on news.ycombinator.com