Hacker News new | past | comments | ask | show | jobs | submit login

At this point the access points and the global data and surveillance system is too complicated and interconnected to dis-intermediate the access point and the network itself

That is to say you cannot even connect to IP addresses unless you’re already part of the global whitelist for APs

Try yourself. Get a VPN and then connect from certain regions and to won’t get very far and you’ll be slooooow

You’ll say, oh, that’s because bad actors and are therefore blacklisted - and now you’re back to the same problem: their resource is being limited by some social structure that is by definition constrained

You cannot differentiate the broader system from the access point, so long as the system controls which access point access then it doesn’t matter what you do

The only solution is to build an entirely new Internet from scratch, not controlled by the people currently google, amazon, cloud flare, respective ISPs, Netflix, etc…

The Internet is 99% controlled by a small set of corporations that most people haven’t heard of

Just try launching a competitive DNS service and see how far you get




> The Internet is 99% controlled by a small set of corporations that most people haven’t heard of

I would love to know more if you have time to elaborate.


See this post that talks about how hosting email yourself doesn't really work anymore, despite it being an open internet standard -- https://news.ycombinator.com/item?id=32715437

Not sure about people never hearing about them, but the internet is no longer a network of sparsely distributed connected machines hosting and consuming content as it was originally designed. The content is all hosted on very small set of networks controlled by just a few companies -- Amazon AWS, Google, MS Azure, etc. The consumers access the internet through clients that are also controlled by these companies, and so it's easy for them to exclude other networks and hosts even though the network specification is technically "open".


personal email failed because email is a flawed system that was initially designed to work in a trusted environment free of bad actors

I'm as worried about massive platforms playing gatekeepers as much as the next guy, but it's also fairly clear to see that any service that isn't constantly moderated will be overcome with a flood of malicious crap. it practically guarantees closed platforms will proliferate


I generally agree with you, but there’s a difference between a closed platform, and a platform that prevents you from using any other platforms.


Every platform that has an controlled access point, which to be clear is the entirety of the public DNS system, is by default a closed platform

There do not exist open HTTPS gateways for all of the major internet services such that anyone can connect - They’re *all* mediated by some access control gatekeeper, it’s like one of the core functions of the current Internet routing system is to not allow for infinite anonymous peers like you would see in TOR or other anonymized routing services

You can say that this should be the case or not but it’s just the facts

It’s precisely the same argument as: “We have plenty of water fountains, As long as you look white then you can use them all you want.”


https://en.wikipedia.org/wiki/List_of_managed_DNS_providers

I'd venture that even most engineers wouldn't know these by name, which is my point. Not that it's secret, it's just that obscure and niche engineering that holds the internet together is barely understood by anyone not directly touching it (like most critical infrastructure)

For example this wasn't even a big story within tech: https://arstechnica.com/security/2024/05/dns-glitch-that-thr...


It all depends what your privacy threat model is.

There are very few organisations capable of identifying the human user behind a Mullvad connection with a VM.

If you're worried about local law enforcement or advertisers it's probably enough.


I'm more worried about future shakeouts and the end of anonymity, one ASN at a time.

We're currently relying on AS' who 'play ball' with anonymous users, but that's not a requirement for AS operators by any means.


Isp often sell data, ip can be deanonymized by buying data from different data brokers and triangulating the ip.


> That is to say you cannot even connect to IP addresses unless you’re already part of the global whitelist for APs

> Try yourself. Get a VPN and then connect from certain regions and to won’t get very far and you’ll be slooooow

Connecting to a far away VPN doesn’t prove anything about a “global whitelist”

VPN addresses are commonly rate limited and block listed because they’re sources of abusive traffic and therefore trigger all of the common defense mechanisms.

Anyone who has run a forum or other service long enough knows that the spammers and the people trying to evade bans love their VPNs.

VPN services are also oversubscribed, leading to poor performance. Connecting to far away locations also causes throughput problems, especially if the country has poor internet infrastructure. Your connection has to round trip into and back out of that country to get to many services, meaning performance can be very poor.

> Just try launching a competitive DNS service and see how far you get

You’ll probably get as far as realizing that it’s an extraordinarily expensive venture to operate with no possibility of income, at which point you’d shut it down unless your hobby is lighting money on fire. There isn’t a conspiracy in this.


That's what they said, yes. Only certain access points are whitelisted to access things. If you aren't using one then you're blacklisted. VPNs aren't one. You claim there isn't a global whitelist and then proceed to explain a global whitelist.


> Only certain access points are whitelisted to access things. If you aren't using one then you're blacklisted.

No, this is not how these terms work. Blacklist and whitelist are not just words for opposite sides of a partitioned set. Both blacklists and whitelists are explicitly enumerated lists. If I blacklist a single thing, I have not implicitly created a whitelist containing everything else in the universe. Establishing that VPNs are often blacklisted is not - at all! - the same thing as establishing the existence of a "global whitelist".


These are usually implemented as actual whitelists. Someone goes through each AS, and decides whether it should have unrestricted access or not. Verizon gets unlimited access, because it mostly provides service to end users. Hetzner doesn't, because it mostly has servers. There are companies that enumerate all networks and tell you if they are user-mostly or server-mostly networks, so you can block all but the user-mostly networks.


I’m glad we agree so throughly though I’m not sure what you’re referencing with a conspiracy




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: