I remember hearing someone complain on HN of their site getting blocked because it shared an IP with an illegal soccer livestream. I can’t imagine they’re doing this to IP blocks owned by CDNs like Fastly, CloudFlare, or CloudFront though. Or are they? Does this regularly break most of the internet for UK customers?
In the case that a blocked site resolved to a Cloudflare IP, it would likely be kicked off of Cloudflare, or geo-blocked for UK users (by Cloudflare).
https://www.ispreview.co.uk/index.php/2025/07/cloudflare-blo...
So why ControlD? Because I don't want to run my own piHole, basically. They maintain ad block lists that you can edit as you see fit to add things or relax things that may cause issues(which you can't do easily with public ad blocking dns servers).
Why ControlD then and not NextDNS? First, because their support was awesome when I had an issue. AFAICT it was the founder actually emailing me back and forth, and it ended up being my ISP's fault, but I only knew that based on research provided to me by support. Secondly, I got a good deal on a 5-year subscription at one point.
Happy to answer any questions, not affiliated but a fan of the service.
Trying to downgrade or strip extensions from any TLS 1.3 connection will simply break the connection.
0: https://community.fortinet.com/t5/FortiGate/Technical-Tip-Ho...
1: https://docs.broadcom.com/doc/symantec-ech-whitepaper (see page 8)
The DNS filter setting on the FortiGate analyzes the DoH traffic and strips out the ECH parameters sent by the DNS server in the DoH response. If the client does not receive those parameters, it cannot encrypt the inner SNI, so it will send it in clear text.
So basically they mess with DoH ECH config and trigger fallback behavior in the clients. I don't think any browsers do this yet but I think this loophole is not gonna last.(My phone screen is too small to look through the RFCs right now.)
Edit: actually totally possible but you need build quantum computer with sufficient cubits first =)
(I remember using quantum algorithms to find prime factors 25 years or more ago, using the Quantum::Suppositions Perl module.)
Yes, any given domain name (or as non-technical people would think about it, "website" -- any website) could be "blocked" (re-routed to a non-functioning IP, claimed to not exist, other DNS error or malfunction, ?, ???) at any level of DNS (ISP, Local, Regional, Country, ?, ???)
A question your statement so excellently potentially suggests, is:
What's the true extent of the block?
Is it merely a DNS failure -- or are inbound/outbound packets to an IP address actively suppressed and/or modified to prevent TCP/IP connections? (i.e., The Great Firewall Of China, etc.)
You have "Bad Faith Actors" (let's not call them "governments", "countries", "nation states" or even "deep states" -- those terms are so 2024-ish, and as I write this, it's almost 2026! :-) )
Observation: Let's suppose a "Bad Faith Actor" (local or nationwide, foreign or domestic) attempts to block a website. They can accomplish this in one of 3 ways:
1) DNS Block
2) TCP/IP Block, i.e., block TCP/IP4/6 address(es), address ranges, etc.
3) Combination of 1 and 2.
#3 is what would be used if a "Bad Faith Actor" absolutely had to block the "offending" website, no ifs ands or buts!
But... unfortunately for them (and fortunately for us "wee folk"! :-) ), each of these types of blocks comes with problems, problems for them, which I shall heretofore enumerate!
From the perspective of a "Bad Faith Actor":
1) DNS Block -- a mere DNS block of a single domain name is great for granularity that is, it targets that domain name and that domain name alone, and something like this works great when a given company's products and services are directly tied to their website as their brand name (i.e., google.com being blocked in China), but it doesn't work well for fly-by-night websites -- that's because a new domain name pointing to the old IP address can simply be registered...
2) TCP/IP Address / Address Range Block -- The problem with this approach is that while it is more thorough than a simple DNS block, it may also (illegally and unlawfully, I might add!) block legitimate other users, websites and services and businesses which share the same IP or IP address range!
Think about it like this... A long time ago, all of the mail traffic for AOL (America Online), about 600,000 users or so, was coming from a single IP address. Block that IP address, and yes, you've stopped spam from the single user who is annoying you, but you've also (equal-and-oppositely!) blocked 599,999 legitimate users!
So "Bad Faith Actors" -- are "damned if they use the first method, and really damned if they use the second or third methods"... the first method is easily circumventable for non-brand name dependent websites and web services, while the second and third methods risk causing harm to legitimate users, sometimes huge amounts of them... which should be illegal and unlawful by any country's legal standards...
In other words, Countries should read their own sets of laws(!) before contemplating Internet blocks on their Citizens... :-) And not just one country either, all of them!!! :-)
Anyway, an excellent point!
Very thought stimulating -- as you can see by my ramblings! :-)
It looks like this is possible with Chrome-based browsers using a command line flag (--host-resolver-rules) or in Firefox settings. Is there a better way?
Quad 1/8/9 isn't optimal alternative (too much centralization if everyone uses those by default) but running your own is easy.
edit: Thinking of it, anyone knows if it's possible to use that for OS-wide DNS resolves, not just for the browser?