Great idea, and good for casual blocking, but I'm almost moving to an "allow list" mindset. This solution would probably work better for that, I wonder if the good parts of the internet would fit into an 140k list.
Since it's not ready, it'll be best if I point you to the sources I've aggregated from:
https://github.com/hagezi/dns-blocklists
https://github.com/StevenBlack/hosts
https://github.com/jerryn70/GoodbyeAds
Link to my in-progress list file hosting: https://lists.uninvitedactivity.com/DNS/
00_Tiers directory:
I've aggregated the lists into four tiers: Base, Recommended, Aggressive, and Paranoid. And I've got two sets of these, one that includes Newly Registered Domains (Full directory), and one that doesn't include NRDs (NoNRD directory, because NRDs are ... heavy: Paranoid list with NRDs: 9.1 million records, without NRDs: 3.7 million records).
11_Allow directory:
A bunch of allow lists sourced from: https://discourse.pi-hole.net/t/commonly-whitelisted-domains.... I also use this list for an Allow list: https://github.com/anudeepND/whitelist
21_SpecificTopics directory:
Aggregation of lists under the topic that matches the file name. Fake News list needs to be updated as it contains a _way_ overbroad list that I need to remove from the aggregation.
For almost two decades I have been using an "allowlist", in the beginning it was via authoritative DNS. When proxy software began supporting mapping the DNS data into memory, I eliminates the need to use DNS
I have enough historical DNS data to assess own needs and it's much smaller than 140k
The popular blocklists can be quite strange if one takes the time to peruse them
They include what appear to be some very "bad parts of internet (www)" that I doubt most users would ever visit. There is more in these files than only common ad and tracking domains. As such one could call them "bloated"
It begs the question of how these blocklist maintainers even know about these bad parts
IMHO, the greatest exposure to ubiquitous ads, tracking and telemetry is through use of third party resolvers (in real-time, immediately preceding associated HTTP requests), where some process on the user's computer can potentially look up _any_ domain
Pretending that one can anticipate every possible undesired lookup, e.g., via a "blocklist", is IMHO a high risk approach
But it's popular. To each his own
Every user is different. Relying on a "blocklist" that attempts to anticipate every possible undesired domain is not for me. By contrast, the "allowlist" approach has served me reliably for many years. It sometimes suprises me how well it continues to work year after year
Philosophically, IMO, the contrast between "allowlist versus blocklist" is the difference between (a) "going after what one wants" from the www versus (b) "avoiding what one doesn't want"
The quantity of (a) can be relatively small and manageable. But (b) only grows larger over time, a tsunami full of garbage
I've always thought "ad blocking" was a misnomer. Instead of "blocking" what we as users want to do is "not send requests" to ad/tracking/telemetry servers
The www (HTTP) works via requests and responses. Generally, users make requests and publishers (including advertisers) receive the requests and (usually) send responses. Users are generally not receiving requests and few users would consciously send requests (including POSTing their private data) to ad, tracking or telemetry servers. It is "developers" who are making these requests for ads/tracking/telemetry, not users, by automatically generating them with the software they distribute to users (including Javascripts)
IMO, we don't "block" ads because we never requested them to begin with. What we do is ensure that the requests _made by developers_ do not succeed, we "block" requests that we never made
Segue into debate about user "agency"...
and every linux user should turn firewall on. and better download CISA hardening guides /best practices . or similar agency guides in countries other than US. they ARE free and mostly unversal, even if guide deals with other distro same principles apply to any system, just use AI to find right commands on your system.
99.999% of linux users who "converted to light of linux" based on nonsensical persuasion from perverts from LTT and other similar perverts, do not follow basic security practices. And these LTT perverts know this and they are laughing at your faces.
Windows in default configuration (as is installed without users turning off updates and other stuff off ) is 1000 milion times safer than linux for DESKTOP in default installed state. Which is state in which 99.9999% of loooosers sorry i meant to say users are using it in.. And im not even talking about totally irrational things to install like Crossover, Lutris and other nonsense based on WINE. I have 80 calls per month about infostealers, spyware in general which stem from using these nonsensical solutions full of old frameworsk (VB, .net etc ) and no security at all. Even official games, on EVERY platform, are scanning whole networks, combing thru your stuff (documents, pictures, webbrowsing activity... ) and noone cares, this is madness.
Distro maintainers should pull their heads from their asses too, how it is possibel to have firewall not enabled in OS in 2026?? HOOOOW. Not even talking about other stuff than like K12 stuff like firewall.