The problem is that UBlock Origin goes way beyond just blocking a list of domains, it's not just a Piihole.
It's also doing CNAME uncloaking, request rewriting, stubbing functions in pages, it has syntax to handle CSS changes, it has rules that allow requests to only go through in certain contexts or if they're originating from certain domains. And most importantly, there's a giant community of people basically standardized on Ublock Origin who maintain all of these lists and who are constantly identifying new threats and proposing new features.
So a competitor to Ublock Origin would need to be constantly competing with it, it would need to be pulling in new features as they got released, and it would need to be consuming the same lists that Ublock Origin uses. It's just really hard to keep pace with that.
If Mozilla came out tomorrow and said they were going to do native adblocking, I would trust them less than Ublock Origin. I don't think even they would be able to keep pace if they were trying to build a community from scratch.
What about blockit[0]? It does implement some of the features you need. It's still a WIP, but I'm actively working on it and it uses the adblock-rust library from Brave, which is already able to compete with uBlock Origin.
Looks very promising! Correct me if I'm wrong, but it looks like adblock-rust also handles CNAME resolution? That's pretty good. And I also like that it directly consumes adblock rules, it doesn't require them to be converted into a new format.
That being said:
> it has rules that allow requests to only go through in certain contexts or if they're originating from certain domains
I haven't personally seen a proposal for server-based blocking outside of the browser that could begin to tackle this problem, and contextual blocking is a huge part of what makes browser adblockers work so well. As far as I can tell, adblock-rust isn't an exception to that.
It's a very difficult problem. In order for you to have a rule like "block 3rd-party requests on a-bank-website.com", you need more information than just the request itself. Maybe this is something where sites could piggyback off of CORS requests? But browsers don't always send CORS requests.
I'm not sure what the solution would be, but regardless I don't think that blockit would be a replacement in its current state. Still looks like a promising project as a network-based blocker though.
