3 ms·
What does "move ad blocking out of the browser and into the OS" even mean? As long as you have websites built by JavaScript, trying to do perfect adblocking is
by stefan_ 8y ago
What does "move ad blocking out of the browser and into the OS" even mean?
As long as you have websites built by JavaScript, trying to do perfect adblocking is solving the halting problem. Can't do that, but we should start fixing the most obvious, glaring abuses (JS need not be able to tell if something actually loaded, no 3rd party host resource, certainly no 3rd party scripts. If I'm putting foo.com in as the address, my browser should only be talking with foo.com)
- elorant 8y agoYou can't turn off third party JS. What if I have an Intranet that's loading an external API. It could be anything from a chat session to analytics. You do that and you break the web.
- perl4ever 8y agoWell, it's feasible if you fork the web instead. If a significant number of people want draconian measures to be taken, and they converge on a search engine that can exclude the non-compliant, then the new standards have a chance of catching on. Suppose I say I just want to eliminate javascript entirely - that's not feasible to enforce, but if I could just search the non-javascript pages out there, then a new way of doing things can be bootstrapped.
- stefan_ 8y agoUhm, external analytics is exactly what we're trying to break here. And if you need an external API, have whoever is serving the Intranet page load that - that's how it's been forever. Websites loading data through JS from 20+ endpoints on a single page is very much a recent disease.
- krageon 8y agoI run all of my browsers (including mobile) with all 3rd party scripts blocked by default. Most sites that you care to casually consume will work without issue. Anything that uses a payment provider or google to login needs 2 greylist rules (which are mind-numbingly simple to click on once). I can set it up for people that don't understand what they are doing (and I have) and they will not notice anything in almost every case. As such, I conclude that this generalisation is grossly overstated and you probably haven't tried it to see if what you are saying is correct.