Hacker News Reader

Comment by gorhill | Hacker News Reader

> getting rid of ABP may be the primary reason for this move

This makes no sense to me.

ABP is already publishing a version using a declarativeNetRequest-like API, it's in the Mac App Store[1], so clearly it's not going to "get rid of ABP".

Google benefits from ABP and AdBlock[2] being the most used content blockers, as they do not block Google ads by default, and do not block trackers/data miners out of the box.

If Google (rightly) deems content blockers as unavoidable, it sure would want people to install the ones which are not blocking its ads and trackers by default.

* * *

[1] https://itunes.apple.com/us/app/adblock-plus-for-safari/id14...

[2] The two most-used content blockers as per Chrome Web Store.

View on HN

Maybe it's just me, I use ABP and I'm fine with it. I don't mind adds in general .. I mind annoying and experience breaking adds.

If I land on some page where annoying adds still get past ABP I close and never visit the page again.

You are probably right, although a low number of rules will also affect them as their acceptable ads program is based on the fact that they block everything else. I still wonder if they can continue to offer that program with 30,000 rules only. My idea that they target ABP came from their arbitrary 30,000 rules limit.

As I have written elsewhere I consider ABP and AdBlock as "infiltrated" by Google, so they definitely want to see the rigorous blockers disappear first. Should have made that clearer in the comment.

The Safari version supports "Acceptable Ads". I counted ~17,000 rules[1] which includes exception rules. Also, keep in mind cosmetic exception filters are not implemented through the declarativeNetRequest API.

(Edited for accuracy regarding rule count)

* * *

[1] https://gitlab.com/eyeo/adblockplus/adblockplussafarimacos/t...

Wow, ok that means 30,000 definitely won't be an existential threat for them.

It's been my assumption that with a little bit of scripting, EasyList could be pruned to land within the 30K without too much trouble. As you have more experience in this fields though, do you have any thoughts on how viable this approach would be, and if the 30K limit will be a problem in the end?

Doesn't any finite length list that is only able to be updated when the addon is updated mean ad networks can use a random string with more combinations than the size of list to defeat all ad blocking?

Yes, that seems plausible.

Reply on news.ycombinator.com