4 ms·

For Safari users who use a native content blocker, we've done some similar testing to see the impact of our ad blocker. - Speed improvements: https://www.magic

• Terretta 6y ago
  Unfortunately, this could look like you’re pretty heavy, if it’s “only” 2x speed up, by comparison to the post article. But more to the point, what does it mean to test “your” ad blocker, if what you’re doing is providing a rules list to Safari? ”Content Blockers are app extensions that you build using Xcode. They indicate to Safari a set of rules to use to block content in the browser window. Blocking behaviors include hiding elements, blocking loads, and stripping cookies from Safari requests.” ”You use a containing app to contain and deliver a Content Blocker on the App Store. ... Apps tell Safari in advance what kinds of content to block. Because Safari doesn't have to consult with the app during loading, and because Xcode compiles Content Blockers into bytecode, this model runs efficiently. Additionally, Content Blockers have no knowledge of users' history or the websites they visit.” https://developer.apple.com/documentation/safariservices/creating_a_content_blocker https://developer.apple.com/documentation/safariservices/cre... - - - Note: I’ve tested dozens of ad blockers on iOS for interoperability with fin techs that users expect to “just work”. Not performance testing, just using as daily drivers and doing banking tasks. I look for which apps let ads through on popular sites while breaking brand name banks. The antidote to this is great active curation by the core team (or curated upstream), plus a convenient way for the end user to block new things, and to allow things. For a tech audience that uses MacOS and iOS native Safari and misses ability to customize rules, I tend to recommend 1Blocker: https://1blocker.com/ https://1blocker.com/ Of the many I’ve tested, 1blocker lets the lowest percentage of ad or tracker requests through to the DNS lookup stage as logged/blocked by pi-hole, while the web still works as normal users expect. We drop from >40% of DNS requests blocked with no ad-blocking to ~5% of DNS requests blocked with 1Blocker enabled.
  • happybuy 6y ago
    Performance changes depending upon the rule list provided to Safari. A better (not necessarily larger) rule list means that more unnecessary items are blocked and web pages therefore load more quickly. This means that there can be a large difference between ad blockers even if they use the same underlying blocking mechanism. As you note, the rule quality is also important. We judge our rule quality based upon: - How many ads, trackers and annoyances it blocks - How frequently rules are updated to meet changes to pages - How large is the rule set and does it include out of date and redundant items (many ad blockers have overly large rulesets that increase memory use and reduce maintainability; including 1Blocker) - Not breaking any required web page functionality