Marko Zivkovic, in an April report for AppleInsider, revealed several new Safari features to debut this year. Some of them, like A.I.-based summarization, were expected and shown at WWDC. Then there was this:
Also accessible from the new page controls menu is a feature Apple is testing called “Web Eraser.” As its name would imply, it’s designed to allow users to remove, or erase, specific portions of web pages, according to people familiar with the feature.
WWDC came and went without any mention of this feature, despite its lengthy and detailed description in that April story. Zivkovic, in a June article, speculated on what happened:
So, why did Apple remove a Safari feature that was fully functional?
The answer to that question is likely two-fold — to avoid controversy and to make leaked information appear inaccurate or incorrect.
The first of these reasons is plausible to me; the second is not. In May, Lara O’Reilly of Business Insider reported on a letter sent by a group of publishers and advertisers worried Apple was effectively launching an ad blocker. Media websites may often suck, but this would be a big step for a platform owner to take. I have no idea if that letter caused Apple to reconsider, but it seems likely to me it would be prudent and reasonable for the company to think more carefully about this feature’s capabilities and how it is positioned.
The apparent plot to subvert AppleInsider’s earlier reporting, on the other hand, is ludicrous. If you believe Zivkovic, Apple went through the time and expense of developing a feature so refined it must have been destined for public use because there is, according to Zivkovic, “no reason to put effort into the design of an internal application”,1 then decided it was not worth launching because AppleInsider spoiled it. This was not the case for any other feature revealed in that same April report for, I guess, some top secret reason. As evidence, Zivkovic points to several products which have been merely renamed for launch:
A notable example of this occurred in 2023, when Apple released the first developer betas of its new operating system for the Apple Vision Pro headset. Widely expected to make its debut under the name xrOS, the company instead announced “visionOS.”
Even then, there were indications of a rushed rebrand. Apple’s instructional videos and code from the operating systems contained clear mentions of the name xrOS.
Apple renamed several operating system features ahead of launch. To be more specific, the company renamed its Adaptive Voice Shortcuts accessibility feature to Vocal Shortcuts.
As mentioned earlier, Intelligent Search received the name Highlights, while Generative Playground was changed to “Image Playground.” The name “Generative Playground” still appears as the application title in the recently released developer betas of Apple’s operating systems.
None of these seem like ways of discrediting media. Renaming the operating system for the Vision Pro to “visionOS” makes sense because it is the name of the product — similar to tvOS and iPadOS — and, also, “xrOS” is clunky. Because of how compartmentalized Apple is, the software team probably did not know what name it would go by until it was nearly time to reveal it. But they needed to call it something so they could talk about it in progress meetings without saying “the spatial computer operating system”, or whatever. This and all of the other examples just seem like temporary names getting updated for public use. None of this supports the thesis that Apple canned Web Eraser to discredit Zivkovic. There is a huge difference between replacing the working name of a product with one which has been finalized, and developing an entire new feature only to scrap it to humiliate a reporter.
Besides, Mark Gurman already tried this explanation. In a March 2014 9to5Mac article, Gurman reported on the then-unreleased Health app for iOS, which he said would be named “Healthbook” and would have a visual design similar to the Passbook app, now known as Wallet. After the Health app was shown at WWDC that year, Gurman claimed it was renamed and redesigned “late in development due to the leak”. While I have no reason to doubt the images Gurman showed were re-created from real screenshots, and there was evidence of the “Healthbook” name in early builds of the Health app, I remain skeptical it was entirely changed in direct response to Gurman’s report. It is far more likely the name was a placeholder, and the March version of the app’s design was still a work in progress.
The June AppleInsider article is funny in hindsight for how definitive it is in the project’s cancellation — it “never became available to the public”; it “has been removed in its entirety […] leaving no trace of it”. Yet, mere weeks later, it seems a multitrillion-dollar corporation decided it would not be bullied by an AppleInsider writer, held its head high, and released it after all. You have to admire the bravery.
Juli Clover, of MacRumors, was first early to report on its appearance in the fifth beta builds of this year’s operating systems under a new name (Update: it seems like Cherlynn Low of Engadget was first; thanks Jeff):
Distraction Control can be used to hide static content on a page, but it is not an ad blocker and cannot be used to permanently hide ads. An ad can be temporarily hidden, but the feature was not designed for ads, and an ad will reappear when it refreshes. It was not created for elements on a webpage that regularly change.
I cannot confirm but, after testing it, I read this to mean it will hide elements with some kind of identifier which remains fixed across sessions — an id or perhaps a unique string of classes — and within the same domain. If the identifier changes on each load, the element will re-appear. Since ads often appear with different identifiers each time and this feature is (I think) limited by domain, it is not an effective ad blocker.
Zivkovic’s follow-up story from after Distraction Control was included in an August beta build is, more or less, a rehashing of only the first explanation for the feature’s delay from what he wrote in June, never once commenting on his more outlandish theory:
Based on the version of Distraction Control revealed on Monday, it appears as though Apple wanted to distance itself from Web Eraser and the negative connotations surrounding the feature.
As mentioned earlier, the company renamed Web Eraser to Distraction Control. In addition to this, the fifth developer beta of iOS 18 includes a new pop-up message that informs users of the feature’s overall purpose, making it clear that it’s not meant to block ads.
It has been given a more anodyne name and it now has a dialog box.
Still, this shows Zivkovic’s earlier report was correct: Apple was developing an easy-to-use feature to hide page elements within Safari and it is in beta builds of the operating systems launching this year. Zivkovic should celebrate this. Instead, his speculative June report makes his earlier reliable reporting look shaky because, it would seem, he was too impatient to wait and see if the feature would launch later. That would be unusual for Apple but still more likely than the company deciding to cancel it entirely.
The August report also adds some new information but, in an effort to create distance between Web Eraser and Distraction Control, Zivkovic makes some unforced errors:
When it comes to ads, pre-release versions of Web Eraser behaved differently from the publicly available Distraction Control. Internal versions of the feature had the ability to block the same page element across different web pages and maintained the users’ choice of hidden elements even after the page was refreshed.
This description of the Distraction Control behaviour is simply not true. In my testing, page elements with stable identifiers remain hidden between pages on the same domain, after the page has been refreshed, and after several hours in a new browser tab.
Zivkovic should be thrilled about his April scoop. Instead, the two subsequent reports undermine the confidence of that first report and unnecessarily complicate the most likely story with baseless speculation that borders on conspiracy theories. From the outside, it appears the early rumour about Web Eraser was actually beneficial for the feature. Zivkovic accurately reported its existence and features. Publishers, worried about its use as a first-party ad blocker, wrote to Apple. Apple delayed the feature’s launch and, when it debuted, gave it a new name and added a dialog box on first use to clarify its intent. Of course, someone can still use Distraction Control to hide ads but, by being a manual process on a per-domain basis, it is a far more tedious process than downloading a dedicated ad blocker.
This was not a ruse to embarrass rumour-mongers. It was just product development: a sometimes messy, sometimes confusing process which, in this case, seemed to result in a better feature with a clearer scope. Unless someone reports otherwise, it does not need to be much more complicated than that.