HN Debrief

uBlock Origin Is Giving Up the Fight to Keep Ads Off Facebook

  • Privacy
  • Advertising
  • Social Media
  • Developer Tools
  • AI

The submitted piece points to a Reddit note from the uBlock Origin side saying they are no longer going to keep chasing Facebook’s increasingly obfuscated feed ads. Facebook is making ad detection expensive by shredding obvious labels across deeply nested markup and changing the DOM structure fast enough to burn out volunteer maintainers. The key distinction people kept making is that this is not “ad blocking is dead” on the web in general. It is a case where one platform that owns the whole stack has decided to spend heavily to make a hostile client experience hard to filter.

If your product, community, or public communication still depends on Facebook, treat that dependency as a strategic risk now, not a nuisance. For users and teams, the lesson is that client-side fixes are getting costlier while platform lock-in and tracking remain even when visible ads are hidden.

Discussion mood

Strongly negative toward Facebook and broadly sympathetic to uBlock Origin. The mood mixed disgust at Meta’s hostile design and surveillance incentives with resignation that the only robust fix is to leave the platform, even though many people feel trapped by Marketplace, community groups, or local institutions that still rely on it.

Key insights

  1. 01

    Why blocked users still matter economically

    Facebook’s incentive is not limited to getting an immediate click. Advertisers buy impressions, upper-funnel brand exposure, and attribution credit, and Meta can often claim value even when the user was already on the path to buy. One commenter with ad-platform experience said big brands and agencies spend differently, but both still care about measurement enough that rendered ads and verifiable exposure remain valuable inventory. Another commenter pointed to FortNine’s critique of conversion attribution to explain why forcing ads onto resistant users can still look profitable on paper.

    Do not assume low click intent means low monetization. If you run acquisition or brand spend, audit your attribution model hard before trusting platforms that can take credit for demand they did not create.

      Attribution:
    • Lammy #1
    • sicktriple #1
    • dewey #1
    • jpleger #1
    • tomnipotent #1
    • rented_mule #1
    • enragedcacti #1
  2. 02

    Facebook survives through civic and marketplace lock-in

    People are not staying because the product is loved. They are staying because PTA updates, library notices, emergency alerts, small-business pages, hobby groups, and especially Marketplace have no equally liquid replacement in many regions. That makes Facebook look less like a discretionary social app and more like local infrastructure that happens to be privately owned and ad-funded.

    If your organization depends on Facebook as the primary channel for community reach, build an exit path now. Own an email list, RSS feed, website, or alternative marketplace presence before policy changes or product decay force the issue.

      Attribution:
    • pavel_lishin #1
    • corndoge #1
    • celestialcheese #1 #2
    • istjohn #1
    • jm4 #1
    • bothers #1
  3. 03

    Userscripts still work where filter lists fail

    Several people said the problem is narrower than the headline suggests because tools like F.B. Purity and Social Fixer still strip much of Facebook for them. That points to a useful boundary. Generic filter lists are struggling against constant DOM churn, but specialized browser extensions and userscripts that deeply target one hostile product can still deliver a cleaner experience for now.

    If Facebook use is unavoidable, test specialized cleanup tools instead of assuming the game is over. Treat them as temporary workarounds, not durable infrastructure, because they depend on a tiny maintenance surface and can disappear quickly.

      Attribution:
    • neuralkoi #1
    • mmsc #1
    • VCFundedGenYer #1
    • damidekronik #1
    • Beijinger #1
  4. 04

    The next ad blocker may watch pixels

    A lot of people converged on the same future architecture. Stop trusting the DOM and classify what actually reaches the screen, then mask it or generate new rules from that observation. The more grounded version was not “run a giant chatbot on every page” but use smaller local vision models or LLMs to assist rule creation and visual overlays. The catch is obvious. This shifts ad blocking from cheap string matching to a much heavier compute problem, which some people accepted and others found absurd for basic browsing.

    Expect a split in browser tooling. Lightweight blockers will keep handling the open web, while a new class of expensive, model-assisted tools emerges for a handful of highly adversarial platforms.

      Attribution:
    • jimrandomh #1
    • npilk #1
    • adam_arthur #1
    • pjc50 #1
    • StableAlkyne #1
    • TaLiTr #1
    • dreamcompiler #1
    • yaportmax #1
  5. 05

    Visible ads were never the whole privacy problem

    Even perfect cosmetic filtering would not stop Facebook from collecting behavior, shaping the feed, and keeping the user inside a surveillance-heavy product. Several comments argued that ad blocking can make Facebook feel cleaner than it really is, which softens the pressure to leave while none of the core data extraction changes. That framing turns uBlock Origin’s retreat into a reality check more than a defeat.

    Separate interface cleanup from actual privacy protection in your own threat model. If the business risk is data extraction or dependence on a hostile platform, hiding sponsored posts does not solve it.

      Attribution:
    • aanet #1
    • burnte #1
    • dofm #1
    • alex1138 #1
    • SubiculumCode #1

Against the grain

  1. 01

    Ad blocking is not theft of service

    One commenter argued that using ad blockers is bad faith because ads are the platform’s chosen payment model and users should either watch them or leave. The rebuttal was crisp and more persuasive for technical readers. Your computer should render information on your terms, especially when ads are bundled with tracking, dark patterns, and malware risk. That view rejects the idea that publishers own the final presentation on a user’s device.

    Do not let the monetization argument blur the user-control issue. In product decisions, there is a real difference between charging for access and asserting control over how a customer’s device must behave.

      Attribution:
    • Buttons840 #1
    • harlanji #1
  2. 02

    Paid alternatives may not scale to mass social platforms

    Against the dominant anti-ad mood, one line of argument said subscription and volunteer models work for small forums but break once you need billions of users and cross-subsidize those who cannot or will not pay. Examples like Netflix’s ad tier and Reddit’s tiny paid share were used to argue that ad-supported scale persists because most people prefer it, not just because platforms are evil. That does not rescue Facebook’s behavior, but it does challenge the easy assumption that “just charge users” is a universal replacement.

    If you are building a large consumer network, be realistic about how thin the market for paid participation can be. Alternative monetization needs to be tested against actual user behavior, not just cleaner ethics.

      Attribution:
    • yaportmax #1 #2
    • nemomarx #1 #2
  3. 03

    Facebook-specific tools still blunt the damage

    The dominant mood treated Facebook as essentially beyond repair, but some users reported that F.B. Purity has remained effective for years and is the reason they barely noticed uBlock Origin’s change. That does not solve the broader lock-in or tracking problem, yet it does mean the immediate user impact will vary depending on the toolchain people already run.

    Before you rewrite workflows or communicate a hard break to users, check what actually changed in your own stack. The practical impact may be smaller for people already relying on Facebook-specific extensions.

      Attribution:
    • VCFundedGenYer #1
    • damidekronik #1

In plain english

DOM ↩
Document Object Model, the browser’s in-memory representation of an HTML page that scripts can inspect and change.
DRM ↩
Digital rights management, technical restrictions used to control how digital content is accessed or displayed.
F.B. Purity ↩
A third-party browser extension focused on cleaning up and filtering Facebook’s interface.
LLM ↩
Large language model, a machine learning system trained on huge amounts of text and code that can generate plausible responses and software.
Social Fixer ↩
A browser tool that customizes and filters Facebook’s interface and feed.
uBlock Origin ↩
A popular open source browser extension that blocks ads, trackers, and other unwanted web content.

Reference links

Primary source and related coverage

Facebook cleanup tools and alternatives

  • F.B. Purity
    Frequently cited Facebook-specific extension that still removes clutter and ads for some users
  • Social Fixer
    Another Facebook customization tool mentioned as an effective workaround
  • News Feed Eradicator
    Extension suggested for users willing to remove the feed entirely
  • NoBook
    GitHub-hosted Facebook alternative client mentioned by name as a possible workaround
  • MaterialBook
    GitHub-hosted Facebook alternative client mentioned alongside NoBook

AI and adversarial blocking ideas

Ad-tech and attention references

Network blocking resources

Books