A word about private attribution in Firefox

Original Article ↗ Hacker News Discussion ↗

Overall sentiment

  • Thread is largely critical of Firefox’s new “privacy-preserving” ad attribution, especially the opt‑out default.
  • Many see it as Mozilla drifting toward ad‑tech, conflicting with its privacy‑centric image.
  • A minority finds the technical design and stated goals convincing enough to keep or re‑enable it.

Consent, defaults, and telemetry

  • Strong objections to enabling ad‑related features by default without a prominent, explicit choice.
  • Several call the “modal dialogs are user‑hostile” justification disingenuous, noting Mozilla uses modals for less important features.
  • Opt‑out is widely framed as inappropriate for any privacy‑affecting feature; some argue for explicit or even double opt‑in.
  • Telemetry in general is criticized as a privacy violation once software is on the user’s device.

Privacy models: differential vs perfect secrecy

  • One side argues that true privacy (invoking Shannon/perfect secrecy) means giving adversaries zero new information; any aggregate signal is a leak.
  • Differential privacy is criticized as still providing valuable statistical information to advertisers; if they pay for it, it must be useful and therefore invasive.
  • Others counter that PPA only exposes noisy, aggregate conversion counts, not identities, and that calling this “volunteering data” is misleading.
  • Some accept the math and design as sound, but still reject the feature on principle or because it benefits only advertisers.

Strategic stance on advertising

  • “Defeatist vs practical” split:
    • Critics: appeasing advertisers won’t stop fingerprinting or tracking; better to keep the cat‑and‑mouse arms race with blockers.
    • Supporters: ads aren’t going away; giving advertisers a privacy‑friendly conversion channel might reduce incentives for more invasive methods and enable regulation.
  • Working with Meta and depending on ad‑company money is seen by many as structurally misaligning Mozilla with users.

Browser ecosystem and alternatives

  • Broad frustration that all major browsers are tied to large ad/platform companies; some call the situation “bleak.”
  • Suggestions include Librewolf, Ungoogled Chromium, Orion, Brave, Serenity/Ladybird, etc., but each has trade‑offs (funding, engines, performance, closed components).
  • Some argue Firefox the codebase can survive via forks even if Mozilla loses trust.

Technical and implementation concerns

  • Questions about who runs the aggregation servers (DAP/MPC) and risks if that “third party” is compromised or captured.
  • Worry that only a few allow‑listed sites will access the data at first, potentially entrenching big players.
  • Users share techniques to diff Firefox preference files to detect new, silently enabled options.