Firefox's move to support privacy-preserving ads backfired.
@itsfoss "The road to hell is paved with good intentions".
It reminds me a whole lot of Do Not Track, which also had good intentions, but ended up just introducing yet another vector to use for fingerprinting.
I also disagree with the claim that internet ads inherently require tracking.
No other form of advertising provides direct feedback about someone's interests or the products someone purchased after seeing or hearing an ad.
Advertising can be unidirectional, it doesn't need tracking.
@fosstastic @itsfoss Firefox aimed to test attribution *without* tracking. It's like putting in a discount code inside a newspaper ad and counting how many people used the code. You only know the total count, not who used the code or what they are interested in.
@elgregor Advertisers can still track me with this preference enabled. The technology does not prevent any other tracking method, so PPA just adds yet another fingerprinting vector to the vast arsenal of APIs that browsers now unfortunately provide to websites.
I fundamentally disagree with the claim that advertisers are eligible or required to receive this information and that this feature is necessary in any way.
Nothing is preventing them from just using traditional, non-invasive methods.
@fosstastic I think the goal was to prove that attribution is possible without tracking, because advertisers keep telling politicians and courts that it's not.
@elgregor This setting has been implemented in production builds of Firefox. This isn't a technology preview, otherwise it should not have been enabled in stable releases, especially not an ESR release.
You can't fight tracking by introducing yet another method of tracking user behaviour. To me it doesn't matter how "privacy-preserving" it's supposed to be.
It has been shown countless times how supposedly anonymous data can be deanonymised with relatively few points of data.
@fosstastic I definitely agree that it should have been opt-in. I just think it's much less harmful than people make it out to be.
@elgregor The degree of harm can certainly be debated, but enabling it on existing installations without user consent is fundamentally wrong. This approach feels like salami slicing tactics—incremental changes that add up over time. What's next?