1. 3
  1.  

  2. 1

    The amount of unnecessary automated callbacks in browsers is increasingly an issue. And it is just a lot of work to address this manually, and keep up with new stuff being added. And to do that in every profile you create in your browser. The process is a bit broken. Who can realistically keep up with continuous new information leaks that allows external passive observers to infer certain information about ones system and behaviour?

    This screams for a flag “thanks, but no calls to any resources other than indicated by the user”. Even better: a command line option which does this instance-wide, not just for a profile…

    Why not make it less cumbersome to perform this basic hygienic task - which is BTW phrased not very positively in the article in the very first sentence where it says “Some people are concerned..”. Some people need to be concerned.

    1. 1

      Who can realistically keep up with continuous new information leaks that allows external passive observers to infer certain information about ones system and behaviour?

      “thanks, but no calls to any resources other than indicated by the user”

      I’d say the devs weighed this up against the current state of browsers and the internet. Compared to the massive tracking networks and companies with their JS payloads and non-consentual data collection: some requests to mozilla.org and page prefetches seem relatively harmless.

      On the other end of the scale there’s people like us that block large swathes of content and try to actively avoid this. But then we’re just as guilty: each of those content-blocking addons we use phones around regularly to get things like blocklists.

      @Taal: there’s an interesting topic of trust and consent here. You explicitly opt into content-blocking addons, but have to opt-out of these browser behaviors.

      Oh wait, maybe Firefox did ask us about data collection somewhere in the chaotic start screen. But I’m not sure that really counts as effective.

      Sidenote: my browser starts much more slowly when I don’t have a working internet connection. I suspect some TCP requests have to time out (browser? addons?) before the browser creates its window.