If you explore Mozilla Firefox's advanced settings, you'll see three options for data sending: Health Report, additional telemetry, and crash reports. In general, these parameters are opt-in, meaning the user must select them voluntarily. However, a Mozilla engineer indicated that the company is considering implementing a model of "differential privacy" backed by Google's RAPPOR project to collect more information, starting with a study that will be opt-out. User feedback hasn't been very positive so far...
We all agree that Firefox still has several details to fix. Its performance leaves much to be desired on some sites, and its memory leaks tend to become cascades. On the other hand, the adoption of multiprocess support was a big step in the right direction, and it already received optimizations in the latest version 55. Now, if we must mention a point where Firefox stands above its competitors, it's in the area of privacy. All its telemetry functions are opt-in, they are well identified in the settings section, and about:config allows us to make extra adjustments if necessary. However...
The Opt-Out Study
In the Mozilla Governance forum on Google Groups, engineer Georg Fritzsche explained that they will start a new study that will evaluate extended data collection without compromising user privacy. The problem is that developers need "sensitive information" to optimize the browser, starting with the main portals they visit, and how certain functions work on them. The solution Mozilla has in mind is the so-called "differential privacy" (obtaining impartial details from a large group of data, with mathematical proof that no one will be able to learn anything about an individual), a technology that Google and Apple are already using. The most popular implementation of differential privacy is the open source RAPPOR project created by Mountain View, and what Mozilla wants to determine is how well its derived implementation works.
The study will be opt-out, and if everything progresses correctly, its launch will take place in mid-September. At the end of the post, the engineer stated that they were interested in receiving feedback. If we leave aside the occasional direct attacks, some of the questions users raised are logical. How will the opt-out process work? Why does Mozilla need to know part of the browsing history (no matter how anonymized) to optimize performance when there is a large amount of public data identifying popular sites? No one in their right mind wants to start a witch hunt, but we are going through a very delicate era regarding privacy and collection, therefore, Mozilla should share more information.
