Skip to content

bugzilla 1358149 [webaudio] backport to ESR? MB? BB?

1358149 Address fingerprinting issues with Audio[Context]

Can't believe it's only been 6 years 😄 Anyway, RFP already sets some context keys which do affect audio FPs (oscillatornode and/or dynamicscompresor) outside of listing context keys + values. The basic audio test (offlineAudioContext sum) has few fingerprints (7 or 8) in gecko AFAICT (this includes data from researches and myself), and is mostly OS equivalency. IDK if the above patch fixes all entropy without RFP but that's not our concern here

That said, MB might like to benefit from upstream, which is not RFP dependent, but I don't think they would do it just for MB. If the fingerprinting (combined with RFP) is tight, we could even relax TB's disabling of the API. Otherwise, it's a big patch (and three releases out)

logging so we don't forget

cc @pierov

To upload designs, you'll need to enable LFS and have an admin enable hashed storage. More information