Suggestions on improvements for the privacy considerations section

• A line “Underlying codecs are often implemented entirely in software” seems misleading given our findings.
• It’s unclear to me how a privacy budget would be implemented without impacting web compatibility.
• In the media capabilities draft they mention how a user agent would provide a subset of features that were common across devices to reduce entropy. This draft should mention the same and it’s not clear where this would go due to the configurations being part of other specifications linked in the codec registry.
• The Media Capture specification is much clearer about the ability to fingerprint this draft should have all the same pieces.
  • Add sections like this spec to make it easier to read individual topics.
• No mention of codec output being uniquely identifying.
  • It’s unclear if the output fingerprinting technique is different to the Media Capture specification but as support for more codec capabilities and input types increase (MediaCapture only supports streams) prevention of this using applying noise will become more difficult.

Start with the privacy considerations section and compare its coverage with the linked Media Capabilities and Media Capture privacy sections. Determine how the requested codec, configuration, privacy-budget, and output-fingerprinting topics should be addressed; done means the section clearly covers the concerns raised in this issue.