Add opSupportLimits() privacy considerations for fingerprinting #881
Conversation
|
It seems to express similar thing as in https://github.com/webmachinelearning/webnn/blob/main/security-privacy.md#questions-to-consider, should we de-dup and reference the spec in the security-privacy.md? |
|
@philloooo, slight preference to have this in context of the specification close to the algorithm so that more people will read the note. Happy to borrow that text verbatim or adapt. Then add a link to the respective section in the specification. Suggestions for concrete text welcome! |
reillyeon
left a comment
reillyeon
left a comment
There was a problem hiding this comment.
I have a strong preference towards referencing the existing "Privacy Considerations" section or a specific subsection within it because as written this text actually contradicts the "No information from the underlying platform is exposed directly." text in that section.
Co-authored-by: Reilly Grant <reillyeon@users.noreply.github.com>
reillyeon
left a comment
reillyeon
left a comment
There was a problem hiding this comment.
These edits look good, but please follow up by revisiting the general privacy section.
|
I will merge this PR now. A follow-up issue to revise privacy considerations: #886 |
SHA: 8bce961 Reason: push, by anssiko Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Discussed at https://www.w3.org/2025/08/14-webmachinelearning-minutes.html#7993
Seeking initial review from @RafaelCintron and @reillyeon to ensure I interpreted and translated your meeting comments the right way into this spec note.
Fix #875
Preview | Diff