- From: jan-ivar via GitHub <sysbot+gh@w3.org>
- Date: Wed, 05 Oct 2016 21:07:34 +0000
- To: public-media-capture-logs@w3.org
@foolip In the current climate, I think there's room for browsers to innovate and experiment with new privacy features, and giving up on fingerprinting outright would make this hard, which I think explains why there hasn't been an official capitulation on the issue. But I speculate. FYI Firefox's Private Browsing with Tracking Protection is marketed as: *"Some websites use trackers that can monitor your activity across the Internet. With Tracking Protection Firefox will block many trackers that can collect information about your browsing behavior."* In any case, these features live outside specs, so let's get back to seeing what we need to do to mitigate @ShijunS use-case. @alvestrand and I think we agree on at least the following: 1. Remove the need for permission in the active tab with input focus (to solve Shijun's case, and based on Martin's [characterization](https://github.com/w3c/mediacapture-main/issues/402#issuecomment-250352996) of USB-fiddling as user input). 2. Mandate that enumerateDevices() return old data until the event has fired. The spec already encourages fuzzing the timing on firing of the event, but does not require it. @alvestrand mentioned to me we might want to restrict fuzzing in the active tab and maybe even tabs with full gUM (not device-info) permission, to solve the use-case where a user yanks their USB camera, and in response to the `ended` event this fires, the JS immediately calls enumerateDevices() to figure out what happened. Returning stale data in that case would be confusing, but maybe that's a separate issue. -- GitHub Notification of comment by jan-ivar Please view or discuss this issue at https://github.com/w3c/mediacapture-main/issues/402#issuecomment-251799884 using your GitHub account
Received on Wednesday, 5 October 2016 21:08:00 UTC