Interestingly enough, a lot of people are making a fuss about it. The situation is actually less bad than it seems at first. This is because it is about permission, as websites have to ask for permission from the user via the browser. Something is the case with many browser web APIs, including APIs: location, camera, microphone, notifications, clipboard, etc. Something we (rightly) think is so common that the browser should ask for explicit permission. It’s no different with IdleDetection.
You can test it yourself by entering the following into the Chrome console:
IdleDetector.requestPermission(),
You will then get the message, when running on Tweakers:
tweakers.net wants to: Know when you’re actively using this device. [Allow] [Block]
@Geistot The claim in the article:
The API is automatically activated in Chrome 94.
So this is not true. It’s not enabled by default, but based on semantics, it’s not disabled either. Ask for permission is set by default, and requires explicit permission from the user, for each website.
Update:
Part of what makes it confusing is also in this sentence in Chrome:
Sites may ask to know when you are actively using your device.
Which can be read in two ways: “… they are allowed to know when you are actively using your device,” or “They are allowed to know when you are actively using your device, if explicit permission is given. The latter is the case.
Also, part of the confusion is the reaction of Apple and Mozilla, who are both against the feature. In my opinion, the incompatibility is hypocrisy, because they themselves used the following web APIs in the same way as Google used IdleDetector: Location, Camera, Microphone, Notifications, Clipboard, which are much worse than IdleDetector in privacy. But that’s okay, because it asked for permission, and that’s also the case with Google IdleDetector.
Apple and Mozilla obviously know this themselves, so I find their response amazing. I think it should be seen mainly from PR: pretending to be better at privacy than your biggest competitor, and being willing to use non-argument in the process.
[Reactie gewijzigd door Cyb op 23 september 2021 17:15]
