Auto-releasing wake lock depending on the user presence

[anssiko]

Modern laptops in the market have hardware that allows the OS to dim or turn off the screen depending on the user presence. Full disclosure: Intel is working on such technology.

It appears auto-releasing wake locks handles this case. However, I'd encourage the spec editors to check that this is appropriately hooked into the normative prose (unless "release a wake lock at any time" is fine).

Also, maybe the auto-releasing wake locks section could add an additional bullet for clarity:

A user agent may release a wake lock at any time, for example when: [...]

• The operating system is configured to dim or turn off the screen based on the user presence state.

It seems the concept of applicable is abstract enough with its "operating system permits" language:

The wake lock is applicable if the state of the operating system permits application of the lock (e.g. there is sufficient battery charge).

@rakuco @kenchris for reactions. I can submit a PR after you've settled with the famous "in parallel" refactor.

[marcoscaceres]

I'm not sure we need this (given how broad "release a wake lock at any time" is), but I wouldn't object to seeing a proposal for changes or maybe added as an example?

[anssiko]

(Connecting the dots: some user presence use cases were discussed in the DAS WG - 2021 Q2 virtual meeting. The Screen Wake Lock API editorial update #310 that landed was to clarify interaction with such a user presence implementation.)