You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Most browsers defer display of notifications to a system-level notification center. This is great for the user's experience, but also means that developers have to take an almost infinite number of feature support permutations into account—especially when considering display sizes of visual assets.
Browsers could selectively expose properties for larger features, but checking with a system-level notification center might be too slow in order to do this prior to page load. The standard defines the maxActions attribute as one tool to aid developers, but that's not exhaustive and subject to fingerprinting concerns (#110). There are some open issues, notably #28, that would help developers provide a device-appropriate experience, but that's not comprehensive either if the developer wants to generate an image.
I'd like us to have a thought about what we can do to better help developers.
Personally I would like to see a method that allows developers to query capabilities of the used notification center, something like the following:
I realize that this is the opposite the direction @hober would like to go in, as it'd make fingerprinting easier. One could argue that these values are already derivable by considering the operating system information contained in the UA string, just with more work. The only exception here is Linux, which has swappable notification servers.
Thoughts?
The text was updated successfully, but these errors were encountered:
Do developers currently have libraries for this kind of thing? It seems like a rather large addition given that apart from Chrome nobody seems to be actively maintaining their Notifications API implementation.
Most browsers defer display of notifications to a system-level notification center. This is great for the user's experience, but also means that developers have to take an almost infinite number of feature support permutations into account—especially when considering display sizes of visual assets.
Browsers could selectively expose properties for larger features, but checking with a system-level notification center might be too slow in order to do this prior to page load. The standard defines the
maxActions
attribute as one tool to aid developers, but that's not exhaustive and subject to fingerprinting concerns (#110). There are some open issues, notably #28, that would help developers provide a device-appropriate experience, but that's not comprehensive either if the developer wants to generate an image.I'd like us to have a thought about what we can do to better help developers.
Personally I would like to see a method that allows developers to query capabilities of the used notification center, something like the following:
I realize that this is the opposite the direction @hober would like to go in, as it'd make fingerprinting easier. One could argue that these values are already derivable by considering the operating system information contained in the UA string, just with more work. The only exception here is Linux, which has swappable notification servers.
Thoughts?
The text was updated successfully, but these errors were encountered: