If IOv2 is used in a cross-origin iframe, then for privacy/security reasons, we probably can't be any more specific than a simple boolean isVisible flag.
But for the same-origin or same-document case -- where the full DOM is available to script to analyze -- then the privacy and security concerns don't really hold. This is similar to the handling of the rootBounds field of IntersectionObserverEntry.
In the same-origin/same-document case, it may be useful to provide more detailed information about why a target is reported as not visible. This could be, e.g., an enum:
enum VisibilityCheckFailureReason {
kOccluded,
kOpacity,
kTransform,
kFilter
};
We could also provide the Element (or one of the Elements) that occludes the target.
I have heard interest in this feature primarily from ad networks, who typically have their code injected into the top-level browsing context. In that situation, the ad network code doesn't know anything about the host page, and they would like to be able to use IOv2 to identify when pages inadvertently occlude or visually distort ads, for the purpose of working with the publisher to fix it.
If IOv2 is used in a cross-origin iframe, then for privacy/security reasons, we probably can't be any more specific than a simple boolean isVisible flag.
But for the same-origin or same-document case -- where the full DOM is available to script to analyze -- then the privacy and security concerns don't really hold. This is similar to the handling of the rootBounds field of IntersectionObserverEntry.
In the same-origin/same-document case, it may be useful to provide more detailed information about why a target is reported as not visible. This could be, e.g., an enum:
enum VisibilityCheckFailureReason {
kOccluded,
kOpacity,
kTransform,
kFilter
};
We could also provide the Element (or one of the Elements) that occludes the target.
I have heard interest in this feature primarily from ad networks, who typically have their code injected into the top-level browsing context. In that situation, the ad network code doesn't know anything about the host page, and they would like to be able to use IOv2 to identify when pages inadvertently occlude or visually distort ads, for the purpose of working with the publisher to fix it.