Re: [mediacapture-region] Should generation of CropTarget from elements be able to fail? (#48)

> It is useful for specifications to reflect abiding reality. This serves Web developers. Developers often read these specs too, and who often treat these specs as authoritative.

@eladalon1983 that's not how "authoritative" works. What you appear to be saying is Chrome's implementation is authoritative and the spec is not and should be updated to match. Am I misreading that?

I oppose this view. I think Chrome's implementation is buggy so I've filed [crbug 1328836]( Let's discuss there. 

> I think the two issues are highly correlated, and revolve around whether the spec should specify that, in reality, the user agent may fail the process, and that spec authors accept this reality.

No they're not related, unless we conflate _"Should generation of CropTarget from elements be able to fail?"_ with input validation, which is already present. Even Chrome [performs]( input validation:
    await navigator.mediaDevices.produceCropId("");
TypeError: Failed to execute 'produceCropId' on 'MediaDevices': The provided value
is not of type '(HTMLDivElement or HTMLIFrameElement)'.

But this is built into WebIDL bindings and happens synchronously. This issue is about whether generation of a _valid_ target can fail.

GitHub Notification of comment by jan-ivar
Please view or discuss this issue at using your GitHub account

Sent via github-notify-ml as configured in

Received on Tuesday, 24 May 2022 17:01:25 UTC