[Webkit-unassigned] [Bug 253186] REGRESSION iOS 16.4 beta selects ultra-wide for facingMode: environment

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Thu Mar 2 05:58:18 PST 2023


--- Comment #1 from connell at zappar.com ---
I'd like to add to this that there's no usuable workaround to maintain the existing (<= iOS 16.3) behaviour of selecting a device with a natural field of view.

To do so we'd have to iterate through the device list and find "Back Camera" or similar, but that list is localised into the user's chosen device language - so what's "Back Camera" in English would be "Rückkamera" in German.

In Beta 2 we have the situation that:

 - the default environment facing camera is ultra-wide; very different from iOS 16.3 and earlier

 - the device list is localised, so it's very difficult to pattern match to find the camera with the natural field of view

 - there's no visibility on the field of view or other parameters of a given camera in the list

In our opinion the best behavior would be (in descending order of importance :-))

 - avoid breaking existing sites that expect a natural FOV back camera by ensuring the environment facingMode constraint returns the same device it did in earlier iOS, 'Back Camera'

 - provide non-localised fields on MediaDeviceInfo that empower sites to select a camera that best fits their requirements in the case that the default camera does not. An example would be 'lensType': 'wide' | 'ultrawide' | 'telephoto' | 'macro' - something that doesn't add any fingerprinting entropy over the device name but does allow developers to unambiguously pick a camera that suits their needs

Finally I'd like to say that the current behavior would break many sites we have in production, plus those of our customers who use our SDK. We'd be very grateful of your attention on this bug :-)

You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20230302/a2dbee12/attachment-0001.htm>

More information about the webkit-unassigned mailing list