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
oldList in enumerateDevices seems to be per browsing session.
I guess it should be scoped by document, as groupId is per-document and we also want deviceIds to be partitioned per origin, which oldList might mess up with.
It is also not clear when oldList is set/built/updated. Is it set the first time enumerateDevices is called? Or is oldList the result of the last time enumerateDevices is called? Or is it a consolidation of all devices exposed during the various enumerateDevices calls?
I am also not very clear about the purpose of oldList as well. It seems that this is either an optimization (unnecessary then), a way to make sure deviceId/groupId are consistent call after call (but this is insufficient for deviceIds which are living longer than the browsing session and somehow redundant with MediaDeviceInfo.deviceId/groupId definitions), or a way for device === when enumerateDevices is called more than once (testing suggests this is not implemented this way in browsers currently).
I would tend to try removing oldList.
The text was updated successfully, but these errors were encountered:
Also, step 8.2 is not very clear in whether filteredList is a deep clone or not of resultList.
If not, the same object would be returned but the label getter would return the label conditionally on the 'device-info' permission. This is probably worth a separate issue.
oldList in enumerateDevices seems to be per browsing session.
I guess it should be scoped by document, as groupId is per-document and we also want deviceIds to be partitioned per origin, which oldList might mess up with.
It is also not clear when oldList is set/built/updated. Is it set the first time enumerateDevices is called? Or is oldList the result of the last time enumerateDevices is called? Or is it a consolidation of all devices exposed during the various enumerateDevices calls?
I am also not very clear about the purpose of oldList as well. It seems that this is either an optimization (unnecessary then), a way to make sure deviceId/groupId are consistent call after call (but this is insufficient for deviceIds which are living longer than the browsing session and somehow redundant with MediaDeviceInfo.deviceId/groupId definitions), or a way for device === when enumerateDevices is called more than once (testing suggests this is not implemented this way in browsers currently).
I would tend to try removing oldList.
The text was updated successfully, but these errors were encountered: