Please raise a specific spec issue for it.

The `id` field itself ties into the permissions management code, which I see you were already looking into, and would more directly expose to developers the browser's internal notion of how real devices map to the identity stored in the permissions system.

There's an upcoming project to add an `id` field similar to what BluetoothDevice has to USBDevice, SerialPort and HIDDevice which will make this much easier to specify as we can simply say that any two objects with the same `id` must also compare as equal.

I swear I added some language when I wrote Web Serial that makes this clearer, but yes, it is guaranteed and should be added to the specification as otherwise it becomes very difficult for developers.

That seems reasonable. Can you file a Chromium issue with links to your research?

(It should get specced too but we probably want to start with not passing values that the OS will reject.)

Can you file a spec issue for exposing both of these properties?

No. It probably wouldn't be difficult but nobody's presented a use case.

I wonder how much trouble I'd get into for expensing the cost of getting an actual human being to create illustrations for my slides.