- From: youennf via GitHub <sysbot+gh@w3.org>
- Date: Mon, 31 Aug 2020 13:49:06 +0000
- To: public-webrtc-logs@w3.org
> Unless there's a strong argument for using a brand new boolean member `panTiltZoom` in getUserMedia
Some arguments have been described above, some more below.
The current algorithm makes it so that getUserMedia({ pan : true } is different from getUserMedia({ pan : 1 }).
Ditto for getUserMedia({ pan : false } vs. getUserMedia({ pan : 0 }). This is not intuitive at all.
PTZ constraints are currently numerical constraints. This entails selecting devices based on a distance.
If a setup has several PTZ cameras, say PTZ1 and PTZ2, the selection of the device will be based on this distance.
Say a page will do getUserMedia({ pan : 100}), the page may sometimes get PTZ1 or PTZ2 depending on PTZ1 and PTZ2 positions. Even though page does not get access to PTZ2, the fact PTZ1 is selected might give information on PTZ2 position.
The current model is adding complexity to web browser implementations.
The current behavior could be potentially exploited (though prompts will probably help) showing a weak design.
I haven't seen any use case motivating this complexity, all examples in the spec and primer are solely using boolean constraints.
This suggests PTZ numerical constraints for getUserMedia is not a great idea.
--
GitHub Notification of comment by youennf
Please view or discuss this issue at https://github.com/w3c/mediacapture-image/issues/229#issuecomment-683790491 using your GitHub account
--
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Monday, 31 August 2020 13:49:08 UTC