Re: [webrtc-encoded-transform] Generalize ScriptTransform constructor to allow main-thread processing (#89)

> negligible or negative.

Negligible is fine, negative is very different.
I would be glad to further dive in any performance issue that Google Meet folks might be facing when adopting Safari WebRTC encoded transform support.

FWIW, WebKit used to have RTC networking hitting main thread and we got evidence this was causing issues to actual users. Using a WebRTC encoded transform on main thread would typically reintroduce these issues that Safari users encountered. Why should we be doing this?

WebRTC engines like libwebrtc typically try to stay off the main thread in their networking path.
It makes sense to apply the same design principles to WebRTC encoded transforms.

-- 
GitHub Notification of comment by youennf
Please view or discuss this issue at https://github.com/w3c/webrtc-encoded-transform/issues/89#issuecomment-1843257687 using your GitHub account


-- 
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config

Received on Wednesday, 6 December 2023 16:36:43 UTC