Bugzilla

So for now, in this bug, we could just minimize the implementation:
We still compute the new start value in [`AnimationInfo::AddAnimationForProperty`](https://searchfox.org/mozilla-central/rev/c34cf367c29601ed56ae4ea51e20b28cd8331f9c/gfx/layers/AnimationInfo.cpp#421) for the main thread. However, we don't use that value for the compositor. As per my previous patches, we send a flag to the compositor, so we just use that value as the new start value on the compositor.

This makes the computed value on the main thread be a little bit different from the actual rendering result on the compositor thread, but the rendering result could be better (i.e. without any lag on the compositor).

So for now, in this bug, we could just minimize the implementation:
We still compute the new start value in [`AnimationInfo::AddAnimationForProperty`](https://searchfox.org/mozilla-central/rev/c34cf367c29601ed56ae4ea51e20b28cd8331f9c/gfx/layers/AnimationInfo.cpp#421) for the main thread. However, we don't use that value for the compositor. As per my previous patches, we send a flag to the compositor, so we just use the previous sampled animation value as the new start value on the compositor.

This makes the computed value on the main thread be a little bit different from the actual rendering result on the compositor thread, but the rendering result could be better (i.e. without any lag on the compositor).

So for now, in this bug, we could just minimize the implementation:
We still compute the new start value in [`AnimationInfo::AddAnimationForProperty`](https://searchfox.org/mozilla-central/rev/c34cf367c29601ed56ae4ea51e20b28cd8331f9c/gfx/layers/AnimationInfo.cpp#421) for the main thread. However, we don't use that value for the compositor. As per my previous patches, we send a flag to the compositor, so we just use the previous sampled animation value as the new start value on the compositor.

This makes the computed values on the main thread be a little bit different from the actual rendering result on the compositor thread, but the rendering result could be better (i.e. without any lag on the compositor).