Provide captions, sign language and/or visual cues that serve the same purpose as important audio content and that identify unimportant audio content.
+Provide audio description, media alternative and/or audio cues that serve the same purpose as important visual content and that identify unimportant visual content.
+Programmatically convey any semantics visually conveyed.
+Use parallel / complementary ways to convey information.
+Viewport rendering customization.
+Text rendering customization.
+Avoid or give control over interruptions.
+maybe something around staying within a particular contrast range.
+Only display additional content at explicit user action to request it.
+Controls with name-based interactions need names that work for them.
+Users can use whichever input mechanism is preferred and available to them
+Interaction customization
+Context of interactions
+Give users the time they need to do things
+Give users the ability to disable or avoid movement/animation
+Do not design content in a way that is known to cause physical reactions
+In-context navigation
+Provide multiple ways to navigate to content or functionality within a larger context
+A multi-modality version of 2.4.5 Multiple Ways
+Multi-part interactions take effect at the latest point in the interactions to allow for mid-operation canceling
+Don’t require fine motor control, but framed positively
+Allow users to switch input modalities mid-task
+Augment text with ways to better understand it
+Keep context consistent and predictable
+Provide similar structure and functionality via consistent paths & interactions
+Use common clear words.