Container for Multi-Media Narravites; SMIL Revisited

[Sannyasin Brahmanathaswami]

Brainstorm: 

It a shame that SMIL was abandoned.  Well maybe not... but working from ground zero is a challenge. 

We want 

Goal: Implementation:

Input: Voluteers or staff can "build a narrative" by  
      - write a script that  fills the "text" nodes (below)
      - source (search for) for images, audio and video resources and times be referenced.

Output: is it XML, Is it JSON? Is it YAML. Or something else? It has to be flexible but parseable.

Does the below provide enough detail that the production or director or the developer can built it? Using whatever format or tools that he-she like?

----------------
Rough spec:

-- A high level container for a multi-media narrative.
-- It should have a structure; so, HTML would will not do (unstructured)
-- It is "presentation" agnostic 
 	 - you could give it to a film producer and it could serve as a  "script"
	  -  you could built a KeyNote which would export to a video
           - Using to the build a "story" as HTML,  as seen on "Medium"
 	 -  use it to build small HTML5 scenes ala Tumult's "Hype 3"
         - you could bring in Livecode and a log scrolling group or "click for next scene" on a card.

We take the model of a play, which has scenes. 
Each scene describes what happens before "the curtain drops"  then next one 
Each scene has scenario, which happens while the curtain is up, but is distinct   

The container which we will call a "Narrative Script" has one or more scenes

 must have (at minimum)

Narrative Script

Scene ID
    Scenerio ID
		 -- Node for text with attributes:
    				  - layer
     			 - title: Literal text: "Oh What a Day!"
          			 - Pull Quote
             				 - literal text entry: "We took a deep drive dive:
             				 - URI form some "locations" or function from which text can be retrieved (url http://someDomain/SomeBook/chap##/?name=###)               
      	 	 - Quote
             		 - literal text entry: "We want for a long drive in...." 
            		- URI form some "locations" or function from which text can be retrieved (url http://someDomain/SomeBook/chap##/?name=###)               
     
		-  Variable length Text
			- long text entry: "Once open a time...." 
            		- URI form some "locations" or function from which text can be retrieved (url http://someDomain/SomeBook/chap##/?name=###)               

-- Node for image 
      -- still
      -- motion/Gif
     	 -- optional Ken burn effect: 
			- begin loc, ending loc, speed, ratio change	 

-- node for video URI
              Attributes: beginning and ending point

- node for audio URI
              Attributes: beginning and ending point

---------------
Anything missing?