Currently, the Encrypted Media Extensions Stream Format and Initialization Data Format Registry [1] is organized such that each Initialization Data Type has its own page, which covers both the Initialization Data format and - where applicable - how to parse the associated stream (i.e. container).

However:
 * The Initialization Data Type and stream format are used by two different parts of implementations:
  - The user agent uses the stream format to extract the initialization data, but does not necessarily care about its format.
  - The CDM only needs to know how to parse the Initialization Data format and does not need to parse the stream/container.
  - Note: A packager, which is not covered by the spec, needs to know both.
 * Initialization Data Types do not necessarily have an associated container (e.g. "keyids")
 * An Initialization Data Type may apply to more than one stream/container format (bug 26738).

We should have a table and page for Initialization Data Types and separate pages for stream/container parsing information. Maybe we can use the MIME type as the index for a table pointing to the latter pages.


[1] https://dvcs.w3.org/hg/html-media/raw-file/default/encrypted-media/initdata-format-registry.html