On the frontend team we are designing the new normalized data structure, which is how the client will organize all data that is received from the API

Account
- [Account_ID]
  - Paths
    - [Path] - empty string for home docs, /[TSID] for CCC, and /human/readable for children docs
      - when type: Document
        latestVersion: [cid1.cid2]
        mentions array:
        resource: [AccountID+Path]
        mention metadata?
        comments array:
        [Resource: AccountID+Path]
        activity: TBD - cache the activity events about this specific resource
        search: TBD - cache search results when your search is narrowed to this resource
      - when type: Comment
        latestVersion: [cid]
        mentions array:
        resource: [AccountID+Path]
  - Profile
    - latestCid: [cid]
search - used to cache search results when searching the whole node
activity - cache the node's global activity events
- will reference the Path Resources: [AccountID+Path] or the Profile: [AccountID]
Immutable
- [CID]
  - when a document change:
    - State - the current state
      - Metadata
      - Content
    - Blob Data (as it saved in IPFS)
  - when a comment, cap, contact, profile:
    - Blob Data (as it is saved in IPFS)
    - Metadata? If only used if the full blob data is not loaded
- [CID1.CID2]
  - for document versions only
  - State - if this is a document version
    - Metadata
    - Content