1. data schema
   1. definition
      1. formal description of structures a system is working with
   2. basic idea
      1. a way of communicating
         1. "there is a Type of something"
            1. it has
            2. properties
            3. int
            4. float
            5. string
            6. ...
            7. references to other Types
      2. usage scenarios
         1. Content creators can
            1. create it
            2. manipulate it
            3. in a property inspector
         2. File system interactions
            1. save data to disk
            2. load data from disk
         3. Load it and use it at runtime
   3. examples of schemas
      1. Unity
         1. script components
            1. public properties
      2. Unreal
         1. properties of UClass
      3. Blind data in Maya
      4. Table columns in a database
      5. Frostbite
         1. Data Definition Format
            1. DDF
            2. schema format
2. three different data schemas
   1. runtime schema
      1. critical for performance
      2. used by
         1. the game for loading the data into memory
         2. the programmers in the code
      3. purpose
         1. performance
         2. patching
         3. loading
         4. nicely packed in memory
      4. optimized for
         1. reading
   2. storage schema
      1. a.k.a
         1. disk version
      2. what you save
         1. the logical peace of data
         2. not data format
            1. xml
            2. block chains
            3. json
      3. used by
         1. tools to save users work on some form of persistent storage
      4. purpose
         1. automatable with Tools
         2. fit into revision control
         3. multi-user editing support
            1. can be split into multiple parts
            2. for simultaneous editing
      5. optimized for
         1. writing to disk
      6. can be very Object Oriented
   3. tools schema
      1. what the content creators' envision
         1. how the data looks in the mind of the content creator
         2. how you display data
            1. graph
            2. property inspector
            3. timeline
         3. related to UX
            1. can be a subset of UX
      2. used by
         1. content creators
         2. engineers
            1. analysis and decision making
         3. tools developers
            1. to create an editing experience
      3. purpose
         1. UX
         2. visualization and understanding
         3. workflows
      4. optimized for
         1. editing
         2. iteration
      5. usually very Object Oriented
3. Examples of data schema setups
   1. Transformation
      1. data schemas
         1. represented in the editor (tools data schema)
            1. something more user-friendly
            2. property grid
            3. translations
            4. euler angles
            5. scale
            6. gizmo for transformation
         2. represented on disk (storage data schema)
            1. 4x3 matrix
            2. not expecting users to skew mesh
         3. represented in code ( runtime data schema)
            1. 4x4 matrix
      2. convert between schemas
         1. using the build pipeline to convert from the storage schema to the runtime schema
   2. Cinematic Tools
      1. Animation Curves
         1. data schemas
            1. tools data schema (tools frontend)
            2. curve editor
            3. keyframes
            4. storage data schema (tools backend)
            5. list of keyframes
            6. array of structs
            7. continent for
            8. reordering
            9. inserting
            10. removing
            11. runtime data schema
            12. lists of key frame properties
            13. struct of arrays
            14. efficient for
            15. searching
            16. traversing
            17. memory layout
         2. using the pipeline to convert from the storage schema to the runtime schema
4. designing data schemas
   1. requires consideration
      1. 1) start with the storage schema
      2. 2) build your tools (data schema) around the storage schema
      3. 3) implement the runtime schema around the storage schema
         1. if the performance isn't up to your standard
            1. 1) measure
            2. 2) change the runtime schema
            3. 3) write pipeline code to convert from the storage schema to the runtime schema
   2. there can be two extremes
      1. a programmer bias
         1. better for the runtime
            1. the data schema will be
            2. flat
            3. packed
         2. not optimal for a user to understand
      2. a user bias
         1. Object Oriented
            1. easier to build tools around
         2. easier for the users to understand
         3. not optimal for
            1. runtime
            2. not efficiently laid out in memory
            3. wastes cpu cycles for iterating through the structure
            4. not cache friendly
            5. storage
            6. too much space
5. the most important data schema
   1. storage schema
      1. the only one that is persistent
         1. when the power goes
      2. most expensive to change
         1. update content
            1. need to process a lot of content
6. things to look out for
   1. lose of data when converting between schemas
      1. example
         1. converting from degrees to radians and back
   2. making more than 3 data schemas
      1. one more thing that the programmers have to maintain
      2. like the network transport schema
   3. UX and data schemas have a strong relationship
      1. if one of them is changed this might lead to the other changing as well
7. formal support of multiple schemas
   1. no known engine supports multiple data schemas
      1. conditional compilation can help achieve this
         1. make sure the tools/storage schema doesn't reach production code