- introduction of scopes reduces jlib performance by ~20-22%. can we do sth about this?
- moving this logic to draft 2019 would solve this, but introduce two different apis unless we can hide it
- a consistent api might lead to performance impacts of legacy drafts
- an inconsistent api would not be manageable (duplicate utils like getTemplate etc)
- there is probably a lot that can be improved (also in compile time)
- add all subSchemas to scope-history as only if and anyOf are tested
- decision on supported draft 2019-09 format-options
- template default options retrieved from draft
- additionalProperties: true per default
- compile needs another parameter for rootschema, in case refs are defined elsewhere
- remove hard coded schema interpretation
- Improve -- oneOf-Error messages (specific errors where possible, instead of one-of-error)
- Add -- Resolve $ref local json-pointer without requiring compiled schema
- Refactor -- move type validation as keyword to validation/keywords
- Refactor -- Use addValidation to setup base validation mappings?
- Features -- latest draft support
- Refactor -- improve performance
- Feature -- Helper to find a json- and json-schema-pointer