Add shared configuration model to future roadmap
This commit is contained in:
@@ -38,6 +38,24 @@ consumers.
|
||||
|
||||
## Ideas
|
||||
|
||||
### Shared Configuration Data Model
|
||||
|
||||
Use one canonical configuration data model for the public settings API and the
|
||||
internal configuration adapter. This could reduce duplicate type declarations,
|
||||
field-by-field conversions, construction-time allocations, and the maintenance
|
||||
needed when configuration fields change.
|
||||
|
||||
- Consider a narrow architecture-policy exception allowing the root facade to
|
||||
alias deliberately public, data-only settings types. A small public settings
|
||||
package is an alternative if explicit public ownership is preferable.
|
||||
- Treat shared settings types as intentional public contracts while retaining
|
||||
loading, validation, and execution logic internally.
|
||||
- Preserve owned runtime capture and independence from later caller mutation;
|
||||
sharing type definitions does not eliminate necessary value copies.
|
||||
|
||||
Selection requires an explicit architecture-policy decision. Inclusion here does
|
||||
not change that policy or expand the current configuration-handoff scope.
|
||||
|
||||
### Schema Bundling
|
||||
|
||||
Allow consumers to author reusable schemas across multiple files and package
|
||||
|
||||
Reference in New Issue
Block a user