Overview
Data schemas are the structural layer that allows Real Science software systems to exchange information without losing meaning. They define common shapes for records, recipes, targets, residuals, and verification-compatible outputs.
Why schemas matter
They preserve structure across tools.
They reduce naming drift.
They make records easier to compare.
They help software remain compatible with verification workflows.
Core schema families
CRL schemas
RealNet schemas
Heliotron schemas
GeoQ schemas
Arcadia schemas
Materials schemas
Common schema header pattern
{
"schema_name": "ExampleSchema",
"version": "v1.0",
"description": "Short description",
"canonical_status": "starter / provisional / canonical",
"notes": {},
"data": {}
}Example conceptual records
RunRecord
TargetVector
Recipe
VirtualResponse
ResidualReport
ClaimRecord
Design rule
A schema should be readable, versioned, and narrow enough to preserve clear meaning.
If a structure becomes too broad, split it into linked schema families rather than making one overextended record type.