Explanation

Background concepts and design context for DuckDB Semantic Views.

• Semantic Views vs. Regular SQL Views – How semantic views differ from standard SQL views and what the declarative layer adds.

• Metric Grain and How Queries Are Assembled – Why each metric aggregates at the grain of its own table, how multi-grain queries are assembled, and which shapes are still refused.

• Snowflake Comparison – Feature-by-feature comparison with Snowflake’s SQL DDL semantic views and where this extension diverges.

• Databricks Comparison – Feature-by-feature comparison with Databricks metric views and where this extension diverges.

• Transactional DDL and Known Limitations – How transactional DDL works since 0.8.0 and the small set of caveats around read visibility, the PEG parser, and concurrent writers.