NORM_GEN: an alternative to ORM
summary
NORM-GEN generates PostgreSQL user-defined types and stored functions from an annotated JSON schema, producing bulk transfer objects that convert to JSON on the database side and cut hundreds or thousands of ORM round trips into a few set-oriented calls. Generated functions include select by id, a MongoDB-like generic search and Norm2db for inserts/updates/deletes. Tested in production patterns (examples on GitHub) against very large datasets—hundreds of millions of rows—and claims much lower latency and far fewer queries, while table definitions remain a DBA-controlled concern.
AI-written summary · about
chapters (8)
- 0:00Introduction and speaker background
- 4:45What is ORM and why developers use it
- 8:38Why ORM is inefficient and slow
- 13:33Introducing NORM_GEN solution
- 22:20Formalizing the contract with JSON Schema
- 27:45Generating database functions automatically
- 33:51How NORM_GEN differs from ORM
- 36:44Call to action and open source contribution
