Never send a human to do a machine's job
summary
Roslyn Source Generators automate repetitive code by emitting partial implementations that replace Reflection and boxing for value types, here generating explicit IEquatable and GetHashCode for partial structs; the generator compiles, unit tests turn green and benchmarks show elimination of allocations and hash collisions so generated structs match C# 10 record struct performance. The migration to Incremental Generators reduces generator runtime by caching pipeline steps. Includes a small benchmarking harness and IEquatable generation examples, all benchmarked and green tests.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 2:24Source generator basics
- 5:27Performance problems in structs
- 13:58Unit testing generators
- 18:20Implementing ISourceGenerator
- 33:22Incremental generators
- 42:52Best practices and wrap-up
related talks
01h 00m.NET Rocks Live – Modernizing .NET Applications
38mAccelerate
54mPerformance Architecture
23mBuilding in fast feedback loops
03mSteve Yegge, Sourcegraph, ETLS Vegas (00)
26mAI-augmented SDLC: Rethinking how we deliver software
38mFrom code to culture: How AI is reshaping the engineering organization
46mThoughtful Software Engineering