Lightning Updates
summary
Claims a working C++23 generic library that pushes immutable state updates to tens of millions of clients every 3–5 minutes by shipping 32‑byte identifiers rather than full blobs. It models updates as a content-addressable metadata/delta/snapshot graph, uses C++23 generic library patterns and strong typedefs for safety, serializes with protobuf, computes diffs with bsdiff, downloads via an application-provided async fetcher, and swaps state with an atomic shared_ptr swap so clients never see dangling data; demonstrated at scale (tens of millions of clients).
AI-written summary · about
chapters (13)
- 0:00Introduction
- 1:32Defining the problem
- 5:38Graph and data representation
- 9:56Strong typing for safety
- 14:51Designing update structures
- 21:11Serialization and validation
- 24:05Validation logic
- 27:21State management
- 30:06Navigation algorithms
- 35:55Asynchronous fetch path
- 41:50Update execution logic
- 46:32User-facing API
- 53:32Conclusion and use cases
related talks
05mLightning Talk: Powered by C++20 Meta-Programming (MP)
01h 08mThinking Outside the Synchronisation Quadrant
01h 16mPolymorphism, Cohesion, and Coupling in C++
53mC++ Horizons
06mLightning Talk: Speeding up code with C++17 Polymorphic Memory Resources
04mLightning Talk: Now in Our Lifetimes Cpp
12mLightning Talk: A Stable Limited LibTorch ABI? How?! (and Why?)
01h 07mEmploying Modern C++ for High Performance Delta-Coding Compression