Algorithmic Optimisations
Before diving into compiler-level techniques, apply algorithmic optimizations. These give the biggest wins and are backend-agnostic. Scalus provides a library of Design Patterns for common cases:
- Withdraw Zero — run heavy logic once via a stake validator instead of per-UTxO (O(N) instead of O(N²))
- UTxO Indexer — pass indexes in the redeemer for O(1) lookup instead of linear search
- Transaction Level Minting — batch minting validation into a single check
- Merkelized Validator — split large validators into smaller pieces, include only the branch you need
See the full OptimizedPaymentSplitterValidator example in scalus-examples/.../paymentsplitter/ for a real-world application of the withdraw-zero pattern.
What’s Next?
- Design Patterns — full reference for every pattern listed above, with implementations and trade-offs.
- Scala Metaprogramming — once the algorithm is right, eliminate per-call overhead with
inlineand compile-time evaluation. - Measuring Performance — quantify the impact of each pattern on real transaction fees.
Last updated on