Videos

One chart is not enough for a production stack decision. Sharding Is A Last Resort Github : https://github.com/prod-garbage-destroyer/sharding-is-a-last-resort-benchmark In this video: - what each runtime/language wins - where tail latency changes the decision - how to pick based on your real bot

Rust can make Flutter-adjacent compute much faster, but the native boundary is not free. This benchmark compares a pure Dart data-processing lane against a direct `dart:ffi` Rust lane using the same seeded `Uint32List` workload, the same checksum path, two dataset sizes, and repeated measurement ro

Java Streams vs Rust iterators under the compiler Pretty syntax is not the same thing as cheap runtime behavior. This short compares a clean Java Stream pipeline with a clean Rust iterator chain, then looks at what survives after lowering: Java bytecode with stream stages and boxing, and Rust code

OAuth is not optional in production, but it is never free. This video compares the same minimal HTTP API in Spring Boot and ASP.NET Core twice: first as a public baseline, then as a real OAuth 2.0 + OIDC protected resource server. Instead of stopping at "which one is faster", the video shows where

That is exactly how you pick the wrong backend. In this v3 benchmark short: - Rust Axum wins read throughput - Go chi wins p99 - .NET AOT wins p99.9 The winner changes the second you stop staring at one screenshot. Full benchmark breakdown, repo, and artifacts are on the channel. #rust #golang #

The easy headline was wrong. This v3 benchmark is not a single-bar victory lap. Rust wins raw read throughput, Go chi wins p99 latency, .NET AOT wins p99.9 tail latency, and Fiber makes its case on recovery. The interesting part is not just who wins. It is why the winner changes once you look past

RabbitMQ dominated this benchmark: 274 rps at 7.57 ms p95, versus Kafka at 74 rps and MQTT at 39. The useful part is not just the headline. This breakdown shows why the result held across 48 measured slots, how the harness validated each protocol round trip, where the trace evidence lives, and when

One chart is not enough for a production stack decision. Java Legacy vs Modern Speed: Is Quarkus Actually Saving Your Enterprise Cloud Bill In this video: - what each runtime/language wins - where tail latency changes the decision - how to pick based on your real bottleneck Watch the full breakdo

Container benchmark takes are easy to get wrong when methodology is vague. Github: https://github.com/prod-garbage-destroyer/rust-container-optimization-clean In this response, I audit the exact harness and rerun the benchmark with a fair setup: - same Rust source and behavior across all variants -

One chart is not enough for a production stack decision. Go Map vs Redis: Latency Tax In this video: - what each runtime/language wins - where tail latency changes the decision - how to pick based on your real bottleneck #softwareengineering #programming #performance #backend #benchmark #architectu

One chart is not enough for a production stack decision. Beyond JSON: Why High-Performance Systems Are Switching to Protobuf over HTTP/3 Github : https://github.com/prod-garbage-destroyer/protobuf-http3-experiment-runner/tree/main In this video: - what each runtime/language wins - where tail late

One chart is not enough for a production stack decision. Carbon Aware Kubernetes Scheduling In this video: - what each runtime/language wins - where tail latency changes the decision - how to pick based on your real bottleneck Watch the full breakdown on the channel for benchmark method, tradeoff