Blog

JVM Heap vs RSS: Why Does RSS Exceed -Xmx on 512 MB?

Go GC + Swap on a Tiny VPS: Stop Thrashing Before the OOM Killer

p50 vs p95 vs p99 Latency: Why Your Average Graph Lies

Average latency hides tail pain. Learn when mean lies, how p50/p95/p99 work, why averaging quantiles fails, and how Prometheus histograms fix aggregation.

How to Read Server Monitoring Graphs

A framework for turning dashboard noise into a diagnosis, based on the four golden signals and the symptom-versus-cause split — summarized from kciter's guide, with the author's hands-on verdict still to come.

How to Read Server Monitoring Graphs

A framework for turning dashboard noise into a diagnosis: symptoms first, causes second, and why averages lie about response time.

How I Cut GTA Online Loading Times by 70%

A reverse-engineering deep dive into why GTA Online took six minutes to load — and how two lines of quadratic code and a misunderstood libc function were to blame.

DeepSeek Harness Router: How to Install dsh-routing-suite

A rundown of the three-part dsh-routing-suite install — runtime injector, thinking-mode router presets, and the graded task protocol — straight from the project's README, with no first-hand testing added yet.

How Do You Set Up OpenBot AI Coworkers? Clone and Run

A first look at CopilotKit's OpenBot: an open-source template for running AG-UI agents as coworkers with their own computers, gated through a single policy-enforcing gateway.

XZ Utils Backdoor: A Technical Deep Dive

How a two-year-long social engineering effort inserted a backdoor into xz/liblzma's release tarballs, and how a Postgres developer's curiosity about slow SSH logins caught it.

What Is CUDA Rust? cuda-oxide vs cutile-rs for GPU Kernels

NVIDIA now lets you write CUDA kernels natively in Rust, compiled to PTX, through two separate projects: cuda-oxide for SIMT and cutile-rs for the tile model.

How to Read Server Monitoring Graphs

A framework for diagnosing dashboards: symptoms first, causes second, and why the average response time metric is lying to you.

Spring Boot vs Quarkus on a 512 MB VPS: what the numbers actually showed

A side-by-side JVM-mode comparison on a memory-starved VPS, run through StatLite's new Quarkus support. The results are messier than a simple 'X uses less RAM' headline.