Newsdesk
Engineering
Data & AI
Industries
Enterprise Systems
Go-to-Market
Longform
DesignIndiaAll stories

Craft & Leadership

03 October 2026

46 new posts across 8 practice and leadership sources — 6 worth your time.

One week of Craft & Leadership, 6 stories, as published.

Incidents & Ops

Article: High Availability Is Not Resilience: Why Cloud Systems Fail When It Matters Most

A routine TLS 1.3 upgrade silently broke Route 53 health checks, causing a CDN to stop routing traffic to a healthy region while internal dashboards showed no issues. The post analyzes the critical differences between high availability and resilience, highlighting how control-plane dependencies create invisible failure modes.

Why it matters — High availability configurations can fail silently due to control-plane dependencies, meaning true resilience requires explicit ownership and continuous validation of recovery paths.

InfoQ Architecture · 03 October 2026 · Read the original →

Architecture

Uber Eats Rebuilds Search Pipeline to Cut End-to-End Latency by 50%

Uber rebuilt its Uber Eats search pipeline to achieve a 50% reduction in end-to-end latency. The overhaul involved implementing Above-the-Fold measurement, reducing retrieval work, parallelizing hydration, and redesigning advertising data.

Why it matters — A 50% latency reduction can be achieved at scale by parallelizing hydration and optimizing retrieval work rather than relying solely on infrastructure upgrades.

InfoQ Architecture · 03 October 2026 · Read the original →

Practice

Bliki: Sensible Default

Martin Fowler introduces the concept of 'Sensible Defaults' as practices that should be used absent overriding context, contrasting them with the often-disliked term 'best practices.' Examples include version control, separating UI from domain logic, and automated deployment pipelines.

Why it matters — Framing engineering standards as 'sensible defaults' rather than 'best practices' acknowledges context-dependency and reduces developer resistance while still establishing clear baseline expectations.

Martin Fowler · 03 October 2026 · Read the original →

Practice

Refactoring Street Fighter With AI 🕹️ — with Adam Tornhill

This episode features Adam Tornhill, founder of CodeScene, discussing how his team refactored the entire 300K+ line Street Fighter 3 codebase. The conversation explores the practical application of AI tools in executing large-scale code refactoring.

Why it matters — Using AI to refactor legacy code at scale requires structured methodologies to safely manage and transform hundreds of thousands of lines of code.

Refactoring · 03 October 2026 · Read the original →

Leading Teams

Healthy Feedback

This post outlines principles for establishing effective feedback loops within engineering teams. It emphasizes that both praise and criticism must be treated as positive, constructive tools to successfully improve human collaboration.

Why it matters — Effective feedback relies on the shared understanding that both praise and criticism are positive mechanisms for growth, rather than personal attacks or empty platitudes.

Martin Fowler · 03 October 2026 · Read the original →

Testing & Quality

Copilot Code Review Gets API, New Default

GitHub announced that Copilot code review now supports API requests, allowing for integration into custom scripts and workflows. Additionally, the 'Balanced' effort level has become the new default for Copilot code reviews, a change that took effect on September 28, 2026.

Why it matters — Engineering teams can now integrate Copilot code review into custom workflows via API, and the new 'Balanced' default effort level impacts review thoroughness for existing users.

github.blog · 03 October 2026 · Read the original →

Looking forward to automating the reviews of the code we didn't write.

6 stories, every Saturday

Published here every week. Follow by RSS to get it as it lands.

← Previous issue Next issue →