Most important tools are a pen and paper

#628 – June 01, 2025

AI, LLM, machine learning but pen and a notebook still dominates

Most important tools are a pen and paper
3 minutes by Juha-Matti Santala

Juha-Matti shares how a notebook is his most important tool. He explains that stepping away from the computer to write and sketch in his notebook enhances his thinking process, helping him design solutions, understand code flow, and expose gaps in his understanding. Writing about code as if explaining it to others helps identify inconsistencies and design flaws, while also creating a valuable record of his thought process that can be referenced later.

Skip PACT Files - AI Finds Contract Breaks Automatically
sponsored by Signadot

Stop writing and maintaining PACT contracts for every API change. Signadot’s SmartTests uses AI agents to automatically detect breaking contract changes by comparing live API responses between service versions. No schema definitions, no consumer-driven contracts to update - just deploy your code and let our AI agent identify what actually breaks downstream services. Forward-thinking engineering teams are achieving zero maintenance as APIs evolve while catching critical integration issues faster.

Breaking the sorting barrier for directed single-source shortest paths
7 minutes by Ran Duan et al.

Breaking the sorting barrier for directed single-source shortest paths presents a deterministic algorithm that solves the single-source shortest paths problem in O(m log^(2/3) n) time for directed graphs with non-negative real edge weights. This groundbreaking result proves that Dijkstra's algorithm, with its O(m+n log n) time bound, is not optimal for sparse graphs.

Harmel-Law's architecture advice process and fast flow
6 minutes by Chris Richardson

Chris discusses the architecture advice process, highlighting how it enables faster decision-making. The process supports team autonomy and works well with Team Topologies and microservice architecture. Chris explains how loose design-time coupling between software elements is crucial for team independence, noting that microservices have an advantage over modular monoliths by localizing technology stack changes.

The software engineering Kanagawa moment
6 minutes by Pasha

Software engineering is undergoing a major transformation as AI-powered coding tools evolve from assistants to autonomous agents. This shift threatens traditional coding roles, especially for junior engineers, and demands a redefinition of what it means to be a software engineer. Pasha argues that while routine coding tasks are being automated, human expertise remains essential.

MCP explained without hype or fluff
6 minutes by Atharva Raykar

Atharva explains the Model Context Protocol which transforms complex M×N integration problems into simpler M+N solutions by allowing AI client applications to interact with various platforms through a standardized interface.

And the most popular article from the last issue was:

newsletters