Best of Infrastructure — July 2025

  1. 1
    Article
    Avatar of hnHacker News·45w

    stan-smith/OpenFLOW: Make beautiful isometric infrastructure diagrams

    OpenFLOW is an open-source Progressive Web App for creating isometric infrastructure diagrams that runs entirely in the browser. Built with React and TypeScript, it features auto-save functionality, offline support, privacy-first local storage, and import/export capabilities. The tool allows users to drag and drop components, create connections, and customize diagrams with colors and labels. It supports keyboard shortcuts, works across modern browsers, and can be deployed to any static hosting service.

  2. 2
    Article
    Avatar of staffengStaffEng·44w

    Adam Bender - Principal Software Engineer at Google

    Adam Bender, a Principal Engineer at Google, shares insights about the staff engineer role, emphasizing the shift from individual technical execution to strategic business thinking and cross-team coordination. He describes how staff engineers solve open-ended problems, reduce system complexity, and mentor junior developers. Bender highlights the importance of communication skills, systems thinking, and the ability to connect disparate teams and solutions across large organizations. He discusses his promotion journey, the concept of 'staff projects,' and provides advice for aspiring and new staff engineers about embracing discomfort during skill development.

  3. 3
    Article
    Avatar of communityCommunity Picks·45w

    stan-smith/FossFLOW: Make beautiful isometric infrastructure diagrams

    FossFLOW is an open-source Progressive Web App for creating isometric infrastructure diagrams. Built with React and TypeScript, it offers offline support, auto-save functionality, and privacy-first local storage. The tool allows users to drag and drop components, create connections, and export diagrams as JSON files. It can be installed as a native app and works entirely in the browser without requiring server connectivity.

  4. 4
    Article
    Avatar of faunFaun·43w

    The Platform Engineer Starter Kit

    Platform engineering teams build and operate Internal Developer Platforms (IDPs) that serve as self-service infrastructure for development teams. They focus on developer experience by creating golden paths, abstracting complexity, providing reusable building blocks, and measuring success through developer happiness and velocity. Unlike traditional ops roles, platform engineers act as product managers for internal tools, designing feedback loops and working closely with SREs while maintaining the balance between abstraction and flexibility.

  5. 5
    Article
    Avatar of techleaddigestTech Lead Digest·41w

    Figma’s $300k Daily AWS Bill Isn’t the Scandal You Think It Is

    Figma's $550 million AWS contract over five years represents about 12% of their revenue, which is completely normal for compute-intensive platforms. The internet overreacted to this disclosure, with many misunderstanding that enterprise AWS contracts include significant discounts and that real-time collaborative platforms require substantial infrastructure investment. Industry benchmarks show compute-heavy platforms typically spend 10-15% of revenue on cloud infrastructure, making Figma's spending reasonable given their 13 million monthly active users and 91% gross margin.

  6. 6
    Article
    Avatar of hnHacker News·44w

    7 Engineers Suspended After $2.3 Million Bridge Includes Bizarre 90-Degree Turn

    Seven engineers were suspended in India after designing a $2.3 million bridge with a dangerous 90-degree turn that went viral on social media. The Rail Over Bridge in Bhopal was meant to ease traffic for 300,000 daily commuters but instead became a safety hazard due to bureaucratic conflicts between agencies. The design changed multiple times over seven years as the Public Works Department and Railways disagreed on land sharing, ultimately resulting in an unsafe sharp turn that neither agency approved as functional or safe.

  7. 7
    Article
    Avatar of kodekloudKodeKloud's Squad·44w

    🔧 Linux is the Real Hero Behind Kubernetes.

    Kubernetes relies heavily on core Linux kernel features like namespaces for container isolation, cgroups for resource limits, iptables for networking, and OverlayFS for image management. Understanding these underlying Linux primitives is essential for DevOps and platform engineers who want to master Kubernetes, as K8s components are essentially wrappers around these foundational Linux capabilities.

  8. 8
    Article
    Avatar of towardsdevTowards Dev·41w

    Implementing Distributed Locks Correctly

    Distributed locks coordinate access to shared resources across multiple processes by ensuring mutual exclusion. The guide covers common implementation pitfalls like single points of failure, missing TTLs, and lack of fencing tokens. It examines various solutions including RDBMS advisory locks, Redis simple locks, Redlock algorithm, ZooKeeper ephemeral sequential nodes, and etcd's Raft-based approach. Each method has different trade-offs in consistency, performance, and complexity. Fencing tokens are crucial for preventing stale lock holders from corrupting state after timeouts. The choice of locking mechanism should match the required safety guarantees and system constraints.

  9. 9
    Article
    Avatar of hnHacker News·45w

    The Email Startup Graveyard: Why 80%+ of Email Companies Fail

    Email startups have an 80%+ failure rate because they try to solve problems that don't exist. Most build UI layers on top of existing infrastructure rather than actual email servers. Companies like Skiff, Sparrow, and Mailbox were acquired and shut down, while successful email businesses focus on infrastructure (SendGrid, Mailgun) or enhance existing workflows rather than replacing them. The core email protocols (SMTP, IMAP, POP3) work perfectly and have massive network effects that make replacement nearly impossible.

See all Infrastructure archives