Tom Bray’s Post

This article on "kebab organizations" is brilliant - from Alex Ewerlöf "The biggest difference is the heuristics that guide the organizational architecture: 🍰 Cake is organized outward: optimize towards the needs of the consumer at risk of higher cognitive load for internal teams. 🍢 Kebab is organized inward: optimize towards the focus of internal teams at the risk of handovers. Every reorg is a trade-off. It makes some things easier and some things harder. " https://lnkd.in/e78eN_zm

Here's Jack Ellis on why Fathom is using Lambda to process billions of events a month - because it's still cheaper in terms of the total cost of ownership! "Until we can get Lambda close to what it'd cost us for 2-3 full-time DevOps engineers, we're not gonna consider (moving off Lambda)" https://lnkd.in/ezzqkjjc

Organize the world's web development information and make it universally accessible and empowering Late yesterday we launched the new version of our frontend code assistant v0. It still gets you from just a prompt to a working, well-designed UI in a single step, but now it can also add interactive functionality, run your code examples, graph algorithms, and answer your questions. All that with up to date knowledge of the newest APIs and best practices. TBH I find it hard to believe how far we've come in under a year from a simple idea of having AI help build UIs. And there still is so much to explore in how to best empower developers get better at their job and ship real products even faster!

New: Prompt caching with Claude. Caching lets you instantly fine-tune model responses with longer and more instructive prompts—all while reducing costs by up to 90%. Available in beta on the Anthropic API today: https://lnkd.in/gmj22PYH. With prompt caching, you can reuse a book's worth of context across multiple API requests. This can also reduce latency by up to 85% on long prompts. Use cases include conversational assistants, document processing, agentic tool use, and the ability to talk to books, papers, documentation, podcast transcripts, and other long-form content.

⛔ Stop Team Topologies ⛔ The highly valued Team Topologies (TT) book provides a clear and practical framework for optimizing team structures in software development. I, too, found the book’s advice and emphasis on effective team interactions very attractive. However, as Dave Snowden and the #Cynefin framework remind us, we cannot simply reuse other peoples solutions to solve our own complex problems. Labelling TT as best practices and restricting topologies to just four team types and three interaction modes without demonstrating a causal effect grounded in proven principles is probably even harmful. Instead I have come to appreciate how to make use of the fundamental principles and laws underpinning the book in a less rigid way. Here is a small sample of laws(not from TT) I have found useful: 1️⃣ Parkinson’s law: Work expands so as to fill the time available for its completion. Do shorter Sprints, cycles, meetings, Timeboxes, etc. 2️⃣ Gall’s Law: A complex system that works is invariably found to have evolved from a simple system that worked. Don't build or fix a complex system as a whole. Find the right granularity and introduce simple elements that work and keep the system working at all times. 3️⃣ Engelbart’s law: The intrinsic rate of human performance is exponential. Start small and slow to "get better at getting better.” 4️⃣ Vierordt’s law: Retrospectively, “short” intervals of time tend to be overestimated, and “long” intervals of time tend to be underestimated. 5️⃣ Larmans Law: Organizations are implicitly optimized to avoid changing the status quo. For more on TT and fundamental laws, frameworks and principles: https://lnkd.in/esFHE-3C