#Announcements

Postman Adds Hub to Centralize Management of APIs - DevOps.com - Postman introduces a way to live edit and manage OpenAPI specs with Spec Hub, handy if you’re not using OpenAPI generation logic.
Upskill your team on Azure Databricks with an on-demand webinar and Microsoft Learn | Microsoft Azure Blog - 4 free learning materials & webinars for Azure Databricks, going through Apache Spark, Databricks, Machine Learning and generative AI.
Microsoft launches usage reporting for Microsoft Graph API - SD Times - There’s now a reporting API for usage against your tennant’s Microsoft Graph API.
o3 and o4-mini: Unlock enterprise agent workflows with next-level reasoning AI with Azure AI Foundry and GitHub | Microsoft Azure Blog - OpenAI o3 and o4-mini are avaliable in Azure AI Foundry, they also go on to explain all the latest and greatest reasoning + security training which has gone into these models.

#.NET

Toolbox Support for Explicit Assembly References in Windows Forms Out-of-Process Designer - Visual Studio Blog - Visual Studio 17.14 Preview 3 now gives a similar dependency visualization experience to Nuget packages for direct assemblies, which is helpful for legacy .NET frameworks or codebases which have downloaded bundled assemblies directly used.
Preview 2 of the .NET AI Template Now Available - .NET Blog - There are now .NET project templates for AI chat bots using vector database and can come with .NET Aspire support.
Implement client assertions for OAuth client credential flows in ASP.NET Core | Software Engineering - If you’re looking at rolling your own OAuth with Duende’s IdentityServer, then this is a great guide on setting up and cosuming the JWTs with secure enrcyption.
The Aspire Compiler. At the heart of Aspire is a resource… | by David Fowler | Apr, 2025 | Medium - David Fowler makes the comparison of Aspire being a compiler for your multi-app services topology, compiling your architecture and hosting to a managed compiled asset.

Master Reactivity in React: Derivations, Effects, and State Sync - Use React’s useMemo hook on derived values, i.e. an uppercased string drived from another string, to avoid re-renders. However, it does feel like we’re compensating for not having singals like Svelte runes.
Conditionally Preventing HX-Boost In HTMX Using An Extension - A cool article on how Ben created a HTMX extension to prevent hx-boost from being added to specific anchor tags. hx-boost converts anchor links to AJAX requests instead, to then be able to use the repsonse like a SPA. However, you might not want to do this in some cases, you actually might want the link to be a browser location change, but it doesn’t allow specifying this by default on dynamic links.

Scaling the Sound: Fleet Management at Spotify - Daynesh Mangal, Spotify - A really good insight into Spotify’s infrastructure managment and a look into their internal tools.
Mastering Fluent Bit: Installing Fluent Bit From Source - This was more interesting for me to learn about what Fluent Bit is, and how it’s a highly scalable log aggregation tool with outputs that data to your observability tooling or messaging system.
Envoy Proxy: Evolved for Serving LLMs - Vaibhav Katkade & Andres Guedez, Google - Google cloud team explain that their cloud networking is based on Envoy proxy, which evolved to serve LLM request and why inference load balancing alogrithm is preferred over basic round-robin.
Platform Engineering for Cloud Teams - How each role contributes to your platform and what their responsibilities are.
Kubernetes: Sort events from older to newer - DEV Community - A handy options for kubectl to list events by timestamps in descending order, but I’d still use k9s :events.

A practical guide to coding securely with LLMs | sean goedecke & Understanding and mitigating security risks in MCP implementations | Microsoft Community Hub - The risks of LLMs joining your company, and how it can quickly become malicious with more access with MCP or jailbroken prompts.
Prompt Engineering with GitHub Copilot in Visual Studio Code - Enhance VS Code Copilot with custom templated prompting, much like Cursor rules prompts.
The DeepSeek Series: A Technical Overview - How DeepSeek changed the AI game with it’s transformation logic.
The original AI model behind ChatGPT will live on in your favorite apps | Digital Trends - How GPT has evolved and all the big apps which are using it, there are some big companies jumping onboard with OpenAI.

Tech hiring: is this an inflection point? - The Pragmatic Engineer - AI has brought challenges to the software hiring process, with many candidates using AI for tests or projects. This in turn bring back in-person interviews and potentially remove remote hiring opportunity.
6 Ways to Deal with Disappointment Strategically - Some real handy tips/tools for handling dissapointment to avoid burn-out or emotional distress.
(My) three core values to set up an organization for success – @hansdeleenheer - An opinionated article explaining what companies should release from over radical ideals and focus on the realistic meaning of core team values.

Microservices to Monolith - CodeOpinion - Why go from Microservices to Monoliths if you’re not architecting Microservices correctly?
Designing software that could possibly work | sean goedecke - Focus on the most important customer journey.
Distributed Systems 101 - This article explains the pros/cons and complexities around distirbuted services/system, a pretty good summary.

Verifying tricky git rebases with git range-diff - Using a handy git tool, git range-diff to diff many batches of commits, a pretty handy tool to enhance git diff.
Testing Will Become More Important, Not Less Filip Hric - AI is here and it’s coding, your testing needs to be bulletproof to allow for code generation from AI or else you’re risking a lot of pain and bugs.
The Network Times: AI for Network Engineers: Understanding Flow, Flowlet, and Packet-Based Load Balancing - A super in-depth technical article on lower OSI layer networking concepts for AI load balancing.

Netflix is revamping search with AI to improve discovery | TechCrunch - Netflix are looking to use AI for search to improve discoverability of new titles via OpenAI models.

What is gRPC and how it works | ByteByteGo Youtube - gRPC simply explained, going through protobuf and service-to-service communications.
What Happens When a Program Calls Sleeps? | Youtube - A real low level explanation around what the CPU and instructions are doing when you call Thread.Sleep, obviously this should not idle a thead, as that’d be a waste of processing, but how does the thread come back to continue the further instructions?