//devdigest

⚡️ Six performance improvements in .NET 11 Steven Giesel breaks down six notable performance improvements in .NET 11, with benchmarks of .NET 10 vs .NET 11 Preview 6 on an Apple M2 Pro. They include: eliminating boxing in generic enum equality checks — up to 86% faster with zero allocations; removing redundant range and null checks in vector operations (~10%); timezone-rule caching that speeds up DateTime.Now by 48%; Guid parsing optimized by 20–25%; faster LINQ Min/Max on small types via vector shuffling (up to 70–75% on bytes and shorts); and overall reductions in generated code size. The key takeaway: the gains come automatically — just upgrade, with no changes to your own code. ♻️ Subscribe for free now!

⚡️ F# Weekly #32, 2026 A new issue of the F# Weekly digest is out. The week's main releases are the FSharp.Data 8.2.0 data-access library and the Mibo 4.0 game framework, whose notably faster PBR pipeline for 3D rendering reaches GPU-bound performance across the raylib and MonoGame backends. The issue also covers fresh updates to FSharp.TypeProviders.Templates 8.11, SQLProvider 1.5.27, SwaggerProvider, linq2db.FSharp, and Polars.FSharp. Notable projects include Plinth (a local-first markdown notebook built with F# and Tauri), Thuja (a minimalist library for terminal UIs), and FsLangMCP (an MCP server that exposes F# compiler services to AI agents). A key theme of the issue is the growth of the F# MCP ecosystem for AI workflows. From Microsoft: NuGet security enhancements, testing-platform improvements, and Agent Framework updates. ♻️ Subscribe for free now!

⚡️ Azure CTO runs DOOM inside Microsoft Paint — 35 FPS and "undo death" via Ctrl+Z Mark Russinovich, CTO of Microsoft Azure, spent his spare time getting classic DOOM to run inside Microsoft Paint and published the source of the DoomPaint project on GitHub. The project delivers 35 frames per second at 320×200 or 640×400. A standout feature is "undo death": with Ctrl+Z you can rewind frames and effectively turn back time after your character dies. The experiment continues the community's long tradition of running DOOM on anything — from test strips and calculators to cash registers and refrigerators; now plain old Paint joins that list. ♻️ Subscribe for free now!

⚡️ Modern authentication in ASP.NET Core: 2FA and passkeys On the elmah.io blog, Ali Hamza Ansari walks through implementing two modern security mechanisms in ASP.NET Core on .NET 10: two-factor authentication (2FA) and passkey login built on the WebAuthn/FIDO2 standards. For 2FA, he covers registration and login, QR code generation for authenticator apps, TOTP one-time-code verification, recovery codes, and account lockout after failed attempts. The passkey section covers credential registration (attestation) and passwordless login (assertion) with biometric or PIN verification, plus key storage and signature-counter validation. It's built on ASP.NET Core Identity, Entity Framework Core, the Fido2NetLib package, and PostgreSQL. The approach markedly reduces risk compared to plain passwords while making login easier for users. ♻️ Subscribe for free now!

⚡️ An architectural view of ML.NET: machine learning without leaving the .NET ecosystem Abdul Rahman breaks down the architecture of ML.NET — Microsoft's open-source, cross-platform machine learning framework that lets you train, evaluate, and deploy ML models in familiar C# without a Python runtime or third-party libraries. He suggests splitting the solution into three projects: a trainer (a console app that loads data, trains, and saves the model), a schema library with shared data types, and a client (Web API, Blazor, or Worker) that runs predictions. At its core are MLContext as the entry point (analogous to EF's DbContext), the lazy tabular IDataView abstraction, and pipelines composed via a chainable Append and executed on Fit(), much like LINQ. Trainers are grouped by task type — classification, regression, clustering, anomaly detection, ranking, forecasting. For ASP.NET Core, PredictionEnginePool provides thread-safe predictions and hot model reload. ♻️ Subscribe for free now!

⚡️ Azure now checks your Linux machines' security against the CIS benchmark automatically Microsoft has added built-in checking of Linux servers against the CIS Benchmark security standards in Azure — and the feature is already generally available. It works in Azure and Azure Arc through Azure Machine Configuration, continuously comparing Linux system settings against the globally recognized CIS requirements. Azure itself finds weak spots in the configuration, flags drift from the desired settings, and produces compliance reports via Azure Policy; rules can be customized and exceptions defined. Previously this required separate tools or manual checks — now it's all built in, making it easier for security teams to tidy things up and demonstrate compliance with regulatory requirements across many servers at once. ♻️ Subscribe for free now!

⚡️ F# Weekly #31, 2026 A new issue of the F# Weekly digest is out. The spotlight is on Microsoft's announcements for .NET developers: the MCP C# SDK 2.0 release, a new agent for unit-testing polyglot projects, integration of the MSBuild binary log (binlog) analyzer into VS Code, Agent Framework updates for discovering MCP server skills, and the July Visual Studio update with agents powered by the GitHub Copilot SDK. Notable ecosystem projects include MongrelDB-FSharp (a pure-F# client for an embedded database with SQL and vector search), the Axial typed-results library, the reactive Fable.Reactive, and Fable.Giraffe. The issue also covers ten fresh package releases, including Mibo 3.0, plus pieces on Result type design and enterprise AI patterns. ♻️ Subscribe for free now!

⚡️ Haiqu applies quantum optimization to wireless network design Haiqu, a Ukrainian startup building quantum software that makes better use of today's noisy quantum processors, has introduced a new algorithm, DCA, for optimizing wireless networks. It solves the connected dominating set problem — finding the smallest number of relay devices that traffic must pass through to keep an entire mesh, IoT, or ad-hoc network connected. Rather than searching for a perfect solution outright, the algorithm gradually improves an existing candidate. The test ran on the 156-qubit IBM Quantum Kingston processor on a 73-node graph: DCA consistently outperformed the standard LR-QAOA and found a more compact network backbone. Tuning it to the processor's specifics cut the number of complex operations and the computation depth by nearly threefold. The algorithm also reuses tuned parameters from smaller problems on larger ones. The results will be presented at the IEEE ICCCN conference. ♻️ Subscribe for free now!

⚡️ AKS without a hypervisor: Microsoft runs Kubernetes straight on bare metal Microsoft has deployed Azure Kubernetes Service directly on physical hardware — with no hypervisor layer. The new option combines the Azure Local Small Form Factor platform with AKS Arc on Linux and connects automatically to Azure Arc for centralized management. Dropping virtualization dedicates all resources to workloads while keeping the familiar Azure experience — portal, Bicep, policies, monitoring, and GitOps. As Thomas Maurer notes, the solution is built for the edge — retail, manufacturing, field sites — as well as sovereign datacenters and scenarios where maximum performance matters. The same Kubernetes and a single interface now run everywhere, from the cloud to the network edge. ♻️ Subscribe for free now!

⚡️ OpenAI cuts prices for GPT-5.6 Luna and Terra models OpenAI has announced price cuts for its GPT-5.6 model family: the fastest and most affordable Luna is now 80% cheaper, while the balanced Terra is 20% cheaper. As of July 30, API pricing is $0.20 per million input and $1.20 per million output tokens for Luna, and $2 / $12 for Terra; the price of the flagship Sol is unchanged. According to the company, Luna delivers the intelligence of last year's frontier-class models at roughly 6% of the cost per task and nearly nine times faster. At the same time, the API introduces Fast mode, replacing Priority Processing: for GPT-5.6 Sol it delivers up to 2.5× higher speed at twice the price, with no change in quality. The lower prices are also reflected in quota usage in Codex and ChatGPT Work, and will arrive on AWS soon. ♻️ Subscribe for free now!

⚡️ JetBrains adds a debugger to ReSharper for VS Code JetBrains has introduced a debugger in ReSharper for VS Code, closing the last gap in its .NET development toolset for the lightweight editor. In the article, Isaac Levin argues that debugging is a "reactive process," while real productivity comes from pairing debugging with proactive static analysis that catches errors before compilation. ReSharper for VS Code offers more than 2,500 real-time code inspections, quick-fixes via keyboard shortcuts, semantic symbol renaming across the entire solution, and fast navigation (Ctrl+T, Shift+F12) — and now a native .NET debugger powered by Rider's engine. According to the company, this turns ReSharper into a unified, intelligent development environment that combines analysis, navigation, testing, and debugging. ♻️ Subscribe for free now!

⚡️ F# Weekly #30, 2026 A new issue of the F# Weekly digest is out. The week's highlights are FsHttp.Studio, a VS Code extension that runs FsHttp HTTP requests directly from F# scripts with rich response rendering (images, JSON, HTML), and the Fable Lit Fullstack Template 3.2, which adds a cross-platform desktop host, hot module replacement (HMR), and end-to-end type safety built on Fable.Lit and RPC via Serde.FS.Fable. The issue also covers fresh library and tooling releases: Fable 5.13 and 5.12, Mibo.Templates 3.1.1, WoofWare.Myriad.Plugins 10.3.1, and CliWrap 3.10.3. Notable projects include the Tree-sitter F# grammar, the Mibo game framework, the Zest static site generator, and the Fedit terminal editor. ♻️ Subscribe for free now!

⚡️ Databricks to run its business on Azure and integrate Genie with the Microsoft ecosystem Microsoft and Databricks have extended their strategic partnership into the 2030s and deepened their integration of cloud infrastructure and AI to help companies build AI grounded in their own data and business processes. Databricks will move its core operations to Azure Databricks and expand its use of Azure Cobalt Arm-based infrastructure, including the new Cobalt 200 generation (up to 50% better performance and default memory encryption). The Databricks Genie assistant and Unity AI Gateway will gain native integration with the Microsoft ecosystem — Entra, OneLake, Power BI, Purview, Microsoft 365, Teams, and Copilot — providing governed access to data and AI directly within workflows. The goal is to tie AI to trusted enterprise knowledge and keep costs under control. Thousands of companies already use Azure Databricks, among them Unilever, Electrolux, and SMBC. ♻️ Subscribe for free now!

⚡️ Rx.NET 7.0 released The endjin team has released Rx.NET 7.0 (the `System.Reactive` 7.0.0 package on NuGet). The key change resolves a long-standing assembly "bloat" problem: previously, adding an Rx.NET reference in self-contained apps with Windows target frameworks pulled in WPF and Windows Forms, inflating output by roughly 90MB (47MB with trimming, 11MB for NativeAOT). UI framework functionality now lives in separate packages — `System.Reactive.Windows.Forms`, `System.Reactive.Wpf`, `System.Reactive.WindowsRuntime`, and `System.Reactive.Uwp`. Support for .NET 7.0 and earlier has been dropped; target platforms are .NET 8, 9, 10, .NET Framework 4.7.2, and UWP. The release wraps up the technical-debt cleanup, clearing the way for performance improvements and new operators. ♻️ Subscribe for free now!

⚡️ Microsoft releases QDK 1.30 with faster quantum program compilation Microsoft has released version 1.30 of the Quantum Development Kit (QDK) — a platform for building quantum programs that compiles code from Q#, OpenQASM, and Qiskit into the industry-standard QIR format. The main addition is expanded capabilities for the Adaptive profile: expressing loops directly without unrolling, calling functions without inlining, and supporting unbounded loops such as "repeat-until-success." In some cases this cuts compilation time and program size by orders of magnitude. The release also adds new editor quick-fixes (auto-import and converting qubits to arrays) and configurable qubit-loss policies in noise models for two-qubit gates. In total, the release includes more than 40 merged pull requests with improvements to QIR generation, Stim compiler support, and the type system. ♻️ Subscribe for free now!

⚡️ Mistral AI strikes multibillion-dollar deal with Microsoft to expand Azure in Europe French startup Mistral AI — one of Europe's leading AI developers, known for its open-weight large language models and seen as the continent's main alternative to OpenAI — has announced a multibillion-dollar partnership with Microsoft to expand computing infrastructure across Europe (no specific figure was disclosed). The deal integrates Mistral's models into the Azure Foundry service and gives Azure customers access to the startup's France-based data centers. Two models are being added to Foundry — Mistral Medium 3.5 (open weights) and OCR 4 for document processing — with deployment options including Azure Local environments to meet data residency requirements. Microsoft President Brad Smith said the deal combines American and European technology. The partnership aligns with Europe's push for "sovereign AI." ♻️ Subscribe for free now!

⚡️ Background tasks in ASP.NET Core with NCronJob and SignalR Damien Bowden shows how to implement a background task scheduler in ASP.NET Core using the open-source NCronJob package together with SignalR to push execution status to clients in real time. Jobs implement the `IJob` interface with a `RunAsync` method that runs on a schedule and supports dependency injection — access to logging and SignalR hubs. In the example, two job types dispatch begin and end messages every five seconds, demonstrating concurrent and non-concurrent execution (the `[SupportsConcurrency(5)]` attribute), while the frontend displays these messages in real time via the SignalR JavaScript client. According to the author, the approach covers most typical scheduling scenarios while staying simple and maintainable. ♻️ Subscribe for free now!

⚡️ F# Weekly #29, 2026 A new issue of the F# Weekly digest is out. The week's main highlights are .NET 11 Preview 6, which brings F# language updates, and the release of the Mibo 3.0 game framework with MonoGame 3.8.5 support plus the Vulkan and DirectX 12 graphics APIs. The issue also covers fresh ecosystem library releases — Fable 5.11, Thoth.Json 13.0, Argu 6.2.5, SQLProvider 1.5.26, and specialized packages such as BioFSharp 3.1 and FsSpreadsheet.ExcelIO 7.0. Notable open-source projects include the Myriad code generator, the Grace version control system written in F#, the Zest static site generator, and the Fedit terminal editor. The issue also features video tutorials, posts on event sourcing and reflection, and Microsoft's .NET modernization resources. ♻️ Subscribe for free now!

⚡️ .NET 11 Preview 5: new Process APIs eliminate deadlocks when reading process output In the latest part of his .NET 11 preview series, Andrew Lock walks through the Process API improvements in Preview 5. The old approach with `StandardOutput.ReadToEnd()` and `StandardError.ReadToEnd()` could deadlock when reading a child process's output. .NET 11 adds safe methods that read stdout and stderr simultaneously: `ReadAllText()` / `ReadAllTextAsync()`, a line-by-line `ReadAllLines()` with `await foreach` support for real-time processing, `ReadAllBytes()` for raw data, and one-call `RunAndCaptureText()` methods that combine startup, output capture, and waiting for exit. The changes remove common pitfalls, cut boilerplate, and make the APIs more discoverable through modern C# constructs. ♻️ Subscribe for free now!

🎥 A Microsoft lecture on the engineering path to a utility-scale quantum computer Microsoft has published a lecture from its educational series "Building the Modern Quantum Architecture," devoted to the engineering path toward a utility-scale quantum computer — one capable of solving problems classical machines cannot touch. The presenter, Microsoft Technical Fellow Dr. Matthias Troyer, explains why the most promising applications lie in chemistry and materials science — from better batteries to sustainable coolants — and how quantum computing will work alongside AI and high-performance computing (HPC) to accelerate scientific discovery. He also explores the idea of teaching AI the laws of quantum physics to predict material properties more accurately. ♻️ Subscribe for free now!