News For Dev-ops

Reader Otul Osan wants to know what the use case for C:\ WINDOWS\ WINSTART.BAT was, compared to C:\ AUTOEXEC.BAT and when exactly during system startup it launches. In Windows 95, you could create a winstart.bat file in your Windows directory. During startup, the virtual machine manager initializes and creates the so-called "System virtual machine" (the "System VM"), which is the virtual machine that all Windows programs run in. But before running the user-mode kernel in that virtual machine, the virtual machine manager runs the winstart.bat batch file if it exists. In pictures: First, we boot up MS-DOS and the command prompt. (Note: All diagrams omit lots of details not relevant to the discussion and are not to scale.) Stuff (unused) MS-DOS The box labeled "Stuff" is a catch-all for random things that go at low addresses, like the interrupt vector table and the BIOS data area. Next, command.com runs autoexec.bat , which might install so...

Starting today, Amazon EC2 R8a instances are now available in Canada (Central) Region. These instances, feature 5th Gen AMD EPYC processors (formerly code named Turin) with a maximum frequency of 4.5 GHz, deliver up to 30% higher performance, and up to 19% better price-performance compared to R7a instances. R8a instances deliver 45% more memory bandwidth compared to R7a instances, making these instances ideal for latency sensitive workloads. Compared to Amazon EC2 R7a instances, R8a instances provide up to 60% faster performance for GroovyJVM, allowing higher request throughput and better response times for business-critical applications. Built on the AWS Nitro System using sixth generation Nitro Cards, R8a instances are ideal for high performance, memory-intensive workloads, such as SQL and NoSQL databases, distributed web scale in-memory caches, in-memory databases, real-time big data analytics, and Electronic Design Automation (EDA) applications. R8a instances offer 12...

Z.ai's GLM-5.2 FP8, NVIDIA's Nemotron-Nano-12B-v2, and Z.ai's GLM-OCR models are now available on Amazon SageMaker JumpStart, expanding the portfolio of foundation models available to AWS customers. These three models bring specialized capabilities spanning long-horizon agentic engineering, efficient hybrid reasoning, and advanced document understanding, enabling customers to deploy high-performance, scalable AI solutions on AWS infrastructure. GLM-5.2 FP8 is optimized for long-horizon tasks and agentic engineering workflows such as full-cycle software development from requirements to deployment. It delivers a substantial leap in long-horizon task capability over its predecessor GLM-5.1 and, for the first time, provides a truly usable 1M-token context window, enabling it to handle project-level engineering context, execute long-running tasks reliably, follow engineering standards consistently, and complete full development workflows in a single task. NVIDIA-Nemotron-Nan...

The next generation of Amazon OpenSearch Serverless now supports up to 10,000 collections within a single collection group, increased from the previous limit of 1,500. Collection groups organize multiple collections and enable them to share OpenSearch Compute Units (OCUs), even when the collections are encrypted with different AWS KMS keys. With this higher limit, you can consolidate significantly more collections into a single collection group and manage them under a shared set of capacity limits. Customers use collection groups to reduce costs by sharing compute across many collections rather than provisioning separate OCUs for each KMS key, while still maintaining collection-level security and access controls. As customer workloads have grown, particularly for multi-tenant applications that provision a collection per tenant, the previous limit of 1,500 collections per group constrained how many tenants could benefit from a shared compute pool. Raising the limit to 10,000 collectio...

We're back, and this is our first minor version bump: 0.1 to 0.2! It's our biggest release yet. This update adds support for local models , lets you run a different agent in each tab with /agent , brings OpenCode into the built-in agent lineup, runs your agent inside WSL where your work actually lives, and makes the Agent Pane richer to navigate and easier to read. It also sharpens Autofix and fixes a long list of reliability issues. You can grab the update from the Microsoft Store , via winget install Microsoft.IntelligentTerminal , or by heading to the release page on GitHub. Let's dive in! [embed]https://www.youtube.com/watch?v=RA3AtxZPPUY[/embed] Local model support is here! We've heard you: one of the most common requests since our first release has been the ability to use local models. With 0.2, we're adding support for local models in the Agent Pane, so you can run your agent workflows against a model on your own machine, keeping them private, of...

We covered the roles your Azure Cosmos DB app needs in our previous post. Now let’s look at what happens after you’ve assigned those roles, switched identities, disabled keys and suddenly staring at a 403 . As we continue this security series, we’ll walk through the most common causes, how to diagnose them, and how to get back to green quickly. The situation You did everything right. You gave your app a managed identity, assigned a data role, switched your code over to  DefaultAzureCredential , and turned off key-based auth like all the good security guidance told you to. Then you deploy, and every single call comes back with this: Response status code does not indicate success: 403 (Forbidden). Don't panic. I've watched a lot of people hit this exact wall, and it's almost always one of a few things. None of them take long to fix once you know where to look. Let me walk you through them roughly in the order they tend to bite. First, read the error the right way Here...