Granite blog

Why generating package-lock.json on macOS silently breaks your Linux builds — and two ways to fix it.

March is here, which means it’s time to recap all the innovations our team implemented over the past month.

As a solo founder juggling product development, customer support, and growth, every hour counts. You're probably already using AI coding assistants like Cursor to ship faster, but are you leveraging the full power of Model Context Protocol (MCP) servers to automate your entire development workflow?

When you're building a static site—whether it's a landing page, portfolio, or documentation site—you eventually hit the same wall: forms. You need contact forms, newsletter signups, or feedback forms, but you don't want to spin up a backend server just to handle form submissions.

We are finally ready to introduce the first update to the cloud platform this year.

«We built Granite to make cloud infrastructure as simple as pushing a Docker image,» said Alexander Gorshkov, founder of Granite. «Our mission is to help developers focus on their product, not their servers.»

Redis is usually chosen as a cache because of its speed and versatility, allowing developers to provide users with a faster and smoother experience. But after its license change, many companies are looking for an alternative. In this post, we will take a look at the most popular options for replacing Redis as a cache.

When you're building an application that interacts with a database, you’ll often need to establish a connection to the database each time your application wants to retrieve or store data. This can be slow and resource-intensive, especially if your application needs to interact with the database frequently. This is where connection pooling comes in.

The article demonstrates that multi-stage Docker builds significantly reduce image sizes. A single-stage build results in a 258 MB image, while a multi-stage build shrinks it to 6.35 MB, making multi-stage builds a highly efficient practice.

Docker packages your application and its dependencies into a single image for use in a container, but these images often become unnecessarily large. For example, a simple Node.js application using an official image can swell to 644 MB, even though the application itself takes up less than 1 MB. Optimizing the image can significantly reduce this excess size. So what can we do to reduce the image size?