January 2022 Summaries

This paragraph provides an objective, neutral, and interesting summary of the text. It covers the key points of the tutorial, including creating a Harry Potter API with Node.js/Express, MySQL, and PlanetScale. The tutorial guides users through setting up their project, initializing the connection to PlanetScale, seeding the database with data, and creating endpoints for querying characters and wands. Additionally, it emphasizes the importance of error handling and provides examples of how to implement it in the API endpoints.

The PlanetScale team implemented a password strength meter in their signup form using an entropy-based approach, which measures a password's strength by calculating the number of attempts it would take for a computer to guess it. This method is more effective than traditional specific rules and provides users with quick feedback on their password strength as they type it in, playing nicely with password managers. The team used the strong_password gem and auto-check-element web component to implement this feature, rendering a meter SVG that displays the user's progress and providing an indication of whether their password is strong or weak.

PlanetScale is an open-source database-as-a-service platform built on Vitess, which powers YouTube and other hyperscale companies, allowing for high-performance queries and connections. Prisma is an ORM that enables developers to write queries in JavaScript/TypeScript, with three products: Prisma Client, Migrate, and Studio. The guide demonstrates setting up a new database, installing Prisma, defining data models, and creating an API route using Next.js to interact with the database. It also covers deploying the database to production using PlanetScale and Netlify, allowing developers to deploy their first PlanetScale database in just a few minutes.

At PlanetScale, our backend API is built with Ruby on Rails. It’s a pretty standard Rails application. We use minitest for our test suite and FactoryBot for creating test data. Our team has invested time into keeping our test suite fast as our app has grown. When working locally, we run individual tests or just one test at a time to keep it efficient. We've also improved the parallelization of tests on CI with 64 core machines, which significantly reduced our test suite running time from around 12 minutes to 3-4 minutes. By auditing and optimizing our factories using FactoryBot, we were able to create fewer objects, reducing our test run time further to around 1 minute. Our approach has improved team productivity by providing quick feedback cycles for feature development.