February 2023 Summaries

PostgreSQL provides a rich and extensible type system that allows developers to create custom data types with validation rules, casting rules, and other features. In this article, we explore how to create a custom `semver` data type to represent Semantic Versioning values, with validation rules to make invalid states unrepresentable. We demonstrate how to store each component of the version as a separate field on a composite type, add validation rules using domains, and use a generated column to persist the text representation of the semver type. The article highlights the importance of aligning business logic with the database to improve performance, throughput, and simplify application code.

In this tutorial, you learned how to create a simple shooting game using Flutter and Flame. You also added real-time communication between players using Supabase's real-time features. The final result is a multiplayer shooting game that can be played with friends in real time. Here are the key takeaways from this tutorial: 1. How to create a basic Flutter project and set up the necessary dependencies for Flame and Supabase. 2. Understanding of Flame's core components such as Game, Components, Scene, and CollisionCallbacks. 3. Creating game objects like Player, Bullet, and Gauge using Flame's PositionComponent and SpriteComponent classes. 4. Implementing collision detection between bullets and players using CircleHitbox. 5. Adding real-time communication between players using Supabase's RealtimeClient and presence features. 6. Sharing game states with other clients to create a multiplayer experience without managing your own infrastructure. 7. Using Flutter dialogs to create a lobby and post-game UI. 8. Testing the game by running it on an emulator or physical device. With this knowledge, you can now create more complex games using Flame and add real-time multiplayer features using Supabase's real-time capabilities.

We have released several new features and updates in Postgres, GraphQL, Edge Functions, and Supabase this month, including the addition of pgvector for storing embeddings, a new doc search feature using OpenAI embeddings, support for Views, Materialized Views, and Foreign Tables in pg_graphql, automatic WebP detection, and improvements to Supabase Clippy. We've also added several new features to Supabase, including support for downloading source codes of deployed edge functions, building a Discord Slash Bot, and deploying on DigitalOcean. Additionally, we've reached 45k GitHub stars and are hiring for several positions, including an Integrations/Marketplace Lead, Cloud Platform/Site Reliability Engineer, and Social Media and Community Manager.

Supabase has released "Supabase Clippy," an AI-powered search feature for its documentation site, built with Next.js OpenAI Doc Search Template. The feature is similar to ChatGPT and is trained on Supabase's own docs. It uses OpenAI's embeddings API and Postgres database to store and query relevant documents related to user questions. Currently in MVP stage, the feature aims to improve developer experience by providing a more interactive search interface. Future plans include expanding the knowledge base, caching queries for faster responses, and incorporating user feedback for relevance improvements.

A new PostgreSQL extension called `pgvector` has been released in Supabase, allowing developers to store and query vector embeddings within their database. This enables efficient storage and search of large amounts of data, such as text or image embeddings, using a scalable and flexible solution. The `pgvector` extension introduces a new data type called `vector`, which can be used to store and query vector data in PostgreSQL. It also provides three new operators for calculating similarity between vectors: `<->` (Euclidean distance), `<#>` (negative inner product), and `<=>` (cosine distance). The extension is designed to work with the Supabase platform, which provides a set of tools and services for building and deploying AI applications. With `pgvector`, developers can build scalable and efficient search functions that can be integrated with their products, making it easier to find relevant data and improve user experience.