March 2021 Summaries

In a joint webinar on April 7, 2021, Pipe and Hasura will discuss how Pipe uses Hasura to help companies grow on their terms. Pipe is a fintech startup that enables businesses to trade their monthly or quarterly contracts for annual value upfront without taking on equity or debt. The webinar will cover Pipe's architecture, the benefits of using Hasura, and their future technical vision. Presenters include Peter Downs, Director of Engineering at Pipe, and Steven Yi, Head of Product Marketing at Hasura.

This article discusses working with dates, time, and timezones in GraphQL and PostgreSQL. It explains how to use custom scalars like Date, timetz, and timestamptz for date and time data types in GraphQL. The text also covers reusing Postgres types in custom code through webhooks, using server timestamp inside a mutation, querying with ranges, and using subscriptions for real-time updates. Additionally, it highlights the importance of having a multi-cloud and multi-region deployment strategy for distributed Postgres.

The Open Web Application Security Project (OWASP) aims to ensure that web applications are developed with a security-first mindset. OWASP is best known for its "Top Ten" framework, which provides guidelines on structuring secure applications. As the industry moves towards microservices and SaaS, it's crucial for organizations to validate their dependencies according to OWASP standards. Hasura, an open-source data and API platform, ensures that users maintain rigorous security standards by providing in-depth training and tutorials. The company follows a security-first approach and offers features such as access controls, HTTPS enforcement, and audit logs. Additionally, Hasura aligns with the Software Assurance Maturity Model (SAMM), which measures a company's security aptitude across five categories: governance, design, implementation, verification, and operations.

The integration of Azure Active Directory with Hasura JWT claims mapping is discussed in this text. It explains how to implement JWT claims mapping when some auth providers do not allow users to add custom claims. The process involves using a JSON object called "claims_map" where keys are session variables and values can be a JSON path or a literal value. A sample implementation process is provided along with the necessary setup instructions for Hasura Cloud, Azure Functions, and Azure AD B2C. Custom claims are added to the Azure AD B2C setup, and environment variables are set up locally and in the cloud app. Finally, the text explains how to use Azure tokens with Hasura instance and provides an example of JWT token claims mapping for non-B2C tenants.

The latest release of Hasura introduces support for multiple Postgres databases. This feature enables multi-tenant architecture and federated data graphs, allowing developers to maintain disparate systems in a unified ops platform. With this capability, teams can add existing databases to Hasura for observability, tracking, maintenance, and migrations within a single console. Additionally, it supports polyglot workloads by unifying various data structures and models into one back-end, leading to increased developer efficiency and operational improvements. Furthermore, multiple Postgres support facilitates the creation of federated graphs, enabling seamless communication across data sets without compromising performance or security.

The Hasura GraphQL Engine version 2.0 was announced in February after six months of engineering effort across the company. This release includes significant refactoring to the core of the product and new feature additions. Some key changes include switching to a "Parse Don't Validate" approach, separating storage of Hasura's metadata from Postgres database, support for multiple Postgres backends, support for SQL Server, and support for inherited roles. Despite these changes, 2.0 remains backward compatible with previous releases after an automated metadata upgrade. The release also includes support for REST endpoints in addition to GraphQL.

In the latest release of Hasura 2.0, users can now add REST endpoints to stored queries and mutations. This feature is significant as it allows developers to use both GraphQL and REST APIs in their projects without compromising on developer velocity or security concerns. By creating a named query/mutation in the GraphiQL API explorer and pressing the REST button, users can provide a helpful name for the endpoint, specify its suffix path, and choose which verb it should respond to. Hasura also supports traditional body parsing for dynamic variables and dynamic URL segments. This update enables developers to maintain and deliver REST endpoints where needed while still benefiting from GraphQL's features.

The latest version of Hasura, v2.0, introduces significant changes to the GraphQL Engine, including support for multiple data sources and generating REST APIs. This post provides a guide on how to migrate from Hasura v1.3 to v2.0 in different scenarios. In v2.0, users can connect to multiple Postgres DBs or even different databases like SQL Server. The metadata is stored in Postgres and can be configured using HASURA_GRAPHQL_METADATA_URL pointing to a separate Postgres DB. For self-hosted OSS Hasura GraphQL Engine, users can update the docker image to point to the latest version. In case of Hasura Cloud, updates will be rolled out automatically to all existing instances. The latest version of Hasura CLI is required for managing metadata and migrations using the CLI.

A new video series has been launched covering the features of Hasura 2.0. The series will consist of short snapshots and longer-form videos to provide a comprehensive understanding of each feature. The first video focuses on creating a new project, setting up tables, and adding relations. Future videos will delve into more advanced features. Viewers are encouraged to share their thoughts and stay updated with the latest releases by following the creators across various platforms.

The text discusses using PostgreSQL as an alternative to traditional cron systems. It highlights that Postgres is great for eventing and can be used to build a flexible and robust cron system with just a few lines of code. The author explains how time instances can be modeled as rows in a table, allowing for the creation of recurring or one-off events. They also mention benefits such as scaling by running more delivery threads, storing additional metadata, and being able to pause/cancel future events.

In this article, we explore the top 7 GraphQL IDEs that developers should be aware of in 2021. These include GraphiQL, Insomnia, Postman, GraphQL Playground, GraphQL Editor, Apollo Studio Explorer, and Altair Client. Each IDE offers unique features to enhance the development experience for GraphQL users. Additionally, a special mention is made of graphqurl, a CLI tool for making GraphQL queries. These tools are essential for developers working with GraphQL APIs, as they provide interactive query building, documentation exploration, and error checking capabilities.

The JAMStack (JavaScript, APIs, Markup) architecture is gaining popularity among developers for building fast-performing web applications. It combines the best of static and dynamic content by leveraging precompiled HTML files and client-side interactions. Key practices for using JAMStack in production at scale include utilizing CDNs for improved performance and user experience, proper caching settings to avoid stale or buggy pages, version control through Git, automated builds and deploys, atomic deploys for consistency, optimizing for Core Web Vitals, and considering headless CMS solutions.

Data model diagramming (ERD) is a discipline of structuring project parts in an easy-to-follow diagram that exposes cardinality, attributes, and functionality. It helps to think through logical dependencies and ensures the best developer experience when exposing schema to them. Modelling is crucial for GraphQL as it directly exposes table architecture. With Hasura's Multi-Database support, modelling becomes even more important in determining where data should live and how they connect together. The right tool for diagramming should have annotation capabilities, fluidly change diagram type, multi-user edit mode, and export options. Entities are the main components of a model, while connections bring semantic value to the diagram. Understanding cardinality is essential in defining relationships between entities. Picking a model grammar/notation helps communicate a commonly understood representation of models. Optimizing your model involves normalizing data by breaking content models into discrete entities. In practice, start with a simple conceptual model and gradually add fidelity as needed.

This article discusses the development of a new feature for Hasura using Haskell. The feature aimed to provide a safe and efficient mechanism for exposing GraphQL capabilities to environments where REST had better prospects for integration. The team followed an iterative approach, with clear boundaries on what could change at each stage. They also emphasized the importance of strong static types, purity, and expressivity in Haskell, which allowed for "fearless refactoring" during development. The author suggests that while there are still technical improvements to be made to the language, Haskell can effectively facilitate feature development and improvement through commoditized contribution.