August 2023 Summaries

Temporal's August newsletter highlights various updates and events, including the upcoming Replay 2023 conference in Seattle, featuring a first-ever Hackathon for hands-on application building with Temporal. The conference, set for September 12-14 at Hyatt Regency Bellevue, will host speakers from major companies like Netflix, Microsoft, and AWS, offering insights into how these organizations leverage Temporal. New technology features include updates to the Go SDK and a new functionality for receiving Certificate Expiration notifications. Several on-demand webinars are available, such as setting up the Temporal SDK and leveraging cloud observability endpoints. The newsletter also announces meetups and events like Temporal Happy Hour, Strangeloop, and GopherCon US, where attendees can engage with Temporal experts.

In Spring 2023, Temporal held its second annual Hackathon where a team of three solution architects developed a trivia game using Temporal and incorporating ChatGPT for question generation. The game was designed to scale to millions of users, support multiplayer and unlimited games, built with polyglot languages, maintain game state and durability when things break, have a timeout mechanism for questions, provide a single source of truth for game logic, and be accomplished in just a few hundred lines of code. The backend infrastructure was implemented using two workflows: GameWorkflow and AddPlayerWorkflow. Two frontend versions were created - the first one used SMS to join games and make one player host on their computer, while the second version added browser support for both desktop and mobile users. Key learnings included keeping logic in one place, using two workflows instead of one, implementing moderation, iterating with small pieces to create an MVP at each stage, and optimizing performance through synchronous Workflow Updates. The game is available on GitHub and a hosted version for playing.

Google Cloud Run simplifies deployment of container-based applications, handling infrastructure provisioning and scaling based on traffic. However, it has limitations when used with Temporal Worker applications that require long polling to process tasks. To deploy a Temporal Worker on Cloud Run, one needs to disable CPU throttling, set the minimum number of instances, and use a sidecar container for metrics collection. The OpenTelemetry Connector can be utilized to send metrics data to supported platforms like Google Cloud Managed Service for Prometheus.

Over the last decade, India has become a significant player in the global IT sector, with an estimated digital economy of $1 trillion by 2025 and a workforce of 4.36 million employees as of 2020. The country's software engineering hub is driven by both global outsourcing and emerging startups, including companies using open source Temporal. In response to this growing community, Temporal Cloud has enabled its service in the ap-south-1 (Mumbai) region. India-based developers are an important part of the Temporal community, and the company's co-founder Maxim Fateev will be visiting Bangalore from August 21-25 to engage with local developers.

Parallelism and Concurrency are two distinct concepts often confused in computing. Concurrency refers to handling multiple tasks in an undefined order, while parallelism involves more than one task running simultaneously. JavaScript and Python, despite being single-threaded, can achieve the illusion of multitasking through concurrency. Temporal provides reliable concurrency and parallelism with a durable, distributed event loop. Concurrency is achieved by breaking down applications into smaller parts and persisting state changes, while parallelism is enabled by running multiple instances of an application on different cores or machines. Temporal mitigates most challenges associated with highly parallel systems by providing fault tolerance, concurrency, and distributed computing capabilities.

A Hackathon is being hosted in Seattle alongside the Replay conference for developers interested in building applications with Temporal or rapid prototyping. The event will take place on September 11th and 12th, with a virtual kick-off on the first day and an onsite portion on the second day. Participants are encouraged to form teams and can submit multiple entries for different categories. Prizes will be awarded in various categories such as simplest example of Temporal, breaking Temporal in interesting ways, touching the most features, most unique example, and founders' choice. The event aims to create more opportunities for community interaction and application building using Temporal.

Time-travel debugging, also known as reverse debugging, allows developers to step backward as well as forward in their code execution. This is a powerful tool for debugging production code because it enables developers to see what happened at any point in the program's history. The technique was first introduced with Smalltalk-76 and has since been implemented in various languages, platforms, and IDEs. There are three main approaches to implementing time-travel debugging: record & replay, snapshotting, and instrumentation. While running traditional debuggers doesn't make sense in production, time-travel debuggers can record a process execution on one machine and replay it on another machine for debugging purposes. This technique is particularly useful when dealing with hard-to-repro bugs or when recording the execution of a program is already part of its normal operation.

This article discusses how to build a customer loyalty program using the Actor Model and Temporal Workflows. The Actor Model requires actors to be able to send and receive messages, create new actors, and maintain state. Temporal differs from other actor frameworks in that it is general-purpose rather than specific to one model or system design pattern. The customer loyalty program example includes a Workflow representing a customer's loyalty status and the ability to send and receive messages, create new Actors, and maintain state. The implementation details vary depending on the language used, but the core concepts remain the same. In this example, customers can earn points by performing certain actions, unlocking rewards as they reach different thresholds. Customers can also invite others to join the loyalty program or gift points or status levels to their guests. The Workflow ensures that messages are received and acted upon, and that the state of each customer's account is maintained accurately. The article provides code examples in Go, Java, and Python for implementing this customer loyalty program using Temporal Workflows. It also discusses handling long-lived customers by using Continue-As-New to reset the Event History when it becomes too large. Additionally, the article covers spawning new customers by creating other Workflows and signaling them with a specific message. Overall, this article demonstrates how Temporal Workflows can be used to build an Actor-based application in a highly distributed, concurrent, and scalable way, making it easier to manage long-running processes and maintain state across multiple instances.

Temporal's Roadmap focuses on Approachability, Adaptability, and Applicability to enhance user experience, integration with other projects, and versatility across diverse use cases. Key updates include improving Temporal Cloud's automation capabilities, introducing a .NET SDK, developing a VS Code plugin for TypeScript workflows, working on crucial aspects like workflow versioning, launching global namespaces on Temporal Cloud, and addressing the need for handling larger payloads. The company encourages feedback from users to shape its development process and will discuss the developer ecosystem at Replay, their annual conference.

This is our monthly update newsletter, sent on July 25th, featuring new updates including the release of Temporal 101 in Python, the .NET SDK now available in Public Beta, and several notable technology features such as certificate generation for Temporal Cloud with tcld and improved logging support in the TypeScript SDK v1.8. We also have upcoming meetups and events, including a meetup in Berlin and the DevOpsDays Chicago conference, where our team members will be giving talks and hosting happy hours. Additionally, we have released a new course on Temporal 101 in Python, and highlighted some community favorites such as articles on saga patterns and distributed systems with Temporal.