May 2022 Summaries
11 posts from LaunchDarkly
Filter
Month:
Year:
Post Summaries
Back to Blog
Edge functions are serverless cloud functions that run on and are replicated across edge nodes, often synonymous with CDN (content delivery network) nodes. They can help remove latency by moving serverless backend code closer to the client. Some key benefits of using edge functions include modifying request/response bodies, enabling A/B tests without flash of rendering, immediate user redirection, and modification of HTTP headers. However, there are potential tradeoffs such as limitations on certain providers and additional latency involved in retrieving data from a single location. Cloudflare offers multiple options for moving data to the edge, including Cache API, KV, Durable Objects, Transactional Storage API, and D1 (still in beta). These tools can help bring your data closer to the edge, improving performance and user experience.
May 31, 2022
2,227 words in the original blog post.
Agile methodologies, such as Feature-Driven Development (FDD) and Test-Driven Development (TDD), are crucial for success in the modern IT landscape. FDD combines industry best practices into a five-step framework that focuses on user needs and breaks features into smaller pieces. TDD emphasizes writing tests before code, encouraging small, incremental tests, frequent commits, and lean code. Both methodologies have distinct pros and cons, with TDD being more suitable for smaller projects and FDD being versatile and comprehensive for larger or complex projects. Feature flags can be used in conjunction with FDD to minimize the risk of deploying features, facilitate testing in production environments, and provide developers more freedom in their code implementation.
May 26, 2022
1,665 words in the original blog post.
LaunchDarkly has integrated with Zendesk, allowing teams across various departments to access feature flag variation information for a user directly from within the Zendesk ticket. This integration benefits support teams by providing them with real-time data on users' experiences without needing to switch applications or communicate with other teams. Companies such as FullStory and Loom have found this integration valuable, improving their customer service capabilities and overall efficiency.
May 24, 2022
477 words in the original blog post.
"Testing in production" refers to running code on production servers using real data from real users without showing the new behavior to most users. This practice is often used during the final stages before releasing software to a broad audience and is not a substitute for quality assurance or unit testing but an extension of testing into the most realistic environment possible - the real world. Testing in production can be implemented in two ways: to prove or disprove a stated hypothesis, or to ensure that changes do not cause negative effects on customers or the environment. While it is not a replacement for other types of testing and does not eliminate the need for a QA team, testing in production provides accurate data by testing against real users and their unique behaviors. Feature flags are crucial in safely testing in production as they allow developers to make smaller changes, easily reverse course, and progressively deliver new features based on region or user group. LaunchDarkly's feature management platform enables teams to test software changes in a production environment at high frequency and scale while reducing the risk of testing in production.
May 19, 2022
1,952 words in the original blog post.
Software development can be risky due to factors such as differences between environments, inability to scale, vulnerabilities in supply chain, buggy code, and insufficient monitoring. However, DevOps practices offer opportunities to mitigate these risks through automation, speed, and efficiency. Strategies for reducing risk include leveraging containerization and infrastructure-as-code, ensuring a secure software supply chain, utilizing static and dynamic analysis, implementing effective monitoring and alerts, and incorporating feature flags. These strategies can help developers deliver safe, reliable, and secure software while minimizing disruption to end-users.
May 19, 2022
1,481 words in the original blog post.
LaunchDarkly, a feature management platform, has expanded its commitment to healthcare customers by enabling Business Associate Agreement (BAA) with clients who require it or wish to use protected health information (PHI) on their platform. This move aims to help more healthcare companies innovate faster and improve patient care. The company already assists many healthcare organizations, including top health insurance providers, in making safer and more efficient software releases. LaunchDarkly does not need personally identifiable information (PII) or PHI and encourages customers to limit sensitive data exposure. It also offers various security features such as encryption in transit, multi-factor authentication, granular access controls, full audit logs, private attributes, secure client-side evaluation, Relay Proxy, and Approvals to help clients meet their security and compliance needs.
May 18, 2022
557 words in the original blog post.
Feature branching is a popular method in software development that allows developers to work independently on specific features or bug fixes without affecting the main codebase. This approach helps maintain high-quality code and enables faster code update approvals. However, feature branches can sometimes cause issues such as merge conflicts, lack of communication among teams, and increased complexity due to numerous branches. To avoid these pitfalls, developers should limit branch lifetime, ensure effective communication, manage the scope of feature branches, adopt a naming convention, and use feature flags for better control over code updates. Feature flag management solutions like LaunchDarkly can help optimize feature branching by reducing the need for additional branches while ensuring stable, ready-to-release code with faster development cycles.
May 17, 2022
1,524 words in the original blog post.
Astro is a static site generator (SSG) that offers several innovative features, such as generating zero client-side JavaScript by default and allowing users to use any framework or none. It also supports partial hydration for selective component rendering and dynamic Markdown content usage. Additionally, Astro makes it easy to include various types of static assets without configuration. The platform is compatible with LaunchDarkly's Node.js SDK and client-side JavaScript SDK, allowing developers to integrate both tools seamlessly.
May 12, 2022
1,434 words in the original blog post.
The Relay Proxy is an open-source project supported by LaunchDarkly that enables multiple servers to connect to a local stream rather than making several outbound connections to their streaming service. It can be configured to carry multiple environment streams from multiple projects. LaunchDarkly uses the Relay Proxy internally to sit in front of its "catfood" instance, which is an internal private instance used for pooling flag information. The Relay Proxy acts as a caching mechanism and backup, allowing services to continue functioning even if the catfood service breaks. Deployment is done with Spinnaker, an app service built at Netflix that automates deployment processes. Upgrades are easy, as they can be managed through a pipeline that automatically deploys updates across all regions.
May 10, 2022
597 words in the original blog post.
LaunchDarkly is good at feature management, but there are certain situations where using feature flags may not be beneficial. These include using API secrets as flag variations, relying on flag values for configuration needed at startup, putting the feature flag directly in the functionality, and using them for email blasts. In these cases, other tools or methods should be considered instead of feature flags.
May 05, 2022
600 words in the original blog post.
LaunchDarkly is a feature management platform that allows users to release functionality into applications in various ways. Some of the most common methods include using the Workflow Builder for process-driven workflows, targeting rule-based rollouts and scheduled changes, leveraging flag triggers for integration with different platforms, and setting flag prerequisites for dependencies on other feature flags. These features enable teams to manage their application's features effectively and ensure a smooth user experience during updates or new releases.
May 03, 2022
993 words in the original blog post.