February 2016 Summaries

In this episode of To Be Continuous, Edith Harbaugh and Paul Biggar discuss the concept of staging servers and whether or not they are necessary in a continuous delivery environment. They also talk about how feature flags can be used to reduce risk when deploying new features. The conversation touches on various aspects of software development, including testing, monitoring, and the role of QA teams.

In this episode of To Be Continuous, Edith Harbaugh and Paul Biggar talk with Jocelyn Goldfein about continuous delivery in software development. They discuss the importance of continuous delivery, how it can be misapplied, and its evolution over time. Jocelyn also shares her thoughts on legacy code and why there is still so much of it out there.

Feature flag-driven development is an effective method to improve product development by allowing teams to release features incrementally and manage risks. However, managing feature flags on a large scale can be challenging. Companies often face the decision of building their own internal feature flagging tools or purchasing existing solutions. While building in-house may seem appealing, it requires significant resources, time, and expertise to maintain and scale effectively. On the other hand, purchasing an enterprise feature flagging platform like LaunchDarkly offers a stable, multi-language compatible, scalable, intuitive, and well-supported solution that can be easily integrated across teams and products.

LaunchDarkly has introduced an audit log feature, providing customers with detailed visibility into changes made to their feature flags. The audit log records all alterations to each flag in the system, allowing teams to filter by timestamps or conduct prefix searches on flag names and keys. Currently, the log displays entries for a single environment, but future updates will enable push notifications to platforms like Slack and HipChat, as well as custom webhooks for receiving audit log events. The feature's dashboard allows users to view a history of changes made to their flags, with options to examine patch details and differences between old and new JSON configurations.

Contextual observation is a method of observing how users interact with a product in their natural environment. It differs from traditional usability testing as it involves real-time observation of user engagement with new features post-launch. This approach provides valuable insights into the performance and efficacy of a feature, as well as qualitative feedback on user satisfaction. Feature flag driven development (FFDD) is an essential tool for contextual observation, allowing developers to control the release and visibility of new features. By gradually rolling out features and targeting specific user segments, FFDD mitigates risks associated with launching new features while enabling continuous delivery of software updates. Companies like Google, Facebook, Amazon, Quora, and Asana have successfully implemented contextual observation using feature flags to improve their products' usability and user experience.

The author shares their excitement about joining LaunchDarkly, a small startup where they will be working on the backend. They appreciate the company's focus on building the best product for customers and its efficient communication style. The team has produced impressive results with minimal resources by leveraging existing tools and services. The author is enthusiastic about contributing to the success of LaunchDarkly, which has reached a significant market traction point. They look forward to continued learning, hard work, and making an impact on this successful product.

In this episode of To Be Continuous, Edith Harbaugh and Paul Biggar discuss the nature of open source software development. They talk about how people's motivations for contributing to open source projects can vary greatly, from a desire to improve their coding skills to a need for additional income. The hosts also touch on the challenges that come with managing an open source project, such as dealing with unrealistic expectations from users and balancing the needs of different stakeholders. The conversation then shifts to the topic of workplace productivity and how technology can sometimes hinder rather than help in this regard. Edith and Paul discuss their own strategies for staying focused while working on software projects, including turning off notifications and setting aside dedicated time for coding tasks. They also consider the potential implications of emerging technologies like Google Glass and Apple Watch, which could further blur the lines between work and personal life. Towards the end of the episode, Edith and Paul delve into the issue of funding for open source projects. They explore some of the different approaches that have been tried in this area, such as asking companies to contribute money based on their usage of certain pieces of software. The hosts also reflect on the broader question of what it means for something to be truly "open source," and how this concept has evolved over time. Overall, this episode provides a thought-provoking look at some of the key issues facing the open source community today. It offers valuable insights into the motivations and challenges faced by both developers and users in this space, as well as highlighting the potential pitfalls associated with relying too heavily on technology to manage our work lives.