January 2018 Summaries

Data science has experienced significant growth, with organizations increasingly using data as a strategic asset to derive meaningful insights, often through machine learning and personalized software experiences. R and Python are widely used programming languages in data science for their ability to clean and explore data, while Git and GitHub are essential tools for tracking changes and enhancing collaboration. GitHub repositories facilitate project organization, integration with editors, task management, and code quality through automated builds and tests. They also enable collaboration with product developers and sharing of findings. The data science webcast series will delve into the applications of Git and GitHub in data science workflows, featuring sessions on best practices, a conversation with Pirelli, and insights from the GitHub data science team, as narrated by Bas Peters.

The text discusses the advantages of collaborative coding, highlighting faster onboarding for junior developers, improved code quality, and thorough code reviews as key benefits that have led to the increased adoption of social coding among developers. Inspired by the positive outcomes of such practices, GitHub has introduced features like pull requests and real-time collaboration tools such as Teletype for Atom. The latest enhancement is the support for multiple commit authors, allowing for clear attribution of contributions in commits and pull requests, which is achieved by adding "Co-authored-by" trailers in commit messages. This feature is available across the GitHub platform, including GitHub Desktop, and ensures all contributors receive proper acknowledgment in their contribution graphs.

Project maintainers can now use multiple issue and pull request templates in their repositories to better gather necessary information from the start of a thread. To implement multiple issue templates, a directory named ISSUE_TEMPLATE/ should be created in the project's root, where various templates like bugs.md can be added. These templates can be accessed by appending ?template= followed by the template name to the new issue URL. Similarly, a directory named PULL_REQUEST_TEMPLATE can be used for pull request templates, and these directories can also be placed within a .github folder to avoid clutter in the root directory. The default templates remain functional when no specific template is mentioned in the URL, and further details are available in the documentation by Bryan Clark.

GitHub has introduced a streamlined process for repository owners, collaborators, and prior contributors to report problematic content, such as comments, issues, pull requests, and commit comments, directly to GitHub Support. This new feature includes a contact form for providing detailed information and screenshots, allowing users to classify the nature of the content as spammy, harmful, or off-topic. Additionally, comment authors can report their own comments if they have been edited abusively by others. For more detailed instructions on how to report content, users can refer to the provided documentation.

In 2017, there were significant developments in integrated development environments (IDEs) and GitHub, including the introduction of new tools like Xcode and major updates to existing ones such as GitHub Desktop. The GitHub Professional Services Team, alongside guest experts, hosted a series of webcasts to highlight these advancements, offering recordings for those unable to attend live sessions. The series covered various IDEs, including IntelliJ, Atom, Visual Studio, Eclipse, and Xcode, all in conjunction with GitHub, aiming to enhance users' daily interactions with these tools. Looking ahead to 2018, GitHub emphasized integrated tools and launched the GitHub Marketplace to streamline workflows, with sessions focusing on products like Codacy for code quality, Black Duck for open source security, and Sentry for error reporting.

Data science teams, although different from engineering teams, can enhance their workflows by adopting engineering best practices, as demonstrated in a webcast by GitHub Senior Solutions Engineer Bryan Cross. GitHub facilitates experimentation by allowing data scientists to snapshot and iterate on their work without fear of losing previous versions, ensuring a comprehensive record of both successes and failures. Discoverability is improved through GitHub's search functions, enabling teams to find and share work efficiently, thus avoiding redundant efforts. Collaboration is streamlined with tools like issues and pull requests, which allow for threaded discussions and code reviews, fostering cross-functional teamwork. Results can be easily shared through GitHub's support for Jupyter and RMarkdown notebooks, as well as GitHub Pages for web hosting, all benefiting from GitHub's snapshotting, search, and collaboration capabilities.

GitHub Enterprise Solutions Engineer Tommy Byrd demonstrates how to deploy GitHub Enterprise as a virtual appliance within thirty minutes, utilizing Terraform for automated resource management and detailing the services included with the appliance. The webcast highlights the flexibility of the GitHub Enterprise Management Console API for configuring settings and managing installations, reducing maintenance overhead for administrators by incorporating core services from GitHub.com. The architecture ensures minimal installation overhead, allowing administrators to benefit from a secure and up-to-date product. Byrd also emphasizes the ease of collaboration with the Enterprise Support Team, which relies on support bundles containing critical logs for troubleshooting, as GitHub Enterprise operates behind a firewall without direct access from the support team. Comprehensive documentation and resources are available to assist administrators in resolving issues, and GitHub's support team is readily accessible for further assistance.

This month sees the introduction of two new apps aimed at enhancing API monitoring and team productivity, available on GitHub Marketplace. Moesif offers real-time visibility into live API traffic with advanced analytics, enabling users to analyze customer usage, understand errors, and prevent major outages through passive monitoring and automatic alerts via Slack and PagerDuty integrations. DeepAffects, an emotional analytics application, helps managers and team leads improve team dynamics and productivity by analyzing emotional cues and tones in issue descriptions and comments, offering insights into team happiness, trust, and conflicts to optimize performance. Both tools provide new ways to build better software and foster more effective team collaboration.

The recent integration of Codeship with GitHub Enterprise simplifies the process of connecting GitHub to continuous integration and continuous development (CI/CD) tools by utilizing a cloud-based platform with native Docker support, eliminating the need to manage CI/CD servers or local installations. This integration allows users to host code on a private GitHub instance while running builds on private, single-tenant Codeship machines in the cloud, ensuring that the code is never stored on these machines. It offers the flexibility required for enterprise-level workflows and private source control, while upcoming features include finely-tuned permissions and restricted IP ranges. Users are encouraged to try the integration and anticipate future enhancements that will further unite popular software development tools.

GitHub has introduced a new feature that allows users to gain a comprehensive overview of their projects with progress bars that track tasks categorized as to do, in progress, and done. This update enables users to track project boards from other repositories and organizations by entering the URL of any GitHub project into the note field to create a project summary card, akin to the recently released issue summary cards. Existing cards will automatically update to reflect these changes, eliminating the need for manual upgrades. Users can refer to the documentation for more information on tracking project progress and how to disable tracking for projects or boards that are not meant to be tracked.