November 2018 Summaries

Steve Winton, a Partner Engineer at GitHub, shares his tools and workflow that help him manage his day-to-day tasks as a remote employee. He relies on various communication tools like GitHub Issues for internal discussions, Slack for quick conversations with his team, Zoom for video conferencing, LICEcap for screen recording, and Keynote for creating visual aids. For productivity, he uses Bear for note-taking and list-making, which syncs across devices and is Markdown friendly. He also utilizes Alfred Workflows to simplify his work with GitHub, Insomnia as an API client for understanding and organizing APIs, and Visual Studio Code for writing code. Steve values the consistency that spaces bring in his coding environment, finding it helpful when sharing code snippets.

Sentry will be at AWS re:Invent on November 26 through 30 in Las Vegas, offering various ways for attendees to meet and learn about the platform. You can visit their booth #611 near the Expo hall entrance or schedule a meeting with sales staff to discuss deeper questions or topics. Sentry is compatible with virtually every language and platform, including cloud hosting with AWS.

Sentry Scouts Meetups aim to bring together software developers to discuss the future of building and using software. The August edition focused on machine learning (ML), a key area of development in the tech industry. ML is distinct from artificial intelligence (AI), with panelists explaining that AI represents a long-term goal, while ML is a more immediate tool for deciphering human behavior. Building effective ML models requires careful consideration of metrics and objectives, as well as understanding of the limitations of the technology. The experts emphasized the importance of respecting individual comfort levels with ML and working together to develop this powerful tool responsibly.

Building a SaaS application should not start with software or architecture, but rather with a monolith that can be evolved over time. A better approach is to begin with a monolithic structure and then focus on the essential building blocks of SaaS applications, including multi-tenancy, tenant isolation, and audit logs. Multi-tenancy is critical in SaaS as it allows multiple customers (or tenants) to operate within a single installation of software. Tenant isolation provides security by allowing code to be written in a way that avoids accidental disclosure of information from other tenants. Scoping operations down to the current tenant is also crucial to prevent security problems and protect tenant data. Audit logs are vital for uncovering what dangerous operations were executed on a customer's tenant, especially when dealing with issues like deleted data. By focusing on these building blocks, developers can build a solid foundation for their SaaS application.

Sydney Ryan is joining Sentry as a Data Engineer, bringing experience from a small startup in Minneapolis to the bay area where they appreciate the milder climate. They are excited about this new opportunity and look forward to working with the team at Sentry. In their free time, Sydney enjoys watching live music, visiting the New Parkway cinema, or spending time with their two cats at home.

The author, a freelance developer, was working on an Electron app that needed proper error reporting due to its cross-platform nature and complexity. When native crashes occurred, he initially relied on workarounds like restarting the renderer process to recover, but this stopped receiving reports from beta testers. He then discovered minidumps, which are crash reports that capture essential information about a crashing process's memory, making it easier to analyze and debug issues. The author integrated minidumps into his Electron app using the built-in crash reporter and Sentry SDK, which provided context and breadcrumbs with the minidumps, helping him identify the source of native crashes and report issues more effectively. With this setup, he was able to provide stack traces, C code to reproduce issues, and even debug information from minidumps, making it easier to analyze and fix bugs in his Electron app.

In this article, Py Surfaces Critical Errors with Sentry, a tool used for monitoring and tracking errors in software applications, has helped the company identify and fix issues with its sign-in process. The error was caused by a misconfiguration of a third-party service integration that was breaking the sign-in flow for users who were brought in through this integration. Py uses Sentry to track and surface critical errors, allowing it to focus on fixing the most significant issues first. The company has fixed the bug and now logs a warning instead of an error when encountering similar issues in the future, while also ignoring non-critical errors from third-party services that are not essential to its functionality.