February 2021 Summaries

Selenium, developed in 2004, is a popular tool for end-to-end testing but struggles with modern web complexities due to its reliance on XPath, which is less effective for today's intricate and dynamic HTML structures. Tests using Selenium often face maintenance challenges because the HTML elements they reference, such as XPath, CSS Selectors, and data-testids, are prone to change, causing tests to break even when they work from an end-user's perspective. As web applications become more intricate, the limitations of Selenium become increasingly evident, particularly when dealing with third-party applications where HTML is not controlled by the tester. The article suggests that end-to-end tests should focus on user interactions rather than the underlying HTML structure, advocating for tools like testRigor that allow testers to write tests from an end-user's viewpoint, using labels or placeholders to ensure stability and reduce maintenance time. This approach helps maintain test relevance and functionality even as the underlying application evolves or changes frameworks.

Test automation tools are essential for ensuring that products function correctly from an end-user perspective, even if existing metrics indicate all systems are operational. The main challenge with implementing functional monitoring is test stability, as unreliable tests can lead to unnecessary alerts. Effective functional monitoring should involve a continuously running smoke test suite that alerts the relevant parties only when there is a genuine issue, ensuring that critical functionalities like purchasing or registration are operational. Notifications should be sent through various reliable communication channels, such as Slack, email, SMS, and phone, to ensure timely responses to issues that could significantly impact business operations.