October 2015 Summaries

Testing a mobile app is crucial due to the vast number of downloadable apps and device fragmentation. To develop apps, it's essential to test on as many devices as possible through mobile app testing. This involves understanding standard requirements such as usability, functionality, performance, edge cases, target devices, and other important elements. Testing stages include different life-cycle stages of an app, with manual and automated approaches used depending on the type of test. Manual testing is suitable for checkups, usability, edge cases, and performance, while automated testing is better suited for regression, end-to-end, and integration tests. A combination of both methods is often necessary to comprehensively test a mobile app, requiring a clear testing strategy and research on key devices in the target market.

### Capacity Planning for Teams` Understanding capacity planning is crucial to lead a team, especially in software development. It involves estimating the duration of tasks, accounting for dependencies and other factors that impact the team's workload. This includes understanding the team's obligations, such as meetings, bug verification, and other tasks that affect their capacity. The capacity changes daily due to design changes, bugs found, dependent tasks, and shifting priorities. To accurately plan a sprint, one must consider these factors and account for intangibles like team strengths and weaknesses. By combining level of effort (LOE), duration, and capacity planning, sprint planning can become more routine and efficient.

A quality strategy is a set of activities that ensures consistent delivery of high-quality software that meets clients' expectations. A well-defined quality strategy blueprint includes various facets of testing and other activities to reduce risk and tighten release cycles. Great user stories are crucial in building better software, as they represent small slices of requirements that benefit users or customers. Understanding the Quality Blueprint increases collaboration and communication early on, providing faster feedback and allowing for iterations during sprint planning. A quality product is not easy to build, but it's essential to address quality issues promptly and rebuild a company's image after an incident. Poor software quality can lead to losing key assets to a competitor with a better reputation.

As a lead on a QA team developing mobile apps, I constantly evaluate risk to ensure the schedule and end users are not impacted. Risks can be identified during sprint planning using tools like Fibonacci poker, which encourages developers, designers, and QA to communicate about potential challenges. A checklist of questions for the QA team to consider is also helpful in identifying risks such as data security, performance, usability, quality, team capacity, test automation, manual testing, infrastructure, painful tests, and schedule dependencies. By understanding these risk types, teams can prioritize mitigation strategies and minimize delays and issues with interactivities, ensuring a successful mobile app release.

The author's team implemented a bug bash, an untapped resource in their testing, to address the lack of manpower for thorough testing. The bug bash brought together testers and people without testing background, including developers, project managers, and product owners, to provide fresh eyes on the system. This approach led to the discovery of under 20% of overall bugs, surpassing automated regression's findings. However, the team learned that overusing the tool could lead to burnout, so they adjusted their frequency and focus. They found that gamifying the process with prizes and live dashboards increased participation, while keeping the search focused on specific areas improved results. The use of group chats and Subject Matter Experts also helped prevent duplicates and no-fix issues. The bug bash proved effective in expanding testing resources and improving product quality.

To get started with test automation, it's essential to prepare by setting up an operating system, learning basic Unix fundamentals, and installing necessary tools such as Ruby Version Manager and browser add-ons. A solid programming foundation is crucial, so it's recommended to choose one programming language and stick to it until mastery. Resources like Lynda.com, Codecademy, and Rubymonk offer interactive exercises and courses to help build skills in Ruby or other languages. Behavior-Driven Development (BDD) and Cucumber training materials are also available for learning BDD and Cucumber. Additionally, attending conferences, webinars, and meetups can provide valuable networking opportunities with industry experts and like-minded individuals. Open-source technology options like Capybara, Appium, Selenium, and Watir can be explored to gain hands-on experience with different testing frameworks and tools.

We use Sauce Labs for cloud-based testing of our web apps on multiple browsers and OS platforms, allowing us to run tests in parallel, speed up our test suites, and save time and money. We write tests using the BDD (Behaviour Driven Development) technique with Gherkin-based language and Python code, backed by tools like Behave Pro, which links JIRA issues to test scenarios and provides requirements traceability. Our server implementation supports saving and restoring previous states, known as fixtures, to set up our test environment. We continuously evaluate and improve the testability of our web apps, ensuring correct, readable, and understandable test coverage, and use macros or custom BDD steps to isolate test suites from third-party app changes. The migration to automated testing required a mindset shift for our team, but with the right tools and processes in place, we've improved the quality and efficiency of our automated testing.

Automating the provisioning of servers to treat them like 'cattle' rather than 'pets', using automation software such as Chef to maintain consistency and automate setup. This approach can also be applied to testing infrastructure, with tools like Sauce Labs providing a richer set of functionality for externalizing interface testing. The use of cloud-based testing solutions requires the creation of a VPN tunnel between the internal application and the testing driver machines, which can be achieved using Chef and the Sauce Connect proxy. Additionally, Chef can be used to manage other aspects of the testing infrastructure, such as Jenkins or TeamCity servers, and deploy applications into container runtimes or virtual machines.

The author, Ashley Hunsberger, a Quality Architect at Blackboard, Inc., is writing an open letter to developers to address the common misconception that QA is their enemy. She emphasizes that quality is a team effort and not just owned by testers. The author criticizes "The Hotshot" developer type who thinks only testers own quality and fails to write unit tests or integrate tests for their code, sending buggy code to QA. Ashley encourages developers to take ownership of quality, work with QA to define tests upfront, and run their own tests before submitting code. She also stresses the importance of testing culture and collaboration between teams. The author invites developers to join her in creating a better team dynamic and working together to achieve quality and Continuous Delivery.