October 2016 Summaries

The automation of quality assurance (QA) and testing adds value and efficiency to workflows. Tools like Selenium WebDriver are widely used, but other frameworks such as Appium, Watir, WatiN, VSTS, JSoup, Sahi, and the cloud can also be considered for automating QA operations. The cloud plays a crucial role in speeding up testing by taking advantage of parallel testing and making testing more agile and scalable. Cloud resources allow for faster deployment of extra testing resources as needed, ensuring seamless testing that does not hold up continuous delivery pipelines.

Dealing With Test Log Data` discusses the importance of test logs in identifying problems with testing processes, debugging tools, and tracking down intermittent or hard-to-trace errors. It also highlights the need for a good strategy to deal with log data, including using built-in logging features, integrating third-party tools, and setting up a system for full integration of all logs along the development and operations delivery chain. A well-implemented testing log system can provide valuable insights into software behavior, help identify regressions, and enable efficient analysis of log data, ultimately increasing the quality and value of software and its delivery chain.

I've encountered creative solutions to common problems people have with Appium while giving conference talks and providing onsite training for organizations. When it comes to having one set of tests for both iOS and Android applications, it's not always possible to share tests between your applications, but if you'd like to give it a try, here's my advice: abstract your tests to the level of user actions and create an implementation of those actions for each platform. This way, you can write generic tests that can be shared across both platforms. However, if the flows are different between applications, sharing a test might result in messy code and you'll probably be better off keeping them separate. If your applications are identical from a UI point-of-view on iOS and Android, then sharing a test is much simpler, as long as you ensure the accessibility IDs match on the corresponding UI elements in each version of the app. Additionally, automating SMS messages on iOS can be challenging due to the limitations of switching between applications, but setting iOS to show SMS messages as alerts can help. I've also assembled a collection of some capabilities that may prove extremely useful for Appium users, such as autoWebView, ignoreUnimportantViews, nativeWebScreenshot, locationServicesAuthorized, Auto[Accept|Dismiss]Alerts, and interKeyDelay. Finally, the network conditioning API in Appium allows you to simulate various states of connectivity while running your tests, which can be quite useful.

To optimize your mobile testing process and increase app quality, outline a comprehensive testing strategy and choose the right approach between manual and automated testing. Automated testing offers higher device coverage, reliability, and ROI, while manual testing provides user experience and flexibility. A balanced approach of 80% automated and 20% manual testing is ideal for most apps, allowing for faster time-to-market, better user experiences, and reduced bugs and issues. By adopting an agile development environment with integrated testing, testers can work closely with developers to deliver high-quality results. This combination enables the benefits of both manual and automated testing approaches, resulting in improved app quality and reduced testing hours.

As a mobile app tester, it's essential to avoid focusing solely on the UI, as this can lead to poor user experience and uninstallation of the app. Creating an in-house testing infrastructure is complicated and may not be cost-effective, with device management taking significant time and money. While emulators/simulators are useful for early development stages, real devices should be used for testing before releases to ensure reliability and effectiveness. Testing everything at once can be overwhelming, so it's crucial to prioritize the most important aspects of the app based on target market data. Randomly testing an app without a strategy can lead to missed bugs and ineffective bug fixing. Finally, understanding the app itself is vital to creating comprehensive test cases that align with user expectations.

Jenkins is an open-source continuous integration (CI) tool that helps orchestrate development processes with automation. To create a CI pipeline, users need to install the Jenkins Pipeline plugin and other necessary plugins, such as Pipeline Stage View and Multi-branch Pipeline. The pipeline script can be written within the Jenkins UI or in a container file called Jenkinsfile, which simplifies the chained CI pipeline by creating one job that reads the container file. The Jenkinsfile orchestrates different stages, including checkout, commit, acceptance, and testing, using nodes, stages, and build steps. Once developed, the pipeline script can be loaded into Jenkins to create an end-to-end CI solution, triggering automation of build, test, and deployment processes.

Avoiding the UI: Why and How to Run Tests With Scripts` The article discusses the benefits of using test scripts to automate software testing, particularly in repetitive tasks. It highlights that manual testing is often sequential and time-consuming, whereas automated testing can be parallelized and saved time. The article explains how scripted tests can be run from other scripts, controlling the test environment and output, allowing for further automation. It also discusses integrating test automation with continuous delivery (CD) systems, such as Jenkins, to minimize UI interaction. By automating testing, developers can save time and improve efficiency, making it a valuable approach in modern software development.