A survey of web and mobile software testers has offered some insights into how they prioritise what to test, and how they test it. Almost a third said they did not test on Apple’s Safari browser, despite its rising popularity.

Sponsored by Sauce Labs, in January 2015 the report sampled 504 individuals from a variety of companies, countries and industries. It found that which web browsers to test, and which versions, varied across the industry. Befitting its status as the most popular browser, 98% of teams said they tested web-applications on Google’s Chrome, against 95% on Firefox and 91% on Internet Explorer. Only 73% reported the same for Apple’s Safari, despite its growing market share, and just 19% tested on Opera. Almost half (43%) said they only tested the most recent versions of the browsers, against a fifth who attempted to test multiple older versions of each one.

The mobile environment was reported to have more quality issues than the desktop. Respondents said this was because of the newness and complexity of these environments, which meant they were less stable and received more updates. One additional suggestion was the difficulty of getting hold of devices for testing. There was no standard approach to this: roughly a third tested only using simulators or emulators; another third tested only on the devices themselves; and the remainder used a combination of both.

Finally, the rapid uptake of new working processes has produced somewhat mixed results. The influence of continuous integration (CI) and agile prompted 61% to say they now tested earlier in the development cycle, with 69% reporting an increase in automated testing. Not quite half (49%) said they produced better software as a result, however, and only 26% said that they now tested in a more systematic or rigorous fashion. Cloud testing has grown swiftly; 43% of test teams run unit and functional tests in the cloud, with the majority of these tested by developers.