In January 2013, SmartBear Software launched an extensive study with software professionals to investigate whether they are performing code reviews and, if so, how they are doing them. We also asked them which bug detection strategies they preferred besides code review, and what was working best for them in their quest to improve overall software quality.
Coming in at the bottom of the list as the least commonly used strategy is beta testing. We asked respondents to differentiate between low-volume beta testing and high-volume beta testing, but both were ranked as the least commonly used methodologies: 23% and 7%, respectively.
We found that the most commonly used bug detection strategies include:
- Unit tests (86%)
- Integration tests (75%)
- Functional tests (71%)
- Regression tests (67%)
Unit tests are great because they find problems early in the development cycle, and when bugs are identified early, they are easier to fix. Integration testing is essential to verifying functional, performance and reliability standards, and regression tests are great for covering your tracks and catching bugs that may have been introduced by code changes.
All of these tests can be automated. If you choose not to automate, then users must manually perform numerous tasks costing time, money and in some cases – sanity. Even when developers are shooting for 50% code coverage, you’re still talking about hundreds, if not thousands of unit, integration or functional tests (depending on length of code).
If you aren’t already automating these tests, try a free trial of TestComplete here.
With features like cross-browser Web testing, HTML5 application and distributed testing, you’ll find TestComplete really simplifies Web and windows testing for advanced and novice testers alike.
- Making Code Review Work
- 7 Silly Programming Challenges to Do for Fun
- Programming on Early Microcomputers: A Retrospective