Regression Testing

Regression Testing

Regression testing is a set of test cases aimed at finding defects in already tested areas of the application. This is done not at all in order to finally make sure that there are no bugs, but to find and fix regression errors. Regression errors are the same bugs, but they appear not when writing a program, but when adding a new program section to an existing build or fixing other bugs, which caused new defects in an already tested product.

Thus, we can say that the purpose of regression testing is to make sure that fixing some bugs does not cause others and that updating the build does not create new defects in the already tested code.

There are several types of regression tests:

Verification tests. Conducted to check the correction of a previously discovered and previously opened bug.

Version verification testing. It contains the principles of smoke testing and builds testing: checking the program’s main functionality in each new build.

Regression testing itself – re-execution of all tests that were written and conducted earlier. They are executed on existing test cases, regardless of whether bugs were found during their passage or not.

Testing in a new build of already fixed bugs in old builds. This is done to check if the build update has reactivated old defects.

Some points about how to do regression testing:

• RT generally does not cover the entire application but only those areas that, in one way or another, “are in contact” with the changes in the build.

Further, bugs already closed earlier are tested.

1) Regression testing is recommended several times (3-5). Therefore, to save precious time (and, perhaps, to get rid of “routine”), the power of test automation is actively used in regression tests.

2) Carrying out the final RT, for which tests are selected by priority, determined by the most significant number of errors found.

Regression testing is also actively used in extreme development.