We pay a lot of attention to quality of our products. With the components becoming more and more complex and broad functionality expansion we can’t continue improving our software without a reliable quality control mechanism.
In addition to continuous testing of usability of our components from end-user and programmer point of view, we use an additional code testing system based on NUnit. By now we have over 370 tests that automatically verify code of our components. Testing code accounts for more than 30% of all software code. All main algorithms such as sorting, grouping, serialization, etc are covered with tests.
This makes us sure that when we modify existing code and add new functionality, we don’t encounter unpleasant surprises.
Any software contains and will always contain errors. When we discover errors we create new tests to make sure that such errors will never appear again. This is advantageous both for us and for our users.
To show you, how seriously we approach testing of our software, we will provide some statistical information:
- Number of tests: > 300
- Percentage of testing code in software: > 30%
- Coverage of data level with tests (Dapfor.Net.Data): > 65 %
- Coverage of thread-safety routines with tests (Dapfor.Net.Threading): > 84 %
- Coverage of various algorithms (sorting, grouping, etc) with tests: > 65 %
- Average coverage of entire code with tests: > 35 %
