While the methodology (test-first or test-last) plays a major role in the efficiency of the invested time and resulting quality of the tests, in this post I’d like to highlight some of the immediately visible improvements that automated testing strategies introduce with regards to development speed.
While developing you continuously roughly verify whether your implemented code works as expected. Considering a web application environment, this means to start the app in the browser and to check whether everything works. This is connected to a lot of effort, you need to open the browser, click through the UI to arrive to the point where your implemented functionality will take effect and then verify the according output. If it doesn’t yet work, the cycle needs to be repeated. This takes time!
Instead, using a test-driven approach you would
- write a small test, instantiating your object, invoking – say – the method on it you wrote and asserting the according result.
- check whether the bar/dot whatever your xUnit framework is, is green or red
- if it’s red, change your implementation, re-run the written test etc…
Whenever you are tempted to type something in a print statement or a debugger expression, write it as a test.
This really depends on the quality of your tests and on the overall test coverage. There are a lot of papers analyzing this kind of relationship. From my experience I can say that it has definitely a positive effect. What I’ve seen is that the kind of bugs change. You’ll experience less source code defects (i.e. missing null reference checks). What usually remains are small bugs that are quickly fixed and those that arose due to a misunderstanding between the customer and developer.
As with every thing, there are also downsides. Usually a rule of thumb is the “less code you write, the less you have to maintain”. Tests help you in the maintenance phase of your product, while fixing bugs or extending with new functionalities in that they provide you with a safety net that ensures you didn’t break anything existing. But on the other side, writing tests means writing additional code that needs to be maintained as well!
Therefore, writing good, maintainable tests needs to be learned and that learning curve shouldn’t be underestimated.
So, give unit-testing a try and don’t give up immediately. The initial learning phase will be quite tough, but once you got it, you cannot live without ‘em any more ;).
This guide will introduce you to the world of Software Architecture!
This 162 page guide will cover topics within the field of software architecture including: software architecture as a solution balancing the concerns of different stakeholders, quality assurance, methods to describe and evaluate architectures, the influence of architecture on reuse, and the life cycle of a system and its architecture. This guide concludes with a comparison between the professions of software architect and software engineer.