Tuesday, 19 March 2013

How and when to raise a defect?

Many times when testing, especially in Agile, it's important to communicate with the team any issues be it environmental or code related. A common misconception of Agile is that it's best to not document anything and get things done as efficiently as possible, often this means cutting back on the amount of documentation that is written. 

Whilst I agree that too much documentation is a bad thing, every project needs some form of documentation, be it hand over documents, a wiki on what it is that's been developed and how it works or even a set of Given When Thens (More on this in a future post! - In the meantime look here) of the new functionality, this all helps with future teams being able to look at and read and will hopefully help any future development around the areas. I am however, all too aware, that far too often people send round documents, that 95% of the time don't get read, and just get forgotten about. So it's definitely important to strike the right balance.

This, I believe can also be applied to raising defects. Often when I notice there is an issue with something I am testing, I will investigate the cause of the issue myself, and try and figure out why it's occurring, I do a Defect Dance. Any information that I can add to the defect will help the developers identify and fix the issue (see What to log in a bug).

Often when I am about to raise a defect and after investigating I will speak to a dev and go through the issues and my findings,  from here I will ask if it is worth raising the bug or if it's a quick fix then get them to fix it quickly and move on. It's important to take the right tone when raising a bug with a developer, as for some reason they can be protective over their code! There isn't any point in swearing at them, or even playing a blame game, the bug will be fixed a whole lot quicker if you help them by supplying all necessary information. 

I feel that all too often, people blindly raise bugs, regardless if there is already a bug raised, or if it's a production issue or even when a quick discussion with a developer and they can fix it there and then, it saves time of the tester, saves time of the developer and I believe it helps the team work more closely together. Obviously, if teams are offshore then this isn't really viable, and defect tracking tools are needed... That however, is a whole different story! (I'm sure I will post a blog post or two about that!)

1 comment: