I have been out of blogspace for awhile, extremely busy on a client site. So what's up, Nunit is almost a religion on the current project I am working on. My observations here are that the nunit code is often four times bigger than the actual business code we are writing. This strikes me a bit, it takes about 3-4 times longer finishing off a task due to this religious nunit practice and in the end I often wonder what did I actually gain here!!. Also most of the time you have to "stub" your way out of things in order to simulate a "real" unit test, this also adds time in the end (you have to be able to stub class on such a big project, otherwise you will never finish your tasks).
Now we are entering a phase where this...what do we call it ... GUI...right GUI :) is giving more focus. This is where Nunit fails imo, still we are trying to simulate mouse-clicks, scrollbar events you name it and it simply "steels" our time IMO. There are some projects out there, that might raise to the occasion one fine day, but for now we're stuck on that part. Anyway as a consultant you don't care, you just do the job :).
I see the points of TDD.. I also the see the cool thing about testing the actual business requirements in the unit test ....fast forward 1 year.... a new developer comes along and changes something that will break the business requirements. All red. This is all great and I like the shift on how to use the class instead of how is it implemented...but it sure do add some extra time to the development plan :) which you should be prepared to pay for.
Now WSE 2.0 has been released and I didn't have the chance to check it out yet, # 1 on my to-do list.
Currently reading C# design patterns and Test-Driven Development in Microsoft .NET, thumbs up for both so far.
"I stub, there for I am"
-- Me.
10:34:54 PM 
|
