Saw an posting on slashdot this morning, followed the links and came up with a few interesting articles. One is about Vista missing its ship date and I'm not suprised. And not because its Microsoft. I firmly believe that it's very difficult to predict the duration of a software development project. And especially one started completely from scratch. As we all know there are a countless number of issues affecting the required development time for a project. So the question is: how do we get better at predicting project timelines? Is realistic or even just possible to give up trying to accurately predict development time and perpetually rely on the spiral development model? Or will we always be forced to rely on empirical models for estimating? Personally, I think both are poor solutions. The never-ending software project is bad for everyone; constant patching is very problematic to say the least. Regarding the empirical model, basing estimation of a new project that has never been done before on something that was done in the past will always be inaccurate. I believe a new approach needs to be developed to solve this problem.
And I also had to throw in one about Vista's security claims. Why does Microsoft continue to say things like this? I don't understand how this helps them in any way.
The World As Best As I Remember It - blogs.msdn.com via slashdot
Microsoft: Vista Most Secure OS Ever - betanews.com via slashdot
