[User23]

I think I get what you're saying and there's merit to it. However, it appears to me that you are arguing against something the submission doesn't say. If I were you I would reread it carefully and without prejudice, because based on your comments here I believe that you would enjoy learning the concepts it's presenting.

[_0w8t]

I did read the article. My reply was about its conclusion. The article itself advocated a toolbox approach of creating useful tools first that could be written and tested independently and then assembling the application from those. But this has little to do with anticipating future changes, more about of flexibility of assembling the application itself. I.e. the conclusion was not warranted.