It seems to be a common problem that, although the will is there to improve practices, persuading higher-ups is almost impossible. Although many of like to thi

It seems to be a common problem that, although the will is there to improve practices, persuading higher-ups is almost impossible. Although many of like to think that cold, hard logic is the only suitable method of reasoning with somebody, the reality is that even programmers are somewhat human. We're all proud, emotional souls. If, like me, you've spent half a lifetime on your own, hacking and tinkering with computers, your skills in handling people may be somewhat underdeveloped. I've recognised this in myself and have been trying to do something about it. I can thoroughly recommend Dale Carnegie's classic book: http://en.wikipedia.org/wiki/How_to_Win_Friends_and_Influence_People http://en.wikipedia.org/wiki/How_to_Win_Friends_and_Influenc... In most cases, it should be possible to win over the stubborn manager or sensitive programmer who's blocking any progress towards sane development practices. There's plenty of material in the Carnegie book suggesting how one can gently persuade people, i.e. how to win arguments. In hindsight, a lot of the book's advice seems obvious and corresponds with my own experiences, but I needed somebody to make it all explicit. I have first hand experience of having to placate a sensitive, confidence-lacking programmer. A few times it's been necessary to reject his ideas or code, and he's taken it very personally, being unable to separate in his mind an 'attack' on his work from an attack on him. Once you know how, it can be fairly easy to reassure him. I've never had the problem of a stubborn manager who thinks he knows better, but I have a good idea of how I'd approach him and try and persuade him of the right way to do things. Then again, I still feel like my people skills are great only in theory, and somewhat lacking in practice.