Better Only If Done Right

Some believe that approach X is clearly better, but only if "done right". Doing X in name only is usually not sufficient. There is a right way to do X and a wrong way to do X, but only experts know the right way.

Cakes tastes better than charcoal if cooked right. Adding too much baking soda is a bad way to bake a cake. Baking a cake only in name is not sufficient. I suspect 'BetterOnlyIfDoneRight' is true of MOST patterns, design principles, etc. both inside the software field and outside of it. There are a great many ways to do things wrongly, especially when doing it "in name only" (which implies you could actually be doing any random thing and calling it X). I expect in software, it's made a bit more difficult since any 'X' is an abstraction that needs to be properly tweaked and decided upon for any given project P - so instead of X, you have X(P), which may or may not offer benefits - perhaps X(P) is only better if F(P) is true. Sometimes a good consultant (they do exist) would be able to look at your project and say: "no, neural networks (or fad X) is NOT the thing for this project. A more traditional approach would work better."

Some suspect that such claims are merely a ploy to get consulting work. Its implied that the knowledge cannot come from a book, but only from experienced and expensive consultants. Such claims may or may not be true, but it is hard to know without extensive background and experience checks. There's lots of room for charlatans in IT, and filtering them out is a gray art.

I expect that by the time you can learn of 'X' and it is exciting enough for you to consider hiring consultants, the information describing X is in a book. Somewhere. And if you could find some people on your team willing to read that book, or who have already read it, and who managed to actually comprehend the darn thing (which is often easier said than done) you probably wouldn't need consultants - your team could tell you whether X even applies to your project, and any benefits one could anticipate from it, and how to properly tweak X so it works for you. But most teams of programmers aren't motivated to learn skills beyond their immediate set unless they can see immediate application.

It is true that there are lots of charlatans in IT. In fact, I think it wise to consider EVERY consultant a charlatan UNTIL such a time as he or she proves otherwise. There's a certain amount of budget you can set aside for potential waste/potential gain in business - a buffer to avoid being burned too badly.

I believe that things like this do exist. You cannot throw Windows coders onto the Unix platform and expect anything useful.