I don't necessarily agree with that. There is good advice in there that makes for better, more maintainable code. Named parameters were introduced to C# to support COM interop, specifically Office, where some methods have a dozen or so parameters, most of them optional. Method overloading is fairly common, but I've seen people get carried away with it. In that case, the design should be rethought because the method is most likely doing too many things.

>But the focus of these guidelines is for *framework* design - AFAICS from a CLS compliant POV.
>For internal C# libraries that are not always the best solution.....