If this is such a good way to run a tech company, why isn't it also a good way to develop software? In other words, why are projects led by BDFLs like Linux and Python so successful, while the design-by-committee approach is usually regarded as producing substandard outcomes?
Depends on the committee. The Ada designers, acting on committee's needs, came up with pretty much what they wanted. Had performance, availability and pricing issues that kept it from taking off as much as it could. Yet, the language's features accomplished their goals very well as proven by a number of empirical studies by military and defense contractors.
So, that's one success story. Anyone have any other examples? I'm asking because I really think this recurring theme is a myth about committee-design languages where the problem really is a prevalence of bad committees.
