5 ms·
Robert Martin seems to cherry pick stats and trends and whatever that fits his world view. In Clean Architecture he cites a study with N=1 as proof TDD works. I
by jmartinpetersen 7y ago
Robert Martin seems to cherry pick stats and trends and whatever that fits his world view. In Clean Architecture he cites a study with N=1 as proof TDD works. I have no hope for him.
- billfruit 7y agoI generally dislike building arguments on top of statistics, rather than on theory which to my eye almost always looks more convincing. With statistics it is as if the burden is passed on to the reader to check the veracity and validity of the said numbers, where as if it is an argument constructed out of theory, the whole of it is present there for the reader to judge intrinsically on its own merits.
- gunnihinn 7y agoThat restricts you to mathematics, logic or philosophy. Engineering, or the rest of the world, operates _at best_ on statistics.
- billfruit 7y agoProgramming in general and Software correctness does not depend upon statistical arguments, so perhaps such a view could be taken for software engineering too. Also even considering aesthetics and elegance, I find use of statistics to power a reasoning, ungainly, compared with other means of arguing the same thing.
- crimsonalucard 7y agoYou're right. Why are you voted down. Programming doesn't need to be based on statistics it's largely a deterministic system amenable to proof and logic. This application of "engineering" onto programming is largely the result of people not truly understanding the system they are dealing with. The system isn't some sub-sonic jet that will be subject to unknown turbulence, it's just axioms and rules.
- meowface 7y agoIdeally you'd have some of both.
- sooheon 7y agoI'm genuinely curious what the stats backed world views you hold are.