The best way to develop software

The best way to develop software doesn’t exist. There’s only the best way that you know of, for a particular problem, with particular people, given their skills, experience, and constraints.


3 thoughts on “The best way to develop software

  1. Isn’t that (i.e., the best *known* way, given certain conditions and subject to a particular set of constraints) what people usually mean when they say “best” in any context? Maybe I interpret these kinds of expressions too charitably in assuming these considerations are already implicit.

  2. I suspect people usually have some context in mind, though they don’t state it, and even their unstated context doesn’t take enough into consideration. I’ve heard too many arguments such as whether Java is better than Ruby, not realizing they’re arguing whether a hammer is better than a screwdriver. Sometimes you try to get more context and people will say something like “I mean for web development.” as if that narrowed things down sufficiently.

  3. Another issue I’ve observed is when people carry their understanding of “best” from one context into another without acknowledging the underlying assumptions about what “best” means in any given context.

Comments are closed.