Worse is better
Also: Gabriel, the New Jersey approach
Richard Gabriel's uneasy thesis that simpler, uglier designs often beat better ones by spreading faster.
In a 1989 essay, Lisp veteran Richard Gabriel set up a contrast he found troubling. The "MIT approach" (the right thing) prizes correctness, consistency, and completeness even at the cost of simplicity. The "New Jersey approach" (worse is better) prizes implementation simplicity above all, accepting an inferior interface and some incorrectness if it keeps the system small and portable.
His unhappy observation: the worse-is-better designs — C, Unix — win, because they are easy to implement everywhere, spread like a virus, and improve later, while the right-thing designs are still being perfected. Quality that ships beats quality that's correct.
Gabriel spent years arguing with himself about whether this was good or bad (he wrote rebuttals under a pseudonym). The essay endures because it names a real and uncomfortable dynamic in how computing ideas actually win — relevant to why the late-bound, elegant systems of Smalltalk and Lisp lost ground to cruder rivals.