''NukeTheSiteFromOrbit'' in the context of LegacyCode means RewriteCodeFromScratch . It derives from the movie ''Aliens'', where Ripley says "I say we dust off and nuke the site from orbit. It's the only way to be sure" [of eliminating the Alien creatures]. When programmers are confronted with the daunting task of slowly refactoring code until it's pretty, they often want to throw up their hands and claim that it's too risky to do so. Sometimes they are right, sometimes not.

''But remember: Before you NukeTheSiteFromOrbit, be sure to check for stray aliens hiding in the landing gear...''

LyleHayhurst: It's probably safe to NukeTheSiteFromOrbit when you find yourself PolishingaTurd.

''Most businesses are not willing to NukeTheSiteFromOrbit - they have deployed systems that need to be supported, glodamnit. But ReFactoring is not the solution because it's too glodamned slow. TestInjection is the only real way forward when you're wading through a CodeSewer.'' -- PeterMerel

"That's odd. It's been my general experience that it's /far/ easier for businesses to accept NukeTheSiteFromOrbit than any sort of incremental redevelopment. The whole doing something big and dramatic vibe seems to be very attractive. Funding 'big project to make Foo 2.0' is a much easier sell than 'start long-term project to slowly improve our big ball of mud'. " -- AdrianHoward
* agreed, in my experience.  I think that it has to do with visibility and short-term metrics - refactoring will result in slower updates, and gradual improvement of things, but no big visible thing, while a rewrite will have a splashy public cutover when it's done, so the business managers will get lots of fortune and glory
----
CategoryIdiom