semver.js

// 1.0

semver.satisfies('2.1.0', '>2.x.x') // true

// 2.0

semver.satisfies('2.1.0', '>2.x.x') // false

It's an odd case, since it depends on what you resolve the first x to,

semver.satisfies('2.1.1', '>2.0.x') // => true seems correct, however

semver.satisfies('2.1.1', '>2.2.x') // => true seems incorrect

This is so much easier with ~ definitions, since it's more deterministic:

semver.satisfies('2.1.1', '~2.0.0') // => true is correct

semver.satisfies('2.1.1', '~2.2.0') // => true is incorrect

I don't know if this is a bug, but would rather argue that > , < and >=, <=, with a non-exact version are problematic and should be considered invalid definitions.