2 ms·
Compatibility is not the only thing version numbers are used for. If we assume bugfixes are a lot more common than new features, then being able to distinguish
by lmm 3y ago
Compatibility is not the only thing version numbers are used for. If we assume bugfixes are a lot more common than new features, then being able to distinguish versions that contain new features from versions that don't is useful, if only for documentation. And if it's not useful to you then by all means treat increments to minor / patch the same when consuming other libraries, and always bump at least the minor version when you make a new release of your own libraries.
- TylerE 3y agoI actually find semver worse than useless for that. Given two versoins, 1.1.10, and 1.2.18, which is newer? Who knows!
- conradludgate 3y agoWhy do you need to know that?
- lmm 3y agoI suspect they meant the other way around - 1.2.18 is always newer than 1.1.10. But if you're running 1.1.18 and see 1.2.10, it would be nice if it were easy to tell wmhether your bug from 1.1.18 is potentially fixed in 1.2.10 or not.
- TylerE 3y agoActually I didn't. Perhaps the 1.1.x branch is for legacy clients and only gets the occasional crucial backport, where the 1.2.x branch is the active you-should-use-this-one-for-new-stuff version, and sees frequent feature releases. It is the very treatment of semver as being equivalent to a monotonically increasing decimal that I object to.