Semver needs to die. Especially for young software and with AI this doesn't make sense anymore. I can't even remember when I released my latest patch release.
Commit
1de76d4
I'm the opposite, I'm at v0.1.739 right now. I think I'm going to switch to CalVer, makes more sense.
Interesting, do you ship features as patches or do you mainly patch things because it is "feature complete"?
A bit of both, honestly.
It's especially visible in pre-launch (e.g. before v1.0) versions. However, even for post-launch software, it doesn't really make sense to release a minor or patch every hour, but that's pretty much the speed at which things are moving right now.
One hour might mean +10 features (minor) and +50 bugfixes (patch). Sure, we can slow things down, but that means customers won't get the "full package" until we decide that enough stuff came together. It also means, that the features and bugfixes that are piling up never really get the real-world stress-test until they are published/tagged, which can be a mess sometimes.
All this was not really an issue before AI, because development didn't move at nearly the same speed as it does right now.
commit 1de76d4
100% agree! I also think that for pre launch versions it is okey to ship features and bugs in 0.0.x versions and breaking changes can go to 0.x.0. Because usually breaking changes can always get migrated somehow. And pre launch software is also the state where the most breaking…