Conversation

The idea that a major version bump reflects a previous mistake in your API design continues to prove surprisingly (to me) controversial I think my position on this issue is best described like this: major version bumps should be rare and apologetic, not careless and frequent
5
22
I think I see what's happening here Will and I are thinking about Python libraries, where it's possible to start a new library (like sqlite-utils or Rich) with a strong idea of the scope of the project Which means aiming not to break backwards compatibility is a reasonable goal
1
5
But there are plenty of other categories of software where the scope is FAR less obvious at the start of the project, and where all kinds of outside factors may force backwards incompatible changes whether you like it or not So I think "FAILS" offended people who work on those!
5
11