Why not actually adopt semver? It sounds like it’s hesitation to run up the major version number to quickly, but if your tool has breaking changes most releases then why not represent that in the version number?
Even things that would be fixing issues compared to the spec would cause theoretical breakage to downstream code.
Granted, that's more of an argument against semver in general than an argument against using it here. Show me a minor version changelog and I will offer a changeset that breaks because of it. I think the only real time this isn't true is when a change is _purely_ adding features
That’s fair, and agreed that semver may just be a bad choice here (which is probably why they avoided it). I think my worry is more that using a system which looks like semver invites assumptions as to stability.
An exciting day for the Mypy team! In the last 2 years I've become a full convert that anything "production" should be typed with Mypy checking -- In my day to day it makes maintaining a large, complex project so much easier, both in catching bugs I missed, but also all the IDE hooks and such that make writing good code in the first place easier.
