On 2014-12-15 20:24, Dan Burton wrote:
Was that sufficiently unhelpful? :P I think my conclusion is perhaps that the PVP should remain undefined on this particular question, since I see no obvious answer. Or, since the PVP already encourages a cautious over-estimate of what constitutes a breaking change, perhaps we should go with this fallback: "when in doubt, call it a major change."
I'd add that the level of caution should definitely also be a function of the number of (published) reverse dependencies that the package in question has. (I wouldn't be too worried about unpublished reverse dependencies; if you keep your code hidden from the world, it's *your* problem to deal with breakage that cannot be foreseen by the authors of your dependencies.) Regards,