@jonorossi, quick question because it's been a while since I worked with .NET in-depth, and I'm not sure I remember correctly: IIRC, the nullable hints get encoded in custom attributes. Does the addition of custom attributes to types and methods cause a binary breaking change that would require downstream code to be recompiled, or not? I don't think so, but if it were, we would probably need to bump the major version when merging this.
P.S.: You can ignore the above question, I just remembered: unlike modifiers (modreqs and modopts), custom attributes are not part of the type/method signatures, therefore the addition of attributes should not cause any breaking changes. I ran a quick experiment which confirms this. So this PR should be safe to merge in a non-major version.
@jonorossi, quick question because it's been a while since I worked with .NET in-depth, and I'm not sure I remember correctly:IIRC, the nullable hints get encoded in custom attributes. Does the addition of custom attributes to types and methods cause a binary breaking change that would require downstream code to be recompiled, or not? I don't think so, but if it were, we would probably need to bump the major version when merging this.P.S.: You can ignore the above question, I just remembered: unlike modifiers (modreqs and modopts), custom attributes are not part of the type/method signatures, therefore the addition of attributes should not cause any breaking changes. I ran a quick experiment which confirms this. So this PR should be safe to merge in a non-major version.