Bug 18310 – __traits(isDisabled) returns true for non-function struct member annotated with @disable

Code: ------- struct S { @disable string toString(); } static assert(__traits(isDisabled, S.toString)); // OK struct T { @disable float toString; } static assert(__traits(isDisabled, T.toString)); // NG ------- Expected behaviour: either annotating a non-function member with @disable should be rejected with a compile error, or __traits(isDisabled) should return true in that case.

Well, the spec is very clear on this one. If the symbol is a function and is marked with @disable then traits(isDisabled) will return true, otherwise it will return false. So either if the symbol is not a function declaration or if it's not marked @disable, traits(isDisabled) returns false. I don't see any problems here, since marking @disabled a non-function is a no-op, so it makes sense to return false in that case. Closing as invalid.