>But I don't understand why one would resist changing a spec which is known to be "wrong" and non-RFC compliant. Is the argument that revising it to be RFC-compliant would risk destabilizing more sites than fixing it?
>
+1
I share the same concern with you.
Jiankang Yao