> Put another way: not all specs are equal. Some are formal, some are informal.

So is the argument that the language reference is as good as many existing language specifications, and my complaint is only one of semantics?