>It's a counter example in the context of a discussion about alternative ways to enforce (or not) rules, but not in a discussion about the complexity of the rules themselves.

I guess I'm trying to distinguish between rules (that are enforced) and standards or principles that are adhered to voluntarily, with the point being that you can in many cases reduce complexity in the rules by moving it somewhere else that doesn't require coercion to operate.

>As for enforcement, I agree it's interesting to look at examples of cooperative processes, but a relevant distinction is that breaking rules on the internet don't usually have consequences to other people.

I would think they generally would: If someone doesn't follow the standard then compatibility is impaired and the users of that software won't be able to interact fully or at all with users of software that does comply with the standard. That certainly negatively impacts them if they desire that compatibility, we just haven't chosen to make that negative impact illegal/punishable in most cases.

>When they do, we do get calls to "enforce" the rules, e.g. in the context of net neutrality.

Network neutrality is probably a bad example of "internet" regulation because despite the name it's really telecommunications regulation, which is a very different animal. The need for network neutrality comes from the existence of other telecommunications regulations: If (hypothetically) Uncle Sam didn't exclusively license the wireless spectrum or provide only select companies with massive subsidies and access to eminent domain to lay cable then there would be no call for network neutrality, but once the government steps in to establish a monopoly or cartel it then has to limit its power.

I would think the better example of your point would be Microsoft being hit with antitrust for breaking compatibility with Java in breach of the license.