I generally agree with you about the necessity of complexity, but the "law" of the internet is actually a counterexample: Nobody enforces it. If you want to start with something that vaguely resembles HTML and make a bunch of changes and call it XML and start using it, nobody comes to your door with a warrant for your arrest and locks you in a cell, or even sues you for breach of contract. It's quintessentially private ordering, where everybody does it the same (to the extent that they do) voluntarily because in those cases compatibility with others is more advantageous than diverging from the standard.
And it is probably worth looking through the laws for instances where consensus-based processes like that could, perhaps not eliminate, but move some of the complexity from the regulatory system where it serves as a trap for the unwary into voluntary standards that work because it's in everyone's interest to follow them.
At the same time, you can still have a set of guidelines and issue warnings (but not penalties) for violations so that people will have notice that if the evil the guidelines are intended to prevent occurs as a result of their violation they'll be responsible for it. Sort of like a compiler warning. But if no evil results because you're doing something sensible but merely unanticipated, or because you've taken some alternative precautions with similar effectiveness, then no penalty would attach as long as no evil occurs.
This approach probably works best where the consequences of a violation don't involve risk of physical harm. Allowing a certain amount of risk of economic harm that the perpetrator is required to compensate for or be punished for causing in the event it occurs is much more palatable than the same when the risk is physical injury, because economic harm can be perfectly compensated for or insured against in ways that physical injury obviously can't.
> I generally agree with you about the necessity of complexity, but the "law" of the internet is actually a counterexample: Nobody enforces it.
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.
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. When they do, we do get calls to "enforce" the rules, e.g. in the context of net neutrality.
>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.
And it is probably worth looking through the laws for instances where consensus-based processes like that could, perhaps not eliminate, but move some of the complexity from the regulatory system where it serves as a trap for the unwary into voluntary standards that work because it's in everyone's interest to follow them.
At the same time, you can still have a set of guidelines and issue warnings (but not penalties) for violations so that people will have notice that if the evil the guidelines are intended to prevent occurs as a result of their violation they'll be responsible for it. Sort of like a compiler warning. But if no evil results because you're doing something sensible but merely unanticipated, or because you've taken some alternative precautions with similar effectiveness, then no penalty would attach as long as no evil occurs.
This approach probably works best where the consequences of a violation don't involve risk of physical harm. Allowing a certain amount of risk of economic harm that the perpetrator is required to compensate for or be punished for causing in the event it occurs is much more palatable than the same when the risk is physical injury, because economic harm can be perfectly compensated for or insured against in ways that physical injury obviously can't.