Carrier's post is very long and I don't have time to analyse it in a whole lot of detail, but do go and have a look. As it happens, I agree with most of what he says, even though he claims that his is a moral realist position. In part, this shows how slippery terms like "moral realism" can be. His analysis does not show that two people who are fully rational but start with different desires will end up acting the same way in what is otherwise the same situation. What it is most rational to do for person A with initial desire set X will not be most rational to do for person B with initial desire set Y - or at least Carrier has not shown otherwise. There may be an objectively most rational thing for person A to do - i.e., the thing that will most conduce to her achieving her goals, obtaining her desires, and so on - and an objectively most rational thing for person B to do. But it is a leap of faith to assume that there is a single most rational thing for any person at all to do in that situation, irrespective of her initial desire set.
So the question is this: When someone says that phi-ing is morally obligatory in circumstances C, does she mean, or does her meaning at least include, the claim that phi-ing is the most rational thing for any person to do in those circumstances? If so, an error theory looms because across a vast range of circumstances no such claim is likely to be true.
But Carrier is right that morality takes a reasonably determinate form and is in that sense not just arbitrary. Yes, given the common human desires for a degree of social peace, etc., it's likely that all societies will develop codes of conduct for their members that have a fair bit in common, and it's worth exploring what it might be.