This thread definitely got some emotional responses from some people, and the commits here are already giving me a headache. I've spent the better part of the day trying to sort out what code between 8841, 8842, 8843, and 8848 was actually needed to fix _just_ the bug in #2659, and what was related to this new feature. I still can't tell for certain what does what. Additionally, this new feature does seem to be digging into the skin of some people, as I do agree that this behavior change and its adverse side affects are modifying retail behavior (even if for the better). But given the fact that the next release will be the new pilot code, which will allow adding all sorts of new keypresses, I'm now wondering if it would be better to back this change out entirely for the time being, and have it rewritten as a new command, instead of modifying an existing command. As it stands, I can't even be certain that .14 is working as intended, because 8848 was the only commit backported, and I'm not sure that was the commit that was supposed to fix 2659 or not (which was filed by Mastadon before his feature patch even hit trunk). I think everyone would be fine with it only cycling attackers, and not all hostiles, as long as the current behavior of the default R action is maintained. Because currently, this would make it such that there is no key you could spam in the heat of battle and be _certain_ it's the closest attacker (or thing to spew missiles at). Getting rid of that one key is bound to bother some people, so I think this is likely to be a good compromise, only made possible with the new pilot code.