The turret code and the ai code are synonymous really. Changing turrets affects AI, so it's really the same thing. Or, at the very least, you have to approach them the same way: make the changes modular, flagged, disabled by default. This means that it's NEVER an easy change, regardless of how small or simple the change might appear. Anything that you do change can quite easily break a mission.
I assume that it should be implied, but I'll spell it out just in case: combo turret support will NOT hit 3.6.10. Consider 3.6.10 feature frozen. Anything that wasn't already slated for it in May, it's going in. Period. It's far too late in the process for that. The goals for 3.6.10 are simply to get what has long since been started finished, get everything cleaned up and tested, and ship it.
You can start putting in requests for 3.6.11 if you want, but I would suggest that you keep a well defined list, because we aren't going to go searching for all of this stuff that people have asked for when the time comes to start work on 3.6.11. We are going to be changing the process for stable builds later on anyway, to make it much easier for us to get things out on a basic schedule: we will switch over and start the new version, take any requests for what people want in it, narrow that list down to what is achievable in the time available, and that will be it, it will be feature frozen after that. Then only changes that relate to features on that list, and any bug fixes obviously, will hit CVS for that build.
