[PlaneEater]

The point about each turret only having a single ShellStart is correct for all ships. This results in an imbalance between ships with numerous single barrel turrets and those with twin or triple barrel turrets.

At present each turret can only have a single zone of fire which is ok for bow and stern turrets but mid ship turrets need two zones of fire to be defined which is currently not possible. This can be seen on Marat where only one of the two central turrets can fire to each side.

For FPS reasons it is still not practical to represent a realistic rate of AA fire from a group of ships. The fire rate needs to be reduced to maintain a good FPS. HSFX have addressed this in part by adopting "effective" ranges so that the smaller guns only join in at close range.

The largest game engine omission at the moment, in my view, is the lack of any sort of AI for ships. If we approach this from the pilots point of view, it is unrealistic that your target ship takes no evasive action to avoid bombs or even torpedoes. Creating an AI for ships should be top priority for any ship related coding work.

Ashe

I thought of both those factors, intended to mention them, aaaand... oops. Good catch, asheshouse . I did a test last night that involved putting a battle line of KGV generics within spitting distance of two Tirpitz, and scattered around a few Fletchers and Z class DDs. The Tirptizes won. My hunch is that it has to do with the KGV effectively having only three guns.

Other AI objects have turrets that fire from both barrels - the Sturmovik and Stuka rear gunners, and the B-29 fires from all four barrels with each of its turrets. My hunch is that correcting the multi-barrel turret issue is feasible. I couldn't say for certain about the restricted fire arcs. Worst case, we can very nicely ask Team Daidalos to include a quick code-based fix in their next patch.

As far as AI, yes, ships need to maneuver, dodge, and evade properly, especially against air attack. That's going to involve heavy AI scripting and pathfinding, both of which I know practically nothing about other than I know I don't know it. Any AI routines added into that will also have to be aware of any new features that are added such as more complex damage modeling. That's all a coding job, so it's beyond my role as an artist. That is where I will need to link arms with some of the skilled coders in the IL-2 mod communities to make that element happen.


For naval AAA fire, I think with a hard look at optimizing effects and making careful compromises, I bet we can come close. As an artist, I can help work on cutting down the visual load of effects to a bare minimum while maintaining the necessary elements of realism. Some aggressive culling of smoke / flak effects and tuning draw distances will probably help. Full AAA activity online is probably going to remain impossible for a while, and that will probably require more compromises.