What if instead of trashing the bunker building, the 1.0 bunker modules were replaced with completely different, more... uh... "modular" modules?

IE: instead of tunnel pieces and walls and stuff you got modules that were basically whole bunkers in themselves - not so much designed to connect to each other in the specific ways the old modules do, but to be placed in relation to each other in a loose way only.

Examples:

> Brain bunker that has two entrances and blast doors and isn't just a room for your brain to be in. Brains are mobile these days anyway which imo changes things.

> Craft landing module with sites for turret placements and barriers against horizontal fire

> Uni-directional watchtower/trench thing for area control specifically (so it's not tall because the focus is on controlling the area in front with machinegun fire, not being a wall) image

> Some pre-fab underground bunkers.

> Mining shaft

> Vehicle hangar

> Artillery

> There could be a bunch with a focus on serious subterranean fortifications, since that's what makes sense for very-entrenched defense in CC maps, where enemies come from three sides.


Mostly these would be either:

1. Half underground (most modules used for defensive troops - the central idea basically being trenches with concrete over top)

2. Fully underground (brain chamber, mining sites (?), armory...)

3. Fully above ground (sniper tower, low wall, whatever)





Let's face it, the terrain in CC isn't interesting or important enough right now for map-specific bunkers to be super necessary.

And tunnels are nice, but in CC defenses aren't reinforced from "the rear", because there is no rear - they're reinforced from space. That's important.

Basically I would really love it if CC were less about throwing a bunch of grenades and doomed soldiers down tiny underground passageways and more about figuring out the smartest way to take out an enemy sniper position or destroy a landing site or clear some trenches with a machinegun in the back


or something


Also I am Coalition so if this seems boring and not explodey lazery enough that's why

Instead of solving this by removing an essential sandbox feature, this should be solved by fixing its fundamental problem, e.g. the AI being incapable of navigation in many cases.

That'd be nice, but as I've discussed in the mod thread, it's practically impossible from a computational standpoint. Certain types of pathing present very serious problems for a realtime AI. The AI's already one of the major sources of CPU use atm and the pathfinder's one of the big choke points the last time I did any stress testing, even given how it works now.

One of the big things that should have been done quite some time ago was removal of the hardcoded pathfinder and total replacement with a Lua-based one using the current codebase, imho; that and some straightforward timer tests would show where the engine's internals need optimization and give people like me a genuine route to try and build a better mousetrap and demonstrate it in public. Right now it's a spaghetti mess and I don't think even Data really understands how it's working or why it's so slow, and the Lua it patches into is reliant on a lot of engine-side code that cannot be ignored. It's one of the perils of incorporating scripting support into a codebase that's this old.

Anyhow, it's all basically moot; we still haven't seen a release with fixes that would allow the mod to operate properly ever again, let alone be reasonably stable... and nobody else is working on a true TC-level gameplay conversion, so I guess it's all just theoretical arguments about nothing useful and will have to remain that way

Pity, though, it was fun to play something utterly different and have a semi-serious discussion about game design issues while it was technically possible