Rockets and decoys cut
Decoys were never really fun, and sensors accomplish basically the same thing (take aggro off the player). Rockets were really cool but not useful. Then I added grenades, which fill a similar role to rockets since they wait for enemies to approach before exploding.
New core design
Destroying six separate modules turned out to be tedious, so I lowered the number to three. People were also confused why they the core was invulnerable until the turrets were destroyed, so I put a force field around it which disappears once the core becomes vulnerable.
Force field changes
Previously, force fields had a short battery life. You could place one near a battery to increase its life, but it would still last less than a minute. The spherical shield itself was invulnerable, but there was a "generator" object inside the field that could be destroyed. If you happened to be inside the field when it was created, you could easily destroy the generator. Minions could also walk through the field to get inside and destroy the generator.
The problem is, now there are a lot more minions in a typical game, since they spawn automatically. I wanted force fields to be big, expensive, and important, but they don't last long with so many minions around. Plus, making them exclusively vulnerable to minions felt a little too rock-paper-scissory.
Force fields now sport an indefinite lifetime and a large amount of health, second only to the turrets. They can now be worn down from the inside or outside via minions, bolts, sniper bullets, grenades, or plain old drone attacks. And they're much more expensive.
Server optimization
The most CPU-intensive task in the game right now is actually ragdoll simulation. If more than 5 minions die at the same time, framerate drops from 300 to 60 or less. Ragdolls are strictly cosmetic, so I disabled them on the server to avoid CPU spikes. During normal gameplay, the server now puts one CPU core at 20-40% utilization, although AI might drive that number higher.
Project status and future
I decided not to teach this summer and instead subsist on money left over from last year. That means there will have to be a Kickstarter and/or Early Access release sometime around fall.
Audio
This is the biggest thing missing for any kind of release. Jack has been excited to work on this project for years now, but is currently crunching on LawBreakers. I started filling out Wwise audio events and spreadsheets in preparation for him to join the project this summer / fall.