This weekend, as a form of relaxation, Zun enjoyed one of his favorite pastimes: optimizing the performance of Colony Survival. He thought of a way to dramatically optimize terrain rendering. CPU use dropped by 60 to 80%. In a test world with max view range, the amount of frames per second jumped from 65 to 178! With medium view range, the effect is roughly +10% fps.
During the week, he spent most of his time writing new transport vehicle code. That's right, it's not specifically airship-code. A lot of people commented on last week's blog that they'd still like to see boats, and we agreed with them. Update 0.7.0 will probably only contain airships, but a later update will hopefully contain boats. We're even thinking about using the transport vehicle code for minecarts :)
I spent multiple days testing and balancing the game. We were afraid we'd encounter an unbalanced mess of glitches and inconsistencies, but we were positively surprised! Problems were way more minor than expected. We've still got good hope of reaching the June/July release date!
A lot of people have suggested either gates or monsters that can destroy walls ('better siege AI'). We've thought about these options sincerely, and we have not found ways to implement these features in the near future with a reasonable cost/benefit balance.
Better siege AI / wall breaking monsters
We'd love to have monster assaults be more siege-like and less like tower defense mazes. We love the sieges in Lord of the Rings and have great memories of defending the walls of Minas Tirith in the LOTR ROTK game. We've spent hundreds of hours capturing and defending cities in Total War.
But programming something like that into Colony Survival is bordering on the impossible, for multiple reasons. Firstly, building a city in Colony Survival is completely dynamic. There are no scripted, prebuilt fortresses. We, the developers, cannot know your particular town or castle, so we'd have to write an AI that recognizes how your defenses work and plans a way around them. That's a kind of AI that we have never encountered in any video game.
Secondly, building in Colony Survival is not realistic. You can easily gather, carry and place hundreds or thousands of cubic meters of stone. And your castle is not affected by gravity: your entire city could float a hundred meters above the ground. The top of your wall does not collapse when the bottom is destroyed.
80% of fortresses built by players is probably decently realistic, with buildings enclosed by a wall with towers and a gate. But 20% is completely unique, built inside mountains or on sea or floating in the air. The siege AI would have to work with these colonies as well.
So writing an intelligent siege AI that dynamically assesses your colony, searches for weak spots and attacks these in a realistic manner with ladders, towers, catapults and focused assault groups is probably impossible to achieve in a realistic time frame. It's something that we haven't seen in any AAA game, and we'll probably not going to invent that with just one competent programmer.
A 'dumb' siege AI, with monsters more or less heading straight for the banner and destroying walls that block their path, is more viable. But we doubt that'd be fun. Instead of being able to plan elaborate defenses, guiding monsters through a path of your choosing, you'd be continuously assaulted from all directions by monsters that are akin to termites. The only valid solution would be to evenly space your guards around the perimeter of your colony. While we can imagine this being a fun twist, it changes some fundamental aspects of Colony Survival gameplay that are certainly going to be missed by a large part of the community.
Gates
Gates are regularly mentioned, and I can imagine why. The idea of a large castle gate with dozens of monsters hammering on the walls is awesome. Again, we love Lord of the Rings :)
There are two major problems:
1.) Queuing. Gates will (temporarily) block the progress of monsters, so a queue will emerge in their path. Currently, monsters and colonists can move through each other. They just go from point A to point B, not caring whether another colonist, monster or player is in their way. We'll have to fundamentally change how our models, movement and pathfinding work to accommodate gates. If we don't, monsters will just clip through each other and it will look ugly and anti-climactic.
2.) Spamming. Imagine gates work brilliantly. You place them between monsters and the banner, and monsters will assemble in front of your gate, hammering on the gate until it breaks. Why not just build a maze with 50 gates? We could add artificial limits on the amount of gates, but that would be weird as well. It's not like wood and iron are in short supply in Colony Survival. Instead of replacing unrealistic mazes with realistic gates, it would probably replace unrealistic mazes with very unrealistic ugly mazes...
Conclusion
This does not mean we are 100% content with the current state of monster assaults and will not change it ever again. We are thinking of ways to improve them, and one of the first significant updates after 0.7.0 will probably contain new guards and monsters. Things like monsters that heal/protect other monsters in close vicinity of monster that can damage guards with ranged attacks could change the current maze-meta. So do not read this as a discouragement for your suggestions! They're all read and appreciated, and even if we cannot agree with a specific solution, it still indicates a problem that could be solved in a different way. We would just like to make clear why exactly we are not working on some popular suggested features.
Bedankt voor het lezen!