Ability to permanently remove a non-constructed wall tile from a room (for an extremely high price)
-
Allowing players to build structures in hostile rooms is a topic we should pursue. I'm not sure whether it's a good idea or not. Just being able to build walls while attacking a room would a be a huge change. However, I think we've found a nice little feature with wall roads (I think the name "tunnel roads" is confusing).
- Allow road construction sites to be placed over walls.
- Increase the build/upkeep costs of wall roads 25 or 50.
- Update the engine to allow movement onto wall roads.
-
@tigga said in Ability to permanently remove a non-constructed wall tile from a room (for an extremely high price):
That puts the cost to build at 37500 energy, and hit points at 625000 which... seems a lot. Maybe the hits could be toned down, but the energy cost seems alright.
Regarding the idea of 25x swamp build costs, I agree that 37500 energy is almost nothing. You eventually have lots of energy in most rooms and in the lifetime of a room you will get through hundreds of millions of energy. So amortised over the room lifetime, 37,500, or even 375,000 energy is basically nothing. It will be paid if necessary (and if there was a mineral cost that would also be happily paid).
The maintenance cost of 0.125 energy/tick on the other hand is definitely something to think carefully about. Let's say a room + remotes has a net energy income of 20-60/tick, if you dropped a few tunnels that could be a substantial fraction of the room's income. This would be the most important statistic of the tunnels to get right. Too low and the tunnels will be used very freely, too high and they would be used very rarely. Neither extreme would make an interesting decision for players.
As to the hits of the tunnel, I actually think that the hit points of most room structures are negligible once boosts come into play, extensions can be 1-shot by a boosted creep, so I don't think the game would be worse if structures started being a bit harder to kill. Especially if there was a sizeable up front cost to building tunnels, you would want a bit of HP for any marauders to chew through to give you some chance to preserve your investment. Tunnels in neutral rooms would admit a new form of guerrilla warfare where you could cause economic damage to someone with hit-and-run attacks.
Because the maintenance cost is the most important balancing parameter of a tunnel, lets have a look at a few rooms on the map I've found by clicking around and look at the consequences of some tunnels being built:
https://screeps.com/a/#!/room/shard2/E5S6 - this is an SK room with some big walls such that if you were harvesting it from the south or west you'd have some really long paths. Imagine someone's harvesting it from the south, an obvious place to put a tunnel would be through the 3 wall tiles at y=40, x=20-22. This would cut short the path of an energy hauler travelling via the bottom right exit by 50 steps. If we assume a 33W17M hauler costing 1250 energy, in a creep lifetime that source requires 12+ visits (4000 * 5/1650). So 12 visits that travel 50 steps is 600 ticks of hauler lifetime that you save, or 500 energy (600/1500 * 1250). So if the tunnels cost less than 1/9 energy per tick (500/1500/3) then you make an energy profit from building them. However, you also make an intent saving and a spawn capacity saving. You probably also would make some intent saving on your source keeper killer creeps which wouldn't have to travel as much between source keepers.. So I would argue that those tunnels would still often be worth building even if the energy cost was higher than 1/9 per tick. A question for the audience: what would you pay, in energy per tick, to save ~half a hauler and spend the cpu/spawn capacity on a different creep?
https://screeps.com/a/#!/room/shard2/W4S8 - this is a non-SK room. Again assume we're harvesting the room remotely from the south. The tunnels of choice would probably be 3 spots at y=45, x=19-21. In this case the path saving is about 40 spaces. Doing the same calculation again we find that we require 9 visits and save 360 ticks of hauler lifetime. In this case we make an energy profit if the tunnels cost less than 1/15 energy per tick.
Clearly, in two somewhat similar rooms, each with 3 tunnels required to improve a path, differing mainly by the presence of source-keepers, we see that it's much easier to justify building tunnels in the sourcekeeper room. But in rooms with thicker walls it might be very hard to justify building tunnels.
The important question is I think, what energy loss would you be prepared to take (if you would take any at all) to reduce the number of haulers you needed and gain more spawn capacity and spend less cpu each tick?
-
likewise with containers the upkeep cost could be different for owned rooms and unowned rooms.
should it be possible to build tunnels through build walls ?
should it be possible to rampart walls to prevent tunneling?
-
Tunnels implementation have been deployed to the PTR: https://screeps.com/forum/topic/2322
