Posts made by Smokeman

Oh yeah. roleRecycler.loot failure on creep recycler8_1-8436 :TypeError: Cannot read property 'structureType' of undefined [3:50:06 PM][shard1]roleRecycler.loot failure on creep recycler7_1-9144 :TypeError: Cannot read property 'structureType' of undefined [3:50:06 PM][shard1]roleRecycler.loot failure on creep recycler12_1-9149 :TypeError: Cannot read property 'structureType' of undefined [3:50:06 PM][shard1]roleRecycler.loot failure on creep recycler15_1-9650 :TypeError: Cannot read property 'structureType' of undefined [3:50:06 PM][shard1]roleRecycler.loot failure on creep recycler6_1-9697 :TypeError: Cannot read property 'structureType' of undefined [3:50:06 PM][shard2]roleRecycler.loot failure on creep recycler7_1-1500 :TypeError: Cannot read property 'structureType' of undefined [3:50:06 PM][shard2]roleRecycler.loot failure on creep recycler4_1-1989 :TypeError: Cannot read property 'structureType' of undefined [3:50:06 PM][shard2]roleRecycler.loot failure on creep recycler3_1-2122 :TypeError: Cannot read property 'structureType' of undefined

Those poor guys just try and try to loot those tombstones. Noop. It's like trying to go in the out door, it's not happening for them.

Were you still shooting it? As long as it can heal itself, it's healing someone. It has to have nothing to heal to flip the off switch.

This could also be done as a powerCreep power: OPERATE_STORE (Not to be confused with OPERATE_STORAGE.)

The effect grants the store the power to use the store.withdraw() and store.deposit() methods on adjacent stores. Duration: 1000 ticks, Cooldown: 300 ticks. Max transfer: 200,400,800,1600,3200 Cost: 20 ops.

The visual effect would have ramps that sprout out of the building.

This limits the scope to 3 per room, and if used in conjunction with OPERATE_STORAGE would completely max a level 5 ops generation capability (100 + 20 + 20 + 20) in the creation of a really impressive Vault / mass processing room, or as a level 3 secondary power that can handle a link in one shot.

Idea for a new structure that can be used to facilitate store to store material transfers without the need for a non moving, interstitial creep to do the actual transporting.

The structure would be non-obstructing and stackable, allowing the structure it sits on to transfer it's store in small amounts (Certainly no more than a non-boosted creep.) in one tick to a neighboring structure with a store.

The Conveyor Belt would look like a square track that goes around the building. One structure, the one at 0 distance from the belt, would be the implied first store, and any store at distance 1 would be the parameter. As such: belt.transfer(p_spawn,'power',1000); would transfer 1000 power to a neighboring powerSpawn, and: belt.withdraw(room.terminal,resource,qty); would withdraw 1000 of a resource form the neighboring terminal. The belt itself would have to be on a tile with an appropriate structure, like a storage.

Basically, it's a glorified non-moving transfer creep that only needs 1 tick to move material instead of 2. The advantages are: It's permanent. It declutters the base core allowing other creeps to approach the central store structure more easily. It saves CPU and ticks by halving the overhead of the transfers..

The survey was extremely short and easy, and very non invasive. I recommend everyone with even the tiniest interest to take it.

I am curious about the one shard being 20 CPU limited. I don't see the reason, as I don't see the need to offer a no sub option. But that's just me.

I don't have a "lifetime sub." Not because I don't think it's a "good deal"... I certainly would have already "saved money" with it, but because I'm a purist that's heavily invested in online games. I just don't think it's a good idea. Players of online games need the monthly "Skin in the game" of a payment.

As such, my strategy will probably be to cancel my persistent world sub after a trial run at the Seasonal world (Assuming I like it, of course.) I DO wonder what the response of people that DO have lifetime subs will be.

The other thing I'm curious about is: "This can be a content event like "go there, do that, fight with this thing... " Where is "There?" Unlike conventional MMOs, Screeps lacks mobility. Will there be special portals to the "Seasonal Challenge Area" that we could send teams of some kind through?

It's actually pretty easy. Each source has a unique ID, they get assigned when originally created (When Shard 3 was created, in this case.) and are assigned left to right, top to bottom. So the top source has a lower ID number than the one below it, as it was assigned first. These IDs are permanent.

If you use "let sources = room.find(FIND_SOURCES);" to find the sources, you'll get sources[0] pointing to the top one, and sources[1] pointing to the bottom one. the sources array returned by find will have a length of 2. The ID itself for the top one will be "sources[0].id"

But if you wander over to the sources and then search for the closest one, you'll just get whichever one is 'closer.' One of them has to be closer, even if they're both 1 away.

So... the problem is how you're searching for the sources.

I thought the same thing at first, but it makes total sense. There is a "Data Legacy" here that needs to be maintained. In the case of the labs, sure... it's TOTAL capacity is 5K, but it's always partitioned off into a 2K segment for energy, and a 3K segment for whatever mineral is stored there.

So if you just say "lab.getUsedCapacity();" which do you mean? Energy or Mineral? Getting both munged into one number is useless, as you can't use that to determine how much MORE of either energy or mineral will fit into there.

You can still "general purpose algo" it with a for loop that iterates the store's keys, of course, if your goal (As yours is here.) is to simply drain everything from it. And... not only does that work for all stores, (Except for the super weird bird, Nukers.) but you couldn't do that before with many of them..

It's a lab. It uses a constrained store that returns NULL if you don't ask what to look for in getUsedCapacity(); You can only make that call with no parameter with a general store, like in a storage or terminal.

You have to figure out what's in it first. It will be energy and one mineral or compound.

Edited to add: Ok, the solution is pretty obvious, actually, the property "mineralType" is not depreciated, so it will still be intended to work.. Simply calling "lab.store.GetUsedCapacity(RESOURCE_ENERGY); or lab.store.getUsedCapacity(lab.mineralType); to get the corresponding used capacity.

Sure, just park one of the miners below and to the right of the top source, there is one point of access to it. Of course, in that room you can access both sources from that point, so you have to be sure to get the id of the one you want to mine when you mine it (Store it in the creep's memory.) Then use "Game.getObjectFromId(creep.memory.yourSavedSourceId); to get the source object.

You can even use just one creep and switch back and forth between the sources.

I thought I was supposed to get the remainder of the Stronghold's 150K tick life cycle to enjoy the Invader free sector after I killed the stronghold?

Apparently not. All 4 of the ones I killed over the last 2 days are respawning right now. Even the one I killed late last night... It has popped up in a new location and is a few thousand ticks from forming.

"Max" is just a number. In a game, there needs to be a reasonable expectation to reach some semblance of that number in your lifetime. Even then, with Screeps, the difficulty ceiling of just coding your colony is, and I quote a poster on the Steam Community: "The difficulty ceiling is so high, that if it shattered today the pieces wouldn't hit us in our lifetimes."

Once you paint yourself into a corner, changing the color of the paint isn't going to matter. I'm ok with people with GCL 30 quitting because they no longer have an advantage... in reality, they're probably quiting because they were really done playing anyway and don't want to bother fixing their colony. Don't let that hold the game hostage.

Me? I would be fine with having the max GCL halved, even though it would mean mine would be halved and I would have to decide which half of my rooms to convert into energy and delete. There is still the PCs to build, and the literally never ending difficulty ceiling. The only reason to EVER run out of coding challenge in this game is because you decided to stop.

I generally agree. I especially like the Math.ceil(hits / 10e6) limitation to ramparts: But I would also like to see their max height reduced to 1/10 of current, so best case... 30M ramparts that decay at 3X rate between 20M and 30M. Just spank them hard. 30M should be the limit to walls, too.

This would heavily encourage active defenses and deter zombies that currently just hide behind walls.

Do NOT like the burst damage. Screeps is a game of deterministic outcomes and pure strategy. It does not roll dice. Nor should it start rolling dice.

The other problem is GCL. It's WAY too high. If you have 30 rooms spread across 6 sectors, you're invulnerable. Even is a coordinated group was to attack you from 3 sides, you could just expand out in the 4th.

What I think should be done: First, the aforementioned changes to ramparts and walls. Reduce max GCL to no more than 15. 10 would be better.

And now the big one: Remove the focus on PvP combat. Still allow it, sure. But add Alpha Predator NPCs that seed in the areas between players. Basically... zombie 'players' that are clearly NPCs and will follow rules specific to their class and danger level.

Have CONTESTS for whom can write the best Alpha Predator NPC in each class. Each class and power level in the class has rules the NPC must follow, some of them would be low level punching bags, others top level horrors you might be better off not getting close to. The number of which level of each class can be tuned by devs to provide a challenging environment.

So, if I claim a room from a player who has respawned, and their ruins have a 500,000 tick timeout... can I destroy the ruins? I mean... it's already destroyed. As far as I know, you can't destroy a tombstone.

Is it just going to sit there for 20 days waiting to decay?

I like the ruins concept. I already liked the strongholds, so this is all win.

Question: If I destroy a structure to create a ruin with a 500 tick timeout, can I use that to essentially bypass the "can't withdraw from a nuker" constraint? With proper placement, it should be possible to yoink the up to 400K energy from the Nuker's ruins to the storage in times of dire need.

But did you see the video of the new Screeps game at the top? That looks GREAT! It's super fast and appears to have a "time dilation" ability that lets you build an entire base fast! And that soundtrack! Awesome! I'm totally going to play that game when it comes out.

@tun9an0 Isn't being able to sell it a form of rebate?

No. A 'rebate' in this context destroys a resource in exchange for not using it. Selling it does exactly the opposite... it encourages 100% of the resource to be used (Thus increasing server load.)

I completely disagree.

Not only do I oppose resource buying with fungible resources (Ultimately, it becomes an RMT scheme.) but this goes completely counter to the direction the game needs to go. CPU is a precious commodity, one the servers need more of, despite being a gigantic farm of industrial grade servers that serve only 1800 people. What happens when there are twice that many players? 2 giant farms of industrial grade computers? Eventually, the entire planet's IT resources will be devoted to Screeps and Bitcoin.

A far better direction is to reward people for NOT using all their CPU, say, a rebate of some kind on unused CPU, so EFFICIENCY becomes a legit goal, not just a way to "have more rooms" once you already have 30 of them.

@wtfrank No, I wasn't even going there. But I could if you want, but for completely different reasons.

No, I was railing against the basic "All the cost up front, all the benefit at the back end" design philosophy. It's like an omlete recipe that allows you to make a 1, 2, or 3 egg omelet, but the difference is you use small, medium or large eggs... all the omelets use 30 eggs. (To be fair, it's hard to come up with a good metaphor for this.)

You could argue (And I'm sure you want to.) that the incremental list of minerals to make the tiers of boosts are the extra cost, but I would counter that that's a mechanic to exponentially increase the return / cost relationship for the higher tiers, not to establish the base cost of the boosts. The 'steps' between the benefit levels (typically a linear scale, like 100%, 200%, 300% 400% for non boosted, T1, T2, T3) is too steep at the beginning.

I love the concept of using boosts for economic purposes. But they're just too expensive. Prohibitively so. You can use them for things that pop occasionally, like mineral deposits themselves, with their 50K cooldown. But in common use, even a small operation will outstrip your ability to mine the base minerals, or convert other resources to them..

Combat is so rare, that you can stockpile boosts for that. But most people don't even bother with that. They never even get to the point of developing effective manufacturing and boost code since it's just easier to respawn or just quit. And there is no incentive to use boosts for the economy. (outside of mineral mining, and the niche of Power Banks with only one access point.)

My rant comes with a suggestion:

Reduce the cost of boosts for T1 and T2 by 66% and 33% respectively. Keep everything else the same, but require only 10 T1 boosts per body part, and 20 T2 boosts per body part. Why should a +100% carry boost cost the exact same amount of Keanium as a +300% carry boost? How does that provide an incentive to learn about boosts and write boost code?

There are 4 essential "costs" in game, and they overlap with each other in many ways (You can sell minerals/energy for credits and use the credits to buy minerals/energy.) The 4 are: Energy, Minerals, Lab reaction time, and Spawn time. The challenge is to balance these. By artificially inflating the cost of low level elements (T1,T2 boosts in this case, but it applies to PC powers as well.) you severely curtail their ability to be used in situations where they're less than economical in their full form, or that full form is simply not an option.

So much wrong with this, but I understand why you think it could work.

First, Screeps is not a system where your code is running all the time. It runs in slices, but unlike a Linux program with a select loop, your access to resources is not guaranteed in the time between slices.

As such, you cannot attempt to insert a realized class into an object you place in a creep's memory. You can "do it"? But it will be a data only version... as such, you most certainly cannot execute something like:

Game.creeps[creep].memory.myObj.run(); out of a creep's memory. Oh dear gosh no. Only pure data can go in memory in this context, certainly not functions.