Bay 12 Games Forum
Dwarf Fortress => DF Gameplay Questions => Topic started by: Kento on September 01, 2011, 12:43:19 pm
-
I think this question has matured from its status in the little questions thread.
The basic idea is this: Small, industry-specific, largely self-sufficient cubes connected by narrow corridors. "Self-sufficiency" would consist of the following:
- Farm plots for food and alcohol production.
- A few female birds and nest boxes for eggs.
- Bedrooms for psychological benefit.
- Dining area (perhaps not a hall, as parties may prove problematic if frequent.)
- One citizen with growing, farming, and brewing labors enabled, so that he or she can quickly learn these skills, use them when needed, then quickly return to the cube's main industry.
Of course there would also be industry specific stockpiles and workshops.
Each cube would connect to a central stairway by a narrow (1 tile wide?) corridor, with restricted designations at the end away from the corridor. Other ways out could be made, for example it might be usewhile to have a high-traffic designation route to the trade depot that can be open and closed with a lever.
Cubes can be placed to minimize the hauling distance of their raw resources, custom food stockpiles can be made to steer preferred meats (probably gotten from the dwarven caravan) to a cube with citizens who like it, multiple cubes can be made for the same industry to control what materials are used.
The reason this is a question, and belongs in this forum, is that I do not understanding the way the game paths well enough to know if this will work. A 6*6*6 cube is only 216 tiles, 9*9*9 only 729, it seems if dwarves stuck to searching inside of these cubes, pathing ought to speed up. At the same time, to prevent them for going for a resource far away, cubes would have to not be directly next to each other, which may cause problems I did not anticipate.
Edited a half hour later to include some extra benefits to the design:
Dwarves would be pretty safe, and so would not need to travel long distances to get armor or weapons. Safe dwarves do not need to go to the hospital. Clothes, I think, are broken, and would only seldom need to be fetched if they were not. The restricted designation only needs to be added after the construction of the cube is completed, so hauling furniture to the cube and resource material for workshops should not be a huge issue.
People who do not like cave adaptation might not be happy with this design, but the idea is to have specialized dwarves that would not ever need to go outside, except to haul to the depot or remove constructions. A completely underground path to the pitting chamber can exist.
-
This should work fine in the way that you intend, but I'm not really sure that it will speed up FPS by any significant amount. Speed up the hauling process as the goods dwarves need are close by already yes, but FPS no.
-
To be honest, I'm not entirely sure why I'm concerned about FPS, I prefer playing between 25 and 50, but making dwarves only search very small areas seemed like an interesting challenge.
It was my understanding though that the most common cause of FPS drop was dwarves searching large areas in search of optimal paths. I thought this method would shorten them in the way that burrows ought to.
-
Shutting down the areas that dwarves have to search does improve FPS minimally, but the largest thing that slows down FPS is just the shear number of creatures you have on your map that are pathing combined slightly with the total number of items you have on your map.
Forming cubes would cut down on the amount of hauling that needed to be done though, effectively giving you more work from your FPS. This would come at the price of more micromanagement from you though.
-
I think a column design is probably the best way to reduce hauling distances without a complicated system of using garbage chutes as one-way pneumatic tubes, though bedroom placement becomes a larger issue.
-
Probably this design,
sswwwss
ssswwwsss
ssswwwsss
wwwxxxwww
wwwxxxwww
wwwxxxwww
ssswwwsss
ssswwwsss
sswwwssx being stairs, w being workshops, s being raw material stockpiles, is optimal. Haulers would only have to move raw materials three steps from the stairs at most, and each raw material is within four steps of the center of a workshop. the center of a workshop is only two steps away from the stairs as well. bedrooms can be made behind workshops to prevent a little traffic on the stairwell, and minimize time wasted traveling from workshop to bedroom and back.
-
I've tried not to harp on it too much, but you'll see in my previous posts a lot of info about FPS efficiency and density of a cylindrical floor plan. My cylindrical design (http://"http://mkv25.net/dfma/map-10684-amusedpillars") was conceived specifically with the same goals in mind as your "pod" plan. I use circles instead of squares because diagonal pathing is 1.4x more costly; I place related stockpiles and workshops in close proximity to each other; and I designate the central shaft as high-traffic to reduce the amount of time the A* algorithm spends looking on each floor for something that's not actually there. Outdoor resources (wood, webs, fish, gathered plants) are stored near the entrance to the cylinder since each item comes with a long-distance hauling task "tax" attached to it.
I went with 15-tile diameter floors so that I could break the floor into four 5x5 workshop areas (workshop + 1 tile 'ring' of input stockpile), four large storage areas, and a ring of storage around the stairwell for products likely to be created on each floor. That way your skilled workers can crank out high-quality goods making short trips... your haulers can quickly empty a workshop of clutter... and then other haulers can move bins up and down the central shaft if they need to be traded away or used as inputs somewhere else.
It does improve FPS slightly, but the real benefit (as iam2roy notes) is that you get much more bang-per-dorfbuck. Your artisans, whose time is most valuable, do very little long-distance hauling; your haulers have short trips so you don't need to use as many of them; and because the fortress is dense, the longest path distance between two points is fairly small, so your FPS at 80 dwarves is likely to be higher than your FPS at 80 on a less-efficient design. Of course, you need to be careful, because your newfound efficiency makes it easy to spam out more stuff, which hurts FPS...
(Edit: Fix'd, thanks.)
-
To be honest, I'm not especially concerned about FPS, and prefer a step-efficient design like the cylinder (though a cylinder will have more defense issues if an intruder actually gets in). I think in so far as step efficiency is concerned, the 9x9 with corners missing (I have a hard time calling the 9x9 one a cylinder, it's more of a fat cross) and bedrooms (for singles, at least) is probably most efficient, though digging extra space for kennels is aesthetically unpleasing.
Of course that is the "problem" with open ended games like this, efficiency is something we can be proud of, but because there is no time limit, may not be completely necessary outside of defense considerations or avoiding easily avoidable alcohol-drought or starvation. "Yes, functioning magma smelters in x days" is a way to benchmark yourself that can be fun, but if you take too seriously that you spent several hours "carefully preparing your journey," forgetting it was fun to do, you'd feel silly. FPS is a metric, while fun to think about optimizing, can get in the way of other enjoyable considerations.
It is interesting though, I had not considered that step-efficiency may allow more to be done in ten minutes with a lower FPS than FPS efficiency allows. There probably isn't much use to my fortress design outside of novelty, and perhaps as something to test to get a better idea of how the game works, and how to get the game to work. For that end, I will test it, as well as to discover any unanticipated consequences of such a design.
-
I
*Your link is misspelled*
http://mkv25.net/dfma/map-10684-amusedpillars (http://mkv25.net/dfma/map-10684-amusedpillars)
-
Kento, I think you're selling yourself short -- 9x9 is a great footprint, and with a clipped corner, the distances work out pretty well: 3 tiles horizontally vs 2 diagonally (~2.8 steps) is pretty close to circular. You could expand outward one tile and have 4 tiles in the cardinal directions and 3 in the diagonals (~4.2 steps).
If you're worried about the kennel, your second-highest floor can have only a single stairway leading up, and your highest floor can have a single centered stairway leading down. You could use the top floor for a combined kennel / farm / hatchery. On the floor below, a farmer's workshop, kitchen, still, and mill, with storage for freshly-harvested plants.
If you can figure out which industries to bundle together, you can make burrows out of each of your 9x9 pods and place them at different depths -- cloth/bags/milling, wood/ash/furniture, metalworking, etc.
-
Kennels are probably best near the pitting area, and you only really need one. I didn't notice a huge pitting chamber in your cylinder, did I miss it? My first real attempt at a fortress (and the only "major" attempt I've made) ended up having space for over 200 cages.
I like the idea of having the surface entrance of a column be stairs, as it would make the hauling distance for wood shorter.
One problem I thought of with the scattered 6x6x6 cube idea is that the 200 seed limit might result in some cubes having no seeds of certain types. Especially with 40 population 5 cubes... it's too bad that you can't have individual dwarves ignore traffic designations, it would make having an "Urist McAppleseed" not nearly as inefficient.
-
How do burrows relate to stockpiles? I assume dwarves only take items from the burrow they're working in, so could be set up so they would not seek items that are "close" but take many steps to reach. But I understand that they seek routes outside of burrows, and I'm guessing that they still look through the entire stock table, so it would not help with the "total number of items" issue-- and possibly make it worse?
-
Kennels are probably best near the pitting area, and you only really need one. I didn't notice a huge pitting chamber in your cylinder, did I miss it?
Since animals have such a heavy impact on FPS and require micromanagement (grazers, egg-layers, trainable war beasts, etc.) I generally butcher everything that's not a dog and drop the rest into magma or a weapon trap from several Z-layers up. I occasionally allow 5-10 female birds to lay eggs, since eggs are nearly zero-maintenance. So I basically ignore most of the animal husbandry stuff.
How do burrows relate to stockpiles? I assume dwarves only take items from the burrow they're working in, so could be set up so they would not seek items that are "close" but take many steps to reach.
I'm not sure -- I've only used them as a temporary measure to pull craftsdwarves into an area with a desirable workshop. The wiki says that ...the most common bug is the "Dwarf cancels Store Item: Item inaccessible" message spam that results from idle dwarves being in a burrow that contains a stockpile but not the item the stockpile wants to have.
That would drive me nuts if I didn't have a good hauler corps.
-
How do burrows relate to stockpiles? I assume dwarves only take items from the burrow they're working in, so could be set up so they would not seek items that are "close" but take many steps to reach. But I understand that they seek routes outside of burrows, and I'm guessing that they still look through the entire stock table, so it would not help with the "total number of items" issue-- and possibly make it worse?
They will only choose destinations inside their specified burrow. They can path anywhere, even if they go through a forbidden area.
-
I just thought it would be cool if it kept a separate hidden stock table for each burrow, it would make things faster -if- the burrows were being used and there were not too many overlapping burrows, and definitely make burrow designation worthwhile if FPS was becoming a big issue.
What causes domestic animals to be so bad for FPS? Does chaining help? Of the three animals that are both shearable and milkable, the sheep, the alpaca and the llama, I'm pretty sure the sheep and alpaca can survive on 3x3 grass/moss. They're really difficult to justify having, since plants are a better source of cloth and food, wool cloth can be gotten through trade and then forbidden to reserve it for strange moods, as can cheese to make expensive prepared meals--- but I like collecting animals.