Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Dante]

    II. A little hard data.

I experimented under vanilla DF and found that with four dedicated competent miners, you could excavate an area of 6600 tiles from the soil layer IN THE FIRST SPRING OF EMBARK. That's 1650 tiles each, and they all reached 'legendary' status in those three months.

Now, a 2*2 embark has 20736 tiles. Even discounting the map walls, if you embarked on a tiny map with a soil layer, then you have over 20,000 farmable tiles, and you can excavate that with 12 initially-competent-miner-dwarf-seasons of work. Say, two and a half years for a dedicated miner.
...Plus you can make more, from rock layers + water.
...Plus there is surface mining to consider.

This is important, because it means that a simple increase of farm acreage is not a real solution. Sure, there'll be feature-filled caves in the next version, but not for several layers down. And if you embark on a standard 6*6 with a soil layer, you'll have as much as 190,000 tiles of soil in which to situate your farms. You probably wouldn't want to make a room that big, but still.

That's a lot.

According to the wiki,
“A 5x5 plot manned by dabbling farmers will provide enough food to feed and water 30 dwarves. The use of boozecooking will stretch this to 75 dwarves. If you have decently skilled farmers, the same 5x5 plot (with boozecooking) will feed over 200 dwarves”

So, it's clear that farm plots probably should be larger, at least for cosmetic reasons, but since there's a huge number of equally-farmable tiles to play with, that's not the whole answer.



    III. What players do and don't want (caveats).

This thread has featured a lot of hostility, but mainly to the proposed solutions rather than the underlying idea that farming needs to see some improvements. I've tried to work out the main things that need to be avoided.

A. Extra game mechanics. If there are extra mechanics to go in, they will either require dwarves to do things, or the player to do things. People tend to be fine with more jobs-for-dwarves. It's another use for all those migrant fish dissectors and trappers, after the military. But when the player has to do things, some people see it as a chore.

B. Regular player intervention. Also know as micromanagement, this is the most hated form of point (A). People don't want to have to tell dwarves to weed and fertilise each farm plot each season. It's busy-work. Now, on the one hand, the 2D version needed annual farming micromanagement. On the other hand, there was a lot more emphasis on farming in 2D; now there are other distractions. Jobs-for-dwarves are more acceptable than jobs-for-players, unless those jobs are one-offs or not completely necessary.

C. Balance between stages, and learning curves. Some proposed solutions would make farming extremely difficult to set up at first, and nigh-impossible to new players, and still not increase the difficulty much later in the game. Again note, though, that it was harder to set up farms in the 2D version. There should be a mechanic for getting a viable farm set up immediately, even if it means you have to spend more effort on making proper/better farms later.

D. Item management, stacking and FPS. Most people don't want unnecessary items floating around. It was mentioned in the thread that growing individual plants, rather than stacks, over a larger area is untenable. This is because plump helmet [1] → dwarven wine [5], which takes up one barrel, the same as plump helmet [5] → dwarven wine [25] takes up one barrel. So, for many solutions, the stacking and cooking rewrites would need to go in... or something would need to fundamentally change in the abstractions used for 'quantity'.

E. Extra jobs/workshops/stages/interfaces. Similar to (D). It would indeed make farming more difficult if you had to make grind bones to mix with potash and small_remains at a workshop to make compost, and then send out your shovellers to spread that on the ground, and then send the mushrooms to a workshop to be dried, and then to another one to have their spores removed, and then to another one for pickling before you could cook them. It would also make farming painful. It's hard enough to get workshop processing and milling sorted out as it is. Some people (me included) relish extra detail and options, but that's for ideas like beehives and compost heaps and scarecrows and other things which would be optional. We don't want to complicate existing processes that the player has to intervene in. And again, it's necessary that the game provides a level of abstraction.

F. In that vein, abstraction. Currently, both time and quantities are abstract. You can stave off hunger with a mushroom, or a ninth of a horse. Or a third of a biscuit made out of three cooked seeds. Also, a game day is so short that if a dwarf required three meals, they'd starve to death while constantly eating. So solutions based on 'real-life' farming requirements – such as the timing for weeding, picking, and rot/wither times – may not work.

[Dante]

    IV. Forms of solution suggested (mainly from the thread).

A. Simple fixes.
Remove boozecooking. Remove seedcooking. Increase growduration. Make aboveground crops seasonal. None of these are perfect solutions. I play with 3* standard grow durations, don't farm aboveground anyway, and don't use either exploit, and it's still just a matter of increasing your farm size slightly.

A. Increased area investment.
Change the mechanics for plants grown per number of farm tiles. As I said, this is really a straw man solution. To make it difficult enough, you'd need a very large area of farm indeed, and that loss of space would annoy many players; once you'd done it, it would pretty much cease to be a problem. Then there is the justification. If plants need 'room to grow', that implies you're growing one or fewer plants per tile, which again brings up those stacking problems.

B. Increased dwarf ingestion.
Change the mechanics for the amount a dwarf eats (at any one sitting). Perhaps a more long-term solution to (A) because it will require more time spent in actual planting, harvesting and processing, as well as larger acreage. However, dealing with things like multi-component single meals would need a dedicated food re-write and probably comes up with more stacking problems.

C. Greater failure rate.
Fewer plants survive. Perhaps a 1 plant – 1 seed ratio. We have a form of this, in that dabbling growers can produce stacks of 1 (and apparently even 0) plants. It would mean you'd need to increase your acreage or improve your growers. However, it's not that hard to train up really skilled growers anyway. Also, upping the number of seeds required has problems. Plants which give back multiple seeds from brewing works, but then you can't cook the plants. There should be more guaranteed ways of securing seeds – such as growing plants for their seed, more reliable caravans, guaranteed underground plant features, etc.

D. Increased resource investment.
Use actual physical resources in farming. The most common suggestion is making fertiliser necessary. This is realistic, but it would definitely need alternative ways of making it – for those people without the trees, or the patience, for woodburning. Composting of remains and bones and suchlike has been suggested. Another idea would be rotting logs for growing mushrooms (dimple cups and plump helmets). For those plants which are plants, not fungi, another solution might be necessary. Resource use ties further into (E) and (F) below.

E. Increased jobs to perform on farms.
Either to increase yield, or to get any yield whatsoever. The most common suggestions are:
e1. Maintain soil fertility (as above). Some have even said require specific nutrients for specific plants, but this quickly gets into the 'infeasible' territory.
e2. Maintain soil moisture. As simple as a return to the 2D idea of mud – and hence farms – drying up if they're not watered once a year, or as complex as specific plants requiring specific moisture levels (which would need, for player sanity, to be something the dwarves worked on automatically).
e3. Problems: Pests and varmints. We have food-eating vermin already; it would be the work of moments to make them eat food from the farms. There could be more complex mechanisms such as scarecrows, dwarves assigned to vermin defence, plagues of locusts, and so on.
e4. Problems: blights and disease. A soil blight could occasionally destroy an area of farmland so you'd have to relocate to fresh soil for a year. Crop rotation, and lying fallow, should both help to prevent diseases building up in the soil. This could eventually be linked to spheres.
e5. Problems: weeds. Nonuseful plants could be implemented, and grow where crops should be growing. Vigilant farmers would need to remove these. Added bonuses would be poisonous/dangerous weeds, and unskilled farmers ripping out crops or accidentally letting weeds be sent off to the kitchens.
e6. General maintenance. Basically, make-work. A requirement that a farmer stands in the field for x ticks per plant. This could abstractly represent a large number of things. Basically, you'd need to increase the required food workforce.

Note that all “increased jobs” can be divided into player jobs (micromanagement), and dwarf jobs (jobs for dwarves). Many – perhaps most – commenters are against extra details if it means micromanagement. Again, I'd raise the point that farming required much more attention in the 2D version, tempered of course by the point that a lot has changed since then.
So, to avoid the realism-kills-the-fun mentality, as many of these increased jobs as possible should be automated, automatable, or optional (for increased yield).

F. Less abstract plants.
Currently, plants grow in a season, and then the entire thing is ripped up and eaten. This is unrealistic. Separate out the plant itself from the crop. Let the maximum number of crops, the minimum time before the plant can grow them, and the time between crops, be specified in the raws.
You could even track each plant's health individually – a simple 1-10 integer value would do – and this could be increased/decreased by vermin, water, drowning, soil quality, fertiliser, farmer skill, and so on. Below a certain level, and the plant won't crop this season. Too low, and it might die. High enough, and it'll put out extra crops or produce a more bountiful harvest in any given crop.
This idea doesn't necessarily make farming harder. But it could easily tie in to required watering/tending of seeds (to make up for not having to replant each year). Also, longer-term crops may take longer to grow before cropping at all. Tree cropping probably ties in here, but that's another issue.
More specific plants will increase realism and add to your options, and incidentally make farming harder. Rice might require much more frequent irrigation. You might be able to grow seaweed in ocean biomes, but only in a certain water depth. Edible cacti would have to be kept sheltered from the rain. Etc.

G. More difficult mining.
This is a controversial one. Plenty of people don't want mining to be harder, and plenty more get contentious over the actual practicalities of mining changes (e.g. spoil and supports). But if mining was made so that dwarves don't cut through rock like cheese, then taking longer to hollow out space for farming would naturally make farming more difficult to set up. Not, however, to maintain. Some have suggested it be very difficult to mine extensively through soil layers, requiring a player to put up supports or walls to stop the dirt collapsing – especially for sand layers. Again, this would make farming harder to get started. Again, not a solution for farming over time being too easy.

H. Less plentiful soil.
An alternative to harder mining. Depending on how deep you think a z-level is, entire layers of soil should perhaps be much less common. This would require irrigation of rock for underground mining. Again, not an ongoing solution, and not necessarily a realistic one.

I. Remove puddle-on-the-stone farming.
This one makes sense. Rock + water = permanent nutritious soil is not a realistic idea. It could simply be removed. With the coming changes to the Underground, patches of soil could be found near underground water features. Bring back seasonal flooding (river sediments have historical precedent – Nile farming – for creating soil, but they would need to be replenished). You could have dwarves create underground farms by implanting soil: that is, fetching buckets of it from the surface, or making loose soil a material left behind by mining. Adding compost could be a solution. Alternatively, simply make potash and mix it with water, slosh it on the ground and grow on pure fertiliser.

J. Bring back hungry animals.
Related to (B), this has been discussed a bit, but seldom in the context of restoring balance to farming. It has additional benefits of removing tame animals as another infinite foodsource. As weird as it sounds, dogs are overpowered. They give happy thoughts, make for an adequate military early in the game, you can eat nothing but dogmeat, and they give you bone bolts / leather armour.
It could be as simple as putting the current food/drink mechanic in place for animals, without the accompanying thoughts, and adding a carnivore/herbivore/omnivore tag. Plant-eaters wouldn't have to eat farmed plants; they could be supplemented by grazing (a use for grass).

K. Non-food-based demand.
Inindo came up with a bunch of stuff here. Mainly clothing-based. If dwarves were more keen on wearing pants, and especially non-rotten pants, then this would put more pressure on farming. At the very least, pig tails would become an important crop, and depending on how important happiness was to a player, dimple cups would be needed for the dye. Another possibility was a cloth requirement for beds.

L. Nutrition and happiness.
Make dwarves increasingly unhappy to drink the same old booze and eat the same old plant. Simple existing crop rotation can stave it off, but it would remove The Plump Helmet As Be All And End All Of Dwarven Farming.
LordBucket suggested nutritional deficiency from eating only one food. This is realistic and could be implemented to varying degrees – from your dwarves slowing down / getting weaker when they are on a one-crop diet, to actually implementing tags for the nutrients various plants provide. Dwarves might require a minimum dose of calcium, iron, potassium, b-vitamin, c-vitamin and whatever each year.

M. More specific biomes.
Change simplifications like 'biome' and 'season' to min/max temp, min/max water, max crops per year in ideal conditions versus substandard conditions, and so on. It wouldn't be black and white; a climate that is only just inside the tolerance of a certain crop would mean that with good growers and good fertiliser you'd still grow a few... but with greatly reduced yields. This relates to (F) and (C).

N. By all means make it harder.
A meta-solution or reminder: there are already many other methods of food acquisition. All of them are harder and almost all require more micromanagement. Farming could easily made more difficult via the nutrients/irrigation/rotation suggestions, and if a player doesn't want to be involved, they can trade for their food. Or hunt. Or fish. Or gather. Or trap. Or...
Anyway, as WingDing suggested, dedicated food caravans could be implemented, perhaps with additional expenses or costs for those who don't want to spend time farming.

O. Changing the nature of the game.
There are other possible solutions. For farming on stone, you could have to build 3-wide plots, and then dwarves scrape all the mud from those tiles into just the central strip. This could be combined with a re-irrigation requirement. You might need to pollinate crops. You might need to keep your fungal crops irrigated betweens depths 1 and 3 at all times. Entirely different ways of doing things are out there. Perhaps Toady has some ideas himself.

[Dante]

    V. Grand unified field theory

A farming pun!  No, but for those interested, here is my preferred solution. It is in no way comprehensive, but I did try to take all the perspectives I identified into consideration. With a dash of realism and a splash of wishful thinking.

First off, it's clear that there's no simple fix. The things like longer grower times, abstract make-work for farmers, and larger plots make things slightly harder, but are more of obstacles to starting a farm industry than maintaining one. Making aboveground crops seasonal, and underground ones non seasonal, should be canon.

Now, especially given opposition to micromanagement (jobs for the player), I think a greater dedicated farming force (jobs for dwarves) will be one of the keys to Improved Farming. But that's just a start. Those dwarves should have actual things to do.

Add soil quality, soil drainage and soil moisture level for any given tile. Change the format of farm plots so they have (abstract) irrigation channels down the middle. Dwarves would lug water to the plot, add it to the channel, and it would get distributed evenly between the tiles. Dwarves would know what sort of level of moisture would be ideal for any given crop in the ground. You might not reach that if you have too few farmers or too high a soil drainage. Inexperienced farmers might be slightly off the mark, too. The soil quality would be store per tile. Quality would decrease slowly over time, and take bigger hits from lack of crop rotation, lack of fallow seasons, vermin, lack of watering, and floods. Quality, water and the skill of the farmer together would effect the yield of a tile.

Fertiliser use would be more common through all this, but it would still be optional, unless soil depleted permanently. Adding compost would be good – generated from a compost heap building, made out corpses, dwarf chunks, vermin remains, rotted clothing, bones, and so on.

Implement vermin, weeds, and scarecrows. Animals should eat again.

I personally think these features would be enough, but the addition of long-lasting plants rather than these seed-to-crop-to-dinner-plate things would be a nice factor. That way poor farming practises could result in low or no yields any season, rather than simply the plant failing to grow or producing small stacks.

I support more difficult mining (for various reasons), which means that farming aboveground would be easier for a while, as you slowly built into the mountain/ground. That way, new players wouldn't have too steep a learning curve. Soil depletion would still require them to fertilise or relocate eventually, though. Underground farming may (should?) become a later-game option, to work towards. Requiring rotten logs for fungi to grow on – and in some way reducing the stone + water = arable land thing – will help here. Underground farms are of course easier to defend.



Now, gather round, ye children of the forum, and tell me... where are we at with farming these days?
(Also is there anything in the upcoming release that'll affect this? I couldn't think of much off the top of my head, except for non-explicit underground water features).

[Draco18s]

Looks pretty decent.  I do have to say though, that z-level height and number of "soil" layers really doesn't matter, as some biomes currently have as many as 5 soil layers, and the real world does have places where bedrock is thousands of feet down.

[Impaler[WrG]]

Very nice work Dante, I agree with most of your assessment but would still like to see the farm-tile:dwarf ratio increased substantially.  The 'irrigation channels' sound like a good way to handle watering and give a bit of flexibility on watering is done.  Soil drainage sounds like it might be a bit much, if it was included it should be derived from the soil type sand/silt/clay and thus dose not actually add more data/memory overhead to a tile. With respect to multi-layer soil, deeper soil is invariably poorer then the top-soil so reducing the initial quality values for sub-surface layers would be appropriate. 

[Particleman]

Let me put my thoughts on making farming horribly complicated and time consuming this way: No.

Would doing this make the game more realistic? Yes, but I'd like to refer you to dwarves, elves, goblins, gremlins, and carp who can singlehandedly murder the shit out of anyone unfortunate enough to walk by, just to name a few.

Would it make the game more complicated? Yes, of course.

Most importantly, would it make the game more fun*? For a few people, yes, but for many it would just be tiresome, annoying, and tedious.

I agree farming is too easy as it is, but requiring a 10x10 plot to feed a single person is overkill.

(*By this I mean the regular definition of fun, not DF Fun.)

I don't really have much to contribute, I'm just saying I'm opposed to the idea of totally (or even mostly) realistic farming in DF.

[Draco18s]

...Farming right now is trivial.  1 farmer can dedicate 1/10th of his waking life to maintaining a farm plot and feed and entire fortress.  All we're advocating is that farming should use up a portion of your workforce the same way the military does, or your mining squad does, or your mason squad, or...

Not to make farming impossible, complicated, or difficult for the player.

There is a middle ground you know.

[Impaler[WrG]]

I agree farming is too easy as it is, but requiring a 10x10 plot to feed a single person is overkill.

Yes it is, that's why my proposal from the first post which mind you was the MOST aggressive tile:dwarf ratio ever proposed was only 12:1 and that's for full food&booze with zero hunting, foraging or trading to supplement it.  Throwing around hyperbolic straw-man numbers that you clearly just pulled out of your ass is not a real argument.  State what level of difficulty or tile:dwarf ratio you feel is appropriate and we can have a real debate.

[G-Flex]

...Farming right now is trivial.  1 farmer can dedicate 1/10th of his waking life to maintaining a farm plot and feed and entire fortress.  All we're advocating is that farming should use up a portion of your workforce the same way the military does, or your mining squad does, or your mason squad, or...

Not to make farming impossible, complicated, or difficult for the player.

There is a middle ground you know.

Yes. And you can make farming more complicated without it having to be an unnecessary toll on the player.

[Dante]

Thanks guys.

some biomes currently have as many as 5 soil layers, and the real world does have places where bedrock is thousands of feet down.

Yeah, I guess this would affect a bunch of stuff, including soil drainage and as Impaler[WrG] said, quality. Thousands of feet is a surprising figure, how common is that really?

I agree with most of your assessment but would still like to see the farm-tile:dwarf ratio increased substantially. 

Yeah, me too, I just don't think it's a full solution, because space isn't really a pressing issue.
Also, are you opposed to soil drainage because it makes it too complicated, or because it could require too much memory and e.g. cause FPS problems? (I don't know how likely that is, but it would depend on the map size, whether you were storing a small integer value, and whether it ever needed to be updated).

Let me put my thoughts on making farming horribly complicated and time consuming this way: No.

And you can make farming more complicated without it having to be an unnecessary toll on the player.

Yeah, the key idea seems (to me) to be that extra work can be either:
done as a one-off to get things started by the player (which doesn't solve the problem because farms are forever);
something the player has to do regularly (e.g. seasonally/annually);
or something done automatically by the dwarves themselves.

Particleman, what you're reacting against here is the second way of doing things. Luckily, that's not the only solution; as Draco18s just said, extra work by dwarves seems to be the most acceptable solution for players. Personally, I think the game could work with a little more infrequent player interaction with farms, but again, that's the contentious one.

I agree farming is too easy as it is, but requiring a 10x10 plot to feed a single person is overkill.

...
State what level of difficulty or tile:dwarf ratio you feel is appropriate and we can have a real debate.

Quite right. Note, however, that the smallest possible embark size has 144*144 tiles, so a tiny map with a single soil layer could still feed more than two hundred dwarves even at a 10*10 plot = 1 dwarf ratio! I think that people tend to underestimate the number of tiles available in a map when they make their claim for their preferred ratio.

Disclaimer: I'm not arguing for a 10*10 plot = 1 dwarf ratio.

[Nikov]

I'd like to throw some chips in. But before we start, keep in mind I am not factoring in booze cooking for my initial food estimates. That comes later.

First stacking needs to be fixed. Once plants can be stacked together and barrels topped off, low per-tile crop yields will no longer be problematic from an inventory standpoint. I've heard that this can't be done and I've heard that this can be done, so I'm throwing it out there that beating the code with a sledgehammer until it begs forgiveness and neatly stacks items in its time-out corner is a worthwhile pursuit, damn the torpedoes. This is not required for my arguement, but it seems to be hindering a lot of ideas and will certainly decrease the demand on barrels and storage space.

Secondly lets dispel all notions that a tile can be abstracted into an acre, or a hundred square yards, or a hundred square feet. A tile is the amount of space a dwarf, a door, a barrel, a bed, or a table occupies. For the purposes of my arguement I assume its about the same as a DnD grid square, 5x5 feet. 25 square.

Being a bit of a farm boy, I can tell you in twenty-five square feet one grows roughly 12-18 corn plants, a figure which is varied based on soil quality. So the current yields aren't really high, they're actually low. Not all seeds germinating is also already in the game, since a peasant who gets one plant out of a tile put the same amount of seed in as a skilled planter who gets five plants out of the tile.

A six by six tile field is 30x30 or 900 square feet if we accept the 5x5 foot tile. Currently you can feed an entire fortress and then some without ever expanding beyond this scale. The field is a pretty sizable area to a dwarf, but not realistic. In reality, one acre of prime American farmland yields 160 bushels of corn (thanks to mechanized agriculture which we shall presume is 'legendary'), or 270 square feet per bushel (I'm fudging some decimals as this is not a scientific debate).  In reality, this 6x6 plot of Legendary planted and fertilized land would amount to three bushels or 170 pounds of grain. If we assume the average dwarf has a slightly less hearty appetite than the average American, this amounts to some 42 days worth of food. Eating three pounds a day comes to 56 days worth of food. A full acre at this rate produces 8960 pounds of food, or enough to feed six dwarves for a year at near-American rates or eight dwarves at three pounds a day. If no changes are made regarding four-season, four-crop agriculture we can feed 32 dwarves a year with a legendary-planted, fertilized, productive-as-Iowa grain farm. The only problem is, assuming the 5x5 foot tile, this is a 42x42 field! It is almost the size of a 1x1 embark zone (48x48). Realism is not possible in the current simulation, unless almost the entire fortress is dedicated to agriculture (and legendary at it, too!)

With a goal of a somewhat realistic, if fun and gamey above-ground farming system, two points arise. The first is that crop yields must be higher than real life to allow a 200-dwarf fortress to endure on less than multiple embark tiles completely filled with farm plots. The second is that dwarves currently seem to eat very little and that nutrition is not a concern.

The problem of dwarven appetite raises three questions.

1. Does a dwarf satisfy hunger all at once? In Dwarf Companion hunger is tracked by a rather large, increasing value. That is, does a dwarf pick up a biscuit and eat it, returning to 0 hunger regardless of if he had 10,000 or 1,000 hunger?

2. Does a dwarf's food have a nutritonal value based on ingredients as well as a monentary value?

3. If meals did have hunger reduction based on nutritional value, how might that affect dwarven eating habits and consequencially, the demand for food crops?

At present I believe the first two questions are answered with 'no'. If that is true, then a fortress can sustain itself entirely on plump helmet spawn biscuits and wine, while another fortress might have the most elaborate *cow meat roast* banquets, complete with sides of dwarven cheese and *dwarven sugar biscuits*. Both fortresses' residents would still visit the dining room just as often. There is no incentive to create a varied diet, only an expensive one.

To find the answer to the third question, we must consider if each food had a nutritonal value as well as a monentary value. Dwarven sugar might be expensive but empty calories, while cave wheat flour is balanced as a happy medium and the basis of most meals. Meat and fish would both provide nutrient-rich substance to the diet. Dwarves could seek out food based on what their current needs were. Players would then have a strong incentive for crop rotations and importing overland foods that weren't local. Consumption patterns could also be altered. Dwarves might be inclined to pick up a sugar biscuit if they felt a bit peckish when they first woke up, stop for some vegetable stew halfway through their day, and have a good meat roast before going to bed. If dwarves required one medium grade 'stew' per month (which seems reasonable), we come up with three plants consumed a month on average or thirty six plants eaten a year, assuming they were all of nutrient-rich types. Assuming a one plant per tile yield, representing dabbling farmers and a poor harvest, this figures to one 6x6 plot to feed one dwarf for one year. In four seasons we have four dwarves fed. With skilled growers managing six plants 'surviving' per tile, we have 216 plants to harvest, feeding six dwarves. In four seasons, we have twenty-four dwarves fed. With legendary planters acheiving 12 plants per tile with all of the associated irrigation and fertilizing efforts, we have 432 plants feeding twelve dwarves. In four seasons, we have forty-eight dwarves fed; a quarter of a mature fortress. A 6x6 farm plot thusly becomes our new 'dwarven acre', feeding on average 24 dwarves assuming all plants are cooked into meals. Yields can be boosted by way of irrigation mechanics (an ajacent channel of water or brief flooding), fertilizers of various sorts, and any other methods added.

At this point four 'dwarven acres', planted by legendaries, irrigated and potashed, can feed an entire fortress requiring some 6,912 plants to be harvested and 2,304 *plump helmet stews* to be cooked. This can be done in a 12x12 area, or 3,600 square feet. But is this acceptable?

Consider: If booze is given no nutritional value, cooking booze effectively turns alcohol into a sauce rather than part of the dish and dwarves will be happier, but will not be as full. Dwarves will still demand ale, wine, rum and beer, meaning an extra field or two may be required to produce booze. Dwarves will still require fiber crops for the production of clothing. And most importantly, the current four-season agricultural system may well be up for changes.

For example, if the field now has its moisture level tracked it can easily occur that a summer dries out the field if the player fails to construct irrigation, or if the water source for irrigation itself dries up. This could lead to significant reductions in crop yield, or possibly wipe out an entire season. Freezing temperatures may also be tracked. Plants which can grow during a tropical winter might fail during a temperate one, likewise, summer may be the only plantable season in a freezing biome!

Above-ground farming thus becomes a matter of storing enough food to survive crop failures and seasonal changes, just as it was in 1400's Europe. Additionally, ambushing enemies may also spoil you crops while woodland critters simply eat the edges. Trampling fields could also ruin seedlings, making it wise to designate your fields low-traffic. With more imperfect agriculture, we may be looking at closer to eight or nine dwarven acres required to support a full fortress; a figure I am sure many players will find a suitable goal. Tending these fields would be labor intensive as we are looking at some 324 planting and harvesting jobs a season, or 1200 plantings and 1200 harvests a year to support 200 dwarves, each of which consume 48 plants a year (3 in a meal, 1 in a drink, twelve months a year). This comes out to a full-sized fortress consuming 9800 plants per year, requiring a yield of at least 8 plants per harvesting job.

If dwarves eat only once a season, divide all figures for crop yield and consumption by 4, however the labor demands for planting and harvesting remain the same. I believe if this general scheme were followed (modified, of course, but followed) we could well see livestock kept as 'famine reserves' and the fisherdwarf and butcher becoming a hero of society for their supply of high nutrient value ingredients. It will begin to make sense to trade with the caravans for food since your demand could well outstrip supply. Sieges which take control of your surface farming operations may well doom your fortress if sufficent reserves of longland grass flour aren't in a tightly-sealed iron vault for just such a contingency.

You might notice I never really touched on cave farming. Cave farming, in my opinion, should be a low-yield and almost subsistence level form of agriculture that you develop as an emergency food source against sieges or to provide some dietary variety. Fertilizers and flooding should all be required on a yearly basis to keep crop yields as high as four or six plants per tile. Assuming four plants per tile, a dwarven acre only feeds twelve dwarves per year, three or four of whom are in your food service industry to begin with. This is the reason for above ground farms to yield so much more; it allows developed agriculture to liberate the population to other pursuits; below-ground agriculture may well shackle them.

In any case, this post 'ran away' from me. I hope it adds to the debate.

[Dante]

First stacking needs to be fixed. Once plants can be stacked together and barrels topped off, low per-tile crop yields will no longer be problematic from an inventory standpoint.

Agreed; it would make a lot of solutions much more workable. And, Toady's said that he wants to get around to restacking at some point, so I'm pretty sure it can be done.

1. Does a dwarf satisfy hunger all at once? In Dwarf Companion hunger is tracked by a rather large, increasing value. That is, does a dwarf pick up a biscuit and eat it, returning to 0 hunger regardless of if he had 10,000 or 1,000 hunger?

2. Does a dwarf's food have a nutritonal value based on ingredients as well as a monentary value?

3. If meals did have hunger reduction based on nutritional value, how might that affect dwarven eating habits and consequencially, the demand for food crops?

Dwarves might be inclined to pick up a sugar biscuit if they felt a bit peckish when they first woke up, stop for some vegetable stew halfway through their day, and have a good meat roast before going to bed.

Unfortunately, calculations based on realistic hunger and time tend not to work out, because of the two levels: creatures moving in real time, and the year progressing in game time. If you try to mix game time and real time, you end up working out that a crossbow bolt takes two hours to fly across the map and hit a goblin. This is made worse because  food is already abstract, all portions size:1 and so on. I think before we can work out ideal yields/plot sizes, in terms of both realism and gameplay/fun, we need to see some big changes made to the entire food mechanism.

That's all working with just a portion size/hunger system - it's not even taking into account your third question, the possibility for discrete nutritional values.


On the whole, I agree with what you're saying. I'd quibble with your numbers a bit, but it'd just be quibbling. Eight or nine dwarven acres, at 6*6 each, is fairly reasonable, though I'd probably make it yet larger, having lower yields as the default. Again, that's founded on the need for improved stacking.

Cave farming, in my opinion, should be a low-yield and almost subsistence level form of agriculture that you develop as an emergency food source against sieges or to provide some dietary variety. Fertilizers and flooding should all be required on a yearly basis to keep crop yields as high as four or six plants per tile.

This bit, I think, is what a lot of people would disagree with. Sure, it's unrealistic that pouring water on rock makes arable soil. It's even worse in the 3D version, where it makes permanent arable soil. It's worse still in that things growing without light will need more nutrients to compensate for the lack of an energy source. But people see underground farming as integral to dwarfiness. So, I doubt that it will be nerfed completely. Requiring you to put more work into fertilisation or 'Nile farming' might be a reasonable compromise between ridiculous gameplay as it stands and harsh reality.

[Draco18s]

some biomes currently have as many as 5 soil layers, and the real world does have places where bedrock is thousands of feet down.

Yeah, I guess this would affect a bunch of stuff, including soil drainage and as Impaler[WrG] said, quality. Thousands of feet is a surprising figure, how common is that really?

I will admit that I did no research before throwing out that figure.  Soil ranges from none (mountains, duh) to 60 feet deep.  Which is certainly far deeper than we'd consider 5 z-levels (that'd be 12 feet for the open area + ceiling/floor).

[INSANEcyborg]

  As much as I like some of these ideas, some of them are just too complicated.  One of the things I like most about dwarf fortress is that its complexity comes from a ton of simple things.   

  Back on topic, I like the idea of having dwarfs eat more.   Make each piece of food a serving instead of a meal.  When a dwarf goes to eat, they claim a table (or spot on the floor if there aren't any), collect x pieces of food, then eat.  The amount they'd eat could vary, about one to three at a time. 

  How much they eat would determine how often they get hungry.  Right now a dwarf eats about once every 40 days I think.  If they were to eat a one unit of food, they wouldn't have to eat again for 30 days.  If they eat two, it'd be 50 days, and if they ate 3 it would be 60 days.   This creates a trade off between how filling the food is and time spent eating.  I don't know how the current hunger system works, so it'll probably have to be changed, but you get the general idea.

  How much they decide to eat could be tied in with the economy, or a setting in the Orders menu.  Early on, dwarfs would still eat one piece at a time, not much different from what we have now.  In older forts, they'd start to eat more per dwarf, somewhat keeping in line with the planters skill increases. Also, how much they eat at a time could affect how much happiness the "ate a good meal" thought gives. 

  Another idea is that nobles, and possibly legendary dwarfs, could ignore orders and eat as much as they want.  Maybe even 5 units at a time for nobles.  Might be too much FUN if your fort is starving and they feel like having a banquet, but then again it might not matter with all the unfortunate accidents that happen.  Anyway, this isn't that important, so it's a "maybe" idea at the moment.

  Finally, and I know this will sound strange, but what if dwarfs could fill a mug with booze and have a drink with their meal.  Weird, I know, but still...

[G-Flex]

There's an issue that a lot of people haven't been bringing up, unless I've missed it:

Dwarves shouldn't be good at absolutely everything.

Obviously, the different races and civilizations in the game have natural (and cultural) advantages and disadvantages, beyond just body shape. To some degree, this is already true.


So why should dwarves be amazing farmers? They live underground, for Pete's sake. There's not even any sunlight. Farming, by all rights, should be difficult for them not just for the sake of realism or basic difficulty, but for the sake of balance and consistency, so that dwarves don't become some sort of super-race within the game's universe.

Also consider things like sieges. Right now, it's way too easy for dwarves to just wall themselves in and become totally self-sufficient. This isn't even something humans could accomplish in their castles/fortresses, never mind creatures with no ventilation who don't even see the light of day, and there's only so much you can do underground without raiding the entire worldwide tunnel system for resources. Half the point of a "siege" is to screw up supply lines, and starve people out until they give up.


So really, I think there are several good reasons for making farming difficult.

I don't think dwarves should have to rely on trade to get all their food, but I don't think that underground farming should be anywhere near as good as above-ground farming, to the degree that having a large excess of food isn't really feasible for a dwarf fortress to accomplish.



On another note, I was talking to someone on IRC about this, and the idea of fermentation came up. Dwarves are obviously already familiar with fermentation (of the alcoholic variety), so it's not too much of a stretch (especially in a fantasy world!) to say that they could ferment things that humans can't normally digest, like the cell walls of mushrooms, into something edible. Hell, maybe that's how plump helmet wine works.

This isn't really far-fetched at all. Of course, you could also say that dwarves are like cows or something, in that they have symbiotic bacteria living in them that does this for them, allowing them to eat mushrooms all day and derive more energy from it than a human ever would, which is also pretty believable. This all opens the door to strange ways of preparing food in general, which seems to fit the bill for dwarves.

There's also the idea, of course, that dwarves rely on alcohol for this reason: They're fermenting junk all the time anyway, and brewing booze out of stuff is one way of getting food energy out of it (although you'd probably be better just leaving it in the form of sugar, that's obviously less fun).


This is what I like about treating weird stuff like this realistically: You get to find ways to keep the weird element while justifying it, and the justification can lead to yet more weird elements that dovetail nicely together into a world that's fantastical, yet feels consistent and believable.