Print

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

[Lord Shonus]

I'm not actually suggesting dwarves should eat more food units in a single meal. I was suggesting to limit the supply of food in fortress mode by dividing all food sources by 10. The end result: a single cow gives 1 to 2 meat and a single dwarf still eats 1 unit of food per meal.

---

Anyway, thanks for your questions. I now see I must rewrite the OP to prevent misunderstandings like these.

That's rather absurd. A single adult cow provides, in the real world, enough food to support an adult human for well over a year, assuming that the meat is properly stored. That does not include organ meats. To suggest that that same cow would only support a dwarf for two months makes little sense.

[NW_Kohaku]

That's rather absurd. A single adult cow provides, in the real world, enough food to support an adult human for well over a year, assuming that the meat is properly stored. That does not include organ meats. To suggest that that same cow would only support a dwarf for two months makes little sense.

Of course, that same cow doesn't eat any food itself to produce that meat, as long as you have uncaged cows to get pregnant, you have a food source where the only labor involved is the actual butchering, and the hauling of body parts.  Plus you're able to milk them, as well.  Right now, we get about 24 usable parts off a cow, plus about 8 fat... on a normal cow, that's if you don't use unnatural selection to make them all gigantic and fatty and meaty.

Still, I'm not fond of the idea of dividing the current meat by ten, but rather just making dwarves eat X units of food at a time, so that we don't have to deal with fractions of food being produced.

I honestly don't see a big problem producing food even very early on in this version.  In fact, I can pretty much get by just by using berry-gathering, which I would do anyway just to get some seeds for aboveground farming without having to hope that the elves will trade me decent seeds.  After that, a simple manual pump next to a murky pool in a walled-off area is enough to get you your first temporary aboveground farm.

[Lord Shonus]

Feeding of pets is already scheduled, and compensating for the lack of it with a clear hack would seem rather shortsighted.

[NW_Kohaku]

Feeding of pets is already scheduled, and compensating for the lack of it with a clear hack would seem rather shortsighted.

Even if we set the target as a cow feeds a dwarf for one year, then we are talking about Jiri having picked the wrong arbitrarily selected number to cut the food rates by.  To fit 24 meat and 8 fat into one year, then a dwarf should be eating four units of food, or the amount of food should be divided by four, not ten.

Still, I think maybe it would be best to make the amount of meat and food produced somewhat larger, and then make the amount of food a single unit eats based upon the size of the creature in question, especially if we are talking about having livestock being fed.  As hungry as dogs might get at times, I still think they will be eating less in one sitting than your tame elephants will.

[Lord Shonus]

Exactly. That would be a reasonable change, rather than a quick and dirty hack.

[Jiri Petru]

Yeah, dividing by 10 was an arbitrary, random number I came up with. If you think 4 is better, then 4 it is! (does a single cow really feed a person for a whole year? wow). The important part here is that I'm suggesting to handle food in a similar way how thread/cloth is handled now. Ie. 1 thread is actually how much... 10000 thread units? But it gets divided by 10000 for most purposes. Food in fortress mode would be similarly divided to make up for dwarves only eating 8 times a year.

The actual numbers would have to be balanced later. There are other things than livestock to take into consideration, such as field yields, etc.

The idea to have the divider/multiplier dependent on size of your dwarves/humans/kobolds/whatever you play as sounds like a good one. But I'm worried about the implementation. No matter how you divide, people should always eat only 1 displayed unit of food per meal. In other words, the division for eating should be the same as the division for display, so the stocks menu works properly. "100 food units" should always mean "100 meals available".

I'm not sure how this would work out for fortresses with different races/castes. We definitely don't want one cast eating 0.8 food per meal and another one 1.4. Perhaps rounding would be in order? Imagine you have a fort with kobolds, dwarves and humans. Kobolds would eat for example 0.8, which gets rounded to 1. Dwarves 1.0. Humans 1.5 which gets rounded to 2. End result: kobolds eat as much as dwarves do, humans eat twice as much. Simple to remember for the player.

1.0 should be the standard for the average size of your race, the most common race in your civilisation or something.

--------

Nil Eyeglazed: Sorry, I don't have much time now. Will respond later.

[jaked122]

 it is still more fun to trade tomato soup rather than pickled tomatoes and candied water.

[NW_Kohaku]

Yeah, dividing by 10 was an arbitrary, random number I came up with. If you think 4 is better, then 4 it is! (does a single cow really feed a person for a whole year? wow). The important part here is that I'm suggesting to handle food in a similar way how thread/cloth is handled now. Ie. 1 thread is actually how much... 10000 thread units? But it gets divided by 10000 for most purposes. Food in fortress mode would be similarly divided to make up for dwarves only eating 8 times a year.

The actual numbers would have to be balanced later. There are other things than livestock to take into consideration, such as field yields, etc.

The idea to have the divider/multiplier dependent on size of your dwarves/humans/kobolds/whatever you play as sounds like a good one. But I'm worried about the implementation. No matter how you divide, people should always eat only 1 displayed unit of food per meal. In other words, the division for eating should be the same as the division for display, so the stocks menu works properly. "100 food units" should always mean "100 meals available".

I'm not sure how this would work out for fortresses with different races/castes. We definitely don't want one cast eating 0.8 food per meal and another one 1.4. Perhaps rounding would be in order? Imagine you have a fort with kobolds, dwarves and humans. Kobolds would eat for example 0.8, which gets rounded to 1. Dwarves 1.0. Humans 1.5 which gets rounded to 2. End result: kobolds eat as much as dwarves do, humans eat twice as much. Simple to remember for the player.

1.0 should be the standard for the average size of your race, the most common race in your civilisation or something.

That's why I'm saying not to divide.

If we had dwarves that just ate, say, 4 units of food at once (and considering how most food comes in stacks of at least that size, that's not a problem), smaller or larger dwarves might eat 3 or 5 food at once.

It still might be better than that to multiply all food by two or three (both how much creatures eat, and the size of the food stacks any food source gives you), and then you could have dwarves that eat 8 food normally in a sitting, but eat between 6 and 11, depending on how large they become (with things like strength adding to mass, by the way...)

If we have fractions, rather than always rounding, it could be like the speed system, and simply have a percentage chance of rounding up.  So someone who is large enough to eat 7.45 food per sitting would have a 45% chance to eat 8 food, and a 55% chance to eat 7 food.


edit:  Also, something about the way that the food system currently works: Looking at Dwarf Companion back in 40d (no reason to assume much has changed on this front, other than when the dwarves actually decide to eat), hunger is calculated like this: 

There is a Hunger variable attached to every creature who might need to eat.  It is a variant of an intiger, starts at 0, and incriments 1 for every frame/tick of the clock.  There are 403,200 frames in a year, so dividing by 8 brings you to 50,400 frames between meals.  Looking at DC data corroborates that dwarves will probably eat at around 50,000 on their hunger meters.  I am not sure exactly when they start to actually starve to death.  Eating any type of food will immediately revert that hunger meter back to 0.

Using this knowledge, we could actually make food that, instead of simply resetting the hunger variable to 0, actually just subtracts from the hunger variable.  For example, given the new dwarven average size of 70,000, and a target of requiring 4 food on average per meal, and a target hunger reduction of 50,000, we have enough to solve for the variables.

50,000 hunger divided into 4 servings is 12,500 hunger per unit of food for a 70,000 size creature.  So, then, a single unit of cow meat could reduce  875,000,000 hunger divided by the size of the creature.

A dwarf who is hungry could run a quick calculation to figure out how large a stack of food he/she would need to pick up to sate their hunger, and if rounding down, it simply means they wouldn't go all the way down to 0 hunger, and would just eat a little more frequently, even if eating less at a single point in time.

A more complex system (but that just makes it more dwarfy).

[NW_Kohaku]

While I'm on a roll, why not make things even more messy by including some of the nutrition suggestions in the game, as well.

To add another reason not to simply let dwarves eat raw meat, the game could put food into different nutritional categories, and encourage a balance of these categories through reducing the hunger-reducing value of a given type of food the more that the given type of food you are about to eat has already been a part of your diet.

To make a nice, arbitrariy round number, instead of that 875,000,000 number in the last post, we could have a starting value of 1,000,000,000 hunger divided by the mass of the eater, provided they have a 0 value for nutrition from one food group.  This 1,000,000,000 would be subtracted by the value in the variable (probably once it has been multiplied by 1,000 or 10,000) for the food group, so that, if we want an average of 875,000,000, then that value should normalize at around 1,250 (assuming we are multiplying by 10,000, so that it would be 1,000,000,000 - 125,000,000 = 875,000,000).  (Sorry for the superfluous math, but I realize this is complex, and want to make sure that people can follow.)

(This food group variable is a integer, since for programming purposes, that would be somewhat easier on the system than a floating point decimal.  Because I'm multiplying it and dividing it around, it needs to be a largish integer, since there is no precision below the ones digit.  Hence, the next part, where eating one unit of food pushes you up a few hundred points in a food group.)

The value for each given food group could go up by a variable amount each time that the dwarf eats food from that group.  (Let's say for right now, just for round number's sake, this variable, which I will call "A", is 1,000.  So every time you eat meat, the meat food group variable goes up 1,000, and meat will give you 100,000,000 less hunger reduction per unit mass your creature weighs.) 

At the same time, after you eat, so that food starts to give you more nutrition back, all the food group variables would need to be reduced.  (Theoretically, this would happen gradually, as your body uses up the nutrients in the food you have eaten, but if this only applies when we are actually eating, then we can save some processor power by making these calculations only take place when the dwarf is going through the eating process.)  The exact weights of every variable depend on the number of foods we are throwing around to be balanced.  However, I can make a little formula:

Food_Group_Variable = (Food_Group_Variable + A (if this is the food group eaten, if not, this is 0)) * B - C;

Where Food_Group_Variable is the amount that the food's nutrition will be reduced (once multiplied), and B is a fraction/percentage (may actually just be a whole number that is subsequently divided by 100 so that there are only integers in the function), and where C is a flat value subtracted every round of eating from the Food_Group_Variable.

B is needed so that this doesn't simply become a process where you just eat one food in a cycle going around the food groups.  Eventually, you would be able to more-or-less get by on having missed some food groups more than others.  C is needed so that the values will eventually start to reach 0 again.

As nutrition from certain food groups reaches 0, negative symptoms or other effects might also be applied to dwarves, making them slower, and more vulnerable to disease or injury.

Keep in mind that the different food groups do not need to use the same variables A, B, and C.  A (the drop in nutrition) might be larger for things like meat, but lower for things like vegetables or fruits food groups.


This whole complex mess could be used as part of a means of encouraging prepared foods of higher quality, which, by mixing food groups, could potentially help balance nutrition a little better.  (You might even have a Chief Nutritionalist dwarf!)  These prepared meals could split the "A" part that is added into different food groups, as well as encourage greater overall happiness in dwarves.


EDIT: I somehow get the feeling I just scared everyone out of the water.

[Ultimoos]

Not at all. This idea is great. Now we can feed entire fortress with strawberries with out having to bother with other food too much. If the nutrition system was implemented that would force providing better diet for dwarfs. There could be a couple of different nutritions. And a lack of something would make dwarfs more sleepy, lazy, less effective in combat or even make them easily fall for sickness. Prepared meals would now be more effective in feeding. And so you would have to import different food types to provide dwarfs with nutritions that you have problems with creating on your own.   

[Grendus]

I would love to see something like this. Currently the only reason to grow dwarven wheat or any of the above-ground wheat varieties is for flavor. In real life, the reason we eat so many grains is, as stated, they keep for a very long time (if kept dry they can last for centuries and still be edible), they have a very high calorie to acre ratio, and they can be grown in a large variety of conditions.

Even with these changes, players who don't want to bother with preserved meat can just farm more plump helmets. So long as the game remains as moddable as it is, players who want a challenge can mod in more realistic plants (low yield edible-raw plants that rot fast with the bulk of the food industry being focused around slightly higher yield grains) while players who don't want to fuss with it can use the stock plump helmets and simply grow them in bulk for their fort.

[Nil Eyeglazed]

I like nutrition as a problem for cooking to solve.

As in real life, specific nutritional deficits could take years to develop, which is perfectly appropriate-- it gives you time to get on your feet.  All of this could be done within the frameworks that already exist in the game, maybe as part of a syndrome expansion, or as part of the development of stat improvement.  I'm afraid I don't quite understand Kohaku's suggested implementation   I can't imagine how quality would increase the number of food groups involved, other than through more easily making lavish meals.

It would, however, hurt some of the more extreme forts out there.  Farming isn't always possible, and neither is trade.

I'm not as fond of the idea of requiring multiple units of food based on size.  It ends up making smaller races more productive, just because they spend less time hauling.  This doesn't seem right to me.  On the other hand, it doesn't really seem right to me either that larger races could store more food in a single square than a small race could.  I think there's room for some abstraction in the game.

[NW_Kohaku]

Prepared foods should be able to be in multiple food groups at once when they are composed of multiple types of foods.  I.E. a stew made of beef and quarry bush leaves is giving you both meat and vegetables in a single serving of food.

The food groups system, however, is just a simpler way of abstracting dietary needs, and this could alternately be done based upon stating out various vitamins that any given food might be able to give you, although this would obviously be Hell in the raws, as you'd have to do more than just say that any given plant is a vegetable or a fruit, but go out of your way to stat out various vitamins (plus for every creature, their meat would either have to use some kind of default meat nutritional value, or else have a stating-out of their meat).

Also, if this means that a fortress that cannot farm, and relies upon exploitation of the fact that livestock currently don't eat to breed endless supplies of meat from absolutely nothing... well, I can't say I'm going to be terribly saddened by the loss.

[Nil Eyeglazed]

I can't imagine it being too difficult to have a scalable nutrition system.

Creatures get something like, "NUTRITION_TICK:PROTEIN:100" and the meat default gets "NUTRITION_SOURCE:PROTEIN:120000".  If somebody wants to get into more detail, they just add "NUTRITION_TICK:B12:50" to their dwarfs and makes a few foods supply B12.

With the way DF works right now, though, I can't imagine much room for a lot of complexity.  There just aren't enough different food sources to get much beyond basic food groups.

Then, we'll need to figure out a way to make polar bear liver poisonous

[NW_Kohaku]

Also, an advantage of food groups over specific nutrients is what it would look like when you find a problem:

Urist McCarnivore has been feeling sluggish and craving fruits lately.

vs.

Urist McCarnivore has been having gum pains related to a Vitamin C deficiency lately.