Print

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

[AngleWyrm]

Ok, horse of a different oder there.

So how is the soil chemistry going to be played? Is it indirect action through planting of various crops on a plot? Will there be any direct user action, and if so what will it be?

Note that hunting and fishing would become much more important in this system, if the map allowed it, and would probably help a lot to reduce the estimates of plot/farmer sprawl suggested. Otherwise... yeah, if you look at medieval tech, the suggestion is (if anything) too generous. I approve.

Hunting and Fishing are currently capped. Population of wildlife runs out and then the rate of hunting/fishing is dependent on the respawn rate of wildlife. In my forts it only takes one or two hunters and two fishermen to completely use up the incoming animals. So if hunting and fishing are to play a more significant role in feeding the population, then the rate at which critters re-appear in the zone is going to have to increase.

The same situation will happen for farming that requires wood as the prime resource. A ceiling will be hit, which is the regrowth rate of trees. It will also cause a big difference in initial productivity between the different levels of forestation. The number of useful embark zones will go down, with the scales being tipped even further toward Thickly Forested as the preferred embark.

[NW_Kohaku]

Hunting and Fishing are currently capped. Population of wildlife runs out and then the rate of hunting/fishing is dependent on the respawn rate of wildlife. In my forts it only takes one or two hunters and two fishermen to completely use up the incoming animals. So if hunting and fishing are to play a more significant role in feeding the population, then the rate at which critters re-appear in the zone is going to have to increase.

The same situation will happen for farming that requires wood as the prime resource. A ceiling will be hit, which is the regrowth rate of trees. It will also cause a big difference in initial productivity between the different levels of forestation. The number of useful embark zones will go down, with the scales being tipped even further toward Thickly Forested as the preferred embark.

Actually, I hope the large wildlife respawn rate goes down.  Currently, any time animals are killed or walk off the map, the next batch immediately walk on another edge.  It's actually silly how quickly the dead are replaced.  The only real threat to an end of hunting's usefulness is actually killing off every single elk in the entire biome, which often has thousands of the creatures.  Having a minimum time between creature spawns would help greatly.

If forested areas become more useful because you have more use for trees... so what?  They're already fairly useful because most people already find plenty of use for trees, and people often specifically embark on forested areas because of that.  What's actually more purposeful is to make players more aware of water concerns, and making sure they will be able to embark on an area with enough continuous water supply to feed their farms, more than trees.  Embarking on a desert should make it harder to grow crops; There's a reason why most areas in the world that practice major agriculture are either in deforested areas or on top of major aquifers.

And yes, the number of trees you can harvest per year is limited.  I specifically mentioned this, and why it is a good thing.

So how is the soil chemistry going to be played? Is it indirect action through planting of various crops on a plot? Will there be any direct user action, and if so what will it be?

I'm a little confused by these questions, if you are asking what I think you are asking, I just answered these questions a couple days ago.  If not, could you specify what you are asking?

Spoiler: Recap (click to show/hide)

Now I would very much support a product cycle that took animal waste -> fertilizer and water needs firmly into account! But then, in terms of resource management, that would be the lion's share of the challenge and screwing with soil chemistry would just be a bit of window dressing.

Yes, the more work gets done on this, the more it moves away from focusing on soil chemisty.  That's not really a bad thing, though, since you can just consider some of the other ways in which this game is detailed/complicated - metals and stones (and wood and bone) now have attributes that make them more or less suited for certain tasks than others.  This requires a slew of complex statistics (many of which don't seem to have any purpose yet), but which don't really matter to the player - all that really matters is that different materials are better or worse for different purposes. 

If we are going to make fertility management (and fertilizer and water management) the focus, then we need ways of making fertilizers different from each other, and that requires having different stats for them all... (And keep in mind that real-life manure has different qualities, based upon the animal and its diet... rabbit and avian manure are more valuable than cow manure, and free-range grazing cattle have worse manure than dairy cattle that are fed grains... theoretically, we might even go the EXTRA extra mile and stat out the values of every creature's manure.   How would THAT be for a tagline?  "Dwarf Fortress - We optimize our bodily waste!")

Anyway, back on track, the purpose of soil chemistry was never to force micromanagement, but rather to force the player to have to balance the system out, so that things are not QUITE as simple as "designate farm, get food".  We don't have to force the player to mess with specific phosphorus rates, we just need to give the player a mechanic whereby phosphorus, even if invisible, plays a part in how one has to pay upkeep on their fortress.  You can see that in the differences of how the nutrients really play out:

Nitrogen: Runs out fast, but can be replinished easily.  Nitrogen-fixation cover crops are easy ways to replace this, but are useless except for grazers.  Nitrogen-fixation food crops (like beans or peanuts) are lower-producers, and restore less nitrogen, but you get some nitrogen back, anyway, and you're still getting more food out overall.  Nitrogen comes in most fertilizers (so you can avoid serious crop rotation), but nitrogen-heavy fertilizers, especially Urea, can acidify soil, forcing even more dependance upon adding stuff to the soil.  (In Real Life (IRL) - Nitrogen is technically the nutrient that encourages growth, especially leafy growth.  Overmuch burns the leaves, and can cause the plant to focus more on growing leaves than fruit.)

Phosphorus: Depletes somewhat slowly, but is difficult to replace.  Biological waste (manure) gives you a spread of all three of NPK, but only some kinds of manure are really high in Phosphorus.  No crop rotation method really adds more P to the soil, either, so you pretty much have to add something to the soil as a fertilizer to replace P depletion.  Bone Meal is higher in P than other types of biological fertilizer (but also contains N and K), and the only other ways of artificially adding P in any serious quantity is through mining out P-containing sediment (good ones are not common), and then breaking up the stone.  (IRL - This is the nutrient you want in as high a quantity as you can get - it's what helps contribute to the size of fruits.)

Potassium: Common in soil, and slow to deplete, but also slow to raise.  You probably shouldn't need to go out of your way to replenish K too often.  Most manures have plenty K, even if composted first, so just keeping up a regular manure schedule should keep K from being a problem.  This should only become a problem if you go for intensive nitrogen fertilization rather than crop rotations, letting K gradually decline.  K is slow to deplete, but once depleted, takes years to replinish. Potash is pretty much the most refined form of Potassium dwarves can make, but it might be better to just compost the whole tree for a broader spread of nutrients than burn it down just to get the pure potassium unless you are desperate for rebuilding K in the soil.  Keep in mind that now, we can grow trees on our farms. It's kind of silly to chop down and burn trees for the ability to grow more trees.  Especially if regular soil has nutrient depletion, and repeatedly chopping down trees can lead to eventually causing the untilled soil to deplete, as well, and you run out of wild trees to cut down.  (IRL - You need as much if not more of it in the soil as nitrogen, but it tends to bind to the soil instead of being as freely water-soluble as nitrogen is, so it depleates at 1/10th the rate of Nitrogen.  This element is critical in root growth, and disease prevention.) 

pH: Like Potassium, this shouldn't move around too much.  It's mainly a problem when first starting a farm, when trying to change from one set of crops to another, or if you use acidifying fertilizers for a fair period of time.  This should be slow to change in general, taking a full year to change + or - 1 pH.  Most real life crops grow around 6.0 to 6.8 pH.  Alkaline (7.0 to 8.0) crops tend to be "desert" crops that prefer low-water environments.  Alkaline soil prevents some nutrients from being available to crops.  Highly acidic soil (4.5 to 6.0 pH) are the sorts of things you see in extremely rainy areas.  Acidic soils tend to hurt crops, but can wind up being better suited for some crops, because the weeds that would choke them out are less tolerant of the acid (and the higher content of iron and other elements) than the crops are.  Soil acidity is changed by liming the soil (with limestone or chalk) or by using elemental sulfur (brimstone).

Water: It's water.  Plants need it, obviously.  They need lots of it.  Plants that are suited for dryer soils have to make do with a lower amount of sugars, and tend to be either less nutritious or just plain slower growing than plants that guzzle water.  This means that your water supply can determine whether you are tilling a broad area to make use of slower-growing or less-nutritious crops that take less water (and labor per tile) or concentrated farmland that take plenty of water.

Biomass: This is undecomposed stuff mainly for underground fungi to grow off of.  Aboveground, this goes up and down pretty freely, depending on how much of the plant gets tilled back into the soil.  While fallow, it depletes naturally as soil bacteria decompose it into nutrients, and plants can become vulnerable to diseases if this gets too high, while others prefer a thicker layer of humus.  Underground, however, this is more important, and represents the need to constantly supply underground farms with more energy in the absence of photosynthesis.  This requires plenty of dead stuff being added to underground farms.  For fully subterranean civs, it might force that magma farming I mentioned earlier as a way of growing some chemosynthetic life to use to decompose for biomass.

There's also the things I talked about with needing to potentially clean air or generate light underground being solved with farmable underground molds that generate oxygen or can glow... these would also require access to the same watering and soil chemistry needs as other plants, forcing a choice as to where to put your resources.

Again, simply making the numbers the right numbers for one field for one crop isn't supposed to be the game.  It's making players take a look at what they can use as resources to fertilize their fields, and trying to work out what combination of crops and systems for irrigation or fertilization or other aspects they can use to make the most of whatever resources they have in shortest supply.

Deserts and glaciers should definitely be very hard to get a fort established in. (Though a small ice-fort should be able to survive by hunting and fishing... sort of.) On the other hand dwarves should not be morons and seek to mass-emigrate to fortresses in inhospitable areas. You'd have to go the extra mile to get dwarves to come over, I'd think.

Yes, I've long supported a throttle on immigration, just a percentage multiplier, so that you can cut immigration to 10% normal or 5% normal or something like that.

Dwarven waste could probably be discreetly added to a stockpile with the animal waste, without having them do the potty-break thing. Could go either way on implementing loos tho.

Heh, teleporting the waste to the stockpiles or sewers seems to break some of the core realisms of the game, although it might be a nice way of not dealing with the grittier details.

I for one hope that he is thinking exactly that. Resolve the food overabundance by reducing the per-tile farm productivity until an acceptable ratio of farmers & land-to-others is achieved. Having 137 farmers/brewers/cooks in a 200-dwarf fortress, all struggling at skill-10 to just barely produce enough food for everyone. It might involve reducing the output of the individual farmer, so that a pleasing plot size can be selected. And it could be done fairly simply by increasing the time it takes to plant and reducing the crop yield bonuses of farming.

Like I've said before, expanding farm size and labor requirements isn't really a solution to making farming a better, more interesting system, it just means that more of your map needs to be cleared away for farming, and that you need more farmers.  That isn't to say that we shouldn't make it a more land and labor intensive process, but that land and labor alone aren't a solution, as the problem is a lack of complexity and hence, diversity and a need to plan in farming.

I like the idea of additional detail, that would ramp up crop yeilds based on some player interaction/knowledge. But I'de like it to be additional fun stuff, rather than a necessary slog. Maybe by spending time applying pH adjusters and additional nutrients, the number of farmers needed is effectively reduced because of an abundance of crops. So the skilled player gets more mileage out of his staff, whereas the beginner can get by wastefully.

Actually, this is supposed to be part of the system... If we force farming to be more dependant upon either adding something to the soil or by careful crop rotation and resource management so that you can get by on just the manure, then we can make the entire need to feed your people be a scalable difficulty.

Your starting seven can waste as much as they want to - just chop down trees, and grow plump helmets on the logs after you muddy some dirt!  Easy!  You just have to take some external resources that are finite in amount. 

As your fortress gets larger, however, if you have to depend upon cutting down trees to feed your mushrooms, then you're going to run out of trees.  You have to work with less fertilizer, and have to stretch your resources out more just because it's finite.  You need to rely more and more on the manure you can get in a scalable amount from your dwarves and your livestock.

If water has a more meaningful finite nature, as cavern water supplies and brooks and the like are less useful supplies of water, then you can waste as much water as you want early on... but as your farms get bigger, you need to take the same finite supply of water, and feed more crops with it.

If you rely upon just using readily-built-up soils from soil layers, then as you expand, you have to start using less and less desirable stone types, and break them and build the soil on them to make them useful as you expand your farms, increasing the overall amount of work it takes to set up any given farm.  (And as you have to expand it, your aquaducts have to move more water further through the fortress... plus you'll have the sewers to build around, as well.)

Then there's the problem of pests.  They're meant to be a way of forcing crop diversity, by making crops more vulnerable to pests the more that you rely upon them.  This forces the player to reach out to perhaps less desirable crops than they would at first reach for as their fortress starts up.

Basically, the longer you go on, the larger the farm, the harder and more complex it gets.  This is the scalable, end-game-is-harder-than-setting-up mechanic that was discussed back around page 31 that set this whole thing in motion... and you just can't do this without some sort of complex background mechanic in place.

Well, to compare it to a RTS game again... In Supreme Commander, you have two resources - Energy and Mass. 

Mass is something you can only gain from limited points on the map, although they have infinite resources, they are limited in how much you can extract at once.

Energy is fairly easy, you just build more power plants, although these power plants are an Achilles Heel in your defense and difficult to protect, and eventually require expanding out rediculuous amounts of land into giant power farms.  The main reason you have to worry about these is that you have to keep building more of them, and the defenses to go with them.

Ideally, there could be some sort of mechanic like the Mass mechanic, where you have to manage resources because there is a finite limit to the amount of food you can grow on a map.  Fertilizer and water should be a limited commodity, and stretching them out may be a priority, especially on maps with little water.  If you cannot water all your crops, you are forced to start deciding which types of crops you need based upon what sort of resources you have available, and what your major priorities are.  If you cannot get enough of a certain type of fertilizer, it means you'll need to heavily weigh the benefits of certain crops if they are heavy potassium consumers on a map with little trees for potash. 

If all else fails, however, there is the Energy mechanic - you want to maximize crop yield and stretch your crops out as far as possible because the actual act of clearing out a farm, and carving the rock, then starting to break up rock into soil (which, if we have to go this route, there may need to be a better mechanic for determining whether it is depleted soil that can be replinished, or if it is the much more difficult task of breaking apart the stone into soil in the first place), and setting up the infrastructure to water these crops will have to be time-and-labor consuming enough that, like setting up magma forges, players may simply put it off, and consider conserving food and taking fields off of luxury goods or fabric farming for food rather than expanding the fields.

As for underground crops taking up more resources, the biomass variable exists for that - it essentially is a measure of undecomposed organic matter that fungi and other non-photosynthetic crops need to survive.  Biomass is something that wouldn't be of much concern in aboveground farming, as it depletes naturally as soil bacteria eat away at them, especially if fields are fallow during the Winter.  Biomass for underground crops, however, needs constant replinishing with undecomposed dead stuff.  (And dropping extra trees and dead bodies should be one of those "Mass Mechanic" type things, where you only have so many dead bodies per year to chuck on the compost pile to turn into mushroom fertilizer.)

edit: oh, and one more thing: The pest mechanic also exists as a way of ensuring that no one crop gets too prevalent, which in turn means that as the fields are expanded, if you want to avoid having too much density of crops, means you have to start expanding into different kinds of crops as you add more plots.  When there are only a small number of plants that can perform certain functions (say, pig tails and rope reeds are the only cloth), and pig tails are easy to plant, but rope reeds are a pain, or only grow in warmer climates, so you have to run magma beneath the farms to keep the soil warm or something, and require far more careful management or more infrastructure, then you are adding a scaling level of difficulty as your fields expand.

Also:  Keep in mind that for sustainability, the best ferilizer would be manure... which we may not get, depending on how Toady rules on it.  If manure is your main fertilizer, then you have a limit on the amount of fertilizer you can spread - you can't get more manure than you have creatures generating it, and those creatures need to be fed... by the crops raised on your manure fertilizer.  And even then, you have a choice between spreading raw sewage on the fields and composting it first, which makes the manure safer (more crops tolerate it, and it is less likely to spread weeds and pests) but releases many of the nutrients (especially Nitrogen) from the compost.  The benefit of this being that it would create a "closed loop ecosystem", where most nutrients stay within the fortress, and also that it would make you choose which fields get priority for the limited supply of fertilizer - do you balance it out, or grow some crops with high nutrient needs and let the other fields suffer lower overall nutrient replinishment, forcing a greater dependency upon Nitrogen-fixation crops and light feeders, and generally sacrificing productivity in one field for the productivity of another.

If we DON'T have manure, then this really screws with the balance of the system, and you have to start looking for other means of fertilizing crops, which would possibly have to come in large part from dead things... either dead trees (finite resource), or dead creatures... which means that you'd have to hunt or level seiges just to dump the corpse in the compost heap.  It would make the fertilizer supply much more difficult to manage, and mean that if a seige didn't come, you might wind up starving for lack of fertilizer some years, but have overabundances of it other years... a very strange position to be in.  (Which is why it really would be best to have manure for fertilizer.)

[AngleWyrm]

And yes, the number of trees you can harvest per year is limited.  I specifically mentioned this, and why it is a good thing.

It sets an upper limit on the amount of food generated from trees. That limit will be different for each level of forestation. It does not remove the question of how much food is the right amount for a fortress to generate, but it partially hands the question off to the forestry division. It will require figuring out how many trees vs how many mushrooms generated. If it only affects a limited number of the total possible crops, then it could be interesting to choose crops based on local conditions.

The questions I asked boil down to How do you play with this? For instance, I can see planting some nitrogen rich crop in order to boost the nitrogen in soil. There would have to be a way to monitor the nitrogen levels in the soil and the plant. Perhaps 'Q'uerrying a farm plot gives an additional menu item for 'S'oil metrics. Then those metrics could be represented by bars in the same way that selecting how much wood you want from a liason is a bar. The bar could have a marker for current reading (two-color bar), and a second marker for the requirements of what is planted (different glyph). The player can then select a crop, switch to soil metrics, and see the fit, escape back to select another crop, and in this way learn the details of the various real and imaginary plants in the game universe. A marker that is at one extreme means the soil is rich in that metric, or the plant consumes much of that metric. A marker that is at the other extreme means the soil is poor, or that the plant returns more than it consumes.

The available soil chemistry readings (Nitrogen, Potassium, Phosphorus, pH, Water, Biomass*) could be presented in one page similar to the liason orders pages. The desired setting is of course the needs for the current crop, and dwarves should automatically go about watering, fertilizing, and treating the soil to reach that target, provided there is a supply of the various nutrients.

The supply of nutrients are going to cost money to bring in from the Caravan Arc. It will be necessary to decide how much money it will cost in order to raise each crop until harvest. Note that the cost will vary according to how close to the crop's nutrient requirements the soil already is. Which also suggests that the various rock and soil types may have different base level set points.

These base level readings per soil/rock type could be the springboard that feeds into a wide variety of imaginary flora, covering unusual rock types as well as more typical plants covering basic soil types. Maybe different undergrowth flourishes in sedimentary layers, due to several metrics being in the same basic configuration. There might only be a few characteristic configurations to start, with refinements and their additional ecosystems being added over time.

*Just a note on dimensionality:If there are a mere three settings for each of six soil metrics (little/medium/lots) then 3^6=729 distinct soil configurations. Increasing the resolution to ten steps each results in 10^6=1 million unique soil signatures.

And just because every good thread needs pictures: A typical day on the farm

[ZebioLizard2]

Currently, any time animals are killed or walk off the map, the next batch immediately walk on another edge.  It's actually silly how quickly the dead are replaced.

I've *Never* had this happen before, what type of area's are you embarking in? I usually have to go several ingame years without animals nearby at times.

[nil]

Currently, any time animals are killed or walk off the map, the next batch immediately walk on another edge.  It's actually silly how quickly the dead are replaced.

I've *Never* had this happen before, what type of area's are you embarking in? I usually have to go several ingame years without animals nearby at times.

That's how it works in my fortress.  Might possibly have to do with how many animals are in the region you embark in?  I'm in the middle of a huge, isolated evil desert so I'm pretty sure there's thousands of these zombie camels I'm constantly killing.

[NW_Kohaku]

Sorry to take a while in responding, but AngleWyrm, I'm going to respond to your main question by just posting the "recap thread", which should hopefully answer you question... it's just been taking me a couple days, because I get distracted, and it's just getting a little "Too long; didn't write" to sit there and copy down what I've already written, and reformat it... It's so much easier to get motivated for writing these long things when it's in response to someone than when I'm just writing to the ether.

Anyway, to respond to the thing about hunting, first, I was talking about hunting, not fishing, I personally never even fish at all anymore, because I could deplete the entire turtle stock in my murky pools in just a couple months, anyway.  Those are typically very low unless you have a lake or something.

For hunting, however, I always see creatures spawn on the edge of the map the instant the last member of a group of animals is removed from being actively on the map.  (I don't hunt, I just set cage traps and wait for them to be filled.)

[AngleWyrm]

I have two fishermen and two fish cleaners in my current fort of 90-ish dwarves. The four of them produce approximately 1/3 of all the food in my food storage area. I measured this by selling all my fish barrels to the caravan, and it left about 1/3 of my food storage area empty. My farming team consists of six growers, and two each of brewer, cook, and thresher.

...And even then, you have a choice between spreading raw sewage on the fields and composting it first, which makes the manure safer (more crops tolerate it, and it is less likely to spread weeds and pests) but releases many of the nutrients (especially Nitrogen) from the compost.  The benefit of this being that it would create a "closed loop ecosystem", where most nutrients stay within the fortress, and also that it would make you choose which fields get priority for the limited supply of fertilizer - do you balance it out, or grow some crops with high nutrient needs and let the other fields suffer lower overall nutrient replinishment, forcing a greater dependency upon Nitrogen-fixation crops and light feeders, and generally sacrificing productivity in one field for the productivity of another.

In order to make a "closed loop ecosystem" it should include the consumption and output of dwarves, which suggests modelling dwarf manure. It could also make chickens a valuable part of the system.

[NW_Kohaku]

Yes, I specifically want dwarves to have manure, as it's the best way to balance the game, as I argued a page back:

Spoiler (click to show/hide)

Yeah, but dwarves are creatures, too... Not that it wouldn't be a little bizzare if dwarves just let it go when they had to go like livestock would.  (Although, they already do that with vomit and cave adaptation...)

Besides, the problem with only making animals produce manure is that you then have to use animals to provide the fertilizer - if livestock produce fertilizer and consume food, while dwarves also produce live-stock and consume food, then it's sort of a trade-off to have the livestock.  If only livestock produce fertilizer, and that fertilizer is a major component of your maintaining your fertility, you would be forced to always ranch, and the ratio of animals to dwarves would be a major portion of keeping your fortress fed.  That's not really the sort of consequence I'm looking to cause.

And regarding manures, even though I said the following...

Spoiler (click to show/hide)

If we are going to make fertility management (and fertilizer and water management) the focus, then we need ways of making fertilizers different from each other, and that requires having different stats for them all... (And keep in mind that real-life manure has different qualities, based upon the animal and its diet... rabbit and avian manure are more valuable than cow manure, and free-range grazing cattle have worse manure than dairy cattle that are fed grains... theoretically, we might even go the EXTRA extra mile and stat out the values of every creature's manure.   How would THAT be for a tagline?  "Dwarf Fortress - We optimize our bodily waste!")

That was meant mostly in jest... I really don't expect even Toady would start stating out every single creature's excriment so that we can have some kind of rare animal with miraculous "golden poo" that is treasured for fertilizing any soil beyond the dreams of any farmer.  (Of course, it is Toady we're talking about...)

I do somewhat hope, however, that we have at least a basic nutrition system, following something at least as simple as differentiating "meat", "grain", "vegetables", and "fruit" (not really sure where mushrooms would fall under that catagory...  Google search generally says "vegetable", but I'm sure with the wacky way mushrooms work in DF, there can be some in every food group), and that if we are tracking that data, maybe we can have different kinds of manure based upon what creatures have been eating...

[Draco18s]

(not really sure where mushrooms would fall under that catagory...  Google search generally says "vegetable", but I'm sure with the wacky way mushrooms work in DF, there can be some in every food group)

Biologically they're a fungus.  Culinary-ily speaking they're a vegetable.

[AngleWyrm]

Modelling Manure:

If we accept the seven-degree measure of soil as (Nitrogen, Phosphorus, Potassium, pH, Water, BioMass) and use a simple 5-point rating system as {'very low', 'low', 'medium', 'high', 'very high'} then there will be 5^7 = 78 125 possible soil conditions, which ought to be more than sufficient variety for creating ecosystems.

Manure would be used to fertilize the soil. In order to do that, the manure would have a seven-degree measure of what kind of "soil" it is. Then adding manure to a plot could be as simple as a vector addition of (soil + manure) / 2. This would allow manure to pull the soil's metrics in the direction of the manure. If the manure is 'very high' in nitrogen, then the combination would be closer to 'very high' Nitrogen. On the other hand, if a manure is 'very low' in a given metric (maybe it is very dry), then the soil's watered condition would be pulled toward 'very low' Water.

[NW_Kohaku]

If we accept the seven-degree measure of soil as (Nitrogen, Phosphorus, Potassium, pH, Water, BioMass) and use a simple 5-point rating system as {'very low', 'low', 'medium', 'high', 'very high'} then there will be 5^7 = 78 125 possible soil conditions, which ought to be more than sufficient variety for creating ecosystems.

Manure would be used to fertilize the soil. In order to do that, the manure would have a seven-degree measure of what kind of "soil" it is. Then adding manure to a plot could be as simple as a vector addition of (soil + manure) / 2. This would allow manure to pull the soil's metrics in the direction of the manure. If the manure is 'very high' in nitrogen, then the combination would be closer to 'very high' Nitrogen. On the other hand, if a manure is 'very low' in a given metric (maybe it is very dry), then the soil's watered condition would be pulled toward 'very low.'

That's not really what I'm talking about at all.

You might want to look at this post, and the handful after that one on the same page.

Those numbers are meant to be very fluid, and I think we need something like 256 points of data for each crop to have any meaningful ability to measure this changing over time.

On a totally unrelated note, Opera crashed, and I was too stupid to periodically save my progress in notepad, so I lost my recap thread post about 3 chapters into it...  I'm too depressed about it to redo it now, so I'll probably write it over the weekend.

[AngleWyrm]

Those numbers are meant to be very fluid, and I think we need something like 256 points of data for each crop to have any meaningful ability to measure this changing over time.

Six measurements (NUTRIENT_N, NUTRIENT_P, NUTRIENT_K, SOIL_PH, WATER_REQUIREMENT, BIOMASS) with 256 possible values each leads to 256^6 = 281 474 976 710 656, or 281 trillion possible soil conditions. That's too much, and adds only meaningless changes over time. For instance changing the pH from 86 to 87 probably results in no real change as far as which plants prefer the soil, and which plants do poorly in the soil. It's deceptive resolution, such as measuring the width of the USA in inches, or taking someone's temperature to the nearest hundredth of a degree.

To use a scale with 256 points on it, there would need to be less dimensions being measured. The thing to decide "how big" is the total number of soil types. Then you can decide if you want to reduce the number of dimensions, or the scale of the dimensions, or both in order to achieve an imaginary supercube of minicubes in which to store the many types of plants that will be created.

The number of plants that will populate the universe is probably going to number in the dozens to hundreds. Making a space of thousands to tens of thousands would be sprinkling them thinly. It's a good idea to trim out two to three of those variables, and reduce the range. If we settled on just four variables, then to get into the space of thousands to tens of thousands would be something like 10^4=10000, or four variables with a range of ten in each.

P.S. Sorry for your loss. I've had that kind of thing happen to me before and it totally sucks.

[Draco18s]

Those numbers are meant to be very fluid, and I think we need something like 256 points of data for each crop to have any meaningful ability to measure this changing over time.

Six measurements (NUTRIENT_N, NUTRIENT_P, NUTRIENT_K, SOIL_PH, WATER_REQUIREMENT, BIOMASS) with 256 possible values each leads to 256^6 = 281 474 976 710 656, or 281 trillion possible soil conditions.

256 points per crop != 256 points per measurement.

[AngleWyrm]

256 points per crop != 256 points per measurement.

Could you go into some detail on this point?
I'm not quite getting what "256 points per crop" means.

[NW_Kohaku]

I linked to this post to show what I meant: http://www.bay12forums.com/smf/index.php?topic=22015.msg1445642#msg1445642

It's in the spoilered-out areas.

I'll just go ahead and quote myself, then...

Today, I'm going to start making up some sample crops, a little suggestion of how to build this in psuedo-code to illustrate the point...

(I am spoiling some things because I'm a sesquipedalianly loquacious SOB.)

Spoiler: Caveats and Assumptions (click to show/hide)

First, I'm going by the assumption that the current seasonality of crops will be reversed - underground crops realistically suffer from virtually no distinction of season, while aboveground crops do.

I'm also using the "biomass" stat for carbohydrates, which means that, while aboveground crops have tricky seasons to worry about, underground crops do require some starter biomass, such as simple sawdust or the like. 

I also want to start off using the previously mentioned suggestion of having all farm plots be something like a 3x3 plot because that gives me more room to manuever with crop yields.  Currently, all crops produce between 0 and 4 (very rarely 5) crops per harvest, but this could be made up into a production of 0 and 20 or 30 or even 40, depending on the crop.  This would mean that some crops just plain produce more food than others (but have more finicky requirements or longer growing times).  Of course, this all gets swept aside if we just have Volume and Mass tracking of objects, as we could simply track specific kilogram values of crop harvests.

On that note, Process To Bag and Process To Barrel (quarry bushes and sweet pods) need to stop making five times as much food as you put crops in.  That's just plain rediculous and unfair.  No crop can compete with something that produces five times as much food as it.

Of note, however, is that not all crops have the same nutritional value.  The an arbitrary mass of onions do not feed the same amount of people as the same mass of rice will feed.  Onions are not particularly nutritious crops (and notably, do not consume nearly as many nutrients as rice will), so may need to have its harvests kept artificially lower than they would really yield as a means of kludging the larger number of onions you would need to eat to get the same nutrition of eating more nutritious foods.

Also, not many of these crops actually do anything with soil acidification, as far as I can tell.  I might be wrong about some of these, but generally, soil acidification occurs because of the fertilizers used, especially when adding Nitrogen (especially manure-type fertilizer).  Nitrogen-fixating legume crops also cause soil acidification (drops pH).  Potash, alongside raising K, also raises soil pH.  Certain types of Peat can be used to raise pH in organic gardens, as well.  I'll include this as a crop for the special purpose of being a fertilizer for other farm tiles.

Other properties for these plants should also exist, like having a specific type of pest, such as a fungus or insect, that is a particular menace to that specific crop.  (These, in turn, might actually have some kind of fungal or insect nemesis that dwarves could cultivate or purchase that would combat specific pests.  Growing, for example, a fungus like a cordyceps fungus that specifically targets a specific crop's known pests as a biological insecticide.)

I'll even try to turn this into a sort of mockup raw-code, although I won't put it all into raw form, as that is too much copy-paste drudgery.

Spoiler: Long section on format (click to show/hide)

The format I will use:  (Assume that nutrients are an unsigned byte, with values from 0 to 255)

[NUTRIENT_N: Preferred minimum: Tolerated Minimum: Adjustment]
[NUTRIENT_P: Preferred minimum: Tolerated Minimum: Adjustment]
[NUTRIENT_K: Preferred minimum: Tolerated Minimum: Adjustment]
[SOIL_PH: Preferred minimum: Preferred maximum: Tolerated Minimum: Tolerated Maximum:  Adjustment] 
[WATER_REQUIREMENT: Preferred minimum: Preferred maximum: Tolerated Minimum: Tolerated Maximum: Adjustment]
[BIOMASS: Preferred minimum: Preferred maximum: Tolerated Minimum: Tolerated Maximum: Adjustment]
[FULL_HARVEST: stack size]
[PEST: creature]
[SEED_TYPE: category]

This is assuming there is no overfeeding mechanism where you can burn plants with overabundant Nitrogen.  This could be easily added in, considering how SOIL_PH works, but it's not going in right now. 

It should be fairly obvious, but "Preferred" is the ideal range, and "tolerated" is the minimum-before-it-stops-growing range.  The range between those two is where the warning signals of poor leaf quality and the like show up, as a percentage between the two numbers.

"Adjustment" is either a positive or negative integer value that pushes the value for that nutrient up or down, depending on whether that plant depletes or replinishes that specific nutrient.  Note that Nitrogen depletion (and replinishment) should be MUCH faster than anything else, triple your Phosphorus consumption and ten times your Potassium consumption.  Likewise, fertilizers that replenish Phosphorus and Potassium should be slower-working.  Nitrogen is the most volatile nutrient in the soil, and crop rotation largley focuses on keeping nitrogen levels high, while potassium is a matter of building up a good base, and maintaining it with periodical supplimentation.

EDIT:  Actually, I'll just start using +3, ++, +, 0, -, --, -3, -4 etc. to talk about these things from now on, as it's easier to just talk about these in relative terms rather than try to punch out calculations on an abstract without really knowing how well this would work out when I actually simulate them.

For soil pH, assume 127 is a value of 7.0 pH.  For roundness of number, assume 20 is the difference between 1 pH and another.  Hence, 117 is 6.5 pH.

BIOMASS is a measure of how many usable carbohydrates and other organic compounds that can be broken apart for energy by non-photosynthetic crops, like fungi.  It's actually useful in some parts, as this can not only give you a usable crop, but also break apart cover crops or leftover husks of other crops to provide you with more nutrients of the NPK variety.  I can also use this as a shorthand for "Humus", which is basically the soil bacteria, fungus and the like that help break down dead stuff for the plants.

FULL_HARVEST is the maximum size of the stack of crops you can grow with this type of crop. This is a necessary balancer, as crops that are edible and replinish soil nutrients (like legumes) would be amazingly attractive compared to other, soil-depleting crops (like wheat) if they simply produced the same amount of crops as the ones that are supposed to be more valuable.  (The way that sweet pods and quarry bushes are basically the ultimate crops in the current version because they can be turned into 5 times as many food items as any other crop, and Dwarven Syrup is one of the most valuable crop-derivatives, to boot!)

PEST is just a name for a creature that is a specific form of pest for this kind of plant, which would need to have stats of its own.  Obviously, much more detail than just a name would be needed.

SEED_TYPE is a nod to Capn_Mikey's statement: potatos bloated tubers currently are a plant that apparently goes extinct outside the wild, although humans seem capable of growing them just fine.  There should be alternate ways to obtain seeds, whether by the STANDARD method of just dropping seeds when they are consumed, or by DESTROY CROP, which means you have to turn one of the crops into another seed, so that you have to use the actual wheat grain or potato crop to create the next batch of seeds.  This should probably be done as a farmer's workshop task to turn a stack of crops into their seeds, which will "preserve" them for the later seasons.  There are also ON HARVEST type seeds, for the plants that don't actually have useful crops, like growing clover as a means of letting the land fallow.  Basically, the harvest IS the seed.

I will also use some existing terms, like GROWDUR, so if you are unfamiliar with them, visit the Plant Token wiki page as a guide.

Plus, I'm adding in "grazer" as a EDIBLE type, signifying that this makes for a crop that can feed grazing livestock, once that becomes part of the game.

Yeah, I'm editing in more crap as a header.

Spoiler: Trees, orchard trees, and flowers (click to show/hide)

OK, growing a tree is one of those things that have been on the backburner for quite a while.

Being serious about growing trees essentially requires the Volume and Mass suggestion to be involved, so that trees can produce more than one just one set "log" size of 50 liters of volume, and better tree gardening can produce larger planks of wood for your lumber industry.  That is, however, a different argument.  What this really requires is the code that makes growing a "crop" across a timespan of more than a single year, so that we can have a GROWDUR of roughly 6048 for the "average" tree, as it stands with current tree code.

Orchard trees are a different matter. (Although perhaps they can be chopped down for wood as well, 50 liters is not a terribly huge tree, but their growth would need to be slow enough that it would not make cutting down orchard trees preferable to cutting down forests planted specifically for lumber.)  Orchard trees should probably have two different sets of data - a growing cycle that has high nutrient costs for a long period of time before they are of any value, followed by an annual upkeep data.  Orchard trees should have a set date or period that they bear their fruit, and are harvestable.  The amount of fruit they bear is based upon the care they recieved in the previous year since the last harvesting (hence, the point where they bear fruit is the time when all their data gets reset for the next year, not at an arbitrary "new year" date).  Becuase they have a harvesting date that is a set period, planting dates should be based upon trying to hit this harvesting date, so a tree that bears fruit in fall, and takes exactly 6 seasons to bear fruit should only be allowed to be planted in Spring. 

There should be some sort of function (such as a "pruning" period) at a point during the start of a growing season that would function to give one farmer the "skill rank" input for that following year's harvest.  Soil conditions, meanwhile, just need to be kept within their preferred regions to keep crops growing at their preferred state.  Pests will become a much larger issue for orchards, though, as you aren't using crop rotations to prevent the buildup of specific pests.  As long as you do not plant overmuch orchards (and make the greater concentration of a specific species a greater lure to specific pests), it should still be generally easier to just maintain an orchard than to harvest seasonal crops, although perhaps producing less fruit overall.

"Flowers" are generally "they look pretty" garden plants.  (I put "flowers" in quotes, because I expect dwarves will spend more time growing particularly interesting mushroom formations, possibly even mushroom formations that create small crystal growth, rather than real flowers.  Flowers are hippy elf things.) I would presume they would be used to either raise room value, or to be admired the way that a fine door tends to get admired right now (meaning they have to have value for still being in the ground, treated effectively like furniture).  Rather than having a set number of crops that "bloom", they can instead simply have a quality rating like most products in this game have quality ratings.  Alternately, they can be picked and put into some kind of urn or vase pottery item as decorative furniture.

As a decorative garden, they would have most of the same properties as gardening plants, but they would be more valuable as a crop that stays in bloom and is never picked.  "Perennial flowers" should be treated in much the same way as orchard trees - blooming at a set period of the year, and requiring some upkeep between bloomings. 

Flowers, especially if we are going by the "each plot has a 3x3 space" method of farming plots, might actually need to be sewn with multiple seedlings/spawns at the same time, so that as one wilts, the already-planted next set of "flowers" will be coming into bloom, and a plot can potentially have some "flowers" in bloom at any given point in time of the year.

Spoiler: Implimentation ideas (click to show/hide)

Now I want to talk a little about how this can actually be coded...

I'm talking about having a "maximum harvest" value, but not specifying how, to hit that number. 

I propose that the various factors in a plants growth be multiplied together.  That way, a period where a crop had too little water will have the functional equivalent of halving or even stoping a crop's growth for that specific period. 

This needs to be split up into "checking periods", where the game does the heavy mechanics of checking for proper soil quality levels (and recording them for the crop's overall harvest), and subsequently trying to adjust soil quality levels.  That the adjustment comes after is important, as it means that if soil quality drops because of a check into a non-ideal state, it gives your farmers the period of time between checks to address the problem. 

Checking periods, however, cause a problem.  They can't really go on every single 100 steps that pass, that would be a bit too frequent, especially since, if we have 12 checks every day, and crops are now going to have periods between harvests of as much as a full (336 day) year in the case of orchard trees, we start talking about having 4032 individual checks adding up to the growth of a single crop, which, when you are multiplying 6 different factors together, you start having fairly huge numbers. 

It also causes problems because different soil nutrients are going to fluctuate at wildly different rates.  Water may fluctuate with every given day.  Potassium content of the soil, however, should only notably fluctuate on a year-to-year basis.  Potassium, thanks to the general difficulty it poses to plants to actually reach them, will be depleted at roughly 1/3rd the rate of Phosphorus, and 1/10th the rate of Nitrogen, which is water-soluble, and hence can flow freely through the soil, so plants can reach it, consume it, and deplete it far more readily.  (This is the problem with using 255 as a maximum value - potassium goes down so slowly that it should only experience a single-digit change over the course of an entire crop cycle, while water needs to fluctuate wildly, so you can't break that small number of -4 soil potassium down into anything less than four -1 incidents.) Water is even more volatile a number than that, and realistically requires nearly daily maintainance  (of course, we need less than daily checks on watering if dwarves are to be expected to be able to actually do anything about it - they can easily spend days off simply on sleeping). 

This means we may need to have counters in the GROWDUR that make only every 12th tick on the GROWDUR check nitrogen, while every 120th tick on the GROWDUR checks Potassium, while the values for the Potassium check are carried over in every check for other nutrients between the actual checks on Potassium. 

THIS, however, creates the problem of making either the problem of low nitrogen occur at very small intervals (12 ticks of growdur is one game day, and water should probably be checked and adjusted more frequently than nitrogen if we really want to be serious), or making checks on potassium so extremely infrequent that they are only checked a couple times during the growth of a whole plant, since some crops will mature in roughly 40 days, while this would make potassium checked only once every 10 days.

The ultimate tally for a patch of land, however would go like this: each checking period gets a score, which gets added to a running tally for all the checking periods.  This gets multiplied by the original farmer's preperation and planting skill check.  The game then figures out what the maximum possible score could have been for these checking periods times the maximum score for the farmer's skill check, and let's call this the MAX_SCORE.  You then multiply the score that you got for upkeeping the plants by the Maximum Harvest value, and then divide by the MAX_SCORE value to find the number of crops you actually get.  (You actually wouldn't need to store a MAX_SCORE variable, of course, you could find that value on the spot, but this is for explanation purposes.)

During these checks, there are different values for each soil type.  In the most simple, we can say that if the crop is in the "preferred" range for that soil nutrient, it gives a value of 2, and the "tolerated" range gives a value of 1, while values outside those ranges will give a 0, and prevent all growth.  This could be altered, of course, the values could be 4, 3, and 0, so as to make the penalties for falling into the "tolerated" range less damaging, and we could make more ranges than just the "preferred" and "tolerated" ones, with a halfway-tolerated value, so that growth could be 4, 3, 2, and 0.  Each of these values are multiplied by one another, so that if we are having a 2, 1, and 0 spread, and the six nutrient values I list, each check could have a maximum value of 64 points to be added to the tally.  With a daily check, and a growth period of 65 days, that means 4160 nutrient tally points are possible.

Pests can also be included with this running tally.  Pests would be able to eat away at that tally, literally and figuratively.  As pests eat crops, they can give birth to more of themselves (spreading especially rapidly if the pest is a disease or a fungal infection), and hence eat more every check.  If the pests manage to reduce that tally to or below 0, the crop is completely destroyed.  Pests should also be able to starve if no crop suitable for them to attack exist, however, reducing their numbers down to nearly nothing (probably best if this is done by geometric growth/decay modeling) if you avoid giving them any crops they can attack for an extended period of time. 

Pests should also be attracted by a concentration of their favored crop - the more of a given crop you have in a fortress, the exponentially more they attract pests that will attack them, and the longer that you go planting the same crop, the more chances that pests have to attack.

With ALLLL that out of the way, I can actually start posting some crap:

Since I already had that one out there:

NAME: Kappa Gourds (Cucumbers)

Spoiler (click to show/hide)

notes: These are gourd-type crops (obviously), so they are heavy feeders on nutrients, but aren't particularly good for feeding large numbers of people, so these are more for being quality crops, good for their nutritional value, while other gourds (like melons) are just plain tastier than most other vegetables.  As such, they are higher value crops.  Plus you can use them to bribe Kappa into not eating you!

PREFSTRING: crisp snap when broken

EDIBLE: Vermin, raw, cooked, grazer
TILE: 161
COLOR: 2 0 1

NUTRIENT_N: 180: 110: -40 (Heavy feeder)
NUTRIENT_P: 110: 70: -7 (P is relatively low)
NUTRIENT_K: 200: 130: -8 (K consumption is unusually high)
SOIL_PH: 107: 137: 87: 151: 0
WATER_REQUIREMENT: 160: 205: 130: 225: -140
BIOMASS: 20: 160: 0: 220: +16

FULL_HARVEST: 30
VALUE: 4

PEST: Gourd Beetle (Cucumber Beetle), Downy Mildew

BIOME: aboveground, temperate or warmer
SEASON: Spring, Summer
GROWDUR: 780

SEED TYPE: consumes crop

DRINK: Cucumber Ice Pepsi Kappa's Beer  (what can I say, I'm a Touhou fan)  Value: 2
PROCESS TO BARREL: pickle Value: 20

NAME: Cross Brassica (Broccoli)

Spoiler (click to show/hide)

notes:  These are relatively fast growing, and can grow in cold seasons (although I still believe a winter freeze should automatically kill all exposed crops) that normally do not offer many crops, so they aren't as valuable/productive as some other crops.  They are moderate nutrient consumers, though, so can work well in crop rotations.

PREFSTRING: miniature-tree-like appearance

EDIBLE: Vermin, Raw, Cooked, Grazer
TILE: 5
COLOR: 2 0 0

NUTRIENT_N: 160: 120: -28
NUTRIENT_P: 140: 105: -12
NUTRIENT_K: 140: 105: -4
SOIL_PH: 99: 127: 79: 147: 0
WATER_REQUIREMENT: 130: 180: 100: 215: -90
BIOMASS: 100: 190: 60: 240: 0

FULL_HARVEST:30
VALUE: 2

PEST: Cross Maggots

BIOME: Non-freezing, non-scorching, doesn't like hot, either.
SEASON:  Fall, Winter
GROWDUR:  528  (Would take about 20 days/240 growdur longer if you planted in Summer or Spring, but this is for a fall/winter-only crop)

SEED TYPE: Consumes crop

NAME: Rust Grass (ryegrass)

Spoiler (click to show/hide)

notes: This is a "rock breaker" crop, I guess you could say.  It's purpose is largely to add more potassium into the soil by breaking apart clay or even stone to access its Potassium.  For flavor, I decided to make it a bright rusty-colored grass along its seed pod.   Technically, the ryegrass I modeled this on can also be considered a weed!  But it's also used for lawn grass, and for soil cover, especially since it grows well in Winter.

PREFSTRING: rusty red tips

EDIBLE: vermin, grazer
TILE: ","
COLOR: 4 0 1

NUTRIENT_N: 120: 70: -8 (light, especially because it is meant to be plowed back in)
NUTRIENT_P: 40: 20: -3
NUTRIENT_K: 60: 20: +9
SOIL_PH: 103: 221: 83: 241: 0
WATER_REQUIREMENT: 100: 180: 70: 230: -100
BIOMASS: 10: 160: 0: 210: +24

FULL_HARVEST: 8 (seeds)
VALUE: 1

PEST: none

BIOME: Any nonfreezing, it's a grass!
SEASON: all
GROWDUR: 336

SEED TYPE: ON HARVEST

NAME: Gracile Goober (peanuts!)

Spoiler (click to show/hide)

notes: A nitrogen-fixation crop that is also a useful and tasty crop, instead of mere green manure!  This should be a part of any serious crop rotation system meant to avoid excessive use of Nitrogen fertilizers, although some nitrogen fixation crops that ARE just green manure would be more effective.

PREFSTRING: delicate yellow flower; delicious paste

EDIBLE: vermin, raw, cooked
TILE: 232
COLOR: 6 0 1

NUTRIENT_N: 40: 10: +90  (The money shot.)
NUTRIENT_P: 120: 90: -8
NUTRIENT_K: 160: 100: -11 (surprisingly high, but apparently, they do)
SOIL_PH: 107: 119: 77: 135: -4
WATER_REQUIREMENT: 170: 230: 130: 250: -400 (remember, it has a long growing season, and needs plenty of watering during this time.  Supposed to take 20-40 inches of rainfall, or equivalent to grow.)
BIOMASS: 130: 220: 80: 255: -30 (benefits from high humus and decaying biomatter concentrations)

FULL_HARVEST: 30 (low for its growing season, but it has other values)
VALUE: 1

PEST: rootworm

BIOME: warm or warmer aboveground, needs plenty of rain
SEASON:Spring, Summer, Fall
GROWDUR: 1500 (125 days, out of a 28-day month, means nearly 5 months (or the better part of two whole seasons) to grow)

SEED TYPE: CONSUMES CROP

DRINK: Earthen Spirits (George Washington Carver apparently made several kinds of "peanut punch", and even peanut ice cream, but "goober punch" doesn't quite sound right)- Value: 1

PROCESS TO BARREL: Goober paste (peanut butter) - Value: 15

MILL: Goober flour - Value 10

In a nod to George Washington Carver, though, peanuts should be capable of making a rediculous slew of products, including: Soap, Rubbing Alcohol, Quinine, Laxatives, Medicines for Venereal Diseases, Glycerine, various forms of makeups, ointments, shampoos, dyes, paper, insecticides, glue, plastic, polish, diesel fuel, coke, and rubber.  At the very least, dyes are practical for DF.

EXTRACT: Goober Oil (value 5, also usable in prepared meals, making this oil will also produce a minor fertilizer as a side-product): MILL: goober dye base (should be a base in the production of dyes) - Value: 15

SOAPMAKER: turn goober oil + lye into vegetable oil soap.

I'm going to cut this off here, because this post is already pretty darn long, and because I've already spent the better part of the day researching crap, and I need to go to sleep.

You realize what I'm talking about, here, correct?  Why I've talked about changing soil over time, especially nutrients like K, which change only very slightly over time?  The number of iterations of possible values is completely irrelevant.  And maybe you don't care how many inches wide the United States may be, but if you are trying to do an accurate computer model of the US, such as something like Google Maps, then you have to use some fairly precise measurements, at the very least.

Look at how Hunger is currently handled: the hunger value is basically just an integer that gets iterated over time, reaching the hundreds of thousands before dwarves eat to reset that value. 

I'm not talking about having a different label for each and every individual iteration of soil, that would just be absurd, and having only 5 levels of actual fertility is needlessly limiting, and would break the system.  I'm talking about having a hidden fertility variable for every dimension of soil fertility.  Frankly, I'd like the numbers to be a little larger, but I'll just work with having the potassium number only checked and reduced every 120 growdur or some such.