Bay 12 Games Forum

Does anyone here want to tell stories about bugs and unintended features that they actually really enjoyed in previous versions of DF that have since been fixed?

I loved the bug that made towers get into disputes with villages, resulting in towers conquering villages, acquiring way too many corpses and resulting in an orgy of tower building. And not only because it meant that there were a *lot* of necromancers out there; it made the political geography of my worlds occasionally become amazing and provided me with more immersion than I've really gotten out of the game since.

This one time, the only area that discovered necromancy was this long, isolated peninsula jutting down from the top of the map. It was home to two human governments, who I imagined as loose alliances of bickering merchant princes due to both their names having something to do with trade. These states tended to have their major cities on the coast and build their tombs inland, which was a rural area chock full of farming villages. The two civs were roughly divided between the southwest and northeast, but in the inland their villages were chaotically dispersed, meaning that nobody had clear control over the region and that disputes were common.

When the necromancers came, the first necromancer-cult began to take over towns in the inland. When this prompted no response from the bickering lords, their conquests accelerated. Soon, they had a tower built. However, since they would constantly get into disputes with nearby towns, including towns they themselves had conquered, the situation rapidly escalated. Attacking their own occupation governments would cause groups of necromancers to schism, after which they'd go conquer their own towns from someone else. By the end of the century the entirety of the inland had been conquered by necromancers. Towers dotted the landscape, occasionally fighting rebellious underlings and stealing villages from one another, resulting in a series of small kingdoms with shifting borders and frequent internecine conflict. The population of the area had plummeted, unless you count the undead, while the vast numbers of dead law-givers who had found themselves in the wrong places in the wrong time meant that tomb complexes were unusually expansive.

You can imagine the desolation and the despair. To the last free humans, huddled in their tightly-packed coastal cities struggling to cope with the influx of refugees, their breadbasket was gone. Though it's not in the game yet, I imagine that to them it was a time of famine, plague and anarchy in the streets. Those unfortunate enough to have fallen under the necromancers gaze have all wound up toiling for their unfathomably dark designs or sacrificed in their seemingly pointless wars against one another, and not even death is an escape from this slavery. They cower in villages with far too many empty buildings, in the shadow of the towers and the tombs, under the watchful gaze of ambitious apprentice necromancers.

So naturally, I set out to liberate what I'd come to term the Dead Marches by building a band of warriors, slaughtering all the necromancers and taking over all the sites. Since DF handled loyalty funny though, before I conquered my fourth I was tracked down by an army of living humans from the surviving free cities and killed. I don't know the implications of that, though it's probably because somehow the administrators I'd killed were still civ members or the towns I'd declared sovereignty over were still somehow considered theirs even though they'd been taken over by necromancers. Anyway, I'll miss that bug. That was the coolest world I've genned by far.

Anyone else have experiences like this?

The idea I'm going to be putting forward in this thread is rather complicated, so please bear with me. Essentially, at present there are 4 tiers of societal organization planned for DF; Entities, Groups, Subgroups and Families. I think that for Toady to achieve everything he's setting out to (as defined in Threetoe's stories), a lot more social groups are going to be needed to be tracked than just four. Splitting up these tiers of social organization will make the game feel more varied and compelling. It'd likely be better to start tracking these groups in a rudimentary form (even if their specific mechanics are unfinished) before we progress into the Army Arc and its convoluted leader-based diplomacy.

The new tiers of social organization I propose are below.

---

Tier 1: International Groups

The groups in T1 incorporate a lot of information that used to be part of entities. It's split apart so that the groups it defines can be treated differently based on whether their group is favored by whoever runs the sites they inhabit. Individuals always belong to 1 of each category of Tier 1 groups, and Tier 1 groups are created at the start of worldgen. Tier 1 groups are:

Cultures: Groups which share art forms, languages, appearance styles (i.e. clothing, tattoos, etc) and value preferences (more or less so depending on personality). Critters will move from one culture to another by marrying into it, allowing for the culture of an area to change over time. Necessary so foreigners are treated as foreigners rather than just strangers.

Religions: Groups which share objects of worship and value preferences (more or less so depending on personality). Can be linked together (i.e. different sects of the same religion). Will come with their own taboos and traditions eventually. Individuals will flip to another religion practiced in their site if the values it prefers are more aligned with their personal values that their current religion. Necessary so that religions can spread between societies and make things less homogeneous.

Ethnicity: Groups who share physical similarities. They're necessary so that the player can observe population movement in a meaningful way if the people who moved/were subsumed into another culture have all already assimilated.

Tier 2: States

The groups in Tier 2 incorporate the rest of the information that used to be part of entities and determines the formation of governments within the game (so that multiple governments with the same culture can exist if things get more randomized). They've also been split into various types to allow more varied societies to form, and altered so that new governments can emerge or be destroyed during play. Each of the following categories of state determines what kind of sites they have and what kind of relationship they'll have with subordinate or neighboring groups. All civilized criters are a member of one Tier 2 group of any type. Tier 2 groups are:

Territorial States: States that own land and build permanent sites within it. Have provinces, which are groupings of sites in proximity subordinated to a sub-capital, that can try to break away and become independent.

City States: Superficially similar to territorial states, but actually quite different. They are lead by one of their subordinate groups exclusively (the inhabitants of their capital city) and treat other population centers they control as tributary colonies - and these frequently try to become independent city-states of their own. They are more likely to try to exact tribute form fellow city states than they are to conquer them.

Nomadic States: States that own land but have no permanent inhabited sites (except maybe religious sites). Their population instead moves around in camps on a randomly defined circuit. Much like city states they're ruled by a single group, and if leadership would pass from a member of that group to a member of another (most likely because that group has been depopulated) each subordinate group becomes and independent group upon succession.

Tier 3: Institutions

The groups in Tier 3 define associations of individuals across various sites of a Tier 2 Group. It's purpose is to allow groups to control or be present in multiple sites at once while maintaining a single hierarchy, and so that smaller groups can work together. Examples that I've thought of for Tier 3 groups are as follows, though more can probably be added. Note that not all Tier 3 Groups necessarily need to be present for a Tier 2 group to continue to function (they should be enabled or disabled in a tier 2 raw file). Individuals can be a part of multiple Tier 3 groups, but not of the same type.

Kinship Groups: Groups defined by relationship among members. Necessary to allow for tribes/clans larger than immediate families to have an effect on politics.

Local Religious Hierarchy: A group defined by shared religion. Necessary for defining the relationship between temporal and spiritual authorities.

Local Business Organizations: Groups defined by shared profits among members. Necessary for trade cartels, business ventures that create sites (such as a bunch of carpenters in a dozen sites founding the creation of a lumber camp).

Local Criminal Organizations: Groups defined by shared criminal intent. Necessary so criminals can have warring subgroups or be spread across multiple sites.

Provinces: Only found in Territorial States. A single group subordinates other appropriate groups whose sites are in its immediate proximity. Necessary for multiple tiers of land-owning nobility and so that larger areas than a single site can revolt at once.

Tier 4: International Organizations

The groups in Ti4 are T3 Groups that are not contained to a single T2 Group to operate. They include;

International Religious Hierarchy: I'm still unclear as to how these should form during play, but these are religious groups that replace the T3 LRH while acting across multiple T2 groups. Necessary so you can have the Pope, missionaries and meddling foreign clergy.

International Business Organizations: Associations of merchants across multiple T2 groups. Necessary so you can have trade routes and giant trading companies. Can be subordinate to T2 groups.

Rebel Groups: Associations of exiles from a specific T2 group living in other T2 groups who want to overthrow their original government and either take control or hand it off to someone else. Needed so Bay of Pigs type scenarios can happen.

International Criminal Association: Groups of criminals. Needed so that there can be a sinister international guild of assassins and whatnot.

Tier 5: Local Communities

Essentially Groups as currently in the game, but they can be part of T3 groups.

Tier 6: Local Subgroups

Subgroups as currently planned by Toady, with the exception that they can be part of a T3 or T4 group in addition to a T5 group.

Tier 7: Families

Families as planned by Toady, except that they can interact with T3 and T4 groups.

Tier 8: Imperial States

T8 Groups are essentially T2 Groups that have subordinated other T2 groups. I'm unsure at the moment as to when they should be allowed to form, but they're necessary so you can have a society with Nomands and City States at the same time, or so that you can have entire nations rebel against an evil empire ate once.

Tier 9: Alliances

T9 Groups are T8 Groups, except they're relationships between equals. Necessary so you can have the Last Alliance between Olm Men and Dwarves fight a legendary battle against Mankind.

---

Thoughts? Are there any other categories of group to track to could enable more stuff? Am I full of bunk for suggesting Toady break his nice little 4-tiered system into a convoluted 9 tier, 20+ unique group type system mess even if it would enable more varied societies and stories?

Seeing as the Tavern Update is approaching, I think it's a good idea to re-examine how ownership works in DF.

I was disappointed by the old economy, because it could have added so much more social complexity to the fort. The fortress ultimately owning everything but the shops (and then still possessing unlimited rights to bulldoze one) and being the employer of everyone in the fort made the economy feel less like a meaningful change to how you'd play and more like an arbitrary infliction of poverty on some of the forts dwarves.

What I'd like to see are as follows;

1. Every room (zones, workshops, farm plots, whatever) has an owner (by default, the fortress). In the future, this will allow us to sell or rent these rooms and their capacity to produce/provide services to the fortress dwarves or some internal entity like a guild. If, in the future, a room is owned by an NPC the player can't just order it deconstructed without buying it back.

2. Relationships between dwarves in some way represented in-game. Just track who is subordinated to who for now - later the economic relationships between these individuals can be defined. But right now, your Chief Medical Dwarf is just a guy who the voice of god has told to diagnose people. There is no in-game relationship between them and whoever appointed them, or their staff.

However the future economy works out, I think these two things are necessary prerequisites. If you want dwarves to be able to be land owners instead of just tenants, to employ one another or to have really any service economy instead of just a material economy, these frameworks will be crucial in enabling this.

Brace yourself for a text wall.

---

Why do I want to redo skills?

Firstly, I feel like there hasn't been much meaningful change to how dwarven skill words since the first release other than adding new skills or tweaking how fast they accumulate experience. The rather gamey system of 'do task, receive XP, level up' that is the core of the system has persisted basically unchanged.

This system has its problems. Firstly, growth is exponential - since task completion speed increases with level, dwarves start to acquire levels faster over time rather than slower. Secondly, all dwarves have the exact same potential to learn any skill. Dwarves who have never seen a tree in their lives, taking no instruction from any skilled individual, can immediately pick up the task of cutting one down and get better at it with. Your fort need never want for personnel, because there's literally nothing stopping you from training a dwarf to perform the desired role. Some may think this a good thing, but to me this is unrealistic and makes the game easier as time goes on - and seeing as boredom is the second highest killer of forts (behind the FPS) for experienced players, the game definitely doesn't need anything making it easier as you go along. Thirdly, generic EXP points in a skill is very gamey. Urist McMason has done nothing but carve doors all his life, but suddenly you ask him for a millstone and he carves one of equivalent quality of the doors he's been producing, despite having never seen a millstone or attempted one before, simply because they share EXP.

---

So, seeing as I'm dissatisfied with the current system, here's a new system.

Dwarves ability to perform a job with speed and quality shall be determined by multiple stats instead of a single skill.

Firstly, dwarves will have talents. Talents will affect their speed and skill (item quality output for production jobs) at any job, though different jobs will be affected by different talents in different ways. Exemplar talents are dexterity (ability to do fine detail work) Aesthetic sense (ability to produce something others will consider beautiful) and learning (ability to learn new skills). Basically, all current physical stats and social skills fall into this category.

Talents will change over time and can change in different ways depending on the talent. For example, learning should be increased through education but should decline with age. This makes new 'training meta' possible.

Secondly, dwarves would have knowledge. Knowledge also affects how well a dwarf can perform a task, but it isn't inborn to them like talents are. Knowledge will be an exponential increase like the old xp system was, but ideally would be a much more gradual curve. Knowledge is not the same thing as experience, though, seeing as you'll need to train talents through repetitive tasks as well.  The key idea here is to make knowledge the most limiting factor early in your dwarf's 'career' and the least limiting factor later.

Finally, your dwarves will have theory. This is intended for skills that don't actually affect how your dwarf does their job, but determine which type of output that dwarf will produce. Right now this is basically limited to knowing a specific artistic style for dancers, musicians and poets. It could however be expanded to various philosophies, theocratic knowledge (can recount the legend of Urist McProphet's Martyrdom in Magma at the hands of Asog McGoblinNoble) or negotiation skills (knows noble etiquette of the Procedural Names civilization, can better negotiate with nobles of that civ. Alternatively, is fluent in the local dialect of The Region of ProceduralNames, and is thus more likely to befriend people from that region).

Or, as I dream of (but don't actually expect), we could make a system where simply knowing a lot about masonry in general doesn't mean you can immediately make a beautiful, artistic stone quern despite having never made or seen a quern before. If we were to do that, you'd need to be taught what a quern is before you can attempt it.

---

So what does this system actually mean?

Firstly, it's meant to limit your options. Not every dwarf will have equal potential for all possible professions. If learning ability also decreases over time, your ability to have a dwarf have multiple professions will also sharply decrease as they won't learn as efficiently once they're older. Finally, lack of knowledge and ability to acquire it will mean that some fields of work will be less efficient to train anyway, meaning each fort will acquire a specific character as it becomes overly dependant on trading or alternative industries until you get that knowledgeable migrant with high teaching ability. Even after that, you'll have a sort of 'tech tree' unless you get a migrant who's done a specific subset of that industry before (i.e. you'll be able to make swords very well with your legendary weaponsmith, but since he's never seen how an axe is produced he won't make very good ones until a metalsmith who has comes along and gives him a little lesson).

Secondly, it'll necessitate some form of education system, which all societies (trust me, I'm an anthropology student) possess in some manner; it's not just schools like we have in the west, but also apprenticeships, learning from elders, tutors etc. There's already been a lot of topics on this forum on establishing various forms of education systems, but all of these presume the current EXP system remains in place, so they're not really compatible with this idea. I'd love to go more into depth here, but it's probably best to just throw in a system where a dwarf will just ask any dwarf with a theory associated with the knowledge group they're learned in to give them the theory when that dwarf is otherwise idle, and that dwarf will automatically give it to them. This'd serve as a placeholder until a better system can be devised (although one could be implemented at the same time as the new skill system in a much bigger update). I'm mostly motivated to not envision a system here because this post is already super goddamn long, to be honest, but how skills work and an education system are impossible to separate; when it's made necessary to have an education system, it means the nature of the skill has changed, and an education system is designed around the nature of the skill.

Finally; this would make the game considerably harder. Your fort would still get easier as you build it, the game would get easier in general slower, and wouldn't even have a guaranteed rate of getting easier as you wouldn't always get that desired migrant at the right time. This would unfortunately make it harder for new players to learn the game, but it would mean that players who do won't just make their perfect fort by year 5 and then run out of things to do that aren't pointless megaconstruction; they'll have long term goals that will actually somewhat depend on the RNG to make possible to attempt.

I think these 3 effects are largely desirable. The game will enter that boring late game lack of challenge later, it'll necessitate a new layer of complexity in how you run your society as you'll need some form of education system, and each fort would gain a sense of unique character alongside acquired realism.
That being said, obviously the system I've proposed might function better if its altered in some way, which I'll leave to y'all to suggest and Toady to test if he thinks this is vaguely worth it.

So I've seen quite a few threads proposing ways to make feeding a large fort more challenging. Most of these consist of making farming harder, more complicated or just increasing the rate dwarves eat. However, I think there's a way to increase the challenge of sustaining a fort without making the learning curve harder for new players; food preservation!

The idea behind this is that in DF, once food has been gathered (and sometimes cooked) it functionally never rots in the stockpile. This isn't to say it never rots, just that it rots so slowly that while you're playing the game it never actually is noticeable. Also, prepared meals stay fresh forever, mysteriously.
I propose raising the rate the food rots in a stockpile. Food, including prepared meals, should only last around 1.5 seasons (to allow for mods where there's a winter season without crops) before it starts to rot and is reclassified refuse. It should also rot faster on warmer maps. To get around this, players can craft a preservative; salt (Inspired by this thread). Salt can be ground at a mill or quern from rock salt, imported in bags from caravans, or gathered from the sea using a tool in a new gathering zone. Jobs are automatically queued up in the kitchen to salt edible, non-prepared meal items. Salted foods will rot slower, and meals that contain them will also rot proportionately slower (plus have higher value - however, this actually makes sense; a caravan would pay more for food that would stay edible on their journey). Food bought from caravans should be pre-salted.
Secondly, vermin should become more of an actual threat to your fortress. It's calculated that today, about 1/10th of human produce is consumed by mice or other animals before it reaches our plates. This was probably worse in ancient times, before pesticides and systematic rat poisoning. Essentially, the game should keep track of how many edible items there are on the map (it already does this so that you can have stocks). As your stocks get fuller, the rate vermin naturally migrate onto the map should increase proportionately to model them breeding more rapidly thanks to greater available food. This causes players to have to redesign their forts; food and seed stockpiles now need to be surrounded by animal traps if you want to remain efficient, and a sizable cat population is much more important than it currently is.

Now, this makes things considerably harder, but only adds challenge as your fort grows larger; early-game forts should be fine by virtue of not attracting large swarms of vermin. It also adds a new, optional industry (salt), adds incentive to trade, and allows for players to combat the added difficulty through design strategies. Based on this, I think it's a reasonable addition to the game; it adds difficulty and complexity, but makes the learning curve less comparatively steep (the game gets harder as you go on as opposed to easier) and is something players will quickly learn to adapt to.

So, I've got a fort on it's 11th year (the FPS, it buuuuurns). I've been trying to get a baron for the past 5 years, without success.

I noticed when I looked at the civ screen around year 2 that we didn't seem to have anyone filling the position of monarch. The title was just absent. Since then, the title has neither been filled nor shown up. I'm starting to think that I don't have a monarch.

Bug?

So I'm finally going to suggest the game arbitrarily inflict serious wounds and even death on your dwarves unavoidably while you play.
No, seriously.

Accidents happen IRL. How many significant historical figures have died in trivial fashions? You can trip and be unfortunate enough to break a bone or even your skull from a standing position.
Essentially I suggest that occasionally (weighted towards happening around once/twice a year) a dwarf will "trip" and receive a random injury. The vast majority are just shallow cuts/bruises, but rarely a bone will be broken, or even a skull.

The benefit of this to the player is that it gives a constant source of practice for the player's doctors. It also provides a good reason to train up multiple personnel instead of over-relying on a few specialists as a dwarf can rarely die of random mishap. Plus, it's a morale challenge: Dwarves won't be happy about being injured, and the rare (should happen only around once every few years for a large fort) death can strain morale as well.