Is there a chance that we will see some changes to town and building generation sometime in the next update/s? The world of DF is truly massive, but towns, fortresses etc. completely lack any flavour - taverns, for example, are just some simple furniture, sometimes a bartender, some drunks, they never feel realistic or alive, and they're always too similar. This makes the adv mode rather unappealing.
Have you considered tile generation like the one used in Cataclysm: DDA? That can create simple variations (it has several thousand basic types and subtypes of buildings with different combinations - like ground floor A, combined with first floor B and basement C, with loot generated from appropriate sets, like house, shop, restaurant, petrol station - so it's always realistic). It combines simple house layouts written in simple code, to create cities that feel very realistic - kitchens always have appropriate cooking equipment, food, drinks, bedrooms have clothes and sheets, bathrooms are very random - some have medicine, some have drugs, they often have basic things like soap or razors. It's all rather simple but the effect is unlike any other game.
I think that DF would instantly become a very different and a much more advanced game with something like this.
EDIT: I'm not a programmer, but I can make several realistic-looking locations for Cataclysm in a single day. I think that something like that, perhaps using JSON like Cataclysm, might make things easier and faster for Toady.
PatrikLundell:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8148322#msg8148322Doorkeeper:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8148694#msg8148694Ekaton:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8148749#msg8148749FAA (op):
http://www.bay12forums.com/smf/index.php?topic=169696.msg8148756#msg8148756Putnam:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8148794#msg8148794I don't have anything additional to add to
my reply from when you asked about CDDA some months back.In fortress mode, will it ever be possible for Lovers to become pregnant instead of having to get married? So can your dwarves get divorced or cheat on each other in fort mode? I know dwarves can take multiple lovers but that might be a bug or something
Egan_BW:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8148435#msg8148435Yeah, as EGAN_BW says, some of those things can theoretically happen now. Divorce should be possible, and multiple lovers (probably) isn't a bug. But looks like I didn't get the world gen pregnancy change over to fort mode pregnancies, hmm. Have made a note.
Has working on the Steam UI caused you to work on or otherwise revisit aspects of the game that you hadn't planned on getting to at this stage of development?
Shonai_Dweller:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8149866#msg8149866PatrikLundell:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8149971#msg8149971Yeah, depending on what you mean by UI, we haven't really gotten to the big chunks (like VPL or whatever, though that would be in the 'planned' category), but even doing the graphical work I found I had to repair some broken issues with heavy branches and consider grass growth patterns and brook flow code. Stuff that'll make the ASCII version better as well (heavy branch growth in particular was totally broken.) I expect that'll just keep happening.
Any new insights on the new graphics architecture (e.g. separating the graphics from the saves)? Are you implementing it in parallel to the general graphics-rewrite, or has that part of the project yet to be started?
Haven't done it yet, since it feels more related to the restructuring of the mods. For now everything just lives in the raw/graphics folder as before, but that'll change once we figure mods vs. saves out.
Toady, have you given anything thought at all about knucking down and working on fixing bugs and improving the games stability?
In my experiance, the game crashes on a regular basis. During ragular playtime, I can expect the game to crash 3 or 4 times, if it wasn't for DF hack quicksave, and LNP seasonal saves, i would have lost my progress every single time.
Please, stop whatever features you're working on now and focus on improving the games stability, you don't have to change the game itself, just make it more stable. Please.
PatrikLundell:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8150103#msg8150103Shonai_Dweller:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8150156#msg8150156Sarmation123:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8150218#msg8150218I don't have much to add to the replies. I'm working on the stuff I have to work on. We're hoping to do at least one stability release before the Steam version even comes out, but I've been having difficulty scheduling that because of the ongoing art work, which must move forward, plus 2020's multiple real-world disasters which have also constrained work somewhat.
Are there any plans to expand the 'd'etail function to some of the older hard coded reactions like prepare meal, mill plant, mill seeds, or dye cloth? It would go a long way in terms of user experience and would save a lot of linked stock pile hassle.
We didn't specifically have plans for this, though it's a reasonable idea.
Toady One, I see you talking about a lot of work with doing identifiers-- I wonder how systematic it is that you're doing the actual job. I expect the job to be rather systematically compling a list of relevant tiles and then linking up identifiers to tile maps, possibly with identifier text files for the images. Can you speak a little about what sort of job you're doing that requires so much work with what I (probably mistakenly) imagine to be manual identifiers?
Yeah, since we're going to allow graphics to be modded as before, everything needs a name out in the text files. There are over a thousand now and growing (considering how many there are just for the brook animations and the trees and ramps, it didn't take long for the numbers to climb -- a lot of the identifiers accept a number or two, as with other numeric raw text fields.) Each of those needs to be linked into the code, and this process isn't just an "add identifiers" process, but a "make the graphics work" process, which requires different systems etc. depending on the stuff (ie layering shadows on dwarf feet vs. advancing multilayer brook animations.) It's not a simple conversion, but we should have most of the graphics we need on the other side.
Tody One, I'm a modder. I use mostly real-world values in my mods for weapons, creatures sizes, etc. It would be quite helpful if we could get more info on the units used in DF, along with features that allow for us to see it, much like how DFHack does. I know that SIZE, BODYSIZE, and contact area and penetration depth in ATTACK is cm3, temp, and weight in Urists (1 U degree is equal to 102 degrees Fahrenheit and 1 U is equal 1 Kg)[1], but I don't know about SHOOT_FORCE, SHOOT_MAXVEL, and *_STRAIN_AT_YIELD. It would be really nice to see if these have some kind of unit or other equivalents.
PatrickLundell:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8154569#msg8154569The strain/elasticity is 1000 * the percentage change in shape when the yield strength is reached. The force/velocity variables for those projectiles all predate the physics rewrite for minecarts, so they don't have a standard unit in mind. It divides the force by the projectile mass to get a velocity, and caps it with maxvel. This velocity is transferred over to the melee attack system when it comes up with the attack momentum (though there's a mess of other stuff that happens at that point.) It's not based on any SI unit currently as far as I remember, and the speed the arrow flies across the map does not use the velocity (they fly at the maximum frame speed, but don't skip squares.)
With the graphical improvements coming being applied to the current version (or something like it), will the map updates coming with the developments that precursor the myth and magic release require new artwork and/or render old artwork obsolete?
Now that we can actually see which way slopes point, can we order dwarves to carve out specifically oriented ramps?
Does it look cool when fluids dribble down slopes or does it sort of ignore them?
Shonai_Dweller:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8155002#msg8155002PatrikLundell:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8155012#msg8155012Buttery_Mess (op):
http://www.bay12forums.com/smf/index.php?topic=169696.msg8155103#msg8155103Eric Blank:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8155170#msg8155170Oh, yeah, regarding your clarification about geology etc., it's definite that we'll need more images as we go. What we're doing now with the artists is really just part of the new normal of DF on into the future, and how that works when we switch back to a "new features" mode remains to be seen!
I haven't changed the ramp jobs at all, though yeah, theoretically, now that they aren't all ASCII triangles, there'd be some cause to do that in the future.
We still just have test images of water at the 1-6 out of 7 levels, and there are various issues on the table there. Having it partially obscure creatures vs. how it dribbles down ramps are one example where they don't play nice and would need extra work, as I recall.
A few reasonably near time technical questions that may not have been decided on yet:
As far as I understand, the CP437 character set is going to be replaced with a wider set.
1A. Is the set going to be the same for actual text as for character graphics?
1B. Is the set going to be fixed size (e.g. 16 bit), or variable size (e.g. Unicode)?
1C. If the set(s) has been decided on, can you tell us what it is?
I have gotten the impression that the current set of 15 colors is going to be replaced by a wider set(s).
2A. I've seen indications that named colors will be usable, would that be in addition to RGB or instead of RGB?
2B. If both versions are supported, will you be able to use one or the other at will, or will each color usage situation have a fixed definition (or some being fixed and some variable)?
2C. Assuming named colors can be used, would they be fixed or defined in a RAW file where the users can add additional ones for use in RAWs?
We haven't decided if we're going to use a wider set -- we're juuust starting to experiment with our first text-bearing buttons, and we'll see what comes out of that. The graphical tiles and the interface tiles are completely decoupled now, so the only relationships between their sizes are just based on what makes a nice button with an animal/item picture next to it, etc. I'm just using the old 8x12 font now. Dunno if we'll end up somewhere else, either in terms of fixed size or fixed width or anything else, though we've had trouble supporting ttf in the past and I'm not sure what issues there are implementing unicode. It's not as crucial as some of the absolutely horrible stuff about the UI, as long as it is legible.
The wider color ranges are used in recoloring the graphical tile set (weapon materials, gems, etc.), and those colors are already defined in a raw file (descriptor_color_standard.txt), but I haven't done anything with text yet. DF text has always been printed using 24 bit color, and it just artificially narrowed the range. So it would just take a few minutes to get it to print in any RGB color, as a technical matter, but it's a matter of where that is used. Certainly there are lots of places where it would make sense.
Having done nothing with it makes it difficult to comment on usage. The descriptor raw file is more for in-game stuff though, so for some UI text unrelated to game objects, something more along the lines of data/init/colors.txt would be appropriate. But if the text is, say, of an in-game material and we want it to be the same color as that material (possibly brightened if it is too dark), then that would still use a raw color definition.
I may be missing the case you are concerned about.
Time for my newest round of stupid, far off, likely already asked, questions for you!
1:When you change world gen for the myth arc, what specific geographic features are you hoping for (either in general, or in particular)
2: With said new world gen, do you have any thoughts on how will biomes work? If you are planning a change on that front, what?
3: How long until you have working evolution (joke question, but still)
Bonus question: 1&2, but with alternate planes.
Shonai_Dweller:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8157663#msg8157663The first step with the map rewrite is to create a structure which will support what we want to do, whether that's with planes or real world land/waterforms, etc. In no particular order, one list we have is different distant/time scales, a general framework for world spanning features like rivers that will allow world trees or ley lines or tortoises or whatever), world edges, connectors like rainbow bridges/multiplanar rivers, multiple view ports, boats and moving map segments/shifting mazes, shadow/veil concepts, recursive/fractal/'levels of reality'/micro-macrocosm concepts, non-rectangular world shapes (giant pillar, tubes, giant pit, etc., though we likely won't attempt traditional spheres, as wrapping abilities are limited by the grid), better geological layers/intrusions/cliffs/eroded features, support for linkages between rivers/caves entrances/etc., support for small sites (logging camps, crossroad inns, etc.), support for nomadic groups sites (the current tents the armies use are horrible), ability to support world-spanning settlements, portals (various), faerie-type border zones, infinite worlds (not that there's memory for that, but the ability to reshift focus), liquid/air/etc. based worlds, auras/fields/mists/wtvr and restructure good/evil-type sphere links, compatability with astronomical or larger universe concepts, support for regional/world natural disasters and resulting map changes (and map repair like forest regrowth), places in the clouds, underground oceans and other improved underground linkages/structure/features, consider support for broader liquid/material types in terrain, large vortices/whirlwinds/etc., z-level buildings/doors/etc., merging/mixing planes, teleporting sites, illusions of various kinds (non-tactile, tactile but impermanent in some way.)
No way we're getting to all or even most of that on the first pass! But we're hoping to create a backbone that support it all. I also have a few other lists around that are mostly in that one up there, but I doubt that's everything.
Biomes relate to a lot of that, how they are contained or how they can change and bleed together, and what sort of ecological simming we can get away with. Ultimately we'll still need to have rainfall, temperature, drainage, present species, and that sort of thing, and the existing classifications aren't bad there.
Two questions:
1) Is it likely that the game will move to an open development with volunteers working on particular parts of the game, like how it currently has external artists for the Steam version? Current development is very slow, I imagine it must be very hard, perhaps that would help?
2) Any chance that the adventure mode will become more like The Sims soon? I cannot think of any roguelikes where you can just live a life in a fantasy world. I’m not interested in combat at all, but would like to play as a craftsman, have a family, build a house, that sort of things. But above all I would love to live in a dynamic, realistic world. Is that, at least partly, the focus?
Bumber:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8157956#msg8157956Shonai_Dweller:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8158003#msg81580031) Yeah, this is not likely to happen or help, though, yeah, we'd also been considering what sort of partial releases we could make if we are in a position to do so.
2) Soon? Nope! But things along those lines are on the dev page at
https://bay12games.com/dwarves/dev.html , where we have succeeded in not working on them very much. So it wouldn't be fair to say that's the focus as we haven't made significant progress. But it's something we'd like to improve.
hi, I just have small suggestion. I am not sure if someone suggested this before, but it would be nice if we could get a description of a civs and groups flags when we open up their history in legends mode. I hate having to hunt them down and what's worst is not being able to see other groups symbols that aren't civs like bandits, criminals organizations, guilds, merchants, mercenaries, and sects.
And also will you be able to implement it in 47.05? Doesn't feel all that difficult since the information of these groups symbols are already in the game, all it need is to be reveal Or will it destroy the whole code if you do?
Shonai_Dweller:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8158033#msg8158033Yeah, please use the suggestion forums! It is difficult for me to keep track of suggestions that are placed here. It's true that it wouldn't be a hard addition, but the list of those already requested is also many, many pages long, and we do them when we can.
Prompted by a recent interview somebody linked to (thanks!), where world gen branching was mentioned: Have you considered getting someone to help you with running a number of profiler and analysis tools over DF, in particular world gen, to track down uninitialized variables and other potential issues the tools may find?
Another grunt task that might be given to others would be crash replication, where the assistant would perform the repeat job of trying to get the crashes to show up, and then hand over the computer with the debugger at the crash location to you. That might not be very suitable at the moment, though, given that it would probably have to be done at your place. It may also require a fair number of computers for parallel crash generation attempts.
I do not have a person I trust to do this, same as ever. MSVC now has static analysis tools, and I found a few zingers with those last year (or the year before, don't remember.) But they don't find everything, especially in instances where two consecutive world generations cause the problem to come up (so it isn't strictly about the initial initialization.)
What spheres can change surroundings into evil? What effect have each of this spheres?
I think we just have blight, death, disease, deformity and nightmares. The first three kill vegetation, while the others sometimes do. The last two get "evil plants" and "evil animals" sometimes. Nightmares get bogeymen.
1. In the steam release will items in our inventory in adventure mode have their tile symbol next to them so we can get a cursory idea of what they look like without having to examine them?
2. Will we ever be able to start out as members of goblin or necromancer civilizations before the magic release when such things become irrelevant?
3. Sometime in the far future will our adventurers be able to further manage the hamlets they create? Maybe being able to appoint things like administrators or law makers or any position beyond hearth person?
4. For me at least adventure mode tends to crash after spending time in large cities when I try to save the game causing me to loose progress. To minimize lost progress will adventure mode ever get an auto save feature that periodically drops down a save of the game?
PatrikLundell:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8160302#msg8160302Shonai_Dweller:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8160317#msg81603171. Messed around with animals displayed on the embark screen most recently, so that's the current way things are headed, with some support now. But I don't want to promise any particular screen/feature early.
2. It is kind of taped together, as PatrikLundell replied. So I'm not sure we'll see that early.
3. There are a lot of angles here, some of them fresh from the villain release, and some of them languishing in the notes for over a decade. So, yeah, it's sort of a far future goal, with some improvements coming in earlier as we add the ability to do more and more with e.g. semi-autonomous party members etc.
4. Ha, yeah, sounds like the crash fix first here is more important. But some sort of autosave at some point would also make sense.
1.Can you briefly speak about how the UX work is going to be organised, what will be general direction/inspirations of the design etc.
2.The http://www.bay12forums.com/smf/index.php?topic=173474.0 thread to discuss graphics is absolutely awesome, could we get same thing for UX yay/nay type of feedback and ideation?
3.Are there any sounds/music to preview yet?
clinodev:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8160678#msg8160678Putnam:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8160719#msg8160719Shonai_Dweller:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8161037#msg8161037Putnam:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8161148#msg8161148Shonai_Dweller:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8161158#msg8161158Thanks for the music link, clinodev!
And yeah, the thread Putnam linked is what we have, to collect issues, and I'm not sure what else we'll be doing after we begin. Many of our larger ideas about UX are still some time off, if we get time to do them at all, where others hardly rise to the level of having a UX philosophy and are based more on feedback and being bludgeoned by horror. For instance, the entire player experience before getting to start fort mode is just baffling and terrible, and we're redoing it all, but that's being guided more or less by input over the years, and stuff that was clear but we never got to: repairing inconsistent key bindings, having a mouse option for everything, reducing some of the detours off to completely different screens, fronting important information and putting less necessary information farther away (embark screen has various weird priorities), otherwise changing the format of some confusing interactions (e.g. F1F2F3F4F5 as some kind of biome viewing interface), not dropping the player back to the title screen halfway through the process, not guiding the player toward an unannounced initial multi-hour world generation experience by default, and so on.
Once we are safely into fort mode, there are another set of straightforward complaints waiting for us -- digging not being plainly obvious as an option, k/v/q/t being four modes instead of one way of interacting more generally, vpl stuff should we get there, inconsistent keys (again), and several more pages of similar. By the time we work through as much of that as time allows, UX will be much improved, though we won't be to the point where we are doing what dedicated UX professionals do. I hesitate to use the acronym at all.
You've mentioned that you don't like how easy it is to feed your dwarves. Do you have any ideas for what sorts of changes you'd like to make to farming/meat production? Is this a mechanic you want to work on before the magic update, or is this something that's further down the road?
In-universe, are intelligent undead supposed to look like regular people? They're treated like normal citizens in worldgen, but is that because nobody actually knows they're undead? Vampires and necromancers look pretty much normal, so they can hide among the populace (until someone notices that they've been 37 years old for the past 500 years). Are intelligent undead in a similar situation?
This gets weirder when some of these guys can be like, eternally bleeding from a massive gaping wound in their throat, but many intelligent undead are unwounded.
Where do Gorlaks learn their useful advise and practice their stimulating conversation? The little guys seem to be just wandering around the caves solo, foraging for mushrooms or whatever. Is there some kind of secret Gorlak society they don't want us to know about??
Shonai_Dweller:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8161391#msg8161391Bumber:
http://www.bay12forums.com/smf/index.php?topic=169696.msg8161458#msg8161458Farming: I don't think we'll be getting to it before the map rewrite. We'd gone as far as floating NPK value stuff for soil, yeah, many years back (and though we generally say that our mom is a nurse and our dad worked in water treatment, as possible explanations for inspirations/formative aspects of the game, we could also say that our mom is an 8-hr-a-day gardener and our dad is a trained botanist, since that's also true, though it's only partially reflected in what we've focused on.) As I recall a long farming thread took off over in suggestions, and a lot of that is in line with where we'd take things (though I'm not sure as I recollect if I'd handle information the same way), but we haven't gotten around to systematizing plans there.
The game has not been remarkably consistent in how it treats obvious supernatural or monstrous things, like raising the dead right in front of people and so forth. There's just a lot that isn't done. It isn't entirely settled what they look like, though there's some bits implicit in the naming, because I didn't add in appearance changes - so currently, I suppose they do not either look rotted or have changes in appearance, assuming the resurrection interaction effect brings all of the tissues back to skeletons (I think it does?) I think, future-canonically-for-as-long-as-that-lasts, they are supposed to be both scary and undead looking - as you say, vampires already cover the normal looking case, and the names also suggest changes. I'm not sure, in the same sense, whether necromancers should physically change. Both are probably appropriate (and ultimately, intelligent undead that are properly normal looking is also appropriate, though the names would need to accommodate this. And the community behaviors of course.) Mummies are also a case like intelligent undead, since they both use the resurrect effect.
We just never did justice to the underground/animal people civilizations/societies/etc., after their height during the 2D days when they rode bats and flew in patterns and organized themselves a bit. Gorlaks never even got that much, since they came later. But I don't think gorlaks are meant to live in societies of any density, or societies at all. But they are supposed to chat and guide and advise, based on their cavern wanderings and social nature as it relates to others gathered underground. Dunno that we'd ever even imagined two gorlaks meeting, though of course that happens now in performance troupes and taverns and so forth. I don't think they are really ultimately meant to petition for residency either, for any permanent period, though of course that can happen now. We probably also vaguely lean toward having them at depth 2:3 instead of 1:3. They are not supposed to be running around outdoors, and the game isn't great at this either.
When you work on the greatly anticipated stress, needs, and happiness changes, will you focus more on fixing the little bugs that made your original plan not work, or more on modifying the algorithm to run stress towards "okay" rather than either ecstatic or misery until well-run, safe fortresses stop breaking around the 5-6 year mark (without extreme micromanagement?) Likely some of each, but have you decided a focus?
Note I'm not asking if it'll be fixed, as that's been well-covered, but it would be neat to see the original plan implemented! There was a moment around August when I was actually somewhat excited the needs bugs might encourage the fixing of long-standing but trivial bugs, like allowing dwarves to: enjoy well-made meals, rather than only (secretly) craving Zebra hearts or polar bear brains; seek out their specifically needed temple; seek out friendships and relationships; grab desired high-value clothes and trinkets from stockpiles, etc.
That might even be mostly the same thing, depending on how much the long-term stress problems depend on broken needs vs. e.g. broken memories. A lot of the notes I've taken down are about several issues with socializing, meal thoughts, etc. I haven't jumped into it yet, which makes it hard to answer your question completely. Issues with long-term memories inside stress-prone dwarves, that kind of thing; I haven't gotten up to speed on all of it yet, and won't until the work begins. I have some threads marked down to look at when the time comes. Ideally, a really poorly run fortress will break over the needs issues, while even an 'okay' fortress won't completely fall apart over the background stressors.
Now that we're a bit closer to that work, could we get a status update? It's my go-to quote for reassuring people that you don't intend for reasonably well-run fortresses to collapse into tantrum spirals from dwarves being being in a state of constant internal rage after rained on when entering the map as migrants, etc.
Since then, we created the stress feedback thread, read the comments, and have collected a bunch of notes. We're planning on addressing a portion of the list for the Steam release (or before, with one of the patches, however that ends up working out), and hopefully we'll land somewhere more pleasant while not falling back to the permanent euphoria days. I don't have a definite list, since the work will be coupled with some numeric investigation - as I recall, players running experiments found that unmet needs weren't actually a stressor, just more of an annoying blaring siren which doesn't actually cause much stress (which needs to be looked at for that reason), and I need to see how the numbers are actually shaking out on the famous rain rememberance example vs. the body hauling work vs. cave adaptation nausea vs. etc. Then there are the more positive ideas along the lines of people comforting each other more and having some more fort-wide happy thoughts when the player does things like win sieges. We still want poorly run forts to have trouble, but forts that are run well should have more isolated/fixable problems, and those should be explained well and feel satisfying.
Author
Topic: Future of the Fortress (Read 3124228 times)