Print

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

[squeakyReaper]

Unless this is going to be an ASCII game (not likely, as there are graphics in place already), fluff will make a difference visually.  Tan Dwarves, Pale Dwarves, ones with beards, ones without- they all look the same right now.  Perhaps the visual difference in this game will make it worth remembering?

[Little]

Good to see development continuing!

[Hugehead]

Doesn't DF plan to have procedural sentient races and animals? Anyways, it's looking nice, keep up the good work.

[Lord Dullard]

Unless this is going to be an ASCII game (not likely, as there are graphics in place already), fluff will make a difference visually.  Tan Dwarves, Pale Dwarves, ones with beards, ones without- they all look the same right now.  Perhaps the visual difference in this game will make it worth remembering?

This already is an ASCII game; it uses libTCOD. I am not artsy (or insane) enough to try and create procedurally generated 2D sprites. If somebody else is ambitious enough to think they can, I'd be happy to bring them on board for the project, but somehow I'm fairly skeptical that anyone would want to embark on such a masochistic venture.

I may or may not design sprites for the game. At present I am leaning towards 'not', simply because procedurally generated creatures/spirits/gods will not necessarily have a description that coincides with any sprites in a set, no matter how many I make.

As for detail fluff vs. mechanically-important details, I am inclined to agree with Servant Corps. There will be some 'meaningless' detail simply to help the player visualize things, but I feel that each world will feel more unique and enjoyable if details are taken into account mechanically and actually make a difference in gameplay. Reading the description of a creature and finding it interesting is a plus. Reading a description of a creature/plant and finding it interesting, THEN discovering a use for that creature/plant based on your observations or local lore takes the level of gameplay a step deeper. And those players who are only interested in the cosmetic fluff need not bother.

Doesn't DF plan to have procedural sentient races and animals? Anyways, it's looking nice, keep up the good work.

In all honesty, I have no clue. Thanks.

[Lord Dullard]

Work on creature gen is still progressing. It's getting complicated, but much more fun now that I'm no longer stuck on trying to organize the algorithm's framework.

The fact that my girlfriend is a bio major has been ridiculously helpful. She's been lending me a lot of input when it's applicable.

[nenjin]

Depending on how you arrange the creature generation, you could assign sprites based on some very simple classes. BI-PED, QUAD-PED, SNAKE, SNAKE-WING, WING-BI-PED, ect...

It could re-color the sprite based on some simple input from the creature file about it's general color. We've kind of played with this recoloring concept in the TOME engine game we're screwing with elsewhere, I don't know if it's applicable to what you're doing. If you really desire some non-euclidean stuff though, classes become pretty much useless

[Lord Dullard]

Depending on how you arrange the creature generation, you could assign sprites based on some very simple classes. BI-PED, QUAD-PED, SNAKE, SNAKE-WING, WING-BI-PED, ect...

Problem is, this would be infinitely more confusing to the player than simply using ASCII characters. Remembering what a purple 'd' and a green 'D' and a yellow 'R' and a blue 'z' mean would be fairly easy. Remembering the difference between those four if they all use the same sprite becomes difficult - especially if a sprite is used enough times that the colors of multiple creatures become similar. Throw fifteen of the same sprite at somebody in the heat of combat and expect them to know what they all represent and the various capabilities of each? I doubt that would work out well.

I'm open to suggestion, I suppose, although I am currently working within the constraints of a 10x10 tileset (although this could fairly easily be changed).

It MAY be possible to mimic something approaching 'procedural' sprites by creating a 1x1 libTCOD console and pasting several layers of body parts over one another to create a simple representation of the creature, but I haven't tested it so I'm not entirely sure if it would work (and honestly, I'm not THAT interested in finding out, since I'm not exceptionally gifted at designing sprites myself).

[Lord Dullard]

Quick update: The screen size is now adjustable. All display functions scale appropriately with user-specified screen size.

[squeakyReaper]

Quick update: The screen size is now adjustable. All display functions scale appropriately with user-specified screen size.

Huzzah!  Tired of being harassed about it so much?

[Lord Dullard]

Quick update: The screen size is now adjustable. All display functions scale appropriately with user-specified screen size.

Huzzah!  Tired of being harassed about it so much?

I figured it was going to be one of those requests that would crop up more and more as time went on, so I'd probably better nip it in the bud.

Thinking out loud:

Creatures are going to be generated procedurally as follows:

1. Determine the creature's predation type. Is it carnivorous, omnivorous, herbivorous, etc.?
2. Determine the creature's biome/sphere. Is it a primarily aquatic, terrestrial, subterranean, aerial, or amphibian creature?
3. Taking into account the two above steps, determine the creature's body structure, with limitations and exceptions in place according to what the creature eats and what environment it lives in.
4. For each limb/body part, determine the specific morphological features of that limb and its functions and/or specializations. Also, attach an intensity gradient to these functions. So, for example, one creature might use 'long, muscular' limbs with a primary function of 'climbing'; another might use 'long, muscular' limbs with a primary function of 'striking its prey' or 'running for long distances'. Each of these have might other dependencies (striking prey might require claws/talons/etc.) and will grant different statistical advantages and disadvantages. Such advantages and disadvantages will fall along an intensity gradient that will determine how great of an effect they have on the creature's overall statistics.

Gradient:
5: (Most) ['extremely', 'incredibly', 'amazingly', 'fantastically', 'unbelievably', 'inexplicably', 'phenomenally']
4: ['radically', 'excessively', 'abnormally', 'exceptionally', 'extraordinarily']
3: ['remarkably', 'conspicuously', 'prominently', 'surpassingly', 'strikingly']
2: ['curiously', 'peculiarly', 'markedly', 'particularly', 'very', 'significantly']
1: (Least) ['quite', 'somewhat', 'notably', 'fairly', 'rather', 'relatively', 'moderately']

So, from the example above, one creature might end up with 'long, phenomenally muscular' legs that it uses for climbing; the other might have 'long, curiously muscular' legs that it uses for running long distances. 'Long, muscular legs' might grant ++AGI, +STR, and -END (this is just an example I'm pulling out of my metaphorical arse, please don't start a debate with me about statistical advantages of morphological characteristics until I have time to plot them all out). The first creature, which generated a descriptive intensity modifier of '5', would get +10% AGI, +5% STR, and -5% END. Thus, morphological CHARACTERISTICS - i.e., the morphological shape and development of a limb - parallel its statistical benefits and detriments. The morphological/evolutionary FUNCTION of the limb - i.e., it is used for climbing - corresponds to a skill, which will also receive a relevant bonus. I will probably not implement negative skill modifiers, except that skills will be modified by attribute levels, so a creature's skills in one area might be negatively modified by morphological characteristics it develops in its various body parts.

Morphological characteristics are sorted by Predation and Biome/Sphere. Some characteristics will be general and evolvable by a creature of any predation or any sphere (i.e., something like a 'long, lithe tail', since just about any kind of creature, vertebrate (monkey) or invertebrate (scorpion) from any sphere can develop a tail). Some will have specific requirements.

Functions, furthermore, will be sorted by Predation, Sphere, AND Morph. Characteristics. Long, spindly legs should be able to be used for 'maintaining balance' or 'climbing' or 'running', but probably not for 'digging' or 'overpowering prey'.

Statistics for any creature of any size can be anything; however, size modifiers will determine the relative physical capabilities of a creature, which should theoretically restrict a really strong chipmunk-esque fuzzy thing from beating the living tar out of an unsuspecting adventurers which outsize them by a hundred-fold or so.

With all of the above taken into account, I'm going to try to make descriptive creature entries seem unique, interesting, AND morphologically relevant. Looking at a creature should give you a general idea about its capabilities, the sort of things it does to survive, whether it's threatening, and so forth. Taking a more in-depth look at it in the World Atlas will help you out more. I may or may not* require adventurers to encounter creatures before accessing World Atlas entries about them in order to fuel a sense of exploration and discovery in the game (if I decide to do this with creatures, it will probably also extend to stuff like myth, history, cultural facts, blah blah).

* - I can't decide whether it would be better to just give open access to all information immediately after worldgen or not. Perhaps basic facts will be immediately accessible, with more in-depth information (stats, skills, abilities, etc.) available after a certain number of encounters.

[Knight of Fools]

It could depend on the availability of that information.  If it's common knowledge that a certain creature exists, let the character know about it - But once someone goes outside of their home town and starts encountering new creatures, then things start standing out.  Players should also be able to listen to locals talk about certain creatures - Locals can be wrong, of course, underestimating or overestimating the prowess of different creatures.  Most won't find anything to say about common animals (Why would you tell someone about the average, uninteresting dogs?), though they may give warnings about dangerous ones ("Beware the blood-sucking toothless mammoth!  It comes out only at night!").

You'd be opening a whole new can of worms here, though - Priorities in conversation and gossip don't have to be done immediately, if at all.

You could even make an option to tweak this for each player's preferences - People who just want to look at weird animals may not want to have to explore the entire world in order to do so.

I'd suggest making marked differences between locations in order to promote this feeling of the familiar (What you grew up with) and discovering new creatures, sectioning off continents and the biomes inside of the continents and putting a number of different, unique creatures in each - While having a fair spread of "common" creatures, like our world's deer, rabbits, and horses.  Even common creatures can be different from other places, though, so venturing from your homeland to a colder place may lead to the discovery of the Wool-Shrouded Deer, which grows excessive amounts of fur to survive the colder climes.  Going to the desert, though, you may not find any species of deer at all, instead finding a completely different range of animals and creatures you've never heard of before, except maybe from travelers, and even then their information was very scant and incomplete.

I'm very interested to see how this works out.

[Askot Bokbondeler]

posting to follow development

[Lord Dullard]

Great post, KoF. Thanks very much for the input.

Just a quick note:
Since this thread is approaching 10,000 views, I've decided to pledge a portion of any donations/compensation Cult generates (when and if it gets that far, but I certainly think and hope it will) to Toady/Bay 12 as a means of showing gratitude for the excellent games that B12 provides to this community for free, not to mention the forum and the awesome resource of the community itself. Thanks, B12!

[nenjin]

Well played.

I'm enjoying thinking through your model at the moment.

I'd suggest a couple hidden classes/body parts so you can create characteristics that aren't immediately deducible through descriptions, or act as generic modifiers for things like body mass, density, magic?, metabolic rate (hot vs. cold blooded), things of that nature.

I imagine you could develop organ structures along these exact same lines as well.

I really like how you've structured things in such a way that they'll provide stats, gamey numbers stuff for players and as mechanics, as well as SPECIFIC advantages. That was my biggest gripe with Spore, is that it only had these generic values adaptations contributed.

[squeakyReaper]

Just had a random thought come up.

When generating the animals, make sure the game considers their environments and habitats.  Say you were creating a new creature in a forest.  It considers the fauna and life there.  Is there a niche for animals that only eat off the floor of the forest, or are they're many of that already?  Are there trees with high hanging fruit?  Better give them the ability to climb, have long necks, or possibly even throw/spit projectiles.  Likewise for carnivores.  You need to generate an animal that could consistently defeat and then devour some animal niche in the forest.  But having too fierce of one wouldn't make since, as that would over hunt and run out of animals to go after.  Might not be worth it, but at least have the generator consider some niches and relative ferociousness of carnivores.