Rym

[Kay]

http://www.fur.com/~ollie/rym3.html

Here's an example of a related bit (actually a whole lot) of furry world-building. I'd like to see an ItE sequel where it's possible to walk around and just explore the world. "Morrowind" has the beginning of a good conversation system for a world like this: that system lets you define a new NPC in terms of race, class, etc. and have them automatically know everything a person of their status should know.

Imagine if you could generate a random ferret village full of characters who automatically know about the local geography and legends and can tell you about them. You could play for hours even if you ignored the main plot!

[FallenWolf]

Imagine if you could generate a random ferret village full of characters who automatically know about the local geography and legends and can tell you about them. You could play for hours even if you ignored the main plot!

Well i think that would go much too far. Don't forget, Morrorwind cost at least a million dollar and took 3 years of work for a development-team! ^^

Plus, i'd not like to see that cracy new colors and effects on "ITE2"... Of course some better resolution then the old game.. but still the same style ^^

But exploring the world was already possible in the main game, and that could be extended i think...

[Mira]

I wouldn't want the huuuuge Morrowind-type world.  The non-linear-ness of Morrowind bothered me a bit (I'm used to Might and Magic XD)

What I liked so much about the way ITE worked was that the characters didn't look "blocky" like a lot of 3d characters do today (remember N64?  Yeah, like that! >.<).  And as we don't see magic at use in this world, there wouldn't be an opportunity for tons-o'-special-effects to come into play.

Subplots might be fun, but what would you get as a reward?

[FallenWolf]

What I liked so much about the way ITE worked was that the characters didn't look "blocky" like a lot of 3d characters do today (remember N64?  Yeah, like that! >.<).  And as we don't see magic at use in this world, there wouldn't be an opportunity for tons-o'-special-effects to come into play.

Subplots might be fun, but what would you get as a reward?

I know what you mean, Mira Non-linear Gameplay can quickly be too complex and takes away the fun.
What i always liked about the graphics of ITE was that it was all made with love. And that is what the 'player' feels when he plays the game and discovers tiny (funny) details in a game.

But i would really not worry about rewards! There are dotzends of possibilities, just think about special items you can use again later in the game, goodies in the game, nice animations, or just secret rooms! Just what came in my mind within 30 secons so there would be ENUGHT stuff for a reward ^^

~FallenWolf

[Threed]

http://www.fur.com/~ollie/rym3.html
I'd like to see an ItE sequel where it's possible to walk around and just explore the world. "Morrowind" has the beginning of a good conversation system for a world like this: that system lets you define a new NPC in terms of race, class, etc. and have them automatically know everything a person of their status should know.

Imagine if you could generate a random ferret village full of characters who automatically know about the local geography and legends and can tell you about them. You could play for hours even if you ignored the main plot!

World Sims tend to be very light on the plot, though, because it is hard creating a static plot in a dynamic world. Take Morrowind for instance: basically all the "sentient" AI had to do was walk around; the animals used some kind of primitive logic loop that dictated that they played the "feed" animation or "walk" animation until the player entered their sphere of awareness.

So, really, you've got two choices:

A limited world that is structured around its plot.
This is not to say the world would not be dynamic in a sense, but that dynamicism would only operate within a limited subset of the game, because if you allow for truly "free roaming" control that means its highly possible that you break the main plot, or break things associated with the main plot, and essentially break the game because the plot can no longer be completed. In the end, the user wanders around unable to complete the game, and becomes disappointed with the product. To achieve stability you need several static, unchanging variables.

It can, however, be "dynamic" in several smaller aspects, such as AI that follow a predefined routine (go to work, hang out, talk to other NPCs) while not engaging in dynamic behavior such as a thief AI randomly wandering the city, looking for houses with open windows and waiting until the street empties out so it can enter the house and remove anything that has a value property that's high enough to warrant its interest. Another restriction would be how NPCs react to other NPCs - sure, it would be great if a group of paladins went hostile on a group of thieves, but what if you needed the thieves to complete a quest? Sure, you could make them "invincible," but then they end up slaughtering the paladins and that, to be honest, seems like a boring outcome.

That does not mean you can have a really interesting, quasi-dynamic world for the player to explore while uncovering the main plot.

An entirely dynamic world.
This is probably easier from a programmatic and creativity standpoint, but not from a 'we need an epic adventure!' standpoint, because when you are dealing with dynamicisms on a scale of this magnitude, you cannot deal in absolutes - only abstracts. "There is a 5% chance of a rat cave being near the outskirts of a village," not, "There is a rat cave outside village A at position X, Y and filled with DELICIOUS NACHO CHEES."

As I've said, you can't deal with a dynamic world like this in absolutes - only abstracts. The world and AI would have to be defined by a set of "rules." For instance, the world is X long, Y wide. But, if I define the world in absolutes - river here, waterfall here - as opposed to a dynamic algorithm - waterfalls have a 5% chance to occur near the end of a river - a serious problem is going to emerge with your AI: it will invariably be "similar" in "similar" places.

Let's say your AI generates random lore and bandit-stomping dungeon-crawling quests depending on its enviroment. This means that if you use a static enviroment, and create a village next to a waterfall, even though you can dynamically generate the AI, you will still encounter a strange similarity from game between game. The AI has to sample its enviroment, most importantly: it can't just tell the player, "There are bandits on the mountain!" and there not be any mountains around for miles, and miles, and miles.

So the AI samples the enviroment, and the enviroment does not change, big deal!
Big deal indeed - let's take a random quest generate.

A has B which I want, and B is at point C, surrounded by D.

A is an NPC of some kind - ogre, frost giant, whatever.

B is an item.

C is a locaiton.

D is a random variable dependent on A's status and race.

Now, if the enviroment is a static factor, C will have to be sampled from the enviroment. This means, always, a limited set of locations to choose from - they're either on the mountain or near the waterfall or at a point that varies from either of these points by a few feet.

Why? The enviroment is static, and, following a set of game logic that says people cannot be inside the ground, or an object, it limits their area to a few points that the game designates as "viable." Since the enviroment is static - keyword here - these viable points will be unchanging, so it will come to pass one or two quests later you'll get a quest that has you going to the same location you visited just recently, because there are only so many points.

Again, why?

Let's say that A is an Ogre, and Ogres only like swamps. There is a high chance of them occupying a swamp if one is nearby, a rule says. If one area of land is always a swamp area, most of the time that Ogre will be present. Suppose the predefined world has a swamp near our little home away from home. Now, almost invariably, if the Ogre is randomly chosen to be A, it will almost always appear in the swamp somewhere. Now, if the world is dynamically created, A doesn't necessary appear in a swamp, because a swamp is not necessarily nearby. Dig?

This sort of methodology applies to everything in the end: the more static variables you have, the higher influence you place on the randomization and AI functions, and the more "predictable" the world will become. Its Fuzzy Logic - you feed it enough similar information and you will almost always get the same result, every time. You feed it only one or two scraps, though, and the output might be crazy difference.

So, for a world sim to be truly dynamic...

Anyway: World Sim - abstract, defined by a series of rules, much like reality (gravity, physics). In this way you can pretty much guarntee that each new play (from scratch, no save/load) the world will have changed significantly enough for it to seem like a completely different product with a similar theme.

A Computer Game - slightly dynamic, but limited in many aspects.

You really have to ask yourself: do you want a story, or a world?

[Kay]

That was a well-thought out response.

if you allow for truly "free roaming" control that means its highly possible that you break the main plot, or break things associated with the main plot, and essentially break the game because the plot can no longer be completed.

Morrowind deals with this problem by giving you an explicit "you broke the world!" message if a key NPC dies. As long as there's an autosave, it's OK to make it possible to break the game.

Travel restrictions are more of a problem. If the plot relies on your not being able to cross the mountains, and a clever player finds a way to steal an airship early on, the plot might be ruined. In Morrowind it's possible to fly right into the final boss' lair with some generic levitation items, for instance. So, if there's a plot, it's dangerous to give the player too much travel freedom. The usual RPG solution is to have a type of terrain you can only cross with vehicle/item X, which you can only get after a certain plot event. It's possible to severely mess up travel restrictions, though. In Breath of Fire III, you go through an arduous desert journey only to discover that you could have teleported across if not for a wooden crate and an eight-foot ledge in the way. That was infuriating.

You really have to ask yourself: do you want a story, or a world?

I'd most want to see a world with hordes of NPCs traveling independently. Failing that (until we all get quantum computers!), how about focusing on the story, but including a significant amount of abstract world-generation?

For instance, look around ItE's Ferret Village. There are a bunch of empty houses and many similar-looking NPCs wandering around, all with pretty much the same dialogue. What if, instead, there were a few floor plans for each house, and a few pieces of furniture placed semi-randomly in them? If there were a system for stealing and getting caught, you'd have a mini-game in place as you explore random houses for loot. Each NPC would be defined in terms of race and class with a few random variables. You could then get a variety of conversations out of them. If there were a larger city, it could be given a randomized "floor plan" so that you're walking through different districts with different scenery.

I agree that basing the whole game around random bandit-stomping quests would be boring, but the more a sequel experimented with flexible, abstract world generation instead of scripted dialogue, the more replay value it would have. Imagine being able to play through it several times as different species and running into different problems because (for instance) all non-foxes assume you're a thief for being a fox.

You don't have to simulate a whole world to get more interactive characters and setting than adventures and RPGs usually have.