Why isn't this open world?

[Angel]

Wouldn't it be much cooler to generate the colony map (using what's given in the world map) as soon as we explore the colony map further. So that the colony map could become infinite in size but certain areas are just not being rendered, kinda like a fog of war?

^^^^^^ TL;DR ^^^^^^

Edit: Because someone had trouble understanding what I initially wanted, (I purposely tried to keep this suggestion short, because there is a multitude of layers that this could turn into.
So here is a very specific possible way this COULD play out. It isn't necessary to read this.
Quoting myself:

The idea is that the World Map would be used to say "okay this is where one tile of the world map is" and the colony view would be able to technically span the entire world map. "rendering" might be an issue here, the same with pathfinding. It's not about "finding uber stuff", it's about making world map and colony map intertwine rather than just having the colony map and thats basically it. The world map doesn't offer much to work with.

For example, you could select your colonists and make him walk over the entire colony map, to idk.. a second settlement you build, or a nearby friendly NPC colony or smth like that. Turning the world map into something you look at and maybe command over, but add a huge amount of depth to the colony map.

Think about it, you could add in caravans into the colony map, in order to trade with one of your own colonies or an NPC colony. And it would have to deal with things on it's way, displayed on the colony map.

You could, maybe in the future, have the ability to wage wars by gathering a little army from all your colonies in a certain area.

Modders could add automatically generated structures like friendly villages, ruins to explore, even quests for your colony or as mentioned before, bandit camps.
You name it! Anything could happen in the lives of our colonists.

At one point they might establish a peaceful co-existence with the indigenous people just to one day rebuild their spaceship.
Or you conquer the entire planet. The possibilities would be endless.

The whole diplomacy thing could develop into so much more.


If the colony map was auto generated, the possibilities would grow in such a gigantic manner.

It's not about what we directly gain from it, it's about what this would make possible.
This game LIVES from the modding community.

[Igzom]

Yeah! Good idea!

[cleeedet]

Once you land

"The fog slowly closes"

everything that is not 7 squares away from your bace in the world map is now hidden

To reveal you must slowly move through the map. Baces will be revealed in the fog if they call. Aswell as places that you need to go to for a mission.

Hm. Sounds like a fun idea, Or do you mean the place you land on. I would not like to have to move my pawns all over the map just to reveal everything when I have to micromanage enough.

But on the world map. Sure!

[corisai]

Because it will overload your CPU very fast.

Don't forget that:
1) Rimworld is a child of one dev, so it's unfair compare it with something like Factorio (and even there a lot of performance issues).
2) Pathfinding. This is a major source of problem. Pathfinding algorithms get overloaded by bigger maps too quickly (for example Factorio almost not using pathfinding at all, except for biters - and they ruin large maps easily).

[INST_4D]

you could try the game Keplerth. looks a little bit like rimworld at first but you have only one colonist. you control him with WASD, you can build a base wherever you like and have open worl with a lot of npc colonies (friendly/unfriendly).
but it's still in very early stage of EA.

[Astasia]

That should have been the way the game was designed, it would have helped with performance a lot with dynamic and more controlled play areas. Instead of scanning a 400x400 map constantly they might only be scanning a 100x100 visible area, and anything in the fog of war would require manual orders to interact with. Expanding on the map could have been controlled similar to how drones in factorio are controlled, some sort of object that grants visibility over an area and allows colonists to "see" it, and that object could have created a link between other similar objects to simplify pathing and create an automation priority. Instead of pathing 400 tiles to cross a map they might path 50 tiles to the next relay and once there they would path another 50 tiles, in fact the relays themselves could create one constant path when placed and the colonists would always follow it with no AI involved beyond that point. The area near the colonist would be scanned nearly every tick as currently, but areas 1 link away might be scanned half as often, and areas several links away would be scanned very rarely. This would have allowed for much more expansion with much less CPU use, and throw in proper multi-core support and it would have been difficult to run into performance issues.

None of that happened though, and none of it will happen at this point. This game didn't know what it wanted to be at first, and just evolved into what it is. The foundation the game was built on was never really meant for what it ended up as. Originally the game was going to be a space station survival game, taking place in a very confined area with no real need for giant maps or the amount of pathing that happens.

[Angel]

Originally posted by corisai:

Because it will overload your CPU very fast.

Don't forget that:
1) Rimworld is a child of one dev, so it's unfair compare it with something like Factorio (and even there a lot of performance issues).
2) Pathfinding. This is a major source of problem. Pathfinding algorithms get overloaded by bigger maps too quickly (for example Factorio almost not using pathfinding at all, except for biters - and they ruin large maps easily).

That does make sense.

[Morkonan]

Originally posted by 4ngel:

Wouldn't it be much cooler to generate the colony map (using what's given in the world map) as soon as we explore the colony map further. So that the colony map could become infinite in size but certain areas are just not being rendered, kinda like a fog of war?

What areas are not being rendered? What?

Anywhere a pawn goes is going to get rendered. This isn't a one-perspective game. You can jump from pawn to pawn no matter where they are on the map and the camera will center on them right there, where they are - It's gonna get rendered. There is no "fog of war" savings when you have a dozen pawns running all over a map.

Imagine the map being infinite. OK, so what are you going to want that for? Exploring? Resource gathering? And, what are you going to do with that? ORDER PAWNS TO GO THERE. So, you find some nice uber-awesome resources on the South side of the map, 400 tiles away, and some really special resources on the North side of the map, 500 tiles away. You order pawns to gather both... Grats you, ginormous map for little gain and lots and lots of pathing, spawning and critter AI issues.

And, how many pawns are going to end up sleeping on the ground, eating without a table, bleeding to death or freezing to death just so they can walk for half an hour so they can bang once on a rock with a mining pick before returning home to watch TV?

[Astasia]

"Rendering" isn't really an issue since it's a flat plane, but it's still very easy to create a system where terrain visuals are unloaded when the camera isn't looking at it, even while colonists are standing on that terrain doing stuff.

Actually unloading the terrain and preventing it from being scanned is what would help performance, and even in a situation where you are dealing with things 400 blocks away that's still the case. A 400x50 area is much less CPU demanding than a 400x400 area. Once colonists are done working in that area it could be unloaded and you could send them out 500 or 1000 blocks in another direction and you are still dealing with a smaller active area than a current 400x400 map. This is without any of the AI optimization options I mentioned in my previous post, just using the current AI and focussing it on smaller active areas would allow for much larger potential play areas.

[corisai]

Originally posted by Astasia:

Expanding on the map could have been controlled similar to how drones in factorio are controlled.

Drones are lack pathfinding almost at all (they semi-limited by power area but it's very, very CPU-friendly solution) and fly over terrain, being unaffected by it.

Originally posted by Astasia:

The area near the colonist would be scanned nearly every tick as currently, but areas 1 link away might be scanned half as often, and areas several links away would be scanned very rarely.

You will barely win anything from this scheme. Many conditional indexing instead of one scheduled = bad.

There's no easy solution to pathfinding problem. Simpy - math behind it against it :)

Same reason why Kerbal Space Program have very limited gravity model (with only single source of gravitation in any moment) - there's no way to solve gravity interference of multiply bodies easily, so home PC simply unable to do such calculations in time.

[Angel]

Originally posted by Morkonan:

Originally posted by 4ngel:

Wouldn't it be much cooler to generate the colony map (using what's given in the world map) as soon as we explore the colony map further. So that the colony map could become infinite in size but certain areas are just not being rendered, kinda like a fog of war?

What areas are not being rendered? What?

Anywhere a pawn goes is going to get rendered. This isn't a one-perspective game. You can jump from pawn to pawn no matter where they are on the map and the camera will center on them right there, where they are - It's gonna get rendered. There is no "fog of war" savings when you have a dozen pawns running all over a map.

Imagine the map being infinite. OK, so what are you going to want that for? Exploring? Resource gathering? And, what are you going to do with that? ORDER PAWNS TO GO THERE. So, you find some nice uber-awesome resources on the South side of the map, 400 tiles away, and some really special resources on the North side of the map, 500 tiles away. You order pawns to gather both... Grats you, ginormous map for little gain and lots and lots of pathing, spawning and critter AI issues.

And, how many pawns are going to end up sleeping on the ground, eating without a table, bleeding to death or freezing to death just so they can walk for half an hour so they can bang once on a rock with a mining pick before returning home to watch TV?

The idea is that the World Map would be used to say "okay this is where one tile of the world map is" and the colony view would be able to technically span the entire world map. "rendering" might be an issue here, the same with pathfinding. It's not about "finding uber stuff", it's about making world map and colony map intertwine rather than just having the colony map and thats basically it. The world map doesn't offer much to work with.

For example, you could select your colonists and make him walk over the entire colony map, to idk.. a second settlement you build, or a nearby friendly NPC colony or smth like that. Turning the world map into something you look at and maybe command over, but add a huge amount of depth to the colony map.

Think about it, you could add in caravans into the colony map, that would have to choose a path to avoid bandit camps or the like (a huge colony map view would make for a lot more possible changes), in order to trade with one of your own colonies or an NPC colony.

You could, maybe in the future, have the ability to wage wars by gathering a little army from all your colonies in a certain area.

Modders could add automatically generated structures like friendly villages, ruins to explore, even quests for your colony or as mentioned before, bandit camps.
You name it! Anything could happen in the lives of our colonists.

At one point they might establish a peaceful co-existence with the indigenous people jus to one day rebuild their spaceship.
Or you conquer the entire planet. The possibilities would be endless.


If the colony map was auto generated, the possibilities would grow in such a gigantic manner.

It's not about what we directly gain from it, it's about what this would make possible.
This game LIVES from the modding community.

[Astasia]

Originally posted by corisai:

Drones are lack pathfinding almost at all (they semi-limited by power area but it's very, very CPU-friendly solution) and fly over terrain, being unaffected by it.

I didn't say anything about pathfinding there. Currently colonists actively scan every single tile on the map looking for tasks. By creating a task region limited to smaller areas that would significantly reduce the amount of scanning each colonist is doing. That's how the drones in Factorio work, areas outside the logistics coverage aren't being scanned by the logistics network for tasks so you can have giant maps full of lots of empty space that don't have an exponential impact on performance.

Originally posted by corisai:

You will barely win anything from this scheme. Many conditional indexing instead of one scheduled = bad.

There's no easy solution to pathfinding problem. Simpy - math behind it against it :)

Again not talking about pathfinding. Limiting task scanning has a very huge and positive effect on game performance. If it was checking a condition on every tile in the world then yes it would be bad, but in combination with the above system to break the world down into bite-size chunks it would save far more processing than is lost.

The solution for pathfinding was this part:

Originally posted by Astasia:

Instead of pathing 400 tiles to cross a map they might path 50 tiles to the next relay and once there they would path another 50 tiles, in fact the relays themselves could create one constant path when placed and the colonists would always follow it with no AI involved beyond that point.

You have again bitesized pathfinding commands that queue up as needed, or essentially a rail system that removes long distance "pathfinding" entirely.

[desrtfox071]

For the OP, game design questions aside, you can achieve the "open world" effect by using a mod called Set-Up camp:

https://steamcommunity.com/sharedfiles/filedetails/?id=932951772

You are still using the map tiles, but you have what amounts to a nearly infinite world, and can use map times to augment your colony or to create additional colonies. Some of this functionality is already in the base game as well, using the default caravan functionality.


So, basically, RimWorld is already open world, it just isn't seamless.

[Henry of Skalitz]

Originally posted by desrtfox071:

For the OP, game design questions aside, you can achieve the "open world" effect by using a mod called Set-Up camp:

https://steamcommunity.com/sharedfiles/filedetails/?id=932951772

You are still using the map tiles, but you have what amounts to a nearly infinite world, and can use map times to augment your colony or to create additional colonies. Some of this functionality is already in the base game as well, using the default caravan functionality.


So, basically, RimWorld is already open world, it just isn't seamless.

Sadly, this mod isn't compatible with 1.0

[desrtfox071]

Originally posted by Smelly Fish:

Originally posted by desrtfox071:

For the OP, game design questions aside, you can achieve the "open world" effect by using a mod called Set-Up camp:

https://steamcommunity.com/sharedfiles/filedetails/?id=932951772

You are still using the map tiles, but you have what amounts to a nearly infinite world, and can use map times to augment your colony or to create additional colonies. Some of this functionality is already in the base game as well, using the default caravan functionality.


So, basically, RimWorld is already open world, it just isn't seamless.

Sadly, this mod isn't compatible with 1.0

1.0 is still experimental and unstable. Don't expect most mods to be compatible yet.

[Morkonan]

Originally posted by 4ngel:

The idea is that the World Map would be used to say "okay this is where one tile of the world map is" and the colony view would be able to technically span the entire world map. ...For example, you could select your colonists and make him walk over the entire colony map, to idk.. ... Turning the world map into something you look at and maybe command over, but add a huge amount of depth to the colony map.

Think about it, you could add in caravans into the colony map, that would have to choose a path to avoid bandit camps or the like (a huge colony map view would make for a lot more possible changes), in order to trade with one of your own colonies or an NPC colony.

You could, maybe in the future, have the ability to wage wars by gathering a little army from all your colonies in a certain area.

Modders could add automatically generated structures like friendly villages, ruins to explore, even quests for your colony or as mentioned before, bandit camps.
You name it! Anything could happen in the lives of our colonists.

At one point they might establish a peaceful co-existence with the indigenous people jus to one day rebuild their spaceship.
Or you conquer the entire planet. The possibilities would be endless.

If the colony map was auto generated, the possibilities would grow in such a gigantic manner.

It's not about what we directly gain from it, it's about what this would make possible.
This game LIVES from the modding community.

There is not one gameplay suggestion that you have made that could be facilitated by this proposed mechanic that can not already be done or isn't already present in the game as it is.

The only reason for the suggestion appears to be a desire to have one contiguous mapspace for the colony map. (Sort of like "Minecraft" in that respect, for instance.)

But, the game isn't structured like that. At least, it doesn't appear to be. So, restructuring it to accomodate this proposal would have to be incited by a very good gameplay reason for it. And, you haven't made one yet...

So, why?

Every gameplay mechanic you've mentioned is already in the game.

What is the "unique" gameplay element that would be facilitated by this and why is it desirable enough to warrant recoding the game to support it?

(Something besides just the desire for a "big colony map.")

A note on Level of Detail: By that, I don't mean graphics rendering, I mean gameplay. The gameplay centers on a small group of people surviving in a harsh environment with the possible goal of building a spaceship to escape or making an equally arduous journey to find one. The game is designed for a relatively small population of characters to be managed by the player at any one time. The immediate play-spaec is suitable for the size of the group to build and equip a base of adequat size to fulfill end-game requirements with the added ability to leave that playspace and, effectively, generate a new one on the world map if necessary. The size of the playspace supports game elements of reasonable amounts, like terrain, resources, wildlife, etc, to support the player's activities in building their colony. Animals, except for the occasional special event like migrations, caravans, raiders and generated spawns are within the same general parameters as the colonists present.

In other words, the scope of the game isn't like "SimCity" or a 4x game. Also, "travel" doesn't have to be a large part of gameplay activity. Most of the events and tasks the player is faced with take place in one playspace and when travel is necessary, it's a simply overland-map tracker with the possibility of an event that generates a temporary playspace. But, the player has access to a multiude of different instances that can be opened up and played, but only one at a time.

That's the sort of "scale" of the game. Contrast that to what you're suggesting, which is basically changing the game from that LoD to one that encompasses an active playspace of "every possible place at once."

If you took a party-based roleplaying game which had six slots in the party for characters and then suggested that it should instead be a 4X map with the ability to upgrade cities, manage complex economics, and start wars with other factions that hadn't been invented yet, there'd have to be a very good reason to do that, right?

So, think up a very compelling gameplay element that fits within the general scope of the game and the player's tasks while playing that would be a worthy enough addition to the game to warrant a rewrite necessary for this feature - Think your suggestion through.