DT World Generation Question

Development
1

Capture
Capture.jpg1698Ă—871 431 KB

One thing specifically I wanted to ask about the recent DT, is how this would look and work with map generation / starting banner location.

  • Would it be like Minecraft where we pick a seed and are dropped into a specific spot and expand out from there?
  • Or would we still have the same map, but one we explored past the edge, it begins generating into something else?

Also, with showing multiple biomes / maps combining into one, would it be possible to have infinite map generation? If yes, I think it would also make a few other things easier.

  • This would allow us those of us that want to have an empire with multiple towns, as we could just…go out and make them.
  • When playing Co-Op Multiplayer, the two players could start together, and either build together or build seperately and it wouldn’t hurt each other.
3 Likes
2

I’m not part of the team but answers so they can just agree in case I’m right.
If they haven’t changed their stance on it recently then the size of the world is limited by performance. The original aim was to have an infinite, or at least a much larger map, but they abandoned that to focus on making the game run well for a map that had room for a town and some wilderness. That’s the main focus of the game after all.
I think the major performance problems was with the AI and keeping track of all the items. There have been improvement and there is probably still things that can be done, but people complain about not being able to build one big town. It’s not likely that we get to a point where we can have multiple towns on the same map any time soon.
Having multiple bioms on the map should have a negative effect on performance too, since is needs to handle more types of stuff.
It’s more likely that we’ll see a solution where you can have multiple maps/saves from the same world and they can trade with each other and things like that.

2 Likes
3

With the fog of war showed off in the last DT there’s a possibility they could unload empty chunks, which could help performance in a larger world.

4 Likes
4

True, but if the big weight on performance is the town itself with all the hearthlings and items, then adding more towns will get heavy fast, even if you delete everything between them.

The fog can definitely be used to simplify the hidden things. The AI for enemies don’t have to think about how to look alive, only if it’s time to attack or not. Items can be removed completely if left in the fog for too long. But the most important improvements will probably happen elsewhere, like a more effective pathfinder or finding and removing unnecessary loops in the AI.

Don’t get me wrong, I am hoping that we’ll be able to have big or even infinite maps at some point. It was one of the big things that made me exited about the kickstarter. I’m just saying that this prototype doesn’t give me a reason to believe that we’ve gotten any closer to it.

1 Like
5

I remember when that was brought up. Some asked to look into it later when optimization was better, but Tom made the statement on one of his streams that no one has filled the map or come close, so they feel the map was good enough for the time being.

That being said, Minecraft overcame the performance problem by unloading chunks within a certain distance, as @coasterspaul stated. This caused problems with Redstone machines…so it was give and take.

2 Likes
6

Unloading distant parts of the world is a common thing to do in games that only need to care about the part of the world that the player can see. Stonehearth can’t do that (as easily) though, since it still needs to simulate what the hearthlings are doing. If you want to unload one town while you look at another and still see progress when you get back to the first one, the game would have to be able to estimate how long everything takes so that it can load the town in the expected state. In a game with as many dependencies as Stonehearth (walking, building, changing the landscape, crafting, being interrupted by enemies, fighting them and so on) making even a rough estimate gets really hard and the error would get bigger the longer you’re away. If the estimate is too slow the player will be annoyed. If it’s too fast players will take advantage of it, especially in competitive multiplay.
Another solution would be to have support for something like SpatialOS that can keep a big, complex worlds running in the cloud. As far as I know it’s mostly aimed at MMOs, but if a stonehearth player wants to run the game in the cloud and have the money to pay for it then it sure would be cool.

1 Like
7

Rather than load and unload chunks based on what the player sees, though, I think Stonehearth could load and unload chunks based on what Hearthlings see - with a few other things, like flowing water or enemies, also loading their chunks. It would still slow down as you had more Hearthlings doing more things over a wider area, but it might still end up being faster.

8

This is probably the most elegant solution. If they did implement that idea though, Inb4 we get complaints because players forcibly “chunk load” the whole map with beacons/watchmen and the game slows to a crawl anyway…

Honestly, I reckon we could make do with smaller maps with denser terrain features – I never get close to filling in the whole map, and I hate it when a cool feature I want to build with (e.g. a bunny statue beside a small pond in a forest clearing) is ages away from the other cool features on the map. Sure I can build something out there if I really want to, but it’ll never get used. If the maps were smaller, it would be easier to justify off-map adventures, and it would mean that things like dungeons and enemy camps would have a larger impact on the game since there’s more pressure on territory/space. Currently, the enemy camps all spawn a (in)convenient set distance from the town so they’re just outside the town’s influence, but as a result they’re not very interesting or threatening – otherwise they’d be unfair if they kept spawning close-by with powerful enemies. If instead they spawned dotted around the map, players would have the choice to avoid them; so they could have more powerful and interesting encounters.

I think that the only idea which benefits from the large maps would be multiplayer, so there’s plenty of room for players to spread out. However, I think there would still be room on a map half the current size for a few towns to exist near each other – near being kind of the whole point, why have a multiplayer game where the towns are on opposite sides of the map? Multiplayer is about the interactions between players, if the towns are isolated from each other it just makes those interactions take more effort.

9

I’m going to disagree with this. I enjoy how the map is, and try to make it a goal to build something here, and there, and then combine them to make a city rather than a small town.

Honestly, when I look at people’s amazing builds on here, I love their artistic talents. I hate though that no one’s thinking outside the box. How many people have roads and not just paths in their towns? How many people actually build trade hubs, libraries, or anything else that isn’t directly part of the game? Thier towns function for what Stonehearth is, but I feel there should be much more.

1 Like
10

Unfortunately, there’s currently no point in trying to fill the whole map with cool buildings – the game simply can’t handle it. If we could reliably do it, then the aesthetic value/sheer awesomeness of such a massive city would be a reason in itself! I love the idea of having far-flung outposts, monuments and grand buildings, ancient ruins, watchtowers and so on all dotting the map and connected by roads, even if half the buildings aren’t occupied or used for anything other than sheer decoration. I’d gladly build a grand library or mighty fortress knowing it will only get used once every few in-game months, if I could be confident that it wouldn’t bog the game down.

However, with the building editor prone to bugging out, and there being a hard-limit on how many items can exist in the world (not to mention a soft-limit on the number of buildings the game can deal with before the memory simply runs out), such massive construction projects rarely get to serve even the most decorative purpose. It takes so much effort to construct them that there’s no time left for gameplay – I’m not even talking about quests and storylines here (if the player is building such massive edifices they’re certainly going to be the kind of player who’s happy setting their own goals); what I mean is that there’s no chance for the hearthlings to live in the town you’re building. They’ll always be in temporary camps close to the latest construction site, or else trekking across the map to get to one/come home for dinner.

Nobody has even filled the current map in using insta-build or other console commands; and with the ability to make templates it’s possible to /IB a hamlet in a few clicks (the entire hamlet can be saved as a single large template, and then variations sprinkled in with each new one after the template is placed.) Even the largest cities we’ve seen have only filled a tiny fraction of the map, and some of those large cities made ingenious use of templates to speed up their construction.

As a final case-in-point, there have been a few chain games where players worked together to build towns and cities all around the map; and in most cases those chain games became too unstable to continue after only half-a-dozen villages were added. There were other factors contributing to that situation, but one of the big disappointments I remember from the one I took part in was that the hearthlings kept wanting to trek away to use a table or chair on the other side of the map, so each city/village had to be walled off during other players’ turns.

Stoneheath simply isn’t built to handle large distances or massive populations. We’re going to need to redefine the concept of a “massive city”, unless we all want to go out and buy supercomputers to be able to play the game. Of course, work continues to optimise the engine to allow larger constructions and larger populations; but even then we know that the expected limit is somewhere around 50 hearthlings. To support more than that would require giving up some of the more in-depth simulation aspects which make Stonehearth unique compared to, say, Gnomoria or Towns or the Impressions games (Zeus, Pharoah et. al.)

I think that from a gameplay perspective, the scale of the current maps only adds possibilities, nothing concrete. However, to access those possibilities, the player has to put up with increased travel time between different scenes of activity, and an increase in the game’s workload to run everything. So, most of those possibilities never eventuate to anything. By reducing the size of the map, we’re only giving up possibilities we never use anyway, in exchange for increased ability to actually do the things we never normally get to try out. And it’s not like we’d be completely giving up the ability to build massive cities, even if the map was 1/4 of its current size it would still be more land than anyone has currently filled up with buildings. If you wanted a single major city in the middle, some impressively long roads winding out from it, a bunch of unique villages and clearly defined regions, and a handful of massive edifices scattered throughout… you could easily fit that on a map 1/2-1/3 of the current map size, with room for wilderness around the edges in which enemy camps and such can spawn.