Multiple worlds / servers?

[Peter]

What plans are there for multiple totally separate game worlds?
There are a few ways I can think of; they could be implemented right into the server program, or distributed to several separate sets of server hardware- which would be more efficient.
Potentially, making the server software itself available (closed-source, of course; open-source is an entirely different discussion) might allow for some interesting systems.

Whether or not to have separate worlds is the real question- it's a legitimate decision to keep it centered on one world. On the other hand, separate worlds, especially when tools are in place to limit who can join (such as playtime requirements: "you must have spent 8 hours playing before you can join this world"), allow players to feel safer from griefers and such.
Other options, such as high-roleplaying servers, allow players to join in like-minded communities without overlapping with other groups (I.E.; no crowds of Urists muttering dwarven obscenities around unrelated Elven towns).

Finaly, there's the option for map parameters to be different in ways that don't work very well for consistent world maps; one map could be grassland punctuated by various other terrains, like the current one, while annother could be normally forest with clearings of various kinds, and annother could be a vast sea with many different islands; and each world could have easy access to a Ring. Of course, large-scale terrain shifts could render this point moot in the face of a large, varied map.

[Yolan]

As far as I know, the reason most big MMOs have multiple 'worlds' is partly because they have many thousands of players. As their worlds are not procedurally generated, making enough land to keep things interesting for such numbers would be a nightmare, hence it is easier to have multiple copies of the same world. Similarly, bandwidth issues caused by such numbers mean that the servers for those worlds are strategically located to cater to different regions.

Frankly, I don't see H&H requiring multiple worlds, for several reasons:

Firstly, even if, in the long run, several thousand players did end up on H&H (and I hope that does happen!), from a design perspective this is not an issue. J&L are working on the concept of a world that is procedurally generated to be elastic. As long as there are hard resources (memory/bandwidth) available, then there will be land. Moreover, the concentric design means that it takes more and more players to keep pushing out the size, as each ring encompasses more land than the previous one. If we had double the number of rings, say, as the current world has, the total land size would be more than tripled I think (might want to check me on this, I've done no calculations).

Secondly, I might be wrong, but I don't see why J&L would want to set up different worlds solely for the purpose of catering to different players gaming desires. As a sand-box style game, it is largely up to the players to decide how things are going to be. Of course, you could have a world where PK is impossible (sounds boring to me), etc., but if people want to play a different type of game, they can just go elsewhere. Unless/until J&L want to start making H&H a profitable business, there is no need to provide multiple visions in order to maximize player numbers.

[Yolan]

I should add, I think it can be profitable on their vision as is!

[jorb]

As long as it is possible it is our intention to have one singular world for all players, as I actually think this does wonders for immersion, lulzfags or not. There are however limits on the number of players our server capabilities will be able to support. I recall loftar mentioning a number of about 300ish players in game simultaneously, given our present server, and assuming there aren't as of yet unknown problems of scale. 300 is about one order of magnitude above the number of people who are online simultaneously today (Our peaks have been in the mid 40 range). It should also be mentioned that our present server definitely is no monster, and that it could be upgraded.

When we can no longer support all our players on one server, we will add new worlds, which will be woven into the mythology, and have unique characteristics. It is also the intention to make travel between them possible. But we're a far ways away from that.

Oh, and, @ Peter -- I don't necessarily want to keep players safe from "griefers", remember?

(And while I sort of like the ambition of supporting a dedicated roleplaying server, I do think that a server like that would require moderators, which I don't like.)

[moonshield]

I hope there won't ever be multiple worlds, if you have to, i hope you will atleast be able to tranfer characters without any restrictions. Though i guess we could just donate a bit for a new server if that happens . And btw, how much bandwith does it take atm (Even if that's not much of a problem in sweden..)?

[Peter]

Hm... 300 is an interesting number of players... most band-and-village societies of history (and to the modern day) have populations that hover around 200-250 people. If they grow much larger, they tend to split.

I agree that this isn't necessary in the short term; in fact I'd prefer a large and diverse world.

[RiverPhoenix]

Hello

Interesting discussion as I've thought about building constantly growing terrain myself (it'd actually be created by the players themselves moving off the edge of the map and created (once) on the fly - if it not visit often then it's not a problem since it'd just be cached on disk and disk is cheap

The reason I mention the above is because in said world I automatically envision that the further south you went the hotter it got until you got as hot as that world does and then back to frozen - north would give you frozen as the ultimate. I'm not sure whether I'd have the world as flat, spherical or toroidal but (since this is all on a computer) I'd probably make it an infinitely growing sphere with (i.e. you could almost but never quite get half way as a player - not that they'd know - in any direction).

As for PK and griefing I seem to have managed to leave some of the griefing behind by going well away from the Ring in the early stages and being in a town seems pretty good protection from PK'ing - as does the tracking stuff (which it occurs to me would lend itself very well to foxes or other animals as well as humans (hunting dog foxes that kill chickens and likewise allow for hunting dogs and the like

Given that the world is already tiled handing off players from one machine to another that each deal with a different part of the world does seem like it would be too much of an issue either (though one or two things might become more complicated e.g. tracking - though you'd just have to send an internal message to each server to find out where the tracked user was when a scent was used.) I guess you could also have different servers dealing with different layers of the system - e.g. avatars streamed to one UDP port, objects to another and the map itself to a third and let the client handle layering them on top of each other. I'd have said different machines handling different tiles is more scalable though - based on usage statistics you might even have them dynamically allocate how much of the world each one manages at reboot

Anyway, enough from me

River

P.S. You can use any or all of these ideas on the basis that you always let me play your game for free or give me a return trip to the Moon, whichever is cheaper! I'm hoping you guys make it REALLY big!

[Peter]

Hello

Interesting discussion as I've thought about building constantly growing terrain myself (it'd actually be created by the players themselves moving off the edge of the map and created (once) on the fly - if it not visit often then it's not a problem since it'd just be cached on disk and disk is cheap

The reason I mention the above is because in said world I automatically envision that the further south you went the hotter it got until you got as hot as that world does and then back to frozen - north would give you frozen as the ultimate. I'm not sure whether I'd have the world as flat, spherical or toroidal but (since this is all on a computer) I'd probably make it an infinitely growing sphere with (i.e. you could almost but never quite get half way as a player - not that they'd know - in any direction)....

Indeed, with a procedural map like the one I mentioned before, it would be unnecessary to save areas that have been unacccesed in a long time; all that would need to be done is regenerate the area with the same seed number next time someone visits. Of course this is slow and causes permanent changes to the area to vanish, so it can't be used to store places with villages and such, but a spot of wilderness that one player walked through, camped in for a night, and continued on from needn't be saved in nearly as much detail as a town.

Stuff like climate shifts can also be done, if hard to define with the current game tiles. Once seasons and weather get in, we can figure out how large climates can change. Part of the problem is that there is actually a very long continuum of climates all the way from arctic through temperate to tropical to desert, and the current engine runs on exclusive climate types.

[theTrav]

I'm also not a fan of multiple worlds, particularly disconnected ones with different rules.

I would have hoped for a distributed model that could scale the current world up reasonably well, or at the least multiple worlds with overlapping borders that are largely transparent to the players

[RiverPhoenix]

Indeed, with a procedural map like the one I mentioned before, it would be unnecessary to save areas that have been unacccesed in a long time; all that would need to be done is regenerate the area with the same seed number next time someone visits. Of course this is slow and causes permanent changes to the area to vanish, so it can't be used to store places with villages and such, but a spot of wilderness that one player walked through, camped in for a night, and continued on from needn't be saved in nearly as much detail as a town.

Exactly, very handy - even if a small amount of "permanent" changes are made they could be stored alongside the seed (e.g. a hut or a drying rack) and replaced if the place goes unused for a while.

Stuff like climate shifts can also be done, if hard to define with the current game tiles. Once seasons and weather get in, we can figure out how large climates can change. Part of the problem is that there is actually a very long continuum of climates all the way from arctic through temperate to tropical to desert, and the current engine runs on exclusive climate types.

If you associate a temperature with the co-ordinate system and decide that say each fifty tiles north changes the temperature down by one degree and have the procedural map make it more likely that you will get pine forests, snow/rock, frozen bog and ice (apple(/oak?) forest, grass/mountain, swamp and lakes) using the temperature's variance from the (temperate) mean as a guide to percentage chance that these other things will be used instead (the hot equivalent could be rainforest, plains/desert, and oasis's).

Just my tu'penny worth

River