Many players of today's virtual worlds may not realise it, but there are actually two strategies for resetting "used" content. The more familiar one is respawning, whereby an object or mobile or puzzle is restored to some variation on its original state after a period of time. Sophisticated versions of this can involve spawning different things depending on the current world state (for example sheep instead of wolves).
The other strategy for resetting is the sudden reset, also known as the groundhog day approach. In this, it's not individual content items (or linked collections of them) that are respawned, but the entire virtual world. The advantage is that quests, puzzles and game-like content can be far more complex and intertwined, because they don't have to be unpicked a thread at a time. World-altering events can occur - cave systems flooding, mines collapsing, volcanoes erupting - which have so many causal effects that it would be impossible to follow every one through to undo it. The disadvantage, of course, is that everyone has to be dumped out of the virtual world while it resets, which inevitably occurs just at the precise moment you really rather wish it wouldn't.
How about a hybrid approach?
One of the "new ideas" of (graphical) virtual worlds is instancing. In this, a group of players get together, press the magic button, and a pocket universe appears just for them. Their characters can go in, explore it, plunder it, whatever, and then when they leave it simply disappears. In other words, it resets when it's played out - classic groundhog day territory. This leads to the possibility that such pocket universes could have much richer and more fulfilling content than the greater virtual world that hosts them. Indeed, if they become large enough they could be considered as bona fide virtual worlds themselves, with the host world a kind of theme park that gives a context to the individual rides that are its sub-worlds.
OK, so this is something which may interest virtual world designers, but it's not really Terra Nova material. Don't worry, it becomes so shortly.
Note that although worlds that reset through respawning can have sub-worlds, groundhog day worlds can't (or rather they can, but the sub-worlds will all reset when the host world resets, so they can never be truly independent). This means we have a hierarchy of worlds:
1) The real world.
2) A non-game, host world (such as SL).
3) A game-like world that resets through respawning (such as DAoC).
4) A game-like world that resets in groundhog day fashion (such as, ahem, MUD2).
Only 1) is mandatory: it's possible to have any of 2), 3) and 4) without the others. The hierarchy is strict, though: you can't have a 2) dangle off a 3), because that makes the 3) itself a 2). You can, though, have a 2) host a 2) or a 3) host a 3).
So we have a tree. We can envisage several non-game worlds, some of which host non-game sub-worlds or game-like worlds that reset via respawn. Not all game-like worlds will necessarily be embedded in a host virtual world, yet some of them will themselves have game-like sub-worlds or groundhog day worlds embedded within them. Not all groundhog day worlds will be necessarily associated with a respawning world: some may be embedded within a non-game world, and others may be stand-alone.
And now, finally, to the point of this post.
Suppose you're heading up a guild that's entered a pocket universe which has been "spawned" from a game-like virtual world. This game-like world has been coded by a bunch of people, open-source fashion, to run in a generic environment operated by a hosting company based in the USA. Code permitting, who gets to say what one of your guild members can and cannot do within that pocket universe?
Richard
Not really sure what you question has to do with instancing.
Besides:
* Instancing doesn't imply resetting.
* "Instancing" is just a conceptualization. You might view it as an infintite static no-reset world. People just loose the keys to the doors, preventing them from going back.
Posted by: Ola Fosheim Grøstad | Jan 03, 2005 at 07:24
>Not really sure what you question has to do with instancing.
Instancing allows for worlds-within-worlds that are genuine sandboxes. Other than character stats and the acquisition of a few objects, nothing you do in them has any effect on what goes on outside them because they're non-permanent.
>* Instancing doesn't imply resetting.
They do if the instance is non-permanent, otherwise when you entered the instance months later it would be just as played out as it was when you left it.
>"Instancing" is just a conceptualization. You might view it as an infintite static no-reset world.
Well "infinite", "static" and "no-reset" are just conceptualisations. What's your point? The fact that the content is gone when the players leave is what's important, not the name "instancing".
Richard
Posted by: Richard Bartle | Jan 03, 2005 at 12:11
In my ravings around May, after CoH was launched, I was one of the few strongly criticizing the use of instances. But this specific to a point of view: virtual words Vs an arcade single/cooperative approach.
My point is that this approach isn't wrong if you deliberately choose *that* strategy and *that* path. I believe that PvE (in its broader meaning) is strictly single/cooperative play. Today we have a market that is considered to lean strongly toward PvE.
Now these facts all produce together the same result: strongly instanced (PvE) games are successful because they know perfectly what they are. Their nature. They are bringing back their gameplay where it belongs -> to a single player/cooperative experience. So they work.
As I wrote long ago these strongly instanced games are good and successful because they are NOT mmorpgs. They know this and they do not "pretend" to be something else and offer gameplay outside of its proper space.
Because, again, PvE (from my point of view) intersects with single player/cooperative gameplay. They are a single entity that simply doesn't work if you try to separate the two parts.
Now, my point of view is that it is never "wrong" to use something. It depends on how you use tools, like instancing. In this case I believe that we can lean toward trying to add more experiences to a mmorpg. So that we have a complex world, a meaningful implementation of PvP AND also a PvE part, where instancing can be used to deliver the best quality possible.
What I mean is that instancing "delivers a lot" in PvE exactly because it's a "going back where it belongs". What can be a choice is about integrating this part (PvE) with something else that incentivates what makes this genre different. Adding different possibilities, so that instancing becomes a way to offer what is pertinent to it, leaving another part of the VW to work under different rules and strategies.
Now all this is weakly tied to what you write here but it's my way to explain what instancing is and how it can be used. In particular your last example seems near what already happens in Neverwinter Nights. And it's here that what I write plugs in the discussion.
P.S.
From a design point of view I also believe that the "resets" aren't the main strength of this technique. What is relevant is that you can choose and fix a range of levels (power) or a number of players, so that you are able to (finally) offer a challenge. Where, instead, the current mmorpgs trivialize every attempt at delivering a decent PvE exactly because you can rip off the starting conditions, bringing to the encounter more players or letting in a strong character that powerlevels everyone else. Instancing simply adds *more control* in the hands of the creator. This means that we are less "sandbox" and a lot more content-driven experience.
And it's here again that we discover what instancing is and how it should be used.
P.P.S.
Instead what you write about using instancing to produce a more interesting world (and more control again, but to create something that can be more dynamic and more original). That's a particular kind of potential that I wouldn't limit to PvE. It's something that can be used for PvP and so completely unrelated to instancing.
Posted by: Abalieno | Jan 03, 2005 at 15:40
> nothing you do in them has any effect on what goes on outside them because they're non-permanent.
That's more or less true for respawning in MMOs too.
Anyway, you question could easily have been applied to the collective of LPMuds. Either that, or I don't understand the basic question...
>>* Instancing doesn't imply resetting.
>They do if the instance is non-permanent,
I meant to say instancing doesn't imply a reset strategy. You can have instancing with respawn, full reset, or no reset (but possibly evolution), fork'ing etc.
Posted by: Ola Fosheim Grøstad | Jan 03, 2005 at 16:01
Just from a historical point of view, there was a period wherein muds transitioned away from groundhog day and towards individual staggered resets. Early on, the way this was done was by resetting entire zones all at once. This had some big advantages in terms of quest and encounter design--the mud never reset the area if people were in it, you see, so you would camp ayt the edge of the area, and thus be guaranteed to have the experience of the area in a particular order. Call it primitive instancing.
It wasn't true groundhog day because the rest of the mud was not reset... this was quite common in Diku mud derivatives in the early 90s.
To address Richard's actual question--I would imagine that the hosting service will have a ToS, and then the creator/administrator of the zone in question may have one as well, assuming that the hosting service granted him any abilities!
Posted by: Raph | Jan 03, 2005 at 16:36
I had an hard time to figure out what was the point of what Bartle wrote but I think I "got" two concepts:
1- Instancing allows more control over world design, so the possibility to script and deliver a lot more dynamic stuff that affects more directly the world. A deeper interaction and dynamic.
2- The potential in (1) could be used to offer the players themselves a "custom" sandbox where they can create (and modify) the content themselves.
That's why I wrote it smells like Neverwinter Nights.
(then I could have also completely misunderstood everything...)
Posted by: Abalieno | Jan 03, 2005 at 16:39
It seems to me like any kind of instancing or resetting implies a very unrealistic world model. Having a world that sizes itself based on how many players are in it (ie adds more depth to a quest tree in order to accomodate more players looking for new quests/areas to explore) seems like a better way to work this content issue. Then you'd have to improvise on the "if a tree falls in a forest and nobody is there, does it make a sound" concept. Areas that have been played out and aren't interesting anymore could auto-reset when nobody's been there for a while, while nobody is looking.
Posted by: Ivan Tumanov | Jan 03, 2005 at 16:58
Abalieno> Instancing simply adds *more control* in the hands of the creator. This means that we are less "sandbox" and a lot more content-driven experience.
This might describe how many use it, but it isn't a property of the act of creating instances... Instancing + _exclusive access_ does give the creator somewhat more control along one dimension, but then you loose something as a trade-off, so it really depends on what you want to control? _access control_ does of course allow you to make more assumptions when you design, i.e. you control the pre-condition, thus you can get away with making the overall design less dynamic.
What instancing basically provides is an easy way to make static designs scalable (avoid crowding). With emphasis on "easy". Just about everything you get with instancing can be handled without instancing (ownership, balanced challenges etc). It just requires a more well thought out design.
Instancing is primarily a cop-out that allows designers of single-user games to cotinue doing what they know how to do in a MUD context. Instancing may seem dynamic, but it lends itself towards more static design stratgies... (those we know from single user games) Instancing has traditionally been used for generated mazes, but even though generated mazes are different, the overall design tends to be pretty static. I.e. what happens in them doesn't really change all that much.
Now, the flow of this argument may not be so apparent today, as most designs are largely static (instanced or not)...
Posted by: Ola Fosheim Grøstad | Jan 03, 2005 at 17:04
Wait a second.
For me "control" is about knowing what to expect. "Instancing" means control because you can define strictly who is going in, how, when etc... So you can tailor the content exactly for that audience you expect. -> exactly like a single player game, where you know that the player won't get the bazooka till the next level. (for example)
What if you could start "Doom" already with the BFG? This is what often in mmorpgs: content broken and trivialized because the devs lack the control over the world itself.
I agree that what instancing does can be achived without it. But it's still a good tool when you know how and when to use it and I really don't understand why NOT use it if it makes something easier (like you say).
So for me the point is simply how and when use it.
> Instancing may seem dynamic, but it lends itself towards more static design stratgies... (those we know from single user games)
That's exactly what I mean that instancing is perfect for PvE and that PvE = single player.
It works because it is used properly.
Games like World of Warcraft, Guild Wars and Coty of Heroes are successful (for a good part) because they offer an expereince that the players know already (single player/cooperative), while they don't offer much as virtual worlds.
But they KNOW this. And they do well what they choosed.
Posted by: Abalieno | Jan 03, 2005 at 17:21
Abalieno> For me "control" is about knowing what to expect. "Instancing" means control because you can define strictly who is going in, how, when etc...
Well, but this is a property of _access control_ not a property of instancing. Instancing might make the imposed constraints less visible for the users though.
> So you can tailor the content exactly for that audience you expect.
Yes, we agree :-).
> I really don't understand why NOT use it if it makes something easier (like you say).
I haven't said that designer's shouldn't use it. I think many users find small-scale controlled experiences refreshing. Instancing to avoid exessive crowding is also a good approach, at least better than throwing out users at random.
>So for me the point is simply how and when use it.
Right. :)
Abalieno> Games like World of Warcraft, Guild Wars and Coty of Heroes are successful (for a good part) because they offer an expereince that the players know already (single player/cooperative), while they don't offer much as virtual worlds.
Well, at the moment they are successful because they are new, but maybe time will prove you right. :-) Maybe this is the most efficient use of human resources at the moment. Translating from single to multi-user is obviously easier than reinventing multi-user.
However, if we can imagine the ideal MMO design then I doubt that the controlled experiences is it. Spontanous interaction with the environment and other users is something special.
I'd tie the concept of controlled design to the grind. The grind is basically killing the more free form exploration potential as players have an idea of how fast they "should" level.
Posted by: Ola Fosheim Grøstad | Jan 03, 2005 at 18:02
Who gets to say what your guildmates can and cannot do? Sounds like ripe conditions for the meta game of guild social politics, which can be entertaining in and of itself.
Ideally, if your guild is comprised of mostly like minded people, you (assuming you are nominally "in charge" in some way), could let the crew have at it and end up with a pocket world that suited the group pretty well.
And as previous posters had mentioned, this is just about what many were hoping for from NWN. A shared host world that has links/portals to ever changing instanced game worlds (modules). And that these modules could be created by players and made available to the general public. It already happens to greater and less degrees with some of the NWN attempts at persistant worlds, but as NWN wasn't really designed to support that it has been an uphill battle with custom content and players/DM/world limitations.
Personally I would love a NWN3 (since the plans for nwnw2 are decidedly underwhelming), have a more generic game engine not married to the D&D ruleset exclusively, that better supported these efforts.
Xilren
Posted by: Xilren | Jan 03, 2005 at 18:08
Raph> It wasn't true groundhog day because the rest of the mud was not reset... this was quite common in Diku mud derivatives in the early 90s.
I believe it was quite common for several MUDs to reboot several times each week to reduce the memory footprint, and get rid of accumulated effects from bugs and so on. In LPMuds the object designers control the reset on the object class level so you can have whatever reset policy you want.
It is interesting to note that ground hog day resets are quite common on large scale web-based games. Proper resets: wiping the characters too.
Posted by: Ola Fosheim Grøstad | Jan 03, 2005 at 18:16
> Well, but this is a property of _access control_ not a property of instancing. Instancing might make the imposed constraints less visible for the users though.
Hmm yes. You could have a cave and put in front of it two giants that let pass only players of a specific level range. This works as a way to control what's inside without using the instance. Or you could set the cave to collapse if too many players go into it.
As I said instancing isn't the only way to offer its mechanics, but it works better and easily. In the example above you'll get an infinite queue of players outside the cave, waiting to go in. So you can build a tons of similar caves (like WoW is doing with quest-resources, building many different spots). It's again an instancing-process.
What I mean with *control* is about the design side. A content builder will know that in that cave there will be (let's say) three characters grouped together, around level 15. Limiting the access means that a dev can control who experiences the content and how. This allows the same dev to plan and deliver that content specifically tailored for THOSE conditions. So he can offer a challenge and a compelling experience because he knows exactly the "flow" of that part he is creating.
There are other solutions. For example I think Vanguard is trying another path that I personally dislike. Instead of restricting the access they want the content to dynamically adapt itself to the conditions.
So if there are more players, more monsters will spawn, with more HPs and so on.
Personally I find this practice even worst. This goes even more directly against the virtual world. It breaks the self consistency, it isn't believable and it offers a world that always shifts in front of your eyes following Out Of Character rules. It cannot be "mapped", it cannot be understood and known. At the end it produces frustration and it is basically unfun.
Without even considering that it's also way, way harder to develop dynamic content than something specific that you can study and script to deliver a specific, known result.
Posted by: Abalieno | Jan 03, 2005 at 18:34
Abaliano> Personally I find this practice even worst. This goes even more directly against the virtual world.
Depends. In AO the alien ships are configured based on how many players are in town during an attack. That could be explained by "an alien commander" trying to make efficient use of his troops etc. However, in my opinion the designer's "presence" can be visible in even the most static designs. I find the whole treadmill with levels on mobs, spawns and what not disturbing. Instancing doesn't make it better...
Posted by: Ola Fosheim Grøstad | Jan 03, 2005 at 20:46
Assuming the question is related to what the characters can do, rather than the ToS regulating the player's behavoir, I would assume that the guild members are now subject to the rules of the pocket universe. Whether these rules complement or replace the rules of the generic environment depends on the scripting model provided by the generic environment. It would seem to make sense from a design perspective for the rules in the pocket universe to be the same as those in the game-like virtual world it was spawned from, as this spawned relationship implies that it is a part of that game. If the generic environment has been designed explicitly to support game-like virtual worlds to be created it seems likely that the coders of the game-like virtual world should have the freedom to override and replace the rules of the generic environment.
Posted by: Al Riddoch | Jan 03, 2005 at 21:36
What I mean is that with dynamic encounters we are just one step away from cheating: "Look, this group is zipping through our cave, let's spawn a big troll and wipe them all out".
Players may try to return to try again and for each attempt they'll always find different conditions, with the game "cheating" actively so that the difficulty is arbitrarily balanced.
When I go in a place I like to think it's something concrete, that doesn't vanish as I look elsewhere to then shapeshift completely the minute after.
This feels everything but a virtual world. It's just a silly trick that upsets me.
Posted by: Abalieno | Jan 04, 2005 at 00:14
Also, the sense of realism follows set rules.
The "suspension of disbelief" works till you give a bunch of set, fancy rules and then play within them.
You believe to this layer till the set rules don't change, till you are able to dip yourself into the roleplay (the layer of the game, even if the game is just a formal system) and start moving the interactive elements in a way you expect. Raph would say: till you are able to identify patterns.
A ruleset that shifts its rules constantly breaks this basic principle of fun and produces *directly* frustration.
Posted by: Abalieno | Jan 04, 2005 at 00:18
Richard Bartle wrote - Suppose you're heading up a guild that's entered a pocket universe which has been "spawned" from a game-like virtual world. This game-like world has been coded by a bunch of people, open-source fashion, to run in a generic environment operated by a hosting company based in the USA. Code permitting, who gets to say what one of your guild members can and cannot do within that pocket universe?
I'm not exactly sure about why you're asking that specific question, but some clarification questions pop into my head:
How does the guild control what users can and cannot do? "Code permitting" reads as a bit ambiguous. Does the guild have control over the code? Or do they have pre-programmed abilities like the right to admit and banish characeters? Do they have their own "laws" that guild leaders enforce with banishment, fines, etc.?
If the guild can control what users can do in their pocket universe, can the guild charge a virtual admission fee to their private sub-world?
If they can charge a fee, can it be in real money? How much of the fee goes to the meta-world host?
As soon as a guild can create a pocket universe that they have control over, they won't want it to ever die and will whinge to the developers until the pocket universe becomes a spawning universe in its own right. The pocket world becomes more of a private theme park or zoo. So how does this fit into groundhog day?
Posted by: Mike Rozak | Jan 04, 2005 at 01:48
Abalieno, you need consistency on the user interface level. On the world level you need to find a balance between predictable and unpredictable. A balance between known and unknown. This is of course the great thing about multi-user systems: the other users contribute a level of "unknowness" to the system. At least until the gameplay has gone stale (i.e. you can play without communication).
It is the unknown, unexpected and unpredictable that makes the world into more than a scenery.
That's what I dislike most about quest based designs: you end up with scenery, not a world.
Posted by: Ola Fosheim Grøstad | Jan 04, 2005 at 04:38
A bit off-topic…
I watched Memento (by Christopher Nolan) again a few days ago and this discussion of resets reminds me of that film. Resets there (local to the main character) are a narrative technique, which is interesting in itself. But more than that, the film explores the idea of how an individual’s free action would develop given a particular character composition, some strong motivations and constant resets. So, I do not think that resets are necessarily taking away from, eh, realism. To me it is an interesting theme: how our actions and thinking depend on temporal continuity and all it brings: e.g. relationships, dependencies, memories, assigned values.
Sorry, this is nothing to do with Richard’s actual question…
Posted by: Stanislav Roudavski | Jan 04, 2005 at 06:36
I knew it was a mistake to describe this from the point of view of instancing. I didn't intend for this to get into a discussion about the various merits or otherwise of instancing, or of the definition of what instancing is. Still, whenever I post something there's always an immediate tangential response, so I shouldn't be surprised.
Coming back to what I meant the post to be about:
Al Riddoch>I would assume that the guild members are now subject to the rules of the pocket universe.
Yes, but what are these rules?
>Whether these rules complement or replace the rules of the generic environment depends on the scripting model provided by the generic environment.
I mean rules of behaviour, rather than the more "physical" rules of the pocket universe. If the universe don't let you fall through floors, you can't fall through a floor whether you want to or not. If the universe lets you beat another character over the head with a mallet, that's within the rules of the pocket universe but is it allowed by the guild? By the host game-like world? By the superhost non-game world? By the real world? Who gets to say, and why?
Mike Rozak>How does the guild control what users can and cannot do?
I'd assumned like any guild. Initially through sanctions ("no guild property for you if step out of line") and ultimately through membership control ("that's it, you're out!"). I suppose some guilds could have other mechanisms (fines, violence, mockery etc.).
>"Code permitting" reads as a bit ambiguous. Does the guild have control over the code?
No, they don't. I said "code permitting" in the hope that this would stop the discussion heading off in the direction of what the code should allow players to do. I'm specifically thinking about what players can do that the code permits, not whether the code should permit them to do it or not in the first place (which is a different, albeit related, issue).
>If the guild can control what users can do in their pocket universe, can the guild charge a virtual admission fee to their private sub-world?
In theory, they could charge for membership of the guild even if they weren't in any kind of a sub-world. High-powered EQ guilds could charge their members an entrance fee if they liked (and they probably wouldn't be breaking any EULA if they did).
>The pocket world becomes more of a private theme park or zoo. So how does this fit into groundhog day?
Because if they did create their own pocket universe of the kind you describe, then when their membership grew and people started treading on one another's toes again, there would be requests for an instancing feature again.
Richard
Posted by: Richard Bartle | Jan 04, 2005 at 06:50
I don't understand how this relates to instancing. Guild events often take place in "subworlds" already, so what's the difference? The rules are that you behave according to group values and if not: you're out. If guild leaders don't respond then good members leave as they don't want to deal with shit in their leisure time... Where's the question?
Posted by: Ola Fosheim Grøstad | Jan 04, 2005 at 07:33
The civil structure that spells out the do's and don't can be similar to the structures used in IRC, parking lots, sand lots, parks, etc.
You can have Majority Rules, Might Makes Right, First Person In, Vote With Your Pocket, etc.
More likely the guild with the more resources will dominate and dictate the rules of acceptable behavior in a given pocket world.
Of course real life organizations will trump any online organizations. The webhost will dictate the use of their resources. Guilds will hide under the facade of a standards-making body with a non-profit org. registered to receive tax-deductibe donations. So on, so on...
I might reading too much in to the question Richard, but I think you are trying to raise the subordinate government question.
God save the Queen and Monarchy rules!!!
Personally, I like democratic dictatorship to rule.
Posted by: magicback | Jan 04, 2005 at 09:53
I don't see democracy working too well in what Richard is describing. Players will vote in favor of the dungeon with the 'free XP' button every time. Even assuming they aren't that corrupt, it's almost impossible to balance an MMO when you have designers who understand the formulae and systems underlying the game.
One idea that I was tossing around with friends a while ago was the notion that players could craft dungeons, but they were in a 'dream state' - i.e. all rewards in the dungeons couldn't be taken with you. Once enough people had visited the dungeon in it's dream state (meaning, it was fun enough to draw people without rewards), a paid employee would be notified. He then could review the Dream State Dungeon to ensure that it was balanced, fun, and not completely destructive to the flavor of the game. If it passed all of these tests, the dungeon could graduate out of Dream State and into 'reality'.
Even then, you'd want some pretty good data mining tools watching these worlds so you can correct anything that managed to get snuck by the paid employee.
Posted by: Damion Schubert | Jan 04, 2005 at 11:17
Another structure is similar to D&D concept of connected planes where the properties of each plane can be categorized and sorted. The Owner, determine by a number of different ways, determine the properties and acceptable behavior.
So in the guild example, the guild leader is deemed the Owner and is given the rights and coded power to determine and manage behavior. In essence you have a platform for a Guildmaster-DM to run pocket/instanced campaigns for their guild/gaming group.
Posted by: magicback | Jan 04, 2005 at 12:49
Abalieno> Instead of restricting the access they want the content to dynamically adapt itself to the conditions.
While the zone designs were the same, I believe Diablo 2 scaled encounters to the dynamics of the group that entered them. GW could do this too, though I don't know if they are. And I don't think this is necessarily a bad thing. However, I do feel scaling the encounter may be best for one-time-use encounters.
Instancing can do that too. Microsoft's Mythica featured dynamic zone modification during E3 2003, showing bridges getting destroyed and what not. They didn't have a solid answer yet on how they planned to handle replayability of the content when I asked though :) The challenges here seem to be two fold:
1) Repeatable content breaks immersion. If the bridge is destroyed, yet magically fixed next time you enter, then it's become just a "game" to beat again more efficiently.
2) Non-repeatable content is an expensive resource investment for the amount of time it keeps the player interested.
One approach is to treat the zone design and the content therein separately. The first time players complete an instance, they've destroyed the bridge they crossed. The second time, they need to interact with the content at the bottom of the chasm they previously used to bridge to cross. The closest reference I can think of is Ender's repeating dream in Ender's Game. Different "game" each time he entered a dynamically modifying "virtual space".
But ultimately, this explores a technical challenge without getting to the "meat" of the article, or at least what I garnered from it:
Player directed content.
Guild-managed pocket spaces are not that far, mentally, from GM-directed D&D modules. While I wouldn't advocate an NWN-level of tools in a pocket space (since design and fun are hard and what-not), *some* control over content within a pocket space may be extremely compelling. It could be as simple as controlling the spawn of an area based on what a guild builds there. SWG does this, though player structures seem to affect mob spawn trigger locations rather than their changes of occuring at all.
Posted by: Darniaq | Jan 04, 2005 at 14:58
Finally!
That's my point as well. And it's linked to what Ubiq wrote recently:
http://booboo.phpwebhosting.com/~ubiq/index.php?p=182
Things like controlling spawn points are exactly "toys". What I underlined here is that in this case we loose the whole potential if this is instanced and limited to PvE. Playing with toys is way, way more fun (but also harder to design) in a PvP environment.
This BEGS to be used in the world. Not the instance. And this for the simple reason that:
PvE -> instance -> single player
PvP -> world -> cooperative
I also underline that, in my definition, PvP is cooperative, not competitive. All the cooperative activities, also when NOT involving combat, are PvP.
Posted by: Abalieno | Jan 04, 2005 at 16:42
so this is something which may interest virtual world designers, but it's not really Terra Nova material.
Well I'd be happy to see the TN remit expand to cover the discussion of game design. It seems a shame to have you on board and not to.
World-altering events can occur which have so many causal effects that it would be impossible to follow every one through to undo it.
These are exactly the kind of causal connections that the sufficient causal ordering in MASSIVE-3 modelled. Combined with the event recording used for temporal links it would be relatively easy to unpick even very tangled causal connections. Another PhD?
Posted by: Jim Purbrick | Jan 04, 2005 at 19:17
Ah,
MASSIVE-3 was the model I was thinking about in regards to structured space for distributed control.
The RPG I was thinking in terms of the design for allocating control was Nobilis.
Posted by: magicback | Jan 05, 2005 at 01:11
Ola Fosheim Grøstad>I don't understand how this relates to instancing.
Instancing is just the bottom level of the hierarchy. I only mentioned instancing to make the point that there is a terminal node to the tree - a kind of virtual world that doesn't have sub-worlds. This wasn't meant to be a post about instancing and only instancing.
>Guild events often take place in "subworlds" already, so what's the difference? The rules are that you behave according to group values and if not: you're out.
So if you're in a sub-world and make racist remarks about a well-known Secretary General of the UN, that's OK if your guild says it's OK?
Richard
Posted by: Richard Bartle | Jan 05, 2005 at 03:05
magicback>I might reading too much in to the question Richard, but I think you are trying to raise the subordinate government question.
I am, yes. The difference is that at each step in the hierarchy the people who "govern" may not want to govern, but are obliged to by the people above them in the hierarchy. A bunch of loose-knit solo developers working in concert to create the virtual world that the guild's instanced world is spun off from probably couldn't govern even if they wanted to, but if they're regarded by the operators of the hosting world as a useful buffer between them and the excesses that guild members may get up to in a pocket universe then they may find themselves with responsibility they don't want that comes with power they have only intermittently.
Richard
Posted by: Richard Bartle | Jan 05, 2005 at 03:10
Jim Purbrick>Well I'd be happy to see the TN remit expand to cover the discussion of game design. It seems a shame to have you on board and not to.
If we discussed virtual world design for its own sake here, it could completely change the character of the site. Were designer types (eg. MUD-DEV readers) to find sufficient material here to visit regularly, they may well post regularly. They're unlikely to be as shy about posting as the specialist social science lurkers we have at the moment.
>These are exactly the kind of causal connections that the sufficient causal ordering in MASSIVE-3 modelled. Combined with the event recording used for temporal links it would be relatively easy to unpick even very tangled causal connections.
I see two potential problems with this.
Firstly, the bounds of "causal" can be quite strained. Suppose there was a small island that could only be reached by a specialist kind of boat. The boat can only be obtained from a cave, guarded by a dragon. When you want to reset the dragon, you also want to reset the boat and the island (alternative: when you want to reset the island, you also want to reset the boat and the dragon). What if there are three boats, though, each with their own cave and guardian? Are they all part of the same "puzzle"? What if the boats could be used for much more than just sailing to this one island?
Secondly, although it may be possible to track every consequence of an event, it's not always possible to undo it. If there is someone sitting in a boat that you want to return to its spawn point, what do you do? Drown the person? Make a copy of the boat (until eventually everyone has their own boat)? When an event has many, many cosnequences the chances that people are standing on the threads you want to unpick are pretty good. Do you partially reset? Or do you continually check until every single thread is clear (which may never happen)?
Richard
Posted by: Richard Bartle | Jan 05, 2005 at 03:25
Richard> So if you're in a sub-world and make racist remarks about a well-known Secretary General of the UN, that's OK if your guild says it's OK?
Yes, why not? It is not in public if it is guild restricted.
Posted by: Ola Fosheim Grøstad | Jan 05, 2005 at 04:19
Hmm, when I think about it. Maybe it is easier to discuss a case? In Active Worlds there was a community with their own worlds which followed the "philosophy of Gore". A simplified version is something like this: women are only free if they are slaves to men. I.e. women should be slaves and men should be rulers. To what extent this was roleplay or real beliefs varied.
The problem is, of course, that many people just cant stand the idea of "women being abused". So even if it was up to others whether they went to the Gore worlds or not, their presence wasn't as such seen as irrelevant for other users in the parent community.
Should the followers of Gore be allowed to have their worlds within the universe, or should they not? Is the mere possibility that some women might be molested by some violent man after establishing contact within the system be enough to deny the more philosophical/role-playing Gore followers their space?
In the end you end up with the problem "does it matter if your neighbour looks at animal porn in his closet". It doesn't matter if you are not particularly aware of it, but if you are, maybe it does?
Posted by: Ola Fosheim Grøstad | Jan 05, 2005 at 05:15
Richard> When an event has many, many cosnequences the chances that people are standing on the threads you want to unpick are pretty good. Do you partially reset? Or do you continually check until every single thread is clear (which may never happen)?
Doesn't this topic belong more to proper computer science than game design? You can obviously design provably dead-lock free systems with partial resets, but to do so you have to lobotomize the design (and make sure all designers have very high comp. sci. skills as dead-locks are tricky beasts). I am not sure if designers should try to reinvent everything that has to do with processes, databases, temporal logic etc... Sounds like a lost battle.
However, it isn't all that difficult to come up with reset strategies wrapped up in fiction when you detect dead-locks. E.g. an area is invaded by trolls scaring away all the players. So you could accept a design with occasional deadlocks, provided you have dead-lock detection and a full reset back-up strategy.
Maybe the more promising reset stragies are PvP realm-based: players of one side are undoing the acts of players of the other side. If you lack players on a particular side then you can use NPCs.
Posted by: Ola Fosheim Grøstad | Jan 05, 2005 at 06:31
Me> So if you're in a sub-world and make racist remarks about a well-known Secretary General of the UN, that's OK if your guild says it's OK?
Ola Fosheim Grøstad>Yes, why not? It is not in public if it is guild restricted.
OK, so what if it's a guild of paedophiles discussing ways to abduct children? Or a guild of hackers comparing ways to bring down the server? Or a guild of employees of a competing virtual world who are aiming to make this one run as slowly as possible?
Do you feel there is nothing that can go on in the pocket universe which should be prohibited?
Richard
Posted by: Richard Bartle | Jan 05, 2005 at 07:11
Richard> So if you're in a sub-world and make racist remarks about a well-known Secretary General of the UN, that's OK if your guild says it's OK?
Ola Fosheim Grøstad >Yes, why not? It is not in public if it is guild restricted.
I guess this depends on the top level of the hierarchy – how the physical world laws impact things. A short time ago someone in the US was locked away for 6 months for saying that if they had a gun they would shoot Prez Bush, if you are in the US and your server is in the US and you start saying things like this I think the authorities would get very interested, even if you were playing a game called 101 ways to kill the president, having said that I’m not sure what formal responsibility a guild, game company or ISP would have for regulating this sort of thing.
Posted by: ren | Jan 05, 2005 at 07:14
http://www.killhim.nu/ ("nu" means now, english explanation)
In Norway you can get away with it for a while even in public... That is... until the American embassy starts to whine.
Btw, you can get prosecuted in Norway for something running on US servers if the content is having Norway as the target... or so I've heard.
Richard, if I know that someone are planning a crime then the law requires me to report it to the police... so I am not sure what options I have in that case?
Posted by: Ola Fosheim Grøstad | Jan 05, 2005 at 07:45
Ola Fosheim Grøstad > the "philosophy of Gore". A simplified version is something like this: women are only free if they are slaves to men. I.e. women should be slaves and men should be rulers.
Ahh, Gore, if it did not exist someone would have to invent it just to discuss ethics.
One small point on Gore - I think [any practicing Gorians, I ask your favour, feel free to help me out here] that women can be Free without being slaves – though if I remember correctly there is still a strong gender asymmetry as Free women are celibate whereas Free men tend to be far from being so – though of course in many online Gorian spaces a woman is free to role play a male and the other way round so the asymmetry is in terms of characters rather than people.
Though this brings up an interesting point as the practice of being Gorian is an interpretation of a set of novels so a purist interpretation might say that women cannot RP males because the whole point is based on actual differences between the sexes, whereas a more RP approach would emphasise the sexual aspects of Dominance / submission which is not gender specific as such. Thus in this case one could say that the interpretation is very much down to the group /guild that is running the servers.
>Should the followers of Gore be allowed to have their worlds within the universe, or should they not? Is the mere possibility that some women might be molested by some violent man after establishing contact within the system be enough to deny the more philosophical/role-playing Gore followers their space?
As a practical matter this might come down to the way that it is done – if there is a big sign saying: this is Gore and this is what you are getting into, it is a different matter from if there is not and one can easily wander in accidentally – and again its male and female presenting avatars we are talking about.
Ola Fosheim Grøstad > In the end you end up with the problem "does it matter if your neighbour looks at animal porn in his closet". It doesn't matter if you are not particularly aware of it, but if you are, maybe it does?
This is a fairy stock ethical question that depends on what code you happen to follow – if you are a consiquentalist and there are no negative consequences then, no, it does not matter; if you are a deontologist then you have to pick your rights and duties – if free speech is top trump then again its probably fine, if rights of animals not to be abused is tops and you think that animal p0rn constitutes abuse then its not etc etc.
Posted by: ren | Jan 05, 2005 at 07:51
ren> if there is a big sign saying: this is Gore and this is what you are getting into, it is a different matter from if there is not and one can easily wander in accidentally – and again its male and female presenting avatars we are talking about.
Well, it is interesting that you think that "gender-bending" should somehow make a difference. Why?
I am not so sure if the sign helps much. Teenagers have access to the system for instance. Knowing what is going on would most likely encourage exploration. Not that you could hide this from teenagers anyway...
ren> This is a fairy stock ethical question that depends on what code you happen to follow
Yes, maybe the Gore example is better.
Anyway, for the owner of the parent system it gets more complicated as he has to take all the "codes" into consideration. It is very unlikely that the Feminist guild is going to accept the existance of the Gore guild. They would very likely see it as a threat against teenage girls with low self-esteem and feel inclined to rescue the suppressed women.
(Teenagers experiment and people with low self-esteem feel "secure" in a sub-role, but may have problems breaking out of the role due to various types of fear. So insecure women can be victims of bad doms.)
Posted by: Ola Fosheim Grøstad | Jan 05, 2005 at 08:39
Trying another analogy: if a "cured" rapist moved into your building would you prefer not to know about his past or would you prefer that to have a big sign on his door with the text "rehabiliated rapist, not dangerous" written on it.
I would prefer not to know.
I am not so concerned about the morale in the virtual worlds context. I am more concerned about "what makes for a pleasant experience". I assume most people would prefer not to know what the Gore guild was up to or at least not have it broadcasted, while the Feminist might want to exercise their "free speech" and get access to the Gore world so they can reach the "suppressed women".
Posted by: Ola Fosheim Grøstad | Jan 05, 2005 at 08:50
s/Gore/Gor/
Sorry :/
Posted by: Ola Fosheim Grøstad | Jan 05, 2005 at 10:27
Richard> I am, yes. The difference is that at each step in the hierarchy the people who "govern" may not want to govern, but are obliged to by the people above them in the hierarchy.
They may be obliged, but they can effectively ignore any demands. The world could be distributed on the net like P2P or be hosted in a country with lax laws. Perhaps off-shore tax havens will find beneficial to be a net haven too.
However, democratic and consensus ideals appear to be predominate among virtual worlders, so organizations/guild members will generally self-organize into a governing network.
In essence, my point is that one can visualize the internet as a virtual world. Illegal sub-worlds in one jurisdiction will find a home in another jurisdiction. Rogue copies will spawn and respawn. You can map the hierarchy and exert control by their physical network connections, but with future wireless network infrastructures the P2P connections you won’t be able to exert RL controls. So it is back to the democratic and consensus ideals that predominate.
Virtual worlds with their shards, instances, resets, respawns are quite democratic and free.
Posted by: magicback | Jan 05, 2005 at 10:29
If we discussed virtual world design for its own sake here, it could completely change the character of the site.
OK, fair enough. I'd quite like somewhere that sat between TN and MUD-DEV though.
When you want to reset the dragon, you also want to reset the boat and the island.
So you define the boat(s), the dragon and the island as the items you want to reset as a group.
If there is someone sitting in a boat that you want to return to its spawn point, what do you do?
In MASSIVE-3, the player getting in the boat would be the first event sequenced on both the boat and the passenger. When reseting the boat this event and all other causally connected events following it would be undone. The player would find themselves standing on the shore where they got on the boat.
I could imagine some event, like completing a quest, which would sever the connection between the player and their causal past, so that only players half way through the quest would be reset with the boat.
Posted by: Jim Purbrick | Jan 05, 2005 at 10:45
I don't understand how that is different from a full reset (i.e. rollback). The main problem with recording events and undoing them is that you have to do the undoing in reverse order which effectively prevents more than one person doing the quest? What you want is to have a staged resets that goes forward so that you can have say 20 different people progressing (and helping each other) through the quest at the same time. That's what respawning tries to do...
How would you deal with a player that logs off and then back on?
Posted by: Ola Fosheim Grøstad | Jan 05, 2005 at 11:18
Jim Purbrick>I'd quite like somewhere that sat between TN and MUD-DEV though.
Perhaps, but I worry there's still a danger we'd drive away the social scientists and lawyers who frequent these pages, so they'd have to set up some alternative blog to air their views.
>So you define the boat(s), the dragon and the island as the items you want to reset as a group.
Yes, but there may be things on the island that in turn unlock other areas, or that can be accessed in other ways, or that have multiple purposes, and so on and so on, until eventually you have to group the whole world together as something to be reset as a group.
>The player would find themselves standing on the shore where they got on the boat.
But what if they got on the boat from another boat that was itself reset some time ago?
Richard
Posted by: Richard Bartle | Jan 05, 2005 at 11:19
> ren> if there is a big sign saying: this is Gore and this is what you are getting into, it is a different matter from if there is not and one can easily wander in accidentally – and again its male and female presenting avatars we are talking about.
Ola Fosheim Grøstad > Well, it is interesting that you think that "gender-bending" should somehow make a difference. Why?
It’s not gender bending as such that I think is important is more the separation of the limits and nature of role play from actual limits of action.
That is, it seems to me that there is a moral distinction between a scenario where anyone is allowed to participate in the role play in any way they want – however discriminatory things are within the magic circle of the play; and a scenario where only certain types of people are allowed to participate.
Posted by: ren | Jan 06, 2005 at 15:43
ren, I think what makes the Gor scenario so interesting is that it shows quite clearly that the concept "magic circle of play" is a rather subjective convinience construct rather than a good characteristic of the situation. I might accept roleplaying Gor with true roleplayers, but the idea of roleplaying a woman in order to satisfy a person who truly believes in that philosophy isn't all that comfortable. Imagine a movie about pedophiles, you might accept that your kid is going to play a role in such a movie if you believe it will create awareness of the subject and that he will be anonymous and not suffer any damage from it, but you probably would not if you knew that the actors and director actually were pedophiles?
In the context of Gor it isn't all that obvious what is roleplay and what is not. Maybe it isn't really obvious to those who participate either? Maybe a man participates under the excuse that it is roleplay while his basic attitudes are actually aligned with the philosophy of Gor?
One of the true powers of roleplay is that you (can) assume a personality and identity that is orthogonal to your own. If you don't your own identity and personality is at stake. That can be (more) painful.
Posted by: Ola Fosheim Grøstad | Jan 06, 2005 at 16:38
现在的青少年由于饮食、遗传等原因生理上成熟早，同时身处在这个浮躁的社会，那些被人们盲目追捧的轻浮的流行时尚，到处可见的成人用品，很容易让人们尤其是青少年产生性冲动，了解一些性知识是有益而无害的。
wow gold world of warcraft gold发表
Posted by: wow gold | Aug 10, 2006 at 22:46