Next installment on virtual worlds history...

Before I get started on a next leg of the journey down our history timeline I want to bring up some  resources on the web which many of you referenced or requested. After this we will pick up where we left off and then take a crack at the key definitional question: is a "virtual world" different from a "game"?

History Resources on the Virtual Worlds Medium

There are some great sources out there, including:

1) Ralph Koster's Online World Timeline

Ralph artfully weaves together very early concepts of online media such as Memex, with the birth of online worlds and richly innovative platforms such as PLATO, MUDs and early microcomputer-based online environments.

2) My own book (published in October 1997):

Avatars! Exploring and Building Virtual Worlds on the Internet

The book has some good first hand accounts of platforms from the 1994-97 period, plus history sections on seminal work including Habitat (more on Habitat later in this post). And just for you TN subscribers... some years ago a Danish university put the entire contents of my book on their server (from the CD ROM) and you can now get this here (steal that book!).

3) Indeed much of the Terra Nova site is itself a huge and valuable repository of history. Last year TN hosted a discussion (since gone a bit quiet) called:
The History of Virtual Worlds

If you know of other history/timeline sites please let me know (put it in the comments)!

Back to the Timeline

Maze War (Xerox Star version)

OK, lets get back on the journey through the timeline.  In our last installment we took a first peek at Maze War, which was arguably the "first first person shooter". Well, "Maze" (as it is affectionately known) established in just a few iterations from 1974 to 1979 at NASA Ames, MIT, Xerox and elsewhere, all the basic features of a heart-pounding first person shooter. These a-Mazing innovations included many firsts: a first person + plan (overhead) view of a 3D (albeit wireframe) virtual space shared over a network (the Arpanet), the first bot (nonplayer characters) in such a space, as well as spoofing (hijacking someone else's identity and chat) instant messaging, avatars (text handles plus simple arrow graphics at first and later on the famous eyeball in the Xerox Alto version), first in-world building (I think it was in the MIT Project MAC version where you could change the side of a hallway to create a fake passageway), and the first levels in a 3D environment (going up and down tunnels).

Legend has it that in the mid 70s, when the entire Arpanet was operating on a couple of 64Kbps leased lines, a letter went out to the main hubs at the time that demanded "those of you who are playing Maze are taking up to half of the packet traffic on the Arpanet so we respectfully request that you limit play to off-hours" (my paraphrasing). It is ironic that by the middle 70s we had the same dual character we see in the Internet today: the staid, serious and sometimes boring work of email, documents, file transfers and network ops versus the alternate reality of chasing your colleagues down the halls in Maze! Today we sneak in a few moments in World of Warcraft when we really should be writing that next funding proposal. Plus ca change...

As you can see from the Xerox Alto/Xerox Star version of Maze above, the user's 'character' is represented as visual icon, which actually turns left/right depending on the direction of travel relative to the onlooker, so indeed with Maze we have the arrival of possibly the first "avatars" in a game, ie: a graphical representation of a player that is a suggestion of an animate form in motion in the space. It is interesting to note that Philip Rosedale commented at a presentation I gave to Linden Lab in 2006 that the first avatars in the pre-alpha of Second Life were also big eyeballs.

And now we must diverge...

Now here we must diverge into a fundamental discourse. We are using the term "avatar" usually reserved for "virtual worlds" in the context of Maze, a "game". Biologist and author Richard Dawkins beautifully described a "pilgrimage back in time through the evolution of life on Earth" in his recent tome "The Ancestor's Tale". In the style of Dawkins, if we were traveling back along the evolutionary timelines of the medium of shared online spaces, we might observe the animal called "virtual world" join the critter called "multi-user 3D game" in a common ancestor called Maze War. In the truest sense Maze exhibited all the properties a gaming world, but something else was possible outside of game play, using the networked interface and the instant messaging channel, you could chat and just hang around in the maze with people who were physically located thousands of miles away. This experience of "virtual presence" captivated users and confused newbies who were trying to understand the language of these few new pioneers who dwelt in a visual dimensional cyberspace (even before the term cyberspace was coined). For a peak inside the culture of this Precambrian virtual world, see Ted Kaehler's cartoon below, drawn around 1980 at Xerox PARC.

Ted Kaehler's cartoon of a user of Maze War at Xerox PARC, circa 1980

Today we witness journalists trying to grapple with the concept of Second Life and being told again and again "well, its not a game, you cant really describe it as a game". So the term virtual world seems to appear more frequently with but separate from the term game. So if there is a bona fide category of "virtual world" what makes it different than "game"?

So are Social Virtual Worlds different creatures from Game Play Worlds?

Without getting too academic about it, the meaning of virtual world could be derived from its own descriptive terms as a "virtual rendition of the world". Like the "real world", a virtual world is therefore a place where you as an inhabitant are expected to make sense of where you are, what objects are there (maybe even rearranging some of them) and communicating with others (human avatars or non human robots) that might be inhabiting the space with you. So the primary function of a virtual world is discovery then the creation of meaning through the manipulation of the world and communication with others within the world. So entering a virtual world is in a way is like entering the family home as a new born baby, as a child in the first day in kindergarten, arriving at college as a freshman, or as the intern in the first day on the job. So lets adopt the term "social and creative virtual world" or more simply  "social virtual world" and see where that takes us.

This might bring us to suggest complementary terminology "game play worlds" to describe multi-player games. As we all know, there is plenty of social interaction and "off game" communication in most online games, and there is often a great deal of creativity as players shape their own characters and their own home turfs. However, I would posit that the primary reason users are in "game play worlds" is the game. Given that all games impose structure on the players with rules, tools and goals, this gives a pre-built context to any game play world. In stark contrast, many social virtual worlds started as true blank slate! Old timers may recall the endless empty plain of Alphaworld when it launched in the summer of 1996 (see below).

User teleporting into the great empty plain of Alphaworld in Summer 1996

Alphaworld (which later became Active Worlds) had no pre-existing structure or even 3D content other than this vast plain with mountains in the distance (which you could never reach). Users built out the entire space using primitive objects and scripts. While World's Chat, (which preceded Alphaworld onto the scene in the Spring of '96) had a pre-built world (a very nice rendition of a space station), it allowed users to go where they wanted and engage in free form chat. Worlds Chat and Alphaworld, and the many other mid to late 1990s platforms that followed, such as WorldsAway, The Palace, Traveler, Blaxxun, Comic Chat, VChat, Roomancer, Oz, Virtual Places, all shared similar properties and were all therefore virtual worlds. In contrast, early Internet-based multi player games like Quake, Meridian 59 and the Realm, Ultima Online and then of course, Everquest, were clearly focused on game play.

Now one could still muddy the waters and say "how about game play that happens in social virtual worlds"? Fair enough, but this tends to be an structured activity within the larger scope of free form social interacting and building. Of course there is unstructured socializing and general fooling around out of character in game play worlds too but when the break is over, players tend to get back to structured competition.

So does this distinction make sense? Is a social virtual world truly a different creature from a game play world?

So finally... back to some history

Ohmigosh its late and I am probably way over my word limit for this posting! Well, lets give you a teaser for next time. For a real must-read classic, take a look at...

The Lessons of Lucasfilm's Habitat

This was written by Chip Morningstar and Randy Farmer back in 1990 for Michael Benedikt's seminal MIT Press book Cyberspace: First Steps and will give you a great peak into how this fearless (some said foolhardy?) team at Lucasfilm successfully shoehorned an online virtual environment into a Commodore 64 running on a 300 baud modem connection (with ungodly 1980s network latency). This is the first environment where avatars were featured by name. Habitat rebirthed the medium of the social virtual worlds on consumer personal computers inhabiting a graphical world in sparkling Commodore colors.

So we now see our destination for the next posting. Chip Morningstar or Randy Farmer, are you out there, and do you want to work together on the next installment?

One last question for the evening from our jaunt back in time: are MUDs, MOOs, WELL-style conferences and IRC also social virtual worlds? Comments?

Comments

[Elle Pollack]

1.

"...are MUDs, MOOs...also social virtual worlds?"

Um, hell yes? (At least those designed primarily for socialization instead of gameplay, a la Lambada MOO). You're a Terra Nova blogger where everything is "didn't we discuss this on MUD-dev in 1997?" and you have to ask that?

"IRC...?"
No...it's "social" but lacks anything "persistant" or "worldly".

[greglas]

2.

Yeah, this is quite interesting...

My (rough) sense has always been that VWs really kicked off with MUD & the various multiplayer games in the late 70's/early 80's. And if I had to go further back, I'd start (I have started) with text programs like ADVENT, Hunt-the-Wumpus, etc.

But I should note that when Dan and I wrote our brief version of this history, I talked Dan into letting me split out the visual from the textual, because they felt like truly separate paths -- for pretty obvious hardware and market segment reasons. I'm personally quite partial to the textual path, which, at least with MUD, got significantly worldly in a very quick way, but I don't see a need to recite that history here. "See Richard's Book."

W/r/t to the visual path, though, I guess I could say "see Jess's book," but Jess is really not about doing the "world" thing and is definitely keen on the game thing. Fwiw, when Dan and I took a stab at a history of the visual side, we started Space War and Pong and PLATO, went through the Arcade craze a bit and then hit Habitat. But there really was a lot going on.

Bruce seems really interested in defining VWs as visual, social, non-games. If you do that, I can see why Habitat jumps out.

Maze War is trickier. It is a great example of early space, but it is a game. And, in terms of its rules, you can almost see it as a kind of hack of SpaceWar (which was the Maze War of its time) which, above all, contributed 3D.

I think, by analogy, the same Space War/Maze War step was performed when coin-op State of the Art went from Tank (1974) to Battlezone (1980):
http://en.wikipedia.org/wiki/Tank_(arcade_game)
http://en.wikipedia.org/wiki/Battlezone

Anyway -- could write all day about this, but can't. Thanks again, Bruce -- and esp. like the pix.

[ron meiners]

3.

I can understand Bruce's point, and question. At what point does textual communication stop being chat and b ecome something more immersive? Or at what point does the general chat channel in the Crossroads ever become something more than... chat?

[Raph]

4.

Well,it's well known that I see litle difference between the textual and the graphical worlds.

http://www.raphkoster.com/2006/03/31/are-muds-and-mmorpgs-the-same-thing/

They both simulate places, the question is mostly one of presentation. You can in fact display a mud world graphically and an MMORPG textually -- it's been done.

Similarly, they are both places that can embed content, such as games, or not. Given enough effort, you can implement DikuMUD within LambdaMO, and Everquest within Second Life. Probably vice versa as well, though with even more effort.

Randy is, of course, at Yahoo, along with Chip. And Randy is also on the advisory board of my company. ;)

[greglas]

5.

I completely agree that space, persistence and society (avatars) are the keys to (both text and visual) worlds, Raph -- but just to clarify a point:

Would you agree that the *history* of visual and text virtual worlds are two (pretty much) different things? Now, I know there has been major overlap. Iirc, UO and EQ teams were made up of plenty of MUDer and MUSHers and ADVENT-loving RPers, but my sense has been that there are lots of folks like Bruce who are heavily into social & visual world things but that don't see much intersection with textual VW history. As proof perhaps, see Bruce's question.

[KirkJobSluder]

6.

One last question for the evening from our jaunt back in time: are MUDs, MOOs, WELL-style conferences and IRC also social virtual worlds? Comments?

I think the answer to this will depend on whether your perspective is oriented towards:
1: Ludic aspects. (Answer, no)
2: Media. (Answer, maybe)
3: Sociability and communication. (Answer, yes.)

But in terms of sociability and communication we already are looking across multiple dimensions of how people can interact, so I find the world/not-world dichotomy to be a bit limiting.

[Ola Fosheim Grøstad]

7.

Would you agree that the *history* of visual and text virtual worlds are two (pretty much) different things?

You can't lump implementations together based on media. Regensis BSX was a graphical LPmud. Meridian59 was a 3D reimplementation of classic MUD design. Everquest was aiming to be a 3D DIKU. Most newer 3D MMORPGs are inspired by Everquest/DIKU/MUD.

On the other hand you have the meta-verse inspired worlds that are VR-ish, "lets replicate the physical world in the computer"... They might deserve their own branch. The online games grew out of MUDs and single-user games turned into multiplayer versions.

[Thomas Malaby]

8.

I really enjoyed this post, Bruce. Thanks so much. My favorite passage was this one:

Like the "real world", a virtual world is therefore a place where you as an inhabitant are expected to make sense of where you are, what objects are there (maybe even rearranging some of them) and communicating with others (human avatars or non human robots) that might be inhabiting the space with you. So the primary function of a virtual world is discovery then the creation of meaning through the manipulation of the world and communication with others within the world. So entering a virtual world is in a way is like entering the family home as a new born baby, as a child in the first day in kindergarten, arriving at college as a freshman, or as the intern in the first day on the job.

To me, from this insight follows a whole set of answers to other questions that have been raised. In short, what you've put very well is the idea that by "world" we would like to capture the presence in a given domain of a wide-enough range of possibility for action that one must *learn* to act in the world. What this suggest to me is that a domain rich enough in performative possibilities may generate a distinctive *culture* (a distinctive disposition for acting in that domain), and that this (not space, or graphics, etc) is what we mean by "world" in virtual world. I think this is exactly right, and the related questions of graphics vs. text, etc, fall by the wayside if we pay attention to whether a wide range of action is present (which can obviously be true of both kinds of virtual worlds). In this respect, the persistence of the domain becomes a necessary condition as well.

Similarly, while I see the opposition of gamey worlds to virtual worlds as a tempting distinction to make, I don't think it bears out as an either/or. That is, I think whether an MMO is gamey doesn't determine whether it makes sense to call it a virtual world. They *can* be reliably distinguished from the other virtual worlds (which to me are all "game-like") by the presence of established game objectives. These are usually architected by the designer, but can also emerge from social practice to become so established as to make the environment sensible to speak of as a "game". But they can still just as well be virtual worlds if they have broad enough affordances for the participants that it begins to feel both rich in possibility and distinctive in its demands.

[Raph]

9.

Would you agree that the *history* of visual and text virtual worlds are two (pretty much) different things?

Well, no. I see far far too many points of congruency. Where would you draw that line?

Certainly there are lots of folks from the graphical worlds crowd of the mid-90s who see little overlap, but there are also plenty who were heavily into the world of MUSHes and MOOs.

Ola pointed out plenty of overlap on the gaming side -- there's even more than that -- a huge huge portion of the initial wave of MMORPGs came from mud folks.

If what you mean by "history" is "lines of evolution," even then I would say that there are points of congruency where either technical crossover or design crossover occurred.

[ron meiners]

10.

@Bruce and Thomas

Very well stated both. I would like to add only that our passion for these worlds and motivation to involve our "selves" in these experiences is an indication not of some aberrant impulse to frivolessness, but precisely because they afford experiences that are meaningful to us in the way that all experience can be more or less meaningful. We engage in them because they let us play out the next few acts of the interior drama, as it were, creating experiences that affect us in (more or less) exactly the way that other experiences affect us. In some sense it's much less a make believe than another manifestation of our interest in interacting with our worlds in the myriad of ways that are interesting to any given individual. Or, in some sense I would suggest that all "make-believe" is primarily this kind of synthetic life experience: experimentation with the various elements we find around us in an ongoing effort to explore, master, and share them.

[F. Randall Farmer]

11.

Bruce,

Feel free to contact me by email if you want to talk about anything. I've got *lots* of material. :-)

Your fellow pack-rat,
Randy

[ErikC]

12.

Raph says:
"Well,it's well known that I see litle difference between the textual and the graphical worlds."
http://www.raphkoster.com/2006/03/31/are-muds-and-mmorpgs-the-same-thing/
They both simulate places, the question is mostly one of presentation. You can in fact display a mud world graphically and an MMORPG textually -- it's been done."
Nope, I don't believe simulating space is creating place. A 3D visual field is not a place.

And if MUDS handle collision boxes, lines of sight, visual occlusion, psychological effects of threshold transfer, territorial phobia, spatial proxemics, enhanced cognitive mapping substantiated by extra hippcampus activity, sense of social presence via body heat, and game-play affecting sense of balance, I would be more likely to believe this statement.
A test between MUD and spatial MMORPG could be similar to http://www.jneurosci.org/cgi/content/full/25/31/7254
Any reference to a MUD that handles all these things would be appreciated.

13.

Bruce said, "Is a social virtual world truly a different creature from a game play world?"

No, I don't think so. There have been completely social virtual worlds (e.g. Worlds Chat) and almost-completely gameplay virtual worlds (say, multiplayer FPSs if you don't count the lobbies and such), but the most successful virtual worlds incorporate significant aspects of both.

When we were developing M59, we looked hard at things like Worlds Chat and AlphaWorld. The former had nothing but chat, and it quickly became uninteresting for most people who tried it out (it's instructive to go back and look at the hype too). It seemed pretty clear that adding "social referents" to the world -- something multiple people could talk about and interact with, say a ball -- would add a great deal to the fast-shriveling novelty that most people found on logging in. Alphaworlds at least had objects, but they were all form and no function -- you could create buildings but they didn't do anything. The world was still almost entirely social. True, you had external social referents, but they didn't do anything. It would take the abortive attempts at VRML and later, Second Life, to add enough functionality to make a social world interesting to at least some people.

Multiplayer gameplay equates to shared contingency: we called this the "circle the wagons" effect. People band together because there's something scary out there, and this acts as a catalyst to socialization. Socialization then catalyzes their ability to take on greater shared contingency (more gameplay), and the virtuous cycle continues.

Separating social worlds from gameplay worlds seems odd to me, like asking which side of the arch is holding up the ceiling.

[Richard Bartle]

14.

Something which disturbs me about this "history of virtual worlds" thing is the lack of an audit trail. A timeline is not a pedigree. Merely having been first to the cut doesn't mean there was automatic influence on future work.

Example (from a paper I wrote for the "Handbook of Internet Research"): Golf was invented in China. There is evidence from the "Dongxuan Records" that a game called chuiwan ("hitting ball") was played as early as the year 945. A Ming dynasty silk scroll, "The Autumn Banquet", depicts a man swinging something with the appearance of a golf club at something with the appearance of a golf ball, having the apparent aim of conveying it into something with the appearance of a golf hole. Golf was also invented in France, where it was known as palle mail. Tax records from 1292 show that makers of clubs and balls had to pay a toll to sell their goods to nobles outside Paris. A book of prayers, "Les Heures de la Duchesse de Bourgogne", contains illustrations of men swinging something with the appearance of a golf club at something with the appearance of a golf ball, having the apparent aim of conveying it into something with the appearance of a golf hole. Golf was also invented variously in Middle Egypt, ancient Greece, ancient Rome (paganica), England (cambuca), Ireland (camanachd) and the Netherlands (kolf). Nevertheless, despite these assorted claims as to the invention of golf, it’s indisputable that the modern game is the product of Scotland . The golf played today is the direct descendent only of the version that the Scots developed. Follow the audit trail from the US Masters back in time, and Scotland is where it ends.

It's like this for virtual worlds. Hitting a ball into a hole with a stick is a fairly obvious idea, so it’s unsurprising that what we now know as golf has been conceived as a game many times throughout history. Likewise, creating a virtual world doesn’t qualify as an act of genius – it was always going to happen. Indeed, the concept has been invented independently at least six times and probably more. Most of these had little or no impact on the future development of the concept, however. Therefore, when it comes to understanding their history, the important question is not so much which one came first as which of the several ones that came first (as far as their designers were aware) produced seeds that lead to the virtual worlds we have today.

The PLATO games, for example, were certainly prototype virtual worlds, and, depending on your definition of "virtual world" and your loyalty to PLATO, some could be regarded as actual virtual worlds. What influence did they have on the development of today's virtual worlds, though? Almost none - they're like Chinese golf.

This timeline stuff is all well and good, but what's the link from Maze to WoW? It's easy enough to track it (via EQ, DikuMUD and AberMUD) to MUD, but how do you track it to Maze? And when you do track it, do you go through other things that would be considered virtual worlds, or do you go through non-worlds (in which case, although its graphics, say, might be descended from Maze, its virtual worldliness isn't).

You need to structure this timeline for it to be any use. Otherwise, people will look at it 20 years from now and conclude that merely because Maze was written before Adventure, the fact they're in the same timeline means the latter must be a descendent of the former.

Richard

[Andrey]

15.

Ёмоё, а нафлудили-то сколько. Молодцы! Продолжайте в том же духе. :)

[Lavant]

16.

TRANSLATION

"Yemoye, and nafludili- that is how much. Fine people! Continue in the same spirit:)"

I have no idea what the first two words mean.

/lamzorz andrey

[Richard Bartle]

17.

Lavant>I have no idea what the first two words mean.

Neither does Google's translation service. It gives: "Much as well as nafludili-to. Well! Continue in the same spirit."

BabelFish gives the exact same translation that you do: "Yemoye, and nafludili- that is how much. Fine people! Continue in the same spirit."

Richard

[Elle Pollack]

18.

>>"I can understand Bruce's point, and question. At what point does textual communication stop being chat and b ecome something more immersive? "

At the point you insert persistance, one of the things that defines a virtual world.

In IRC, or even over IM, I may emote something like "Elle flops in the corner" or "Elle steals from her friend's tea chest." The only thing establishing the presence of a tea chest, or even a corner, is my say so. Others may elect to play along, but when I/they leave the room or cease the play-acting, it may as well go *poof*. Some items may acquire a limited persistance in the form of running jokes ("Elle tosses the resident pervert into The Closet again"), but the code running the medium - the IRC server and client - have no mechanism to support this sort of thing.

Now, as I think of it, it would be possible to turn an IRC chatroom into a very kludgy sort of world. I could create a bot called TheCloset and program it with a set of commands...one that will cause the bot to spit back a description of what it is and what it "looks" like, one to "throw into The Closet", some funny responses and random behaviors, and so on. I could make a "room" bot that could oversee the rest of the bots and provide a description of the room we occupy. Find a way to easily "move" people among a cluster of such rooms in IRC and you start to have something resembling a text world.

Now MUDs, MOOs MUCKs (MUCK being the one in my experience) and so on, are built for the ground up for such things, so it's a lot easier.

@dig My Room
@desc My Room=A comfey padded room, 10x10 feet square
look here
> A comfey padded room, 10x10 feet square
@door [Exit];exit;My Room=[E]lle's [R]oom;er;Outside
(@door is specific to Nameless MUCK Code (NMC), lets you make both exits at once)
er
> You enter Elle's Room
@create Tea Chest
@action open chest=Tea Chest (and a bunch more to set the sucess/fail messages)
@desc Tea Chest=A Tea Chest
@drop Tea Chest

And so on...the real thing is of course much more verbose with object ID numbers and sucess/fail messages and such. And of course there are descriptions and all sorts of properties that can be set on your character/avatar such as age, gender, species...there are often permission systems for rooms and objects (that works MUCH better than some that I know *cough*Cory*cough), and eventualy you get to a point where it would be hard to distinguish by function a place like Nameless MUCK (an open-genre world and the NMC testbed) from Second Life were it not for the graphics and some other implimentation choices.

*looks up at Richard* Think I've got that about right?

[Elle Pollack]

19.

EricC>>And if MUDS handle collision boxes, lines of sight, visual occlusion, psychological effects of threshold transfer, territorial phobia, spatial proxemics, enhanced cognitive mapping substantiated by extra hippcampus activity, sense of social presence via body heat, and game-play affecting sense of balance, I would be more likely to believe this statement."

And the die-hard text fans would tell you that that's just so much eye candy and some of that is, like, wait for the VR goggles and tactile simulation cyber suits, man. Most games/worlds have simplified representations of reality, picking some elements to emphasize and not others. It's a very concious series of design choices that keeps both the designers and the players from being overwhelmed by detail.

NMC has some very rudimentary proximity stuff (a la +nearby which lets you "see" into adjoining rooms) and there ware some rather interesting ideas being thrown about during the development (not currently active) of NMC2 but the primary representation of space is still node-based (some of the ideas involved nested nodes that were part of the same "place" for example). Graphical worlds are much easier to make contiguious, but in practice most of them tend to use combinations of linked nodes (think "loading screen") and contiguious space.

[Richard Bartle]

20.

Elle Pollack>*looks up at Richard* Think I've got that about right?

That's what I think, yes. Others disagree, and say there is something fundamentally different between a graphical world and a textual one:
http://terranova.blogs.com/terra_nova/2005/06/ce_nest_pas_un_.html

Richard

[ErikC]

21.

That is an 81,456 word thread!
@Elle, the technology you refer to now costs around the same as a current generation game console.
Sorry, what is NMC? If you can give me an example of the most spatially developed/suggestive(?) MUD I'd appreciate looking at that. I haven't played that sort of thing for a long long time.

@Richard, Cory's post and Second Life (if that is his main example) is not what I would argue and not what I would use as an example, at all!

Raph said "Well,it's well known that I see litle difference between the textual and the graphical worlds."
I wouldn't argue that the design of many MUDS are inferior or different to the spatial design schematics of graphical worlds or that many designers have different design considerations when they design both. I would however argue that a well developed graphical world has different design considerations and different cognitive influences on the player.

[Richard Bartle]

22.

EricC>I would however argue that a well developed graphical world has different design considerations and different cognitive influences on the player.

Well yes, obviously. One has pictures rendered on the screen and one has pictures rendered in the imagination.

The thing is, though, that (interface aside) textual and graphical worlds are one and the same thing, in the same way that today's movies are in essence the same thing as silent movies.

Of course time moves on. Fundamentally, though, all virtual worlds have a "virtual worldiness" to them that separates them from other forms of computer software. In the same way that Casino Royale and Farenheit 911 can both be considered movies, WoW and SL can both be considered virtual worlds. Yes, the shoots for a feature and (let's call it) a documentary are different, and there are subsequent design differences, but there's something about them medium itself which is inviolate and which has been there since before Charlie Chaplin made his name.

So it is with virtual worlds. Yes, graphical worlds do have different considerations to textual ones, but they're all part of the same medium. Those design differences don't go near the heart of what it is for something to be a virtual world. They may have got their graphics from (ultimately) Maze, but they didn't get their virtual worldliness from there. They got that from MUD1, Aradath, Islands of Kesmai and Monster.

Richard

[Elle Pollack]

23.

@ErikC:
>Sorry, what is NMC?

NMC is an open source MUCK server distrubution, a few forks down the family tree from first Fuzzball and somewhere further back, TinyMUCK.
http://nmc.limitless.org <- project page
telnet://nameless.muck.limitless.org:2600 <- Nameless MUCK
telnet://redwall.muck.limitless.org:2600 <- Redwall MUCK, the most active and longlasting (since 1996) MUCK I know of running on the code (they're run by the same person). Though if children's fantasy with animal characters set in a quasi-mideavl setting doesn't strike your fancy you'll probably just want to download the files for NMC1 to play with and poke around the NMC2 documents (not all linked to the page, some might still only be in the CVS repository). Heck, if you are or know coders...all the primaries on the project are too busy with Real Life (tm) to put work into it anymore.

>"If you can give me an example of the most spatially developed/suggestive(?) MUD I'd appreciate looking at that"
Well, +nearby more or less just allows for some messages between rooms and for you to look "thorugh" exits, fairly rudimentary. Because the MUCKs tend to be user-built not all the features will always be used to their fullest. Furthermore, the really advanced stuff planed for NMC2 is still vaporware if I didn't make that clear enough the first time. And to be honest my circle of interest never really expanded much further than that...I tried some other MUCKs early on but didn't get far into them. NMC is a pretty small blip on the radar of text worlds but that said, it does some cool and usefull stuff.

>"the technology you refer to now (tactile feedback suits) costs around the same as a current generation game console."

If the current generation console in question is the PS3 than they're really in trouble *snerk*. But the people who buy hardware at that price point ($500 graphics cards, $600 consoles...) is a pretty slim precentage. Not to mention then you have a chicken and egg problem getting developers and players to adopt it, a problem a lot of PC perephrials run into. So I don't forsee their widespread use anytime soon.

[Dr. Cat]

24.

While it's true that most of the major MMORPGs today don't trace any of their influences back to Plato, I don't think you can say that's 100% true of all of them. The main line of descent, I would say, is D&D to MUD to DikuMUD to Everquest to WoW. But I know the Plato games led directly to Wizardry, and thence to the Wizardy-influenced games (Bards Tale, Might & Magic, etc.) that still are a lesser influence in some corners of the genre today. (Shadow of Yserbius was even online.) As well, there's people like Gordon Walton out there, who was very involved in Plato early on & has been a major figure in MMORPG development. I hope my own work shows some Plato influence too, though admittedly not as much as I'd like it to have!

[Raph]

25.

What I am getting in saying that there is little difference is that the client is purely a representation of the serverside sim.

It isn't whether the CLIENT supports "collision boxes, lines of sight, visual occlusion, psychological effects of threshold transfer, territorial phobia, spatial proxemics, enhanced cognitive mapping substantiated by extra hippcampus activity, sense of social presence via body heat, and game-play affecting sense of balance" but whether the SIMULATION supports those things.

The client is just parsing output and displaying it.

Many of the effects you describe as actually happening in the human wetware, and some of them happen regardless of visual perspective (1st person, 3rd person, overhead or chase cam).

[ErikC]

26.

Perhaps I wasn't clear enough.
First premise, a nodes-based abstraction on the server isn't the way we actually experience or professionally design a RW 'place'.
Therefore even if text and spatial / graphics based worlds can be explained or designed this way does NOT mean the latter should be designed this way or that for an optimal simulation text and graphics based worlds (or 'places') are fundamentally similar even if they typically are.
From this I am led to believe if place affordances are fully supported on the client side, then a rethink of the server side simulation as node-based may be necessary.
It may be that a node based schematic abstraction of place is efficient and suitable for text based worlds, but I haven't yet seen compelling evidence that it is for a 'platially evocative' spatial world.

NB:
1. a spatial environment doesn't have to be pretty pictures (imagery) or visual at all in order to be spatial.
2. The above claims do not mean I think spatial worlds are inherently superior, it worries me how often their spatial nature does not actually adds value.

[Richard Bartle]

27.

Dr Cat>While it's true that most of the major MMORPGs today don't trace any of their influences back to Plato, I don't think you can say that's 100% true of all of them.

That's correct, but the influence is only very peripheral. As you say, Gordon Walton cut his teeth on PLATO (as did Andy Zaphron at SOE), so there is some influence through them. It's next to impossible to point at something in one of today's commercial virtual worlds which is the way it is because that's how it was in a PLATO virtual world, though.

There are a few honourable exceptions to the general MUD-rooted family tree, of course. For example, DAOC comes ultimately from Aradath, which was conceived independently of MUD1. As I said earlier, we were always going to get virtual worlds, so it's unsurprising that they were invented independently several times. What irks me is when "first" is presented with the implication of "and what follows descends from it". That's not true. MUD1 came several years before Aradath, but DAOC doesn't directly descend from it (there may be influences, but on the whole DAOC is Aradath via Dragon's Gate). I can't - and don't - claim that DAOC is a descendent of MUD1, even though MUD1 predates Aradath.

Whether or not the PLATO games are or aren't what we'd now call virtual worlds is an interesting question from the point of view of defining what we mean by the term. Even if they do become accepted as a examples of early virtual worlds, though, we're back to golf being invented in China.

>The main line of descent, I would say, is D&D to MUD to DikuMUD to Everquest to WoW.

Undoubtedly D&D influenced DikuMUD, but actually it was only of marginal influence on MUD1 (Roy Trubshaw hadn't played it; I'd played it a lot, but in co-creating MUD1 I drew more from another role-playing game I'd invented myself well before I came across D&D). The Lord of the Rings was more of an influence than D&D. Adventure was an influence on Roy, and therefore if D&D was an influence on Crowther and Woods (I don't think it was, but you never know) then that would be a way for D&D to be regarded as a seed for MUD1.

>Shadow of Yserbius was even online.

Yes, and it deserves more attention than it usually gets, at least from designers.

Another line which did reasonably well but reached a dead end before it had any lasting impact is the one with Islands of Kesmai at its root. We saw several early 2D graphical worlds inspired by it, eg. Kingdom of Drakkar, but they weren't the inspiration for even UO (also tile-based) as far as I can tell.

>I hope my own work shows some Plato influence too, though admittedly not as much as I'd like it to have!

Well, if you feel it influenced you, fair enough - you know your own influences. Indeed, we could find a few years from now that some spin-off from your work has become the dominant model for virtual worlds, and then the MUD line can be consigned to a footnote in history.

Either way, though, what's important is the pedigree, not the timeline. Furcadia descends from Avatar whether or not Avatar (or some proto-Avatar) predates MUD1, and this is something you just don't get from a timeline.

Richard

[Richard Bartle]

28.

EricC>It may be that a node based schematic abstraction of place is efficient and suitable for text based worlds, but I haven't yet seen compelling evidence that it is for a 'platially evocative' spatial world.

Some early textual worlds (mainly the ones on the "IO World of Adventure" system, eg. MirrorWorld) used a spatial model based on a grid system. They weren't popular, as the players felt the rooms were boring and samey - they preferred the world to map to conceptually significant locations rather than to geographically precise ones.

I've written a paper on this subject, by the way, but it's still winding its way through the publication process. I could email you a copy if you like, though.

Richard

[ErikC]

29.

Yes please Richard I'd like to see it, my address is linked from my name.
Some architects have designed places using honeycombs, others using overlapping circles from which they then create the perforated overlapping rooms. I don't know of commonly used abstraction systems apart from nodes etc but interestingly Lynch's 5 mapping elements are often cited in VR navigation literature but not, as far as I know, used as inspiration for game design. Similarly, Christopher Alexander's pattern language, although I am sure there are abstractions out there.

[Raph]

30.

Ditto on what Richard said -- there have been several examples of true coordinate systems used in text environments, as well as many examples of coordinate systems embedded WITHIN nodes, of course.

I find it curious that you would be resistant to this, Eric, given that virtually all graphical virtual environments can be considered as nodes with embedded coordinate systems. WoW works this way, as does EverQuest, and just about all of the multiplayer FPS games; true seamless environments are definitely in the minority. A sense of boundedness is common.

That said, I dispute the notion that "a nodes-based abstraction on the server isn't the way we actually experience or professionally design a RW 'place'." I am no architect, but it sure seems to me that we do indeed commonly design spaces with a sense of "rooms," bounding of discrete spaces, and so on. The concept runs all through Alexander's writing.

[Matt Mihaly]

31.

Richard Bartle wrote:

Some early textual worlds (mainly the ones on the "IO World of Adventure" system, eg. MirrorWorld) used a spatial model based on a grid system. They weren't popular, as the players felt the rooms were boring and samey - they preferred the world to map to conceptually significant locations rather than to geographically precise ones.

Some text MMOs/MUDs (like ours) actually mix these two, using node-based arrangements sometimes and grid-based arrangements other times. We use grid-based stuff when we want to create distance easily but as you say, most players spend most of their time on the node-based map.

I'm surprised we still have to have this conversation in 2007, on Terranova of all places. If a text MMO isn't a "virtual world" because it doesn't support the visual sense in the same way as "graphical" MMOs (ie it lacks something in common with the physical world) then we may as well give up on calling anything a virtual world until you can smell a flower and feel a lover's touch in them.

After all, the physical world is precisely as "real" for a blind person as a sighted one.

--matt

[Elle Pollack]

32.

Well for kicks and giggles, here's a summary of the nested node idea we were tossing around for NMC2 so far as I can remember it.

Take any given "place", say, a clearing in a forest. It's a node in the series of nodes that make up the forest. The clearing is fairly large so the builder decides to make the clearing two nodes within the larger one, a north and south end. In each end he adds two more subnodes...one for the thick underbrush surounding the edge and one for up in the trees.

In the way that you would normaly link rooms, the nodes are linked together by a variety of different relationships; they might be traversable, visable, audible or any possible number of custom-defined relationships. Say the builder links the two halfs of the clearing with visible, transversable and audible_yell; A character standing in the north of the clearing could see anyone in the south end and be able to walk over to them but would have to yell in order to be heard. The trees might be set transversable with a 'if' condition: the characher has to have the property denoting they can climb. Someone in the underbrush might be heard but not seen from the clearing, but can see the people in the clearing plainly.. He might create a custom "broadcast" relationship; when someone transverses in or out of the underbrush, a message "Something rustles in the underbrush" is broadcast into the adjacent clearning node. And so on.

[ErikC]

33.

@Matt who said
"I''m surprised we still have to have this conversation in 2007, on Terranova of all places. If a text MMO isn't a "virtual world" because it doesn't support the visual sense in the same way as "graphical" MMOs (ie it lacks something in common with the physical world) then we may as well give up on calling anything a virtual world until you can smell a flower and feel a lover's touch in them. After all, the physical world is precisely as "real" for a blind person as a sighted one."
regards sentence 1: Terranova is a place? Which a person can exist "on"?
regards sentence 2: I said"It may be that a node based schematic abstraction of place is efficient and suitable for text based worlds, but I haven't yet seen compelling evidence that it is for a 'platially evocative' spatial world." I never said text-based MMOs cannot be worlds.
regards sentence 3: I already said "a spatial environment doesn't have to be pretty pictures (imagery) or visual at all in order to be spatial."

@Raph: I am not disputing that architects draw up rooms or that Alexander discusses rooms or that you can denote them with nodes. The notion of "eternal space" and "dissolved architecture" and "destroying the box" are recurring themes in architecture over the last 90 or so years.
I am suggesting that a node is not the only way a notion of space can be abstracted. If node (and I suspect you have an elastic notion of node) is best for text-based worlds and prevalent for spatial worlds that doesn't mean it is best for the latter. Are "true seamless environments" not just definitely in the minority in games but a dwindling minority? If so, you may be right. If not, I think this is significant.
And no, I am not saying worlds have to be "true seamless environments."

[Richard Bartle]

34.

EricC>Yes please Richard I'd like to see it

I've emailed you a copy.

>Some architects have designed places using honeycombs, others using overlapping circles from which they then create the perforated overlapping rooms.

We had an architecture panel at one of the State of Play conferences. They were very disappointed by what they were shown. Although this was in part due to their not understanding the medium (not having gravity can make a big difference!), they did have a point in that what people were creating didn't really read consistently from the perspective of any archetecural theory. The structures had every appearance of being exactly what they were - the creations of naive amateurs.

>I don't know of commonly used abstraction systems apart from nodes etc but interestingly Lynch's 5 mapping elements are often cited in VR navigation literature but not, as far as I know, used as inspiration for game design.

Ha! Well as it happens, I read Kevin Lynch while researching my PhD, so was familiar with his work when I wrote MUD2. We already had the makings of most of his elements in MUD1, but only by happenstance. The exception was districts ("areas" in MUD1), which were deliberately created to be consistent collections of rooms sharing a collective atmosphere, so I didn't have to do anything different there for MUD2 than I was doing anyway.

For the new areas of MUD2, I planned the major paths and nodes so they hung together in a les haphazard fashion. Edges were far less common than in a city, but apart from the borders that stop you falling off the world I did deliberately insert a major edge (a river), and then later (during expansion phases) other edges. This gave me something (along with paths) to use to separate the districts, plus some pinch-points (nodes) where it was crossed.

In some ways, landmarks are harder to do in text than graphics, but in others they're easier. You can't simply see them and judge for yourself whether they're memorable, but then even in graphical worlds you can't do that at a great distance as it fogs out before you get there. At least in text I can reference a landmark from as far away as I like (although if people don't read the long descriptions of rooms they're not going to pick up on it).

So yes, there isn't a lot of awareness of Lynch's work in virtual world design, but there nevertheless is some. People who managed to stagger through to page 482 of my book may find his name rings a bell, for example.

Richard

Similarly, Christopher Alexander's pattern language, although I am sure there are abstractions out there.

[Richard Bartle]

35.

Matt Mihaly>Some text MMOs/MUDs (like ours) actually mix these two, using node-based arrangements sometimes and grid-based arrangements other times.

Yes, this is how the Mirrorworld games did it, rather than be entirely grid-based. I do have a copy of a design for a text MUD from an adventure game developer, Level 9, that was entirely grid-based. It was never implemented, though.

Richard

[greglas]

36.

Richard> Although this was in part due to their not understanding the medium (not having gravity can make a big difference!)

Actually, if I remember that panel right, Richard, it was just the opposite. One of the best criticisms of the Second Life architecture was that it was *not* adapted to the medium. So, e.g., why on earth have doors at ground level in a world where everyone flies? Why have a slanted roof where there is no rain or snow? The point of this one commenter (dead on imho) was that form follows function, hence virtual form should follow virtual function in Second Life -- yet instead of native forms, we see simulations of familiar, and virtually unsuited architectures.

Prior post by Betsy
here.

Quoting Ann Beamish as saying:
"Why is it that we end up with the very familiar, the hyper-romanticized simulacrum, or the dreary?...I’ve been really quite disappointed in the way this has been going and I’ve been trying to understanding why is it that we’ve been reproducing the physical world as we know it."

As to virtual realism generally, old thread here, where I say I'm reading your book:
http://terranova.blogs.com/terra_nova/2004/01/the_value_of_re.html

[greglas]

37.

I took a look at the old thread, Richard, and, no surprise, I was actually repeating something that *you* said about the sloped roofs. :-)

But you did seem to have a slightly different take on that panel than most of us...

[Matt Mihaly]

38.

ErikC wrote:

regards sentence 1: Terranova is a place? Which a person can exist "on"?

You have got to be kidding me.

--matt

[Richard Bartle]

39.

greglas>I took a look at the old thread, Richard, and, no surprise, I was actually repeating something that *you* said about the sloped roofs. :-)

Oh well, so long as I get to be right somehow!

Architects have theories, but those theories seem to have a strong grounding in physical reality. If the physical rules are different, the theories need to be picked apart so that the rationale behind them is considered. For example, I know architects have theories concerning natural light, but in a world where one light is no more natural than another, is this an issue? Is the fact that natural light means a view of the outside important? If so, we can pick that part from the theories of light, but leave behind the thing about feeling more comfortable in natural light. We could use some of the relationship of light and space, but as we don't even have to have shadows we don't need to take it all.

So there are (or at least there should be) architectural theories that apply to virtual spaces. The panel seemed to want to see these new theories in action but weren't happy when they did because they grated against some existing theories. Second Life houses shouldn't look like real life houses in the sense that they have redundant features, but they should look like them in the sense that there are some functions of architecture that apply in general. A SL house would still have walls, not because it needs load-bearing structures to keep the roof on, but because it delineates private space from public space.

Given the lack of interaction between architects and gamers in the past (we still see medieval-style castles in role-playing games where there are multiple creatures that can fly or burrow), it's not surprising that gamers are having to learn the principles of architecture anew, nor that architects criticise them for their naivete. SL is getting enough publicity now that we could see some actual bona fide architects there creating structures that make sense both for SL and architectural theory. That should be worth seeing.

Richard

[Ola Fosheim Grøstad]

40.

BTW, if one are going to include gamey worlds then there are a lot to choose from: nethack, crossfire, x-pilot and a plethora of 4-player arcade hall worldy games. I can't believe that comp-sci students that went into game development weren't influenced by some of those.

[Ola Fosheim Grøstad]

41.

Just want to point out that the term "graphical mud" yields over 200 hits prior to 1996 on Google Groups (which are incomplete btw). Inevitably some of those are about the merits of graphics versus text, which has been discussed repeatedly and extensively on MUD-Dev too, of course...

Here is the non-exhaustive list of graphical worlds by Hans Mikelson from a 1995 USENET post (for the URLs, archive.org is your friend):

GRAPHICAL MUDs
Dragonspires, Kingdom of Drakkar, Crossfire, Chiba MOO, Regenesis, Neverwinter Nights, Shadow of Yserbius, Netrek ][ Paradise, Chaco Multi-Media MUD client "Pueblo"...

MULTI-PLAYER GAMES
Marathon, DOOM/Heretic/Quake, DWANGO,(no name yet) [M59], KALI

GRAPHICAL CHAT ROOMS
Worlds Away on CompuServe, Worlds Chat

[robusticus]

42.

Well, I echo my eastern european counterparts.

I read your book, Bruce, back in the day. So naturally I am a fan. Cool pics of Alphaworld unclaimed. By the time I got there I happened upon a very populated world including "the stairway to heaven", which, was remarkable to me that I remember it from so long ago.

Mazewar. Oldskool, baby. I can't claim to be quite THAT old, but today I had the game "Tempest" mentioned as someone's personal arcade collection, and I recall playing a 3D wireframe pacman in monochrome. And I browsed a bit on RPG history (circa Nerfbat) and found a wireframe game at that genre's roots.

I do so love the oldskool threads because it reminds me of all these games I played which I had forgotten, like Tempest.

Oh, last note... Eyecandy. Well, eye candy, ear candy and.... touch candy. Lightweight example: crobar > grav gun. I use that example to contrast graphics lumped as sensory perception versus text-only debate (authors versus painters versus directors?). I think I lack the grammar to explain that properly, however. Gravgun was gimmicky. Is that proper grammar? Meh. My trouble with that is I can't lump it into "physics".

Not to offend the text crowd, just that here we are on TN a very remarkably admirably FREE TO PLAY text service. :) The key is cost, I think, but again, I don't know much about the process nor the client technology (Cost Centers A & B, collectively).

Oh, last last note... I totally got a kick out of that old TN thread. Great stuff. LFMAO "yeah yeah yeah, viva la revolucion..."