Worlds.com Sues NCSoft Over MMO-Patent

Lulfas writes "Worlds.com today sued NCSoft over its patent on a scalable virtual world, filed in 2000 and granted this February. This is a very broad base patent, and there is no reason to expect they will only sue NCSoft, when they should be able to use the same patent against other companies. 'Specifically, the suit claims that NCsoft has infringed on patent 7,181,690, "System and Method for Enabling Users to Interact in a Virtual Space" through its games, including City of Heroes, City of Villains, Dungeon Runners, Exteel, Guild Wars, Lineage, Lineage II, and Tabula Rasa.'"

Comments

Prior Art?

[LabRat]

Filed in 2000? Um...shouldn't be too hard to show prior art to overturn it if the patent indeed is applicable to operating MMOs. UO and EQ for sure...probably even text-based MUDs.

Correction...

[aztektum]

It was granted in February of '07

Anyway, this takes some gigantic balls. Granted I only read the abstract, but Ultima and Everquest were active before this shit patent was even filed.

<melodramatic rant!>
Only when we can throw patent examiners in prison for such gross negligence will we have true patent reform!
</melodramatic rant!>

EQ might be better than UO for this one. . .

[JSBiff]

If you look at the link to the actual patent, and begin reading the claims, this does not apply to the (original version of) Ultima Online, or to text muds, because the patent specifically describes a 3-dimensional graphical world as being part of the claims. UO is (or at least was, last time I played it around 2001 or 2002) 2-dimensional. Right about the time I was leaving, they introduced an expansion called Third-Dawn, which still didn't make the world truly 3D, but it did make player avatars and monsters 3D, IIRC.

EQ, as I recall, was true 3D (I only played a trial account for like 10 days once, so my memories are rather vague), so it might be a good candidate for prior art.

It should be noted that the patent does not appear to cover (I don't know for sure; I'm not a lawyer), the idea of a 3D MMO, per se, but rather a few necessary client rendering techniques (which, in reality, almost any 3D MMO would be likely to employ) for determining what other users' avatars should be displayed by the client. It appears the idea they are trying to patent is that, in a 3D world, when you turn the camera to look a given direction, you should only see some avatars, and not others (that is, only the ones in your field of view). Additionally, if there are a lot of avatars, this patent claims protection for the idea that the client can implement a maximum number of avatars to display, and to use the knowledge of the maximum number to display, combined with the position information, to determine some subset of the avatars to display (presumably the X nearest avatars, where X is the maximum number to display, though the patent doesn't specify this explicitly).

I'd be shocked if EQ and Meridian59 didn't both do these things several years before this patent app was filed.

I'd also like to point out, that the patent doesn't specify 'camera orientation' or 'client view orientation' (even though that appears to be what they are trying to cover), but rather 'avatar orientation' (which suggests to me that this patent would only apply to MMOGs where the camera orientation is locked to the avatar orientation). Based on my 3+ years of playing CoH, I can tell you that the CoH client doesn't determine which other avatars to show on screen based on the orientation of my avatar - I can spin the camera freely to point in any direction, even look completely backwards from the direction my avatar is facing, so I suspect that NCSoft could claim that as a defense, if they had to.

Also, I think they could, maybe, make a defense against claim 6 (I'm not sure though):

6. A method for enabling a plurality of users to interact in a virtual space, wherein each user has a computer associated therewith, wherein each computer has a client process associated therewith, wherein each client process has an avatar associated therewith, and wherein each client process is in communication with a server process, comprising:

(a) monitoring, by each client process, a position of the avatar associated with the client process;

(b) transmitting, by each client process to the server process, the position of the avatar associated with the client process;

(c) transmitting, by the server process to each client process, the positions of less than all of the avatars that are not associated with the client process; and

(d) determining from the positions transmitted in step (c), by each client process, a set of the avatars that are to be displayed.

Now, I could be wrong here, but I thought most client/server 3D game protocols do *not* have the clients transmit the position of the avatar to the server, which is part (b)? Don't the servers already know the position of the avatar, and the clients just send a vector, that is, a request to move a certain number of units in a particular direction, at which point the server calculates a new position from the original posti

Re:EQ might be better than UO for this one. . .

[Fluffeh]

*cracks knuckles*

Being an avid MMO player from UO to current, as well as being a 3D developer on the Unreal Engine, I can make for some useful input here.

The original UO was indeed a 2D isometric client. Third Dawn brought 3D aspects to the game, though the world was still rendered in isometric view. It just looked more 3Dish. There were however mods/ports to UO that rendered a 3D world! They were buggy as heck though, but you could (in theory) play in the UO world with a 3D client. (Though it was a third party client).

Now, EQ was a true 3D world. No if's or but's about it.

The whole bit about the number of avatars to display, that's more really related to engine/hardware performance. Basically, when developing an environment developers need to look at polygons on a screen. This is one of the main restricting factors in developing a world. The more polygons a graphics card has to render, the longer it takes. It's not the only thing, but at the time that this patent was filed, it was certainly one of the most important things.

Now, avatars in a game generally have a much much higher polygon count than the environment around them, so naturally restricting the number shown might be very beneficial to rendering a game world nicely, however, soon after this patent was originally filed, a bright spark came up with the idea of not removing entire actors, but adding a LOD factor into models. Basically, it means that the further something is away from the camera, the less polygons it will use. This can also be ingeniously used to reduce the polygons per actor/model in the camera view if the number starts getting too high.

To use the obligatory car analogy, if I want to limit my game engine to displaying 100 polygons on the screen at a time, I can render a car with 100 polygons, but also allow the code to reduce these polygons to 50 if a second car comes onto the screen. Should I need to have 10 cars on the screen at once, they would each be reduced to 10 polygons.

Don't the servers already know the position of the avatar, and the clients just send a vector, that is, a request to move a certain number of units in a particular direction, at which point the server calculates a new position from the original postion + the vector?

The section you write about actors and client/server relative positions is sort of right. It's not far off anyhow. Here is how things work in just about all client/server applications now.

The server sets the original position to the client - when a new level loads for example. The the client sends data as to where it wants to go and both the server and client move the character around. Now, due to a number of factors such as latency, packet loss as well as a number of others, the two locations will become out of sync. However, rather than the server being updated with the client's location, what happens is that the client is updated with the server's location of where it things the character is. This sometimes leads to what is known as the "slingshot effect" where characters (or other actors) suddenly update in the client view and appear to slingshot around the screen to catch up. A notable exception to this rule is World of Warcraft, which does actually have servers that will quite happily allow a character anywhere in the world that the client lets them get. This has resulted in some rather funny "exploits" where people altered their clients and walked past mobs to get to the final boss in an instance and then just started attacking it. This may have been fixed in one of the expansions however. While I am not entirely sure, I think that D&D Online may have also suffered from the client updating the server with actor locations, as I recall there was a considerable number of movement hacks and exploits in that game.