the foreground lights dim as the camera closes in on philip linden (portrayed by christopher lambert.) viewable only in silhouette now, we see an exhausted fighter, his sword already falling from his hands. the decapitated heads of vivaty, metaplace and forterra are still rolling as the bodies that once supported them fall in slow motion to the ground. lightning dances around the industrial setting, lighting philip's face for a moment. queen's "princes of the metaverse" plays in the background as the words of mitch kapor (played by sean connery) echo in the ears of our protagonist... "there can be only one."
or at least that's the image i sometimes get when i hear when about virtual worlds failing. a lot of people have been trying to extract cash from the virtual world, but it's a tough market. and a number of people have commented lately that in the post-hype world, second life™ may be hurting, but it's at the head of the pack and has the market for social virtual worlds more or less locked up.
but is that true? can there be only one?
just to clarify, i'm talking really about public, social virtual worlds. i'm not talking about virtual world platforms or small, departmental virtual meeting rooms. i'm talking about full-on massively multi-participant experiences.
OpenSim is an excellent project; as is SimianGrid. both are open source projects intended to replicate the functionality of linden's server software. the former seeks to maintain compatibility with linden's existing protocols while the later is more focused on next generation VWRAP protocols.
but neither are virtual worlds as much as they are open source software projects. they are both excellent projects, but they both explore the "solution space" of the virtual world domain and not the "problem domain."
no... what i'm talking about is public virtual worlds, with's lots of space for a bunch of people to hang out in world and do those things that virtual people do. this sort of knocks out Croquet / OpenCobalt (which is also mostly a "project") and (sadly) vastpark and teleplace. they're not bad concepts, they're just not massive.
if you look over at the OpenSim Grid List, you see that there are several grids out there, but the number of people visiting them is tiny. why is this? why is it that OSGrid, the largest of the OpenSim grids listed, is only attracting one 200th of SL user base?
my guess is it comes down to functionality, economy and community.
kvetching about features OpenSim lacks out of the box seems to be a popular sport these days. but it probably doesn't do much for the world to repeat some of the more extreme arguments. at the end of the day, some people like the second life permissions system and some people don't. some people like in-game currency, other people don't. but it seems that of the people willing to pay money for virtual space, most people either want those features (so they can do commerce) or don't really care if it's there or not. it seems that few customers are actively opposed to operating in a virtual world with cash and permissions.
without a turnkey system for starting a grid with cash transactions and a robust permissions system, the cost of implementing these features falls to the grid operator. without the virtual economy, it's a little harder to attract people to your virtual experience, and thus your community remains small.
so what can you do to grow OpenSim or Simian or OWL based virtual worlds into mega-monster avatar playgrounds? simple. stop trying to be second life™.
i think the answer to the question "can there be only one second life?" is yes; there really can only be one. but you can go on to have a lot of virtual experiences that are noticeably different from SL, and in many ways product differentiators. so let's not talk about things in terms of second life, and start thinking in terms of "next generation virtual worlds."
the obvious one that everyone is talking about is, "make second life run in a browser." there are enough people out there talking about Unity3D, WebGL and OnLive that a quick round of googling can find you a depth of opinion. (for extra fun, search for "+rezzable +unity3d +opensim".)
the next obvious one is, "make your virtual world more social." it's clear from linden lab's recent moves that they've heard this one. they interpret it to mean "follow the money, integrate with facebook."
another differentiator that is near and dear to my heart is "let info flow in and out of the virtual world." Linden Lab introduced an amazing new feature recently with it's media on a prim and shared browsing experience. but the truth is, putting a web page up in a virtual world isn't exactly new. heck, even sirikata did it. NanoCosm was doing it back in the late 90's. but that's the simplest way to get info into the virtual world.
how 'bout we let in-world groups map to facebook groups and twitter groups? why can't i easily see my facebook friends and twitter followers in world? for that matter, why can't i just log in using my twitter credentials. they _do_ support OAuth, after all.
and assets. Second Life's asset system requires you to upload images and what-not to their servers. why couldn't i just point to a texture or a COLLADA .dae file out on the web somewhere and say: "when i drag this item out of my inventory, i want you to go out to the web to get the data used to render it."
so i think my key point here is... don't try to recreate second life. it's been done. it attracted a fair number of people and a lot of hype. move on, do something better.