Finally, after enough involvement over enough time to become "that guy" to more than a few people, I've stopped using, reading, and caring about Second Life. It's not a very forward looking system, its being locked up by a single vendor preventing the development that could make it great.
Starting from the basic fact that they're both San Francisco startups, I've always seen parallels between Linden Lab and Six Apart. That makes it hard to actually argue what I just wrote without slipping into or having to defend its business model: selling an internet service on which it has a natural monopoly through source code control. I could argue that my employer's business model is not something I can control, but thinking of it more locally, my "business model" rests solely on the company's. At least, that's so if we were speaking morally instead of economically—and if economy is the science of scarcity, doesn't it have great moral implications?
Instead, I would argue actual differences. We aren't the only vendor of blogs; in fact, we're in a pretty competitive space, competing not only with other blog services but other supplementary products (photo hosting, social networks, etc).
Really, though, Second Life is not a community to which I belong. I might find it useful to start doing my own thing, cultivating my own little basement metaverse—it's an unfilled facet of my life as of a few months ago—but at this point it's hard for me to want to replace Second Life, too. I thought I had some good ideas about building a new internet community (in a larger context than, say, the social MUD theses I wrote a while ago, I mean). I started writing this post probably a month and a half ago, with a very different focus. I had a plan, a rudimentary program design. I even imagined a brand identity, and bought the domain name.
This is the crest of a wave, though. Had I acted when I was first developing the idea, I could have helped lead a movement toward a more open system. Garages are starting to hum with the activity of cloning and refining the formula Linden Lab is selling so hard. The basement pirates are already building (not to genericize jarodrussell's term of art too much). At this point even the bloggers are realizing a closed metaverse does not sustain (despite my best efforts I lost the link I was going to cite—I hope I didn't imagine it).
However I like to think duplication is not my cup of tea. Plus things are going better at work, so the fantasies about my own startup are firmly back in the realm of fantasy. (Good thing too, because going into business up against so established a player as Linden Lab in a space about to be so crowded is surely crazy talk.) Plus I spend barely any free time alone at home, so I don't have any time I could build such a thing, anyway.
I guess I should talk about the program plan I was developing. The idea was guided by three precepts:
- Distribution. Linden Lab has shown that you can get amazing horizontal scale by sharding geographically, connecting servers at the physical edges. However, that doesn't let you scale where you'd most like to: inside one place. To step that up, you'd have to model the world in a differently (you might say "more") distributed architecture. You need to solve the problem of how to increase computational density of a space, and scale inward, not outward.
- No invention. There's a bevy of free infrastructure parts for delivering data and user experience. It happens to currently all be optimized for experiences shaped like web sites. Use them. (I think this is what drove me from MUDs: you can't reuse components when your system is some ancient C monolith, instead of a pluggable n-tiered service. Hell, split the difference and be a gooey UNIX pipeball, just give me options.)
- Modest goals. Even using commodity parts, Linden Lab has a big head start (in marketing, if nothing else). Unless you have some field-changing advantage, you can't compete in their niche--mdash;but then, you shouldn't, anyway. Imvu is in a good place here by being differently good.
Based on these axioms, there were three possible products. The first was I (one), and it was the latest evolution of the MUD I always wanted to build. So you have a universe of objects; make each object its own service on the internet, exposing an API by which they can communicate with each other. "Hi, Room, I'm a Thing. May I move myself into you?" "Hello, Thing. (I see you're operated by Player, who is trusted by my operator with that permission.) Sure! Here are my description, exits, and the URIs to the other objects I contain..." Make a gateway service that's a telnet server, which presents itself to the world as an object you inspire, and converts your commands into messages to other objects and messages from other objects into text output for you to read. Very distributed, and so modest I doubt I would bother making it for its own sake.
But from that idea, it's a small leap to the product I called 3D: render the world not in text but as a 3D scene. It's the exact same thing. You can pull textures and rich content over HTTP to leverage web infrastructure parts. (The main reason I started considering seriously the entire possibility of competing services to Second Life is because I found with a little research that the core renderer for Disney's Toontown kids' MMORPG is open source, co-maintained by Carnegie Mellon for use in computer/media coursework.)
In the middle of the two is the path I considered most seriously, and which I was calling 2.0. This is something like the same thing, but instead of using a that speaks MUD, or a 3D program, the interface is a dynamic web site.
The truth is—and I realized this most only just now when I wrote these here descriptions of I, 2.0, and 3D—you could make one contiguous metaverse that supports all three interfaces. It would lay the burden of making three interfaces for every object on content creators, which you wouldn't want to do, but it would be possible at least.
Ultimately, without the work, these ideas are worthless. The ideas of anyone who produces running code are greatly more valuable. As I learn I don't have to do everything (because, as you might expect, I can't do everything), I feel more like it's fine if I don't participate in this movement. On the other hand, if this new energy truly is a rising tide, and if I aspired to compete, then the name I picked would be eerily prescient: I chose Waveshark.