Posts Tagged ‘virtual worlds’

Lively, and the Braid

November 20, 2008

Today’s tech news is that Google is killing off Lively, its “Second Life killer” project. This isn’t wholly astonishing — while it looked very intriguing at first blush, there just wasn’t much There there.

That said, I don’t think this says much about the virtual-worlds idea in general. Yeah, there’s some faddishness to it, and yes, it continues to be marginal. But seriously: I still don’t think anyone has done a very good job of implementing it. Heaven knows that Lively wasn’t very impressive, precisely because it was too limited — basically just self-designed chat rooms with 3D avatars. That’s amusing for a little while, but it provides little benefit over and above text chat rooms, with massive overhead.

There is a moral in this, and it’s why I’m posting about this here — conversation systems need to be as natural as possible.  The form and technology strongly affect the conversations.  I’ll talk about this more at another time, but the immediate point is that you can’t build a successful conversation system that is a hassle.  No matter how cool it is, people just won’t use it.  (This is why I spend at least half my time on CommYou trying to make it easier to use.)

Anyway, it does surprise me a little that this hasn’t been done right yet; my best guess is that this is simply because there isn’t an obvious business plan in doing it well. But I’m still quite clear on how virtual worlds *should* work: it’s more or less what I described as the Braid a number of years ago. High concepts of the Braid:

  • It should be 100% decentralized — anybody can run servers that host however much “space” they want.
  • It must be based on completely open protocols, and there should be straightforward if basic reference implementations for all the pieces.
  • The “space” absolutely, positively *must* be extensible in some fashion, without a simplistic Euclidean model constraining it. This could be done with either Mark Pesce’s Cyberspace Protocol or (better, IMO, but weirder) my Portals Protocol.  (I can’t find the full description, sadly, but here is my original thinking about it, and more can be found in the www-vrml archive.)
  • You have to be able to *build* things in this space. People should be able to create objects, and those objects should be fully tradeable in a trusted way.
  • You have to be able to *do* things in this space. There should be at least a clear baseline concept of physics, and an ability for spaces to define extensions to cause-and-effect. The space needs to be programmable: the owner of a space should be able to define how that space *works*.
  • Combining the above two, it should be possible to build *stories* that happen in this space. At that point, you have a distributed system for writing at least simple RPG-oriented videogames, which I believe is the initial killer app to spread the thing around.

Granted, that’s a tall order — when the VRML project started in ’94, it was probably unrealistic for me to expect that this was where it was going. But at this point, it’s by no means rocket science: all the necessary elements exist, and many of them are commodity code by now.

A lot of projects have sprung up over the years, and many have gotten some of these pieces right, but nobody has put all the pieces together yet. The best progress has been made in for-profit ventures like Second Life, but most of those are proprietary (or at least started that way) with the result that they don’t *think* about the problem ideally — they weren’t designed for open protocols, so they aren’t as powerful and flexible as one might wish.

It’ll happen one of these days, though. For all I know, it’s possible that someone *has* done it, but hasn’t gotten the killer app part needed to get lots of people using it. I’m still quite sure that the potential is out there, but I don’t think it’s going to be realized until people think of the “space” — what I call the Braid — as a commons, in much the same way that the Web is. When it is open enough that *anybody* can build in it and do cool things in it, whether for profit or for free, you’ll see the explosion of creativity that will make it real.

Have you seen virtual-world systems that you think have real potential?  Do you think the whole idea is a dead end?  What do you think would make a virtual world interesting and worthwhile enough to spend ongoing time in?  Do you already do so, and if so, in which systems?  This is a bit of a tangent from our usual discussion of conversations, but it’s an important one: if virtual worlds do become more important, they’ll eventually be an important medium of conversation unto themselves.