3pointD linked to an interesting article on Ogle Earth entitled "Metaverse 2.0" which talks about the fact that Google Earth and Second Life are modeled on the real Earth and a flat Earth, respectively. It argues that SL in particular suffers from a failure of imagination in mapping its world onto a boring, 2D grid and offers both hexagonal, hyperlinked topology and a hyperbolic space as solutions. It is a pretty nice piece and covers a lot of territory that we considered early in SL's history -- back in the "Linden World" days of early 2001 -- so I thought that I'd talk a bit of history as to how we ended up with the 2D grid. Also, I find it interesting that it is completely possible to create the desired world within SL, despite the underlying topological choice.
So where was Linden World in early 2001? At that point we had block-compression working for land transmission, the ability to edge tile simulators, the ability to fly around the world, a spring-membrane water simulation that also crossed region boundaries, and a Jos fluid for the wind. Most of our thoughts at that point were around the combination of a living ecology and cellular automata simulations of various systems, and the topic of overall topology came up quite a bit. Among those that we considered were a grid, a hex-grid, a sphere, a cylinder, a hyperlinked grid, a hyperlinked hex-grid, and even the hyperbolic place talked about in the article. Each of these had strengths and weaknesses.
The simple grid had the advantage of already being done, was conceptually simple, was easy to stick CAs and height fields on top of, and mapped well onto the real terrain that we thought would make up most of Linden World. On the other hand, it isn't very sexy and put sim centers relatively far apart for a given area -- damn Pythagorus! The hyperlinked variant was dismissed because by the time we thought of it quirks in how we managed agents would have made it a pain to handle loops, which is shame because it offers up some interesting features that we'll have to do the hard way via portals and remote cameras.
A hex grid would have been nice for all the same reasons that table top games use them. Edge connected CAs across them would have been slightly more difficult and of course it wasn't actually written yet. In the end, hex grids lost out to a Cartesian grid because we didn't have the resources to tackle problems that weren't absolutely necessary, and the benefits of a non-hyperlinked hex grid over a non-hyperlinked regular grid are marginal. Similarly, the hyperlinked variant is only a bit better than a hyperlinked regular grid. Yes, you get to take one step to 6 neighbors rather than 4, but the regular grid actually gives you 8 neighbors since diagonal moves are allowed, even if it is a slightly longer walk.
The cylinder and spheres were rejected because unlike Google Earth we don't know how big Second Life will actually become and once you start filling in a fixed geometry you have the unfortunate problem of having to either grow the object -- which does work for the Universe, of course -- or having to duplicate it, neither of which seemed like the right option. From a rendering standpoint it might be nice to add some curvature to Second Life, but that is independent of the topology of the land allocation.
The saddle is probably the geekiest option and was suggested to me by a math professor friend at Berkeley. I think it's very cool to see someone else think of it! As mentioned in the article, what's nice about curving a plane onto a saddle is that you get completely different tiling properties, including the option to allow more than 4 squares to meet at a corner -- column 4, row {5,4} I believe. This keeps the world closer together so you can walk to more neighbors easily. The problem is that rendering it in a way that makes sense turned out to be more work than we wanted to try to come up with.
So, that's why we have a 2D grid. But are you really stuck?
In Second Life today you have global teleporting with local, recognizable geometry. I completely agree that we need to make teleporting -- and semantic connections -- easier to create and to use, but they are there. Hell, 2 seconds after we fix llHTTPRequest to use HTTPS I'll have my del.icio.us tracking script posted here, although in the long run we probably don't want to rely on external sites for that sort of tagging. But that aside, you can create nearly all of the desired benefits of the article in Second Life today. In fact, the balance between mainland and island purchases explore those questions and opportunities every day.
Moreover, the question of the value of space and place is a deep one. The web is not a communal place, so hyperlinking may be a better fit than for Second Life. Obviously, we thought so, because from the beginning we sought to create a world that had place. A lot of the impact of that decision is lost in short draw distances, delays on draw-in, and the like, but as we solve those problems, it will be interesting to see how travel breaks down between teleporting and actual travel between places. Teleporting is superior today, but it isn't clear that it is a global maxima. Instead it may simply be a local one created by technological limitations.
(Just noticed that Avi already hit some of these points as well)
NASA has recently announced that they are seeking offers for educational programs "for creating and managing innovative activities, events, products, services, or other types of formal or informal education methods for the purpose of disseminating information nationally about NASA’s projects and programs."
While the synposis clearly is targetting large companies, given some of the work I've seen in SL this year on space science and education, I bet that someone in SL could pwn this.
Given how often the amateur versus expert issue raises itself in relation to Second Life -- whether in terms of content creation, medical research, game creation, etc -- it seems somehow appropriate to watch Raph and Prokofy slug it out over on Raph's blog. In the blue corner we have Raph Koster, till recently SOE's Chief Creative Office and Big Thinker (tm) on Virtual Worlds. In the red corner, Prokofy Neva, undisputed heavywieght champion of extended posts, big thoughts on Second Life, and continuous challenges to authority. Go read it -- it's interesting.
I haven't felt the need to post, but one sequence caught my eye. Raph posted the following:
I personally believe this is a mistaken take on things. I’ll be bold and say:
- no, it’s demonstrably less fun, to the vast majority of people, in SL than in most any of the gameworlds
- the gameworlds have historically been what has driven adoption in virtual spaces
- the gameworlds have historically been what has driven lasting innovation
- the future of the metaverse is going to come from gameworlds, not freeform social worlds
I don’t want to overstate this, but in short, I’d issue a plea to folks like Prokofy, Jerry, and others, to let go of social world exceptionalism
To which the only reply, of course, is:
Would you be willing to give up on online game exceptionalism? :-)
Since entering into the argument at this point would be silly, I'll simply pour gasoline all over the place by asking a couple of questions:
- How many people use the web versus gameworlds? (Being generous to the games, it's 20 to 1)
- Raph is quick to dismiss Second Life as "just a MUSH" -- and leaving aside the separate 3D versus text debate that I've sometimes ventured into -- I still find the claim that innovation is happening in gameworlds somewhat challenging. Given the number of announced gameworlds trying to pivot onto Second Life innovations -- user creation, collaborative creation, user markets, content markets, pay-to-play, no subscription, digital delivery, single shard, streaming content delivery -- what are the innovations in the gameworld space?
As to the question of fun or usage, debating it does seem a little silly. Second Life has show gentle exponential growth since it launched. Either that trend will continue or it will not. If it does, exponential growth combined with the fact that we address a much larger market than games means that we will dwarf gameworlds. If not, we won't.
So, do you want to bet on Linden Lab's employees plus Anshe, Perkofy, Walker, and the hundreds of thousands of other SL residents? Or do you bet on Blizzard, Raph, Daniel and the hundreds of other really smart folks making online games?
Hopefully, you won't have to make the choice, because if we do our job right, many of them will ultimately be building games within Second Life.
It seems like we're reaching some sort of critical mass for mixed-reality events. Within one week there are/have been:
USC's Center for Public Diplomacy's Awards Ceremony, which was attended by more folks in Second Life than in the real world. 3 out of the 4 finalists were created within SL. The very polished Peacemaker won, but Hydro Hijinks made it a tough decision.
The Berkman Center's Beyond Broadcast conference is happening right now. Focused on the future of public media, it has been very interesting watching virtual worlds recognized as legitimate option.
This weekend will see BBC Radio 1's One Big Weekend within SL. Two days of free, great music and now you don't have to get to Dundee to enjoy it!
Of course, the real questions all resolve around how valuable these mixed reality events are when compared to a web cast. The mixed reality event allows remote participants to also connect with each other but at the cost of more technical requirements and bandwidth. The ultimate success of these events will be determined by whether that tradeoff is worth it. I was smiling during the USC event because there were a ton of side discussions going on in the SL audience. It reminded me of being in the back row of a session and making comments or jokes with other folks sitting nearby. Audience chatter can become loud enough to constitute real-world griefing yet it can be among the most important and enjoyable parts of attending. I've had more than one long-term friendship instigated by a joke or flippant remark made during a talk. I suspect that the same will become true in mixed reality events, where it's the unplanned interactions of the audience that lead to new social connections and fun.
Virtual memorials in Second Life. Think about it . . .
I was chatting with some smart people at the Berkman Center last week and talk somehow blundered into personal memorials and it hit me that Second Life would be the perfect place for this. A little web searching revealed that a thousand people are trying this on the web, but they're missing the point. Memorials, both as part of the grieving process and family history, are meant to be experienced with people. With family, friends, loved ones. Going to a website to see pictures of a grandparent is not the same as being able to stroll through a recreation of their garden, to watch a home movie on the television from a copy of their living room, or to do a family gathering on their birthday with relatives from all over the world to spend an evening sharing memories. Maybe even a regular update to the memorial to blend in the lives of their children.
I suspect that this might be best done in Europe first -- given a somewhat more pragmatic approach to end-of-life care and decisions -- but with a graying world population, data indicating that Second Life is more appealing the older you are, and an increasingly tech savvy elderly population, why not spend the next several years working with your parents to make a place to remember them? Why not let people help create the places that will define their memory?