What would make a good competitor for Second Life?

What would make a good competitor for Second Life? What can you think of that would make you drop SL?

1. Havok 3?
   
2. Mesh avatars?
   
3. Custom avatar skeletons?
   
4. Javascript instead of LSL?
   
5. Java (Jini/JVM) instead of LSL?
   
6. Mono instead of LSL?
   
7. Client-side XML+Javascript GUI?
   
8. Client-side XML+Javascript avatar control?
   
9. Scripted animations?
   
10. Custom bump-maps?
    
11. Multiple cameras per client (eg, 1st-person with 3rd-person picture-in-picture)?
    
12. Server cameras (mirrors/reflections, remote cameras)?
    
13. Private spaces (basements, lands)?
    
14. Download/upload for full-permission content?
    
15. Local space (rez your av and full-perm content inside your own computer)?
    
16. Shared local space (allow a client on the local LAN to log in to your local space)?
    
17. Scripts on land, with events like "access request" handled by your own code?
    
18. Local applications viewable in local space?
    
19. Local applications viewable in shared local space?
    
20. Local applications vieweable in-world?
    
21. IM/Email direct to character name (mailto:argent.stonecutter@resident.secondlife.com)
    
22. What else?
    

I mean, the question is not "Would a better competitor kill Second Life?", it's "How good would a competitor have to be to kill Second Life?".

What are your hot buttons?

I say Garry's Mod. GMod's biggest weakness, the one thing that keeps me from completely abandoning SL, is the lack of decent building tools. The second someone creates better controls for building that the Physics Gun, SL loses the race.

GMod already gives you the ability to run your own sim, with the server option. It lets visitors come by your sim. It lets you build inside your sim; you can both save what you build and save the state of your sim; and best of all it's very possible to both import and export objects to-and-from GMod.

take everything good about SL and make it where we can operate our own sims with our own hardware

(heh i didnt even pay 1200$ for my own computer)

More Cowbell.

Agreed, my view would be a distributed system something akin to how Neverwinter Nights worked; You can run a persistant server and then set up "links" that drop you to other people's servers. In the other words, the grid model with sims.

The only down side is there's no asset/login/data server to link everything together, so it becomes more like a P2P network than a centralized client-server model. (I'm talking, between sims, not from the user's viewpoint.) So, sending avatar data around might be rather cumbersome, since unlike NWN your character sheet won't probably fit in a typical ethernet frame. Also, how do you handle an inventory like we have in SL?

I'm feeling the cowbell.

being able to host your own islands and not having to deal with their non-existant tech support would be cheaper, and provide better support.

HTML + GUI API + import meshes / other 3-D program files + easier tool for new users

A place where there is more people and more money.


I don't think technology is important. If people could earn money by posting crap on these forums then Second Life the program would be no more.

One absolute necessity for any really large-scale virtual world would be that land prices must be many orders of magnitude smaller than in SL. The big CPU hogs here are physics and scripts, so a lot of this would have to be done client-side, and some could perhaps be distributed.

Another thing is that it would have to be possible to make a good search engine. I doubt this will ever be possible with SL. The two current systems that have some potential in this regard are Croquet and VOS(Virtual Object System).

Prims, or something similar, could very well be used for a lot of things, but high-precision 3d data is also necessary. There's just too many things where it's absolutely crucial with a much high precision that can ever be possible with prims. How to do this however is another discussion, one simple way would just be to require predownload of data before going to certain locations.