Time For User Owned Sims?

I think second life is so big now, and the user base split so far into their own camps, that Second Life is ready for Epoch II, the age of independently owned sims.

In this epoch, creations you make can be exported as XML files, and are stored on an asset server run by people who run a cluster of sims, so some people would have one sim and an asset server, some four sims and an asset server.

I know, you're all thinking the owners of the asset servers would steal your inventions, but it's no more of a worry than worrying that the person hosting your PHP or ASP site will swipe your PHP or ASP code.

To get around these worries, we'll just have to do it the way we do it with web hosting in real life. Face the facts that a busy host with a lot of customers probably doesn't have time to care about every little thing going through their servers, and if your creation is totally valuable, get your own server.

In time, maybe the big hosting companies will start offering "MS Windows, Linux or SL managed servers".

And the software to run a sim should be called Second Life Sim Management v.x or whatever, sold with a license for use any way you see fit, on a per server basis. Or maybe per sim, but per server would be nicer for the customer.

But anyway, it's time for SL to split up and become an Internet, as per the vision of the CEO Philip Rosendale, in the beginning.

This might work. LL will only have to absolve themselves of one great myth for this to happen. If they will simply admit that one centrally managed asset server does not in any way secure our rights to shapes, textures or scripts, the door will be open for other companies to host parts of the grid.

Baby steps, I like this.

I wonder if their license agreement with Havok allows them to do this.

I don't think it is ready.

Software quality needs to improve. SL should look and perform like a product into which someone invested $20+ million dollars.

Is SL database independent? Can I run it on MS SQL, Oracle or DB2? Because you don't know what DB a customer has or wants. If you want to sell this to MS, it had better support MS SQL.

It needs to switch to a nonproprietary 3D building format. Proprietary formats are "teh suck." A big change we aren't going to see in SL2 even.

Can I hand the install files to a person that is not familiar with the server software to install it with minimal support or do I have to hold their hand the entire way?

Is the software code and DB architecture documented? The latter is more important than the former unless they are going FOSS (Free Open Source Software).

What state are the tools in to administer SL servers?

And so on and so on.

As far as the internet is concerned, SL has some scalability problems in its design aside from the asset server. For example, depending on first and last name for login is hardly scaleable.

Plus, how will SL handle asset transfers between different sites? This is an interesting scenario where each site has full control over their own asset IDs and user IDs. We have distributed asset system that we designed where I work that has a final, final, final fallback solution of big lawyers with big guns. But for most little indie players out there, this isn't a real option.

Right now the most valuable thing about SL is its residents. If SL wants to be the next internet, it got a ways to go.

SL is not the new Internet, it's the new Usernet.

It's a visual platform for trading data. Chat, messages, items, images. The internet does none of this BY ITSELF.

Browsers allow you to look about and see data stored on web servers, while news-readers allow you to browse data stored in large databases of communication.

Or am I missing something?

And the SL client enables you to experience a 3D interactive environment. Something wich was promised with VRML but never really took off.