The future of SL: Why not help make it happen!

Let's now talk about scalable administration and what it means to be a Linden...

Lindens are an interesting species. They are very similar to the moderators on any forum. Some have the rights of a server administrator. Overall, they are those that maintain both the technical and social peace.

They are also the most visable target for the wrath of the users, but that's another discussion.

Now, in a traditional system, it normally only takes one admin to maintain the system, since most of the varied functions are automated. Even massive systems can be maintained by a minimal number of techs.

Now, the problem is that SL is not only a technical system, it's also a social system. Since people are not as easily maintained as predictable technical systems, you have a problem where you need to have a minimal number of moderators based on the size of the 'world' and the number of users.

Now, the current system works, but not in a truly scalable fashion. Many think we don't have enough Lindens to fill the demand, and it's even possible that they are right. So, how can we fulful the need for lindens without putting undue stress on Linden Labs?

Let's go back to the feudal system.

Think. Private sims are the future. Land barons own vast tracks of land. They are profiting based on them, but still have the same problem the rest of us do.

They don't really control the sims.

Now, what if every Sim owner was ALSO effectivly an Administrative Linden? Not everywhere, but only on their own land. What if they ALSO have the ability to make moderative lindens to help mediateproblems in thier sims?

Sound good, but what does that mean?

Well, they would first of all have to have access to information that would make some people scream. They would need access to full logs for the sims they were lindens in. They would need to be able to do all of the tricks normal lindens do, and would have access to potentially embarrising or hurtful data.

On the plus side, barons would be able to designate their own 'knight lindens' which takes the tress off of LL. On the minus, you would have to agree to losing another level of privacy.

<grins> Oh well!

Sim owners would have access to ALL of the tools of a admin linden. Logs of the sim function, scripts running, and even reboot ability would be possible. This does mean that sim owners would have to have a technically trained individual on staff, but this is no differant than owning any other kind of server.

Now, let's open the floor to your opinions. Is it a good idea for the evolution of SL, or a nightmare come true?

You present that choice as if those options are mutually exclusive.

Eventually servers will not all be owned by Linden Lab, and anyone who runs a server will be effectivly god in their sim.

The question is what powers will simulators have? How do you build a massive network of untrusted systems?... The Internet!! Of course...

Most of the problems have already been worked out. Simulators don't have access to your system, or other details except those registered with them.

So if you store your objects with yahoo, you send a request to yahoo to squirt your objects at the other server..

That server decides if you have high enough permissions to rez an object into their world.. If you do ;0 There it is!

Maybe that object is scripted.. Does it run on the simulator? Maybe, or maybe the script for that object is run by your own client. Or yahoos servers... Or some other server somewhere.



The question.. What powers over a simulator does the owner need to have? They need to right to decide who accesses it. Or who can view it... who can interact with it.. Who can use that bit? who can use that other bit?

Here is another good question that came to mind when talking about SL to someone.

When SL goes to a distributed model, and becomes a platform distributed across the internet, what will happen to the eclectic community that is Second Life?

Will SL as it currently exists continue to be developed? Will it continue to have the number of users that it had before, or will the total number of players slowly dwindle as servers appear across te internet? Will communities become insular, with the elves having one set of servers, and the furries having another?

What will make the new SL platform into a COMMUNITY?

As of the moment, I'm not sure. Perhaps a distributed system to keep track of user identity would be a good idea after all.

Think of it this way: The web has DNS server that keep track of sites and attach normal-sounding names to IP addresses so that you can find them easier.

What if we had SLidentity servers that allowed us to keep trach of USERS. They contain the user profile and secured information for purchasing in SL, and SL servers contacts THEM for needed info.

It could work like this: The SLID server has your Credit card information (yes, they're insured, let's move on), and you want to buy something. You buy it in the SL server, and your client sends an authenication code to the SLID server as well. The SL server asks for money from the SLID server, and it matches the request and security code from your client software. If they match, then the transfer of funds and items goes through, with the SLID server mediating the transaction.

This means the SLID sever acts as an online banking system, securing funds for a few seconds and then when the transaction is confirmed moving them on to the proper parties.

Thoughts?

Foolish, you never stay with any topic for very long eh?

The way you describe Second Life ID servers it sounds a lot like paypal. I really don't think any service like that would be linked specificly to Second Life, when you can offer such a service to anyone using the internet for any purpose.

As to the eclectic community of Second Life; I have always been able to find that community wherever I travel online. Sometimes you just have to look harder.



Second Life has terrabytes of content already. It's an online shared virtual space where the users can interact with other users and the surrounding areas in real time. Nothing is stored clientside except for a small ammount of local cache.

It's like a thin client. Everything you have in Second Life is stored somewhere else. If 3D graphics and content can be done real time over the internet with an acceptable level of detail, could Second Life become the next platform for 3D games as well as the 3D web?

And by game I mean a game on the scale of World of Warcraft or Doom 3.

<looks about>

Baba?

Did you kill off all the other attendies again?

I think that I am so thorough and obviously correct that they feel they do not need to reply

Explain a little more about the part "the next platform for 3D games" and "on the scale of World of Warcraft or Doom 3".

It's mostly speculation anyway ;0 If the streaming 3D content and related technology were to be advanced enough, they could be used to develop games similar to World of Warcraft or Doom 3. A player would enter a simulator and accept the rules of the sim which would define what the user was allowed to do in the game.

;0 As I said mostly speculation to get people thinking.. What do you think about it?

Absolutely. I defintely would love to experiment with such stuff. I mean, you can do that now but only so far. As time goes so many more things will likely be possible.

<refills drinks>

<pulls flask out, takes a swig, and hides it again>

I swear it's medicinal...

<looks at the crowd, thinks on it, and takes a second swig for the flask>

Aye. Back to an older subject, but one I think will solve a lot of issues with updates:

Modular construction of plugins.

<looks at the words above, and sighs>

Looks like buzzwords. I hate that.

Anyway. The idea of making the SL server into a framework server that has features 'plugged into' the main software would allow for some of the following:

- Easier upgrades to features, since there would be less chance of code-bleed leaking into other parts. This would prevent odd situation where changing the physics code breaks groups, for instance...
- Faster development time, since building a small program to effect physics and add a few commands to LSL would not mean a total rewrite of the sourcecode, just building and compiling the plugin and adding it to the system.
- The plugins can contain the Client plugin as well, allowing for easier updates of the whole. When a person logs into the upgraded server, just have the server offer the client plugin to the user to take advantage of the new features.

How hard would it be to add this kind of tech to SL?

Very.
Very.
Very.
Hard.

In fact, you would not be upgrading SL to do this, as much as building 2.0 and ripping some of the code from the old to build the framework and plugins. It would be a side project, in fact, to the main one.

So, what would be plugins and what would be framework?

Well, the basic user database would be framework, as would the general 3d engine. It's not really possible to plugin the 3d engine, as it needs to be intigrated too tightly or else it will run poorly.

Physics, on the other hand, might be a plugin OR framework. Either way, it should be admin moddable. Gravity should be able to be changed as needed, in all ways. Inertia and weight... Well, all of features should be able to be changed. Think of a space station where the orbiting gravity ring actually works by having prims that generate their own gravity!

Of course, even the framework would have to be upgraded sometimes. Especially when someone wants a plugin that is not supported by the framework.

How would plugins be made? Compiled sourcecode and scripting, a suppose. Som plugins would be able to run using just basic server-side scripting. Groups, message of the day, etc don't need massive amounts of speed to run. Physics and such would have to be compiled source to keep up with speed requirements.

So, what do you think, what would be framework in such a system and what would be plugins?

You're right, it would be Client version 2.0... They would have to drop the current client and start from a new base. The client server protocol would be the same. Only, it's different all the time, because they're in the process of changing that too... They're trying to do it one peice at a time, but the problem is, they're constantly having to mesh new code..

The community might be better served if they just went ahead and started working on Second Life 2.0, but you would never hear the end of the complaints... Why isn't SL updating anymore?! I hear all about Feature XYZ but it's been 6 months! ;0 OH NOES.

When will old content start to break? Should they imlement a new system and then slowly drop the older stuff as it's no longer used?