The thing is that you all are invited to participate in this effort. It is not about just making the architecture right now a standard, it is about thinking about ways to enhance this architecture in order to make it more stable, have less lag and esp. allow interconnectivity.
And again, community participation is needed!
And to answer the question: They have about 4000 servers with nearly 10,000 regions but there are many reasons for lag and of course Linden Lab also tries to cut these down.
Well, I am not sure what Metaplace is all about yet as not much information is given and one needs to see how it's implemented, how the protocol looks like.
I see some difference to Linden Lab's approach though:
Linden Lab is seeking participation already in the conceptional phase of the project and does not simply publish a complete specification
Linden Lab has already a working virtual world which has proven to exist. OTOH of course Raph Koster has a lot of experience in this field, too.
Linden Lab is committed to Open Source. On Metaplace there is no real statement that it's going to be open source. I am also not sure what you need to run a server or if it's just a webserver. Still a bit unclear to me.
Regarding the term "proprietary", isn't Metaplace not as proprietary as Second Life right now?
I think one important thing is also that Linden Lab is also seeking for people developing parts of the architecture to make sure each of this component works with the rest.
Moreover the proposed new SL architecture might allow for more, like not only doorways between worlds but eventually also contigous landmasses like it's happening right now. Of course neighbouring regions would need to support both the same region design.
Well, it's all more into the blue now regarding Metaplace as nothing can be really seen yet. In the end it might be nice if all this could be bundled together in one standard instead of ending up with several of them.
The currency problem is not solved yet I think but my idea would be that it's puggable and you can support multiple ways of paying for stuff. It might be with L$, by credit card, PayPal, whatever. And yes, there will probably be free grids then. And others will probably compete. But this will happen nevertheless. I think Linden Lab's goal is to be part of this instead of letting other people implement it.
So the proposed architecture from Linden Lab actually allows for regions hosted at home. So in fact you can have avatars visiting your land which is hosted on your server. There is a concept called "Region Domains" of which there can be many and one might think about Region Domains as sort of continents.
The goal should be though to also allow arbitrary region sizes and forms in the protocol while of course first the existing concept will be implemented. But the protocol should be extensible.
Moreover those Region Domains are more or less black boxes. This means that you can implement it as you will, with a cluster or on a laptop. There should also be support for complete disconnected use where agent and region domain both run on your laptop.
Slashdot Mirror
The thing is that you all are invited to participate in this effort. It is not about just making the architecture right now a standard, it is about thinking about ways to enhance this architecture in order to make it more stable, have less lag and esp. allow interconnectivity.
And again, community participation is needed!
And to answer the question: They have about 4000 servers with nearly 10,000 regions but there are many reasons for lag and of course Linden Lab also tries to cut these down.
I see some difference to Linden Lab's approach though:
I think one important thing is also that Linden Lab is also seeking for people developing parts of the architecture to make sure each of this component works with the rest.
Moreover the proposed new SL architecture might allow for more, like not only doorways between worlds but eventually also contigous landmasses like it's happening right now. Of course neighbouring regions would need to support both the same region design. Well, it's all more into the blue now regarding Metaplace as nothing can be really seen yet. In the end it might be nice if all this could be bundled together in one standard instead of ending up with several of them.
The currency problem is not solved yet I think but my idea would be that it's puggable and you can support multiple ways of paying for stuff. It might be with L$, by credit card, PayPal, whatever. And yes, there will probably be free grids then. And others will probably compete. But this will happen nevertheless. I think Linden Lab's goal is to be part of this instead of letting other people implement it.
The goal should be though to also allow arbitrary region sizes and forms in the protocol while of course first the existing concept will be implemented. But the protocol should be extensible.
Moreover those Region Domains are more or less black boxes. This means that you can implement it as you will, with a cluster or on a laptop. There should also be support for complete disconnected use where agent and region domain both run on your laptop.
Look at the Proposed Architecture page on the wiki for more information: http://wiki.secondlife.com/wiki/Proposed_Architecture