As I’m not a developer I don’t have the time - and am not inclined anyway - to get involved with the AWG. My impression that OpenSim developers are really ‘on board’ with the AWG and OGP came from what I’ve read, both in the mainstream press and the blogosphere.

These reports often talk about Linden Lab’s attempts at interoperability in glowing terms, without critiquing whether or not this is a good thing. Whilst I really do admire how Philip Rosedale’s vision and creativity have made an incredible contribution to the development of the metaverse, I am also wary of his stated intention to make Second Life, or at least Second Life architecture, the 3D Internet. I was concerned that OpenSim’s involvement in the AWG was tacitly endorsing this approach.

You are the first person involved in OpenSim I’ve seen address these issues, and I feel re-assured with your response, so thanks.

I feel these decisions fall into two categories. First come those which are important but not truly fundamental to interoperation, such as avatar attachment points and inventory formats. Secondly, there are decisions which really limit the types of platform (such as client-server based or peer to peer) which can participate in the protocol.

The first type of decisions could be worked out if other platforms were represented and had significant influence over the OGP process. The fact is, though, that Linden Lab (and to some extent OpenSim developers) are the only real platforms around at the moment who are prepared or able to engage in the OGP. I would love to see other companies or open source efforts (such as Sun’s Darkstar) represent their platforms in order to improve the applicability of the protocol. However, I believe these can happen later on in subsequent revisions of the protocol.

The second type of decisions are ones that can’t be easily reversed. However, I believe that in some of these cases meaningful interoperability is very difficult anyway. For instance, I find it difficult to see how a Croquet system (which relies on replication and synchronization of the environment on each participating peer could ever interoperate with a client-server based system (where authoritative state is only kept on a single server) except at a very basic level. Having different companies or individual participating in the protocol process would not help very much.

I would also say that it’s a mistake to think that OpenSim developers are really ‘on board’ with the AWG and OGP as such. OpenSim aims to be a general virtual environment platform, and if someone (such as the OGP) comes along with a grid protocol then I think that we’re happy to accomodate modules which allow interoperation using that protocol. Equally, we’d be very happy if some completely different software group came along with their own protocol and wanted to produce modules that could be bolted on to the core OpenSim kernel. In fact, that would be a great thing since it would be a push for us to genericise our core internals to take us beyond our Second Life roots.

Of course, the difficulty will come if there does come a point where the OGP somehow requires platforms to adopt a certain way of doing things, where that way reduces the number of alternative protocols that could be adopted. However, I believe that we will always try our utmost to avoid going down that path. I also think that such a scenario is actually pretty unlikely.

I’m glad to hear that protocols can be (and are being?) developed independent of Linden Lab. That’s not how it’s being reported though. I was under the impression that OpenSim developers were completely on board with the AWG and the OGP.

And isn’t it possible that LL might use their first mover advantage to dictate the protocols anyway?

Opensim isn’t between IBM and Linden, Opensim was made by regular people who liked SL but wanted to get free and create their own virtual world platform which is what is happening at the moment. The involvement with Linden isn’t needed to create protocols, if you want to create protocols do it between you, the programmers, don’t get involved with Linden as they really aren’t relevant for Opensim.

I noticed Sakai is turning away from Opensim to work closer together with RealXtend to get more advanced technology to his grid, who can blaim him, that is why so many people are interested in Opensim, to explore new technology, new options, going further than what is possible in SL. Why spend time with Linden on protocols if you don’t want to make a Second Life clone software or extentions to SL with all the intertop attention that you are giving to this.
The importance of an independant opensource platform is crucial.

I’m going to post here a reworded question that I was going to post on Dusan’s post that you link to, but decided against when the the conversation seemed to come to a natural.

I was left wondering about the issue you raise about the importance of certain fundamental decisions about OpenSim architecture and protocols and how they will shape the coming metaverse.

When it comes to the AWG and the Open Grid Protocol how can we trust a process that is sponsored by a commercial entity that has an interest in directing the process in support of its own commercial interests, and is heavily supported by employees of IT companies with an equally strong invested interest in influence the direction of the coming metaverse?

Can someone convince me that the OGP isn’t biased towards virtual worlds based on a Second Life model. As you say, there is a huge difference between virtual worlds based on the server/client model and the peer-to-peer model. Convince me that if a group of the worlds smartest and experienced virtual world developers got together today, and built the best possible architecture and protocols from scratch without historical encumbrances that they would come up with exactly the same protocols. Convince me that the protocols being developed are open enough that it will mean any (open) virtual world platform that comes along in the near can future will be able to plug into it, irrespective of it’s architecture.

You seem non-partisan enough to be trusted (I’m sorry - and I don’t mean offense to those involved - but I can’t trust employees of said organisations), so if you or Adam Frisby say that my concerns are unfounded I’ll shut up.

Regards. Sean