I like Second Life a lot but always thought that its 2D social backend (friends, groups etc.) is rather poor and certainly no match for modern 2D social networks like Facebook. I think instead of redeveloping the wheel inside Second Life, it would make much more sense to develop a tight integration with Facebook and other social networks. Perhaps this idea could be included in the open source OpenSim development roadmap.
This depends of course on Second Life having poor interoperability features and closed server code. And since the Second Life system is a proprietary walled garden that belongs uniquely to Linden Lab (and I wish to say that they have done a great job by making the metaverse available to the masses), there is not much that can be done. But the OpenSim project is progressing fast, and I think in the second half of 2008 it will approach the maturity of SL server technology. This will open the market of metaverse hosting based on open source SL-like technology, and some service providers are already occupying this niche (see for example Central Grid and watch this blog for our own service announcement). Since OpenSim is an open system that belongs to its developers and users (and every user with time and suitable skills can become a developer), there will be the possibility to develop server-side features to facilitate the integration with Facebook (and why not also with other social network interoperability standards such as Google’s OpenSocial, adopted by nearly all other major social networks). This would permit focusing the OpenSim development effort only on the unique VR features, while offering better social networking features to OpenSim users.
I think leaving the social networking for OpenSim’s users to OpenSim developers is the wisest thing. I would hate to see OpenSim integrate any 3rd party service into the login/id management process. It will not help the metaverse, but rather open it up to more security/privacy issues than it would have otherwise.
A huge portion of OpenSim is social interaction. It is the nature of the project. To use other services to handle the social aspects would be like owning shopping mall without collecting rent from the shops.
OpenSim would be best served by developing its own social system, designed specifically for a 3D virtual environment, not a poorly secured website known for privacy and security issues.
Posted by - 01/05/08
Agree with Burnman above, I’d add that OpenSim could perhaps work in OpenSocial to interact with any of the social nets that may support it.
And of course, anything BUT that horrendous Facebook. Being trapped into a single metaverse has been an irritant for 3 years now, please let’s not have the social architecture be the same way.
Also, Facebook bans for use of pseudonyms, among so much else.
Posted by
Eric Rice - 01/05/08
Well, I am using Facebook mainly as an example, and I am not implying that it should be used as the only login and identity management system in OpenSim. I think OpenSocial still has to really prove itself, but at some point the integration of Opensocial may permit to “login via Ning”, “see MySpace friends online”, “rightclick on avatar to send her a private message on Bebo”, etc.
Facebook may theoretically ban pseudonyms, but does not appear to enforce the ban - I know many people who use nyms on FB, often the same nym they use in SL. FB is a business, so if using nyms is what their users want they will have to lift the ban (or else). Eric, aren’t you being a little elitist here? At this moment, it appears that FB is what people want to use. It may change of course, but right now people are “voting” for FB.
I agree, of course, that we should get past walled gardens to interoperable systems and think OpenSim, OpenSocial and the FB platform are all interesting steps in this direction. As a user of online services, my main requirement is doing more things in less time.
Posted by - 01/06/08
Nope, not being elitist at all, merely bringing up the issue of one network’s lockin to one metaverse lockin. As virtual worlds grow, more and more will probably not interop, and we’ll have the same thing all over again but in 3D.
Second Life is as much of a walled garden as Facebook is, and that’s what we need to be careful of when building out future infrastructures.
Then again, if being elitist = wanting open interoperability and not looking at the virtual world/metaverse evolution as an SL-only silo, then yup, I’ll wear that badge proudly. :)
Posted by
Eric Rice - 01/06/08
Hi Eric - What I want (I am sure we agree here) is open interoperability AND a critical mass of users with a very rich set of popular features (in the sense of features that people actually want).
Posted by - 01/06/08
As noted already on the mailing list I think the best route to go is to use open standards and not a walled garden such as Facebook.
There are actually efforts to do this on both ends, the social networking and the virtual world end.
At the social networking end it’s not so much OpenSocial (because it is mainly about embedding widgets on webpages and nothing more) but in standards such a OpenID, OAuth, FOAF etc. (the DiSo project or dataportability.org are good places for discussions).
At the virtual worlds end of course there is OpenSim but also the Second Life Grid Architectue Working Group which’s aim is to open up the SL Grid and to enable you to host your own server or grid and most importantly make them interoperable. This is done on two fields: The social networking area and the 3D area (which is where OpenSim is likely to come into play).
We don’t know yet how Social Networks will look tomorrow and probably it’s not FaceBook which is the cool thing tomorrow. The world goes Open Standards and so should we.
(beside that I don’t see a clear way on how to get FB and SL play together without some sort of partnership between LL and FB which I think is very unlikely).
I also blogged today about this issue but related to Content Management Systems, I also mention virtual worlds in there though:
http://mrtopf.de/blog/web20/what-is-the-future-of-content-management-systems-related-to-social-networks/
Posted by
Christian Scholz - 01/06/08
Hi Christian,
In principle I agree with you - someday the world will go Open Standards, and so should we.
But unfortunately the adoption of open standards is slow. I see a lot of interesting things happening in the OpenId world, but no _significant_ adoption yet. Perhaps once Google starts turning profiles to OpenIds.
This is exponential growth, at some point it starts to go very fast but at the beginning it is very slow. I am sure that in, say, 5 years, we will have an open standard online identity management system that everyone is using, but I do _not_ see it happening much sooner.
We often make the mistake of thinking too much about tomorrow’s business and not enough about today’s business. 7 or 8 years ago I did not join an interesting (and later profitable) print magazine initiative because e-magazines were “evidently” the future. Of course there are countless similar example.
In summary, I agree that open standards are the future, but do not think Facebook is out of the game quite yet, and it will remain a dominant force in this market for 2 or 3 years at least if not longer.
Great post on Content Management Systems btw.
Posted by - 01/06/08
Yes, unfortunately it is slow and in some areas it’s also not actually clear how it should work.
But for now it would actually make more sense to connect with things like slprofiles etc. Linden Lab could actually open up some APIs and I think people would do great things with it.
But one of the main problem already applies here: Privacy.
It’s maybe even more important in virtual worlds where people do not necessarily want to be connected to real life identities (this is also what I see as a problem with the FB link for many).
One step might be that the profiles (which I think are visible in search now) would be made up as hCard etc.
Posted by
Christian Scholz - 01/07/08
I look forward to a distributed social web based on open standards and open code, with things like OpenId and OAuth to handle logins and permissions, a portable distributed profile (XRI?) and contact system, and two-way interoperability with VR worlds.
Sooner or later such a thing will exist, but it may take time and probably Google will get there first (Shared Stuff, Profiles etc. being gradually integrated with Gmail etc.) with a system that is open _enough_ to make most of the users happy (except perhaps a few purists like us).
I am now thinking that even those who love open standards like we do should recognize that in a technology development ecology there is a place for open system and a place for proprietary systems, each with its strong and weak point. I must write something more about this.
Posted by - 01/07/08
