Gentle reader,
today I want to talk a little about the realXtend project which came upon my radar recently and impressed me great lengths. Their goal is, put in simple terms, to develop the InterGrid about which Gwyneth Llewellyn spoke recently in one of her own blog posts. Well, she even mentions the project in her post.
The InterGrid, put in simple terms again, is the ability to make your own region/regions/grid and take your avatar on grid A, B or C and also take the inventory of it with you. This means your avatar becomes quite more flexible, it can travel around the different grids if done correctly, but also there's quite much work to do until that goal is going to be achieved. This project is using OpenSim as their platform for serving regions, they've already enhanced it with quite much advanced and sophisticated features and also forked the Second Life viewer into an own thing, although this can still connect to the Second Life grid in a compatibility mode.
The realXtend project is backed up by two Finnish companies and around 20 people working on it, programmers, content creators and graphic designers, so a good size but still small compared to the staff of Linden Lab.
Contrary to the culture of Linden Lab, though, they've got a roadmap, and since they're contributing to OpenSim, parts of it overlap with the roadmap of that project, too. Having a roadmap never hurts, on the contrary, it is always nice to have and a good thing for all participating people.
In a recent interview one of the driving forces behind realXtend, the CEO of one of the companies backing it up named Tony Manninen, gave us a very interesting peace of his mind and his over all vision for the project:
Me: And how will the work you have done on the avatar server alleviate this problem unless SL, WoW and other cooperate on interoperability?
Tony: Think of it more like the 3d web. realXtend/OpenSim is like the Apache of virtual worlds, rexViewer is the Mozilla or Firefox of whatever. When "surfing" the web, you are not constantly required to prove and change your identity when loading different pages.
And this line is quite interesting for all of us. Apache is today the work horse of most web servers on the planet, its market share is around 51% in April 2008. But what many people don't know is how Apache started and how it became the king of the hill. In former times, when Apache was non existant, there's been another 800 pound gorilla of webserving software called NCSA httpd. This was back then the leading webserver under an opensource license. Apache just started as a patchset (Apache was just the nickname for "a patch" first only or more precise "a patchy server") way back then for NCSA httpd, adding features many people wanted but the maintainers of NCSA httpd were unable or unwilling to include. So over the time the patch set became more and more important, popular and turned into an own piece of software, winning big grounds against its father until NCSA httpd became obsolete and went into insignificance.
So, what does that mean when talking about Second Life? Simply: realXtend could be the nail into the coffin of Second Life.
So, what's in realXtend viewer and OpenSim already, that would be nice to have in Second Life, but isn't there (yet)? Among already implemented features those biggies:
- the ability to host your own region somewhere on a server of your choice and to connect it to the grid (many would like that since the tiers you've got to pay for Linden Lab are quite expensive),
- the use of a more advanced opensource renderer named OGRE, which also is going to support DirectX rendering on Windows platforms,
- coming with this renderer real time lights and shadows of objects,
- web on a prim,
- builtin VNC viewer for desktop sharing,
- VOIP client and 3d audio rendernig,
- meshes instead of prims - this means you can build far more advanced structures, also build stuff in normal programs instead of the client and import them, which adds to a good graphics experience quite much, but also means a slightly longer loading time perhaps, but still many would applaud them in SL and with right,
- quite more sophisticated avatar meshes, everything can be an avatar, e.g. also mushrooms (this example is included) or a bad snowman,
- Python scripting,
- teleports between realXtend and Secondlife,
- script controlled teleports,
- centralised avatar storage to move the avatar between different grids,
- multiple streaming URLs per parcel,
- and others,
but those are the real biggies. If you also take into account that it just took realXtend to implement those features around four (!) months of time you'll really have to wonder why Linden Lab hasn't done that themselfes already!
Among the roadmapped features you'll find those things:
- Direct3D rendering on windows platforms,
- support for OGG Vorbis,
- support for video codecs beside Quicktime,
- Weather support,
- inverse kinematics,
- avatar face/head animation based on live video camera data,
- lip sync for VOIP,
- cloth physics,
- vehicle support,
- the ability to hold more than 100 avatars at the same time in one region by splitting up the region on several hosts and letting them do their work,
- and others.
So what we'll have here is a real ambitious project to build the InterGrid with nice goals, but they're not only having a roadmap, seems they've already been able so far to deliver their planned features and are going to be in the future, too, in many parts they're already ahead of Second Life quite much.
To put it short: what we've got here is a major competitor emerging for Second Life and even more so on a very rapid speed! Linden Lab is still ahead of its competition somewhat, but realXtend is gaining ground and its gaining it quickly so that Linden Lab should really be make up its mind now what they're planning with the platform in the future, otherwise it is quite possible that they are going to face the same fate as NCSA httpd or Netscape: the technic will remain, but innovations are coming from other sources and the people behind the initial project are loosing the grip on it.
