The gang over at  realXtend (Rex) recently posted a pre-alpha version of their new mesh viewer under project Corus. They have been hammering away on this for 6 months or so, so it will be interesting to see what it does. They seem to have started from scratch after having tried to rework mesh into a SL-viewer version.

We think mesh is a critical next step for virtual, onilne environments. Why? Widespread adoption requires more realistic virtual enivironments. Prims and rigid avatars are too limited. Sculpty prims are too difficult to make. So mesh and prims seems like a logical next step for enhancing the user experience and making it more enticing for mainstream online users.

We have been working to get photo-realistic sculpties into OpenSIm for that last 6 months. In fact they look amazing as you can see on our Heritage Key grid. So the issue is not the end product quality, which is very high, but the difficult, time consuming process to make these objects. Effectively to make a sculptie you need to 1) make a model in 3D-Max/Zbrush/Maya 2) add the photo layer 3) burn the shadows to make it interesting 4) break it into sculpties pieces 5) import to SL/Opensim 6) reassemble . There are even more steps to worry about if you are really gonna try to do this!.  In order to tweak anything you need to restart the entire process. Some people have created tools for this, but it is still quite hard work--especially if you need to be precise and even more so if it something that people know well in the real world.

Next big question is how many polygons are enough if you can put mesh-objects on a grid? The more detail the bigger the file, the more lag. So what is the right trade-off? Are the more efficient download/streaming possibilities to consider? Would it be useful to have lower res versions that people could also see if they are on lower-bandwidth connections? Or maybe I should just buy an Xbox?