SL Objects and 3DS

The reason is because unlike any other 3D environment, SL is 100% user created, and 100% streamed. There is no permanent geometry whatsoever. The world changes literally every moment.

There are many problems that would need to be solved before arbitrary meshes could work in SL. The biggest is level of detail. Most amateurs are not going to have the time or the know-how to create meshes with pre-set variable LOD built in, so we'd need a way for the system to automate it. Unfortunately, there's really no good way at present for the system to intelligently judge polygon reduction over distance. It could be done strictly by math of course, but the results in many cases, if not most cases, would be less than pretty. For example, you could model the Statue of Liberty in exquisite detail, and from up close she could look great, but from any significant distance she'd probably look like nothing but a lumpy green blob as her polygons begin to reduce as the camera moves further away. That's a hard problem to solve.

The alternative would be not to use LOD, but if you think SL's laggy now, imagine what would happen if poly counts were never reduced by distance. There would be very few computers on Earth that would be able to handle SL in any semblance of real time if that were the case.

Another problem is bandwidth. If arbitrary meshes were to be used, the instructions for how to create every single polygon would have to be streamed to every single client for every single model. That would be an incredible amount of data.

By using a common set of parametric objects as building blocks for everything as SL does now, the process is streamlined. Since every client computer enters the world with foreknowledge of ow to make each type of primitive, the only data that needs to be streamed is where to put the prims, what size to make them, how to rotate them, etc. That's a relatively tiny amount of information.

Also, how to you calculate physics with an arbitrary mesh? There have been many suggestions on this, but none have been ideal.

LL has said many times that they do want eventually to allow mesh modeling in SL, and when that day comes, we'll all rejoice. However, these technical obstacles have to be overcome first, and none of them are easy.

In the here and now, what you need to do is learn how SL's present modeling system works, the same way you had to learn how Max worked back when you started with that.

As a professional 3D modeler myself, I can promise you that getting good at modeling in SL will make you better at modeling in Max. Because SL is so limited, it forces you to use a specific type of problem solving intelligence that more sophisticated modeling applications let you bypass. The result of exercising those problem solving skills is you become a much more efficient modeler, both inside SL and out.

Once I started getting good in SL, I found that my average poly counts per model in Maya were going way down while the visual quality was going way up. Also, my texturing skills increased at least a hundred fold.

Give it a little time, and I'm sure you'll say the same thing.