Archive for the ‘jira’ tag
If you only vote for one SL Jira issue this summer…
Please vote for Mesh imports for Second LifeĀ® (If you feel like making a second vote, getting the licensing issues with the SL Viewer fixed would be another good target).
While RealXtend have managed this feat a good year ago, it would be nice to get this feature into the mainline standard viewer. Now sure, going by their track record Linden are probably going to manage to cock it up and implement it badly (Can you say “Nonstandard Custom Format” or “Stupid limitations”?), but a feature like that is going to be incredibly useful even in a broken state (Look what’s been achieved with sculpties for example).
It also has some real benefits for content creators - first the largest is it permits real world artists with real world skills to use those skills effectively. Second big reason is it adds the capability to use well developed editing suites with Second Life seamlessly - this is going to dramatically ‘raise the bar’ on content quality and potentially bring SL into line with games only two years old (from the 5-6 it is now).
The first argument I’m going to hear against this no doubt is “By allowing professionals to create content for Second Life effectively, the inworld tools are going to be worthless” - which is basically saying “Let’s keep the tools retarded.” - the fact is the SL building tools are not the best, and by adding complementary support for out world editors, it means that content can be designed both more efficiently and with better visual quality. If the implementation is done similar to the RealXtend implementation, it will end up being like sculpties - you can manipulate them in world, but the core shapes are developed out world.
Really the biggest argument above is a simple case of protectionism - and like real world protectionism, the people hurt the most are going to be the people who don’t adapt and see their countries (platforms) lose economic growth to competitors. Sculpties didn’t kill builders, and neither will meshes, but yes you will probably need to learn a new skill or two.
The next one will be “This will cause additional lag” - the irony here is that it actually will do the inverse. Mesh objects are significantly more efficient computing-wise than primitives. Not only can you cluster texture space significantly more efficiently using UV maps, but you can additionally render less triangles with more detail. In terms of network downloads - it’s not going to be a case of one primitive vs one mesh, you are comparing lots of primitives vs one mesh, and like textures - you can cache them effectively.
As an ammendum to that - people often bring out the idea of million polygon meshes. The answer here is that meshes can be artificially limited on upload (lock em to a certain filesize or number of polygons) - the ultimate irony here is however that primitives are no better - the average second life furry avatar fully rendered is around ~100,000 triangles, or approx 2x that of the average complete Doom 3 scene.