Saturday, October 16, 2010

Linden Labs really needs to finish Mesh support.... Correctly

Well the other night a friend was telling me excitedly mesh was in beta, I don't really have anything currently done in mesh format and was kinda disappointed it would go so far in development before hitting public Beta, but I started making an inquiry about it asking several questions of some things it should have very basic at minimum that it turned out to be lacking, not being very knowledgeable of it yet and having not tried it out myself yet I refrained from posting a Jira article Until I knew a bit better, the very late the next day a friend gave to me a Jira article she had written, I read it said it was a good idea but was half asleep so I said I would read it again and vote for it later, well I've read it again and it's a damn great Idea. Really with that said it would be crazy and stupid for Linden Labs to release mesh format in it's current state without implementing the suggestion which can be found here as this suggestion suggests some very basic things that make using the Collada format all the worth while:

Pretty much the standard avatar mesh with the sliders and all that, you can replace that with a custom mesh with Mesh import, however Linden Labs was pretty sloppy with allowing users to set a replacement for that mesh, One issue I had been preaching the whole time time this was in closed beta was that Linden Labs needs to allow meshes to be anchored to the avatar shape in some way. In it's current implementation Linden Labs has failed to do so, which is a very basic thing and it is essential for Mesh to be relevant and useful for avatar meshes. In the current implementation of mesh when you replace the standard avatar mesh, all those sliders you get in Edit appearance, don't mean much to the mesh itself, thus meaning if your a human avatar and go to a store and Buy a mesh from a content creator, or are a furry who buys this cute fox avatar you like, your going to look exactly like in shape as everyone else who has purchased the same mesh, thus taking away your individual appearance and not making mesh useful beyond being able to structure rezzable Objects, this Jira Topic addresses that issue and beyond and suggests allowing Morph targets, which content creators can bind to shapes and various other aspects, which will make the sliders in edit appearance useful once again when using a custom Mesh in place of the default avatar mesh, and giving power to tweaking your shape once again with options such as torso length and such, who knows the user might buy a mesh to make them look like a super model and decide the breasts are too big, I'm guessing most will assume not big enough, or maybe they might want to be a little thicker and heavier set or they might like the basic shape of the avatar mesh their looking at but may wish to tweak the sliders so that their weighted a little more or less realistic, this Jira article above puts that bit of customization back in users hands, as with the current implementation of mesh they lose the power to tweak their shape like that when using a mesh, maybe you might just be using mesh so that your avatar looks like it has toes rather than blended together stubs on it's feet, whatever the reason, this Jira article adds back in the flexibility of that, also aside from that it addresses further issues as well, as many of you know, when you wear a jacket clothing layer the default Mesh for avatars morphs a bit so that the sleeves are wider than your arms and it falls into sliders, Yes content creators being able to set morph targets also addresses this as well, as they will be able to set their avatar up to widen in the arms or pants, simply put clothing the current way it's Implemented may be able to wrap around digigrade legs if the modeler took it into consideration with this Jira article in effect, or in the case of a human avatar, maybe that more muscly shape. some people think that standard clothing will become irrelevant with the introduction of mesh clothing, however mesh clothing may not fit everyone's mesh properly without this capability to support morph targets, also creating clothing the old way and being able to use it with custom meshes, enables indefinite flexibility with clothes as standard clothing will be able to fit just about any mesh that is made keeping it in mind. After all you users the community have pushed hard to get Linden Labs to use Collada format, you better ask them to make the most of it as well, otherwise we may as well of been using OBJ without what this Jira article suggest, So vote for it now, this is an absolute must if you care about the presentation aspect of secondlife at all!

and Onto the next point about mesh...

Also my friend made another Important suggestion, not everyone wants to be human, also the way mesh is currently setup the way a Collada mesh is anchored to the skeleton of an avatar kinda makes avatars with custom meshes move as though there a stuffed teddy bear rather than a person or thing that has bones in it, a topic addressing this can be found here:

This topic suggests that content creators be given more Power over how the skeleton works and allow content creators to customize the skeleton which not only allows people to define extra limbs such as tails and the such within their mesh, but it also extends this capability so content creators can tweak the skeleton of content they create, thus allowing the avatar mesh to respond to animation better where as the body moves more stiffly in some points and looser in others, so the importance of this topic though seemingly focused on furry also extends to people not into the furry scene as important to human avatars as well, as content creators will be able to extend their skeleton to move and respond better to animation.

If you would like to see mesh in it's current state, It's not bad, but its lacking a lot of absolutely needed potential as you can see here attachment meshes can be associated with parts of the skeleton to bend with it in the hair:

but take notice of the boot, while it bends well this could be greatly improved upon in realism of the motion.

also if you watch this below video closely around the bunny avatars midsection you can see there some flaws in how its the torso is anchored to the skeleton, custom skeletons in the mesh would allow it to move and respond better to animation, By giving content creators who make meshes more power over how things flex and respond to animation.

really mesh in its state is incomplete without the above suggested functionality, while it is much better than what we had before its lacking a lot and I find it to be truly incomplete and in this incomplete state it might take something from the end result of your own personal look without it, so if your reading all this, drop to the web addresses below, sign in with your secondlife name click votes and click the button to vote for the issue.