#SL Mesh Update Week 33

No new news on the Deformer. Karl, the one programming the Deformer, is usually a guest on MetaReality. Gianna that runs the show has been busy with other things. So, for the last 2 weeks no MetaReality podcasts and that means no updates on the Mesh Deformer.

Content & Mesh Meeting Aug 2012

The Redpoly Process is available for use, but it always was. AFAIK, no one has built a tool for use in Blender to make using the Redpoly process easy. So, that seems to be on hold.

The Creators Content Improvement User Group (CCIUG) has been meeting and will meet tomorrow, Tuesday. The last two meetings have been mostly about redesign of the Build Panel. There has been almost no discussion of Morph Targets. So, there are no updates on the proposal process for them. Plus they have had no news on the Mesh Deformer.

So, in those regards, we are left to speculate. Presumably Karl is working on the Deformer and the Lab is figuring out if they can give him the item he asked for that has to come from the server side. But, who at the Lab can provide a status on that is a mystery. Since Oz has not been around for the last two public Open Source meetings and today’s (Monday’s) meeting was canceled…

Frustrating to say the least… May be there will be a MetaReality podcast this week.

New JIRA Feature Request

Tiberious Neruda explained a problem people are having with the avatar skeleton and rigged mesh. I’ll try to explain it and eventually get it into a JIRA item.

This stems from the agenda item: Will there be any way to offset and resize the collision volumes with a rigged mesh? – Tiberious Neruda

Tiberious elaborated in the Content & Mesh meeting:

Okay… bit of background to this question. Whenever you use a mesh with offset joints, the collision skeleton remains unchanged and since you need to use the ‘default shape’ for accurate offsetting, you end up with collision bones either too small or too large, depending on the avatar size. This in turn affects how the camera focuses on an avatar and in extreme cases, can even cause more strangeness like focusing on something behind a mesh, seeming to aim ‘through’ it.

If you have experienced clicking your avatar or another’s to cam around it, only to find some distant point behind the avatar was selected by the camera and you camera goes flying off, you’ve run into the odd collision skeleton volumes of the avatar… or ‘missed’ the avatar collision skeleton is more accurate.

Tiberious  continued:

Apparently, the client seems to assume some basic things about where the collision volumes are, which is fine if you’re not using a rigged mesh with offsets, but throws things out of whack when you are.

Nyx Linden sees this as a solid user case for allowing modification of the physics skeleton to match uploaded rigged mesh joints. Nyx thinks it would be good to see a feature request for it in the JIRA. Documenting the user case is important. While the Lab does NOT have cycles for it now, with a JIRA it has a good chance of getting handled when they do. This problem is causing some issues so it has a better than average chance of getting handled, or so I think.

Filing a JIRA

I would probably title the JIRA: Allow the modification of the Physics Skeleton Volumes to Match Uploaded Joints.

I would write the description of the feature as:

As a user, when I create a mesh avatar intended to be larger or smaller than the standard avatar by a significant amount, there is a problem when selecting the avatar.

Whenever a rigged mesh item is made with non-standard offset joints, the avatar’s collision skeleton cannot be changed to match. Since we need to use the ‘default shape’ for accurate offsetting, we end up with collision bones either too small or too large, depending on the avatar size. This in turn affects how the camera focuses on an avatar and in extreme cases, can even cause more strangeness like focusing on something behind a mesh, seeming to aim ‘through’ it.

An additional user case is in the combat games using ranged weapons. Game moderators have problems deciding which rigged mesh avatars have an appearance representative of the collision skeleton. This creates controversy by players and unneeded complications for game managers. This problem could be resolved if the appearance and physics shape were more nearly the same.

A reasonable solution would be to either allow the physics layer to be used to upload a collision skeleton that matches the custom joint positions. Or in some way allow the physics joint positions to be taken from the uploaded rigged mesh.

If anyone has a better idea for the wording let me know. If I’ve got some technical point wrong, by all means correct it. Thanks.

Content & Mesh Meeting

The last couple of meetings have had short agendas. If you have a topic or question you want discussed or answered, next week’s meeting would be a good time to add it to the agenda. See: Content & Mesh.


It is summer and vacation time in the USA. So, things seem to be slow right now.

I’m playing with Pathfinding and Navmesh. I’m sorting my way through some problems. Once I figure out what is going on I’ll have an article on what I find. I suspect it is mostly the way I am trying to do things to save prims.

Leave a Reply

Your email address will not be published. Required fields are marked *