From: Guillaume L. <gui...@gm...> - 2005-07-22 19:48:49
|
Hello everyone. The past week has been quiet on the ColladaMaya front. We're actively fixing all the small regressions and other issues found: the instantiation of a mesh and that mesh as a target for a controller, a small regression in pre-Maya 6.5 versions with the skin cluster import, the export of relative vertex positions and their animations and the scale transforms with physica= l entities. You can get the new weekly development build: 0.66, here: http://prdownloads.sourceforge.net/colladamaya/ColladaPluginForMaya-0.66.zi= p?download Siggraph is coming very soon! We will be present at the conference. There is a tech talk that will interest all of you, hosted by Sony Computer Entertainment about Collada: "COLLADA: An open Digital Asset Exchange Schema for the Interactive 3D Industry". This will be a great opportunity to hear about the upcoming developments in Collada as well as meet the people behind it! During this tech talk, companies using Collada will be mentioned and thanked for their support. If your company uses Collada and you would like to demonstrate your support for Collada and our translator, please contact us! The talk is Wednesday, August 3rd, from 1h00pm to 3h00pm, in Hall J, Room 1. I hope to see all of you there! Sincerely, Guillaume Laforte Feeling Software Inc. http://www.feelingsoftware.com ------------------------------------------------------------------ Recent Changes: 22/07/05 (Release 0.66) (cla...@fe...) Physics: Fixed another bug reported by Stephen Hatcher: initial velocities= =20 belong under <rigid_body> instance, not as siblings. (gla...@fe...) SourceForge #1238600: Relative Vertex Position Animation. It seems this=20 feature is used also when applying manual vertex tweaks, which is very=20 common, so it now removes from the mesh positions only the non-animated=20 vertex position tweaks. 21/07/05 (cla...@fe...) Physics: Added a warning for rigid bodies that are exported with non-unit= =20 scale. Also removed code to export inertia tensor. It's not really used=20 anywhere yet, anyway, and I plan to rewrite that code once Nima is=20 completed. 20/07/05 (gla...@fe...) SourceForge #1241963: No animation of re-import. The given problem files=20 work perfectly on Maya 6.5, but some connection is not correctly done in=20 Maya 6.0.1, althought the DG graphs look perfect. Using the Maya Ascii file= =20 outputs of both controllers to find differences, I found that the problem i= s=20 that the "inputComponents" value plug for the connection nodes: "groupPart"= =20 is not set. Used the API classes to assign the correct value to this plug= =20 rather than the previously used MEL command. 18/07/05 (gla...@fe...) SourceForge #1238600: Exporting the difference between the current vertex= =20 positions and the current vertex position tweaks, instead of the current=20 vertex positions. 16/07/05 (gla...@fe...) SourceForge #1239214: Unable to instantiate both controller and controller= =20 target. The original geometry is instantiated, but it is also marked as=20 hidden so that the skinned mesh only shows. The following need to be done i= n=20 the right order for it to work: when a skin cluster is attached to a mesh,= =20 disconnect the mesh from the scene graph as well as mark it intermediary=20 only if there are no known instances of the geometry. Otherwise, reconnect= =20 that geometry when the other instance is created and reconnect the skinned= =20 mesh to use the wanted transform. Ensure that all meshes get the per-face= =20 shaders attached to them, only if necessary. |