|
From: Dirk R. <dre...@ia...> - 2006-02-27 17:30:11
|
Hi Antonio, On Mon, 2006-02-27 at 10:54 +0100, antonio_lioy wrote: > Hi guys, I need to know where the SceneFileHandler puts the global > transform data (i.e. the data that place the whole model on the scene) > when we load a VRML model using its read method. In other words, does > the loader build a graph that has a default initial structure (so I > can easily find the data), because it seems to me that the data are in > different position for different models (I mean global poitioning > data, if it in different positions, how may I find it?) There is no global transformation data. It all depends on how the model is structured, which depends on the exporter used. I would assume most systems to have a transformation node close to the root to move the whole object around, but it is just as valid to have no transformations at all and change all the object coordinate to move objects around. But if I understand your problem right (grabbing objects and moving them around) you don't need absolute transformation anyway, all you need is the relative transformation from the hand/pointer to the object at the time the object is grabbed. You can get that by multiplying the transformation of the hand to world with the inverse transformation from the object to world. Then when the hand transformation changes, you multiply it with that stored matrix and you get a new transformation to apply to the object. Hope it helps Dirk |