This is kind of a follow up to a problem I had many months back which I can't remember the topic name. My basic problem is that I'm trying to render something on screen that has text labels. I need to make sure to adjust the view so that all the text fits on the screen and doesn't run off the edges. (But I don't want too much empty padding so I want my values to be deterministic.)
I moved away from SCREEN_COORDS mode to OBJECT_COORDS mode because it seemed that SCREEN_COORDS mode couldn't deterministically tell me what the bounding box of the text would be until after it was rendered. Since I need to compute my view range before I render, this didn't work out. So now I am back to OBJECT_COORDS mode and I also have to recalculate my character height and aspect ratio manually. I have implemented this and this part of the code seems to work well. My problem though is that I'm still not consistent or correct values for the size of my text via computeBound(). The values seem to be more correct than SCREEN_COORDS mode, but I noticed my values are still off slightly when I have a large scale projection change. I'm trying to understand what's causing this. I haven't yet determined/proved if the problem is in my stuff or how osgText works, but I'm hoping to get a better explanation of how computeBound() works for regular osgText modes. Basically, I just want to specify a string via setText(), and a character size and aspect ratio via setCharacterSize(), and call computeBound() on the text to get the width and height of my text. Under OBJECT_COORDS, shouldn't this do the correct thing? If the answer is yes, then I know the problem is in my code. If the answer is no, I need to understand how it actually works a little better and how to work around it. I am under an orthographic projection (2D) if it makes a difference. Thanks, Eric _______________________________________________ osg-users mailing list [email protected] http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
