dear list,

i'm trying to build a patch which basically emualtes a karaoke-like coloring of a text in given time segments. my approach is to display the text in several jit.gl.text3d objects (generated out of a xml file in a javascript object) and doing the coloring via a moving texture on a plane (jit.gl.gridshape, blend_mode 2 4), lying on top of the scene. my problem is that this seems only to work when the jit.gl.gridshape object is instantiated _after_ the text objects, which is a problem as i generate text objects anew all the time.

any thoughts on this? find the isolated problem in the patch below.

also, i'm not shure if my general approach is a very clever one, i still have problems with positioning the various text objects in relation to each other (as they can't output their size). until now i use a workaround where every new text object gets rendered in an offscreen matrix once and measured with jit.findbounds... dirty and works only approximately. more thoughts?

any help/comments/ideas appreciated! thanks, chris