Thanx to all for voicing their concerns, and for keeping them reasonably sotto voce ;o) Just to clear up exactly what my concern in this thread is, here's the key paragraph from the link I give above. I have re-written the key passages verbatim in uppercase for emphasis : "One of the simplest ways to create a derivative model is to convert a DAZ model (typically OBJ format) to another non-Poser format, and distribute that. Beyond this, a DAZ model may be altered by adding and/or removing geometry in order to create a derivative work. EVEN MORE DIFFICULT FOR MOST USERS TO RECOGNIZE, BUT STILL DETECTABLE TO TRAINED MODELERS, ARE METHODS INVOLVING TOOLS THAT CAN CREATE A DERIVATIVE MESH WITHOUT TRANSFERRING THE POLYGONAL LAYOUT OF THE ORIGINAL. (MANY OF THESE METHODS CAN RESULT IN A MESH WHICH SEEMS LEGITIMATE, OFTEN WITH NO VERTICES COINCIDENTAL WITH THE ORIGINAL. THESE TYPES OF TOOLS/METHODS MAY INCLUDE: NURBS/poly conversions, subdivision/smoothing or triangulation operations, de-resing, SHRINK-WRAPPING/FITTING and randomization operations.) The distribution of models created using any of the above examples is strictly prohibited, and none of these methods will result in a mesh that is not still subject to DAZ's copyright." The method I describe above, and on the link given (to the Amapi Forum), would constitute a "manual shrink-wrapping" method. Like all other methods of making a mesh that fits another, it relies on having the target mesh (i.e. the DAZ figure) in one's 3D app as a reference. It yields a new mesh "with no vertices coincidental with the original", as will naturally be the case with a totally new mesh. This kind of mesh would still be a copyright infringement, if the above text is read literally. I'm still sure that DAZ did not intend the wording of their EULA to be interpreted thus, but I'd prefer to know for sure. I'm sure a lot of other people would, too. So let's wait and hope for a sensible DAZ answer, and keep cool while we're waiting.