> I would think that any use of both forms of scaling could also be > adequately described in terms of only ScaleFactorX+ScaleFactorY as a > pair. [Perhaps that's safer than giving UA's a choice. Karsten]