Hi, Has anyone running NeXTSTEP any experience with the Trillium speech product? This is the basis for GNUspeech. I am curious as to the quality of the speech produced and if it is worth pursuing (and if I should hunt for a NeXT cube or just follow the port onto different hardware) In addition to the auditory quality, I am interested in just how much of a CPU-hog it is (was). I'd be interested in any realistic (i.e. un-tuned) sound samples. Thanks! --don