jknowles@binky.arc.nasa.gov (Jim Knowles) (06/15/91)
The posted timing on the Doctor Laser benchmark for Ghostscript 2.1.1 running on a Sun SPARCstation 2 is WAY wrong, and it's my error. I ran the test while I had some mondo fonts installed (TEX fonts that had been rendered at 2400 dpi then converted to BDF and then to Ghostscript fonts). These fonts were installed to support the fax email application I run (thanks to Dan Newman at Innosoft for supplying them), and I forgot to reinstall the standard Ghostscript fonts before the test. With the standard fonts in place the timing is consistently under 2.1 seconds, rather than the 13.6 I reported. Big mistake. I apologize for the misleading figures to all involved, especially L. Peter Deutsch, author of Ghostscript, and Ralph Carmichael, who collected the data. By the way, Ghostscript 2.2 runs the same test consistently at about 1.6 seconds.