Path: utzoo!utgpu!jarvis.csri.toronto.edu!clyde.concordia.ca!uunet!motcid!murphyn From: murphyn@cell.mot.COM (Neal P. Murphy) Newsgroups: comp.lang.postscript Subject: Re: *COMPLETE* Postscript Description Message-ID: <619@cherry5.UUCP> Date: 22 Dec 89 18:28:03 GMT References: <28@macuni.mqcc.mq.oz> <1989Dec21.000312.3330@ico.isc.com> <17480@rpp386.cactus.org> Reply-To: motcid!username@uunet.uu.net Distribution: comp Organization: Motorola Inc. - Cellular Infrastructure Div., Arlington Heights, IL 60004 Lines: 24 woody@rpp386.cactus.org (Woodrow Baker) writes: >... >my desk at home. 90% of the time it is NOT doing printing. It is just >sitting there. Why not offload some computing work on it. It is, afterall >... I agree. At my previous job, I had to generate a fair number of linear-regression graphs on log-log paper. I wrote a postscript function, sent it to the printer. Then I sent sent the data as a matrix of reals, and various labels to the printer. Next I called the function. It scaled the graph, drew it, labelled it, calculated the slope and intercept of the line, and plotted it and the data on the graph. If I remember, the printer ran at or near its rated speed (8 or 40 ppm). The point here is that computing that is directly related to producing the printed page should be downloaded to the printer, but not to the point of overloading the printer. For example, I wouldn't even *think* of considering asking the printer to calculate a non-linear regression curve, or a non-linear interpolation curve. But the printer should not be loaded down to 100% computing 100% of the time. This is akin to asking the same of the computer in your car; after all, it does spend most of its life shut off. NPN