Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cornell!uw-beaver!tektronix!reed!kamath From: kamath@reed.UUCP (Sean Kamath) Newsgroups: comp.sys.apple Subject: Re: Why Keep the //.... Message-ID: <12140@reed.UUCP> Date: 23 Mar 89 03:57:53 GMT References: Reply-To: kamath@reed.UUCP (Sean Kamath) Organization: Reed College, Portland OR Lines: 53 OK, folks, this will probably be my last set of blathering for a while. This is the last half of my second senior semester, and at Reed they make you write a senior thesisas a *undergraduate* to get out. I'm doing color digital image enhancement, and when I'm done, I will send the best images I have to ncsa. In anycase, don't expect to hear hide nor hair from me till mid-may at the earliest. . . :-( In article jm7e+@andrew.cmu.edu (Jeremy G. Mereness) writes: > >I wish some people at Apple Co. could read this. Some do. >The Apple // may not built like a workstation like a Mac, designed to handle >raw power with little overhead, [omitted] >jeremy mereness Unfortunately, the Mac has just a tad too much overhead to have any real fun with. I decided *not* to do my thesis on a mac ][ (5 megs or Ram, 80 meg internal drive (28 ms), 68881 and lightscpeed C) simply because of the imense overhead I had to deal with. Look, the thing couldn't keep up with the digitizer running at 38.4Kbaud doing *nothing* else. I got hardware overruns *always*. In anycase, when the little application I wrote would work just fine with about 400 lines of C code, I had well over 3000 lines, i.e. 2600 lines of "user interface", which did little else then show me what was happenning. And no, it's not just sloppy writing, rather, I did *all* the things the guidelines tell you to do, except use the palette manager (I stuffed the pallete myself to show the images). I wouldn't even really call a Mac][ a "workstation" any more than I'd call a IBM PC a "workstation". A Sun is a workstation. the Tek4317 I'm typing at is a workstation. Not the Mac. Least, not till it get's VM :-) And at $4000 *MINIMUM!* It's priced like a workstation, anyway. Sean Kamath BTW: I finally got Dhrystones to run on the Mac. Results? About the same as a Vax with a load average in the 3-5 range (1500). Intel 386 running Sys V: 2500. NeXT machine (optical drive): 5000. Of course, dhrystones are pretty meaningless, so think about this: A fellow slave--er thesis student, is running time slices of fluiddynamics problems (the magnus effect). Forget programming this behemoth on a Mac (it really need gobs o mem and process power.), but it takes about 6 hours on the vax, 3 on the intel, and a little over an hour on the NeXT. Same code. Identical. -- UUCP: {decvax allegra ucbcad ucbvax hplabs}!tektronix!reed!kamath CSNET: reed!kamath@Tektronix.CSNET || BITNET: kamath@reed.BITNET ARPA: kamath%reed.bitnet@cunyvm.cuny.edu US Snail: 3934 SE Boise, Portland, OR 97202-3126 (I hate 4 line .sigs!)