Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!elroy.jpl.nasa.gov!lll-winken!uwm.edu!ogicse!pdxgate!eecs!hal
From: hal@eecs.cs.pdx.edu (Aaron Harsh)
Newsgroups: comp.sys.amiga.advocacy
Subject: Re: NeXT Press Release
Message-ID: <2361@pdxgate.UUCP>
Date: 14 Apr 91 22:38:08 GMT
References: <2323@pdxgate.UUCP> <11007@uwm.edu>
Sender: news@pdxgate.UUCP
Reply-To: hal@eecs.UUCP (Aaron Harsh)
Distribution: comp
Organization: Portland State University
Lines: 64

In article <11007@uwm.edu> gblock@csd4.csd.uwm.edu writes:
>From article <2323@pdxgate.UUCP>, by hal@eecs.cs.pdx.edu (Aaron Harsh):
>> In article <10922@uwm.edu> gblock@csd4.csd.uwm.edu writes:
>>   My term program isn't slowed down by PostScript.  The serial port is the
>> bottleneck here.  My spreadsheet isn't slowed down by PostScript.  Processor
>> speed is the bottleneck here.  These are really horrible examples.  I wonder
>> what Gregory has against DPS?  I wonder what he wants to run at blinding
>> speeds that DPS won't let him do?
>
>Ya, right.  Okay, let's take the spreadsheet on in particular.
>DISPLAYING the spreadsheet.  It's the processor's fault?  It's the
>fault of the slow, obtrusive, and generally dumb 040.  Maybe you
>should rethink this.  I'm talking about anything display intensive.
>And frankly, it's faster on the amiga because there ISN'T dp.

  The terminal emulator I'm running on the '030 NeXT in front of me scrolls
back and forth between my previous screens in real time.  I don't know what
advantage you'd get on an Amiga.  (Faster than real time? :-)  I haven't felt
the need to use my copy of Improv yet, but I don't see any reason why it would
be any slower drawing a couple hundred columns of text than this would be.
I don't know how you got this idea into your head that you'll be twiddling your
thumbs while you waiting for a block of text to pop up, but it's completely
false.  It probably does take longer to draw a page of text, but it's very
hard for a human being to tell the difference between 10ms and 100ms.

>Flight simulators don't display much.  They just calculate.  And since
>DP takes up little cpu time if it's NOT displaying, it won't get in
>the way of calculations, will it???

  Which is more display intensive: animating a hundred polygons or drawing
a couple hundred lines of text?  Spreadsheets and terminal emulators don't
display much.  They just calculate.  (Well, I suppose terminal emulators read
data from ttys).  So... DPS won't affect the spreadsheet or terminal emulator
user.  (At least it won't affect it more than it will affect the flight
simulator user)

>>   Well, yes I have run Mach without DPS.  Why?  Because I can.  When I first
>> got my system I decided to test it out and see how much DPS would slow
>> everything down.  I wish I had saved the exact timings, but a 45 minute ray-
>> trace took about 30 seconds less in single user mode (with no Window
>> Server running).
>
>Oh, and I imagine a raytrace is just SOOOOOOO display intensive, huh?
>I didn't say that DP took up cpu time like a pig when it WASN'T being
>used, did I???  Duh.

  So what you're saying is that unless you're running video games in the
background or furiosly waving your windows around the screen, your programs
won't be slowed down by Display Postscript.

>Interpreted language slows it down.  Period.  No, you can't draw as
>many polygons per minute, and I doubt it will, unless you're running a
>68090 over a stock 68000.  And whether or not it brings it to a crawl
>doesn't mean it doesn't help to degrade the performance of an
>otherwise lightningly-wicked computer.

  Try comparing Quickdraw (not interpreted) on a 25mhz Mac IIfx agains DPS on
an '030 NeXT.  The Amiga is faster than either one of thses because it has
special graphics hardware.  The Mac is not much faster than a NeXT.  For a
~10% speed decrease, you get a graphics environment that's powerful, easy to
program, and consistent across devices.

Aaron Harsh
hal@eecs.cs.pdx.edu