NetNews Usenet Archive 1992 #19/NN_1992_19.iso/spool/comp/sys/mac/misc/15616

Internet Message Format | 1992-08-29 | 7.2 KB

Path: sparky!uunet!ogicse!uwm.edu!linac!convex!convex!ewright
From: [e-mail] (Edward V. Wright)
Newsgroups: comp.sys.mac.misc
Subject: Re: RE-MACS COST TOO MUCH (NOT!)
Message-ID: <[e-mail]>
Date: 28 Aug 92 16:36:46 GMT
Article-I.D.: convex.ewright.715019806
References: <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Sender: [e-mail] (news access account)
Organization: Engineering, CONVEX Computer Corp., Richardson, Tx., USA
Lines: 122
Nntp-Posting-Host: bach.convex.com
X-Disclaimer: This message was written by a user at CONVEX Computer
              Corp. The opinions expressed are those of the user and
              not necessarily those of CONVEX.

In <[e-mail]> [e-mail] (price carl wayne) writes:

>>Oh?  What government agency spent thousands of dollars on a toilet
>>seat?  I know the press *reported* that the US Air Force spent
>>thousands of dollars on a toilet seat, but that was dead wrong
>>-- the bill was actually for a toilet *cover*, an injection-
>>molded plastic part covering the entire toilet and weighing
>>600 lbs.  So, do you know of an example where a government
>>agency actually did spend that much on a toilet seat.  I'm
>>just trying to see how careful you are with your facts here.
>>Establishing your credibility, so to speak.

>Back about 5-7 years the pentagon was laid into by congress for spending
>Thousands of dollars on toilet seats, along with a number of other fiscal
>fiascos.  I have personally seen the waste in goverment, so I tend to
>believe this when congress jumped on them and the press reported it, your
>beliefs may differ of both the government and the press.

Yep, that's the difference, I guess.  I believe the facts -- the bills,
the receipts, the photographs -- rather than United States Congress and
the press.  I also believe that there's a difference between a 600-lb
injection molded plastic cover and a toilet seat.  Of course, I know
for a fact that some members of Congress reported the true story at
the time.  But, even though you express such strong belief in the
credibility of the US Congress, that belief only seems to extend to
Congressmen who say things you want to believe.  I suspect that your
belief in the press is equally selective.  Extending to PC magazines,
for example, but no farther.  Well, we've seen how much ou care about
facts, anyway.


>If your decision was not made on cost, I wonder if the "need" for buying
>color and the "undependibilty" of monochrome monitors was made up so that
>your people could have the whiz-bang color monitors.

No, the undependendability of monochrome monitors was "made up"
because when we tested the networking software and it didn't work.
(Why a network driver should even care about the video display is
an interesting question, but this is symptomatic of a major problem
in the PC architecture -- tehre is no such thing as device indendence.)
No other manufacturer would guarantee that their software would work
with a nonstandard monochrome monitor either.

Why don't you just *ask* why things were done instead of assuming
things and making accusations?  You would make less of an ass of
yourself that way.


>And yes, If you buy
>top of the line monitors you'll pay. I know, I bought a new monitor for
>both my PC and Mac this year, and compared prices.  All and all prices are
>about the same.

True, the problem is not at either the low-end or the high-end.
You can get a very good, expensive monitor for the PC, which
will probably not be more expensive than an equivalent monitor
for the Mac.  You can also get very inexpensive low-end monitors
that aren't really suitable if you need to use them for any length
of time.  It's the mid-range monitors, especially large-screen
monochrome, that are hard to find.  This forces many PC users
to either pay for more expensive hardware they don't really need
or try to get buy with low-end stuff that doesn't meet there needs.


>I have never used hypercard or Toolbox, so I can't say.  I do use Pagemaker,
>which is priced identically on both platforms.

Yes, and so is Ventura Publisher.  But both Pagemaker and VP are
high-end programs.  Many people need a good desktop publishing
program but don't need all the features of Pagemaker or VP. On
the Mac, they can choose a program like Publish-It Easy for $100.
On the PC, they will probably be forced to go with VP or Pagemaker
because there's nothing else worth buying.  This is where the
extra expense tends to come in.  Not that high-end programs on
the PC are more expensive than high-end programs on the Mac,
but that low-end programs generally don't exist.  I ran into
a similar situation when I went looking for a bitmapped paint
program that ran under Windows and found nothing but a high-end
program from Fractal Designs.

>All apps that I have looked at buying are competitive on both platforms,
>the little nick-nacks that come with an OS not included.

Little knick-knacks?  For $12, Apple will sell you a Developer's CD
with approximately 600M of documentation, source code, utilities, and
system software.  That sounds like quite a knick-knack to me.


>>No.  Instead, the OS will come along while my paint program is "wasting"
>>time in tight loop, tracking the mouse and updating my drawing.  It will
>>hit the program over the head with a hammer so that it can give time to
>>some more "important" process.  Eventually, it will decide to give time
>>back to me -- by which time my hand and the mouse have moved, leaving a
>>big ugly gap that ruins my drawing.  Yes, sir, I can't wait to get
>>preemptive multitasking on my PC!

>No.  This will not happen, in the multi-tasking scheme that Apple uses
>now (cooperative) it happens, but with pre-emptive, this problem doesn't
>occur.

No, in a cooperative system, it can't happen, because unless the
paint program *cooperates* by giving up the CPU, no other program
can come along and take it.  So when the program enters a critical
code such as that painting loop, the programmer can be sure of holding
onto the CPU as long as he needs it.  That's not possible in a preemptive
system, at least not with a single CPU.

>I can (and have) run 4 ray-tracing boxes in the background on my
>system and could still do forground tasks without a problem, even download
>and not lose characters.  Try it, you may find a new way to work.

I don't doubt it.  If you set the foreground application's priority
high enough, you can certainly get enough CPU cycles to do the job,
even in critical routines.  The problem is, there will be other times
when the foreground program's demands are not as high.  But it still
has high priority, which will deprive the background processes of cycles.
Your raytracers will still run, but not as quickly as they might.  When
to give up the CPU is a decision which can be made either by the application
program or the OS, but the application is better qualified to make the
decision because it knows what it's doing at any given time and the OS
doesn't.  The downside of this is that the programmer can really blow
it and write his program so that it keeps the CPU longer than it should.