Multitasking
K8DO at aol.com
K8DO at aol.com
Tue Aug 22 10:59:37 EDT 1995
Hi steve...
Thanks for the copy of the Email to Howie...
I truly don't know much about the newer, super sophisticated, operating
systems, but I have been around awhile and I do have a gut feeling for the
reality of the cold, cruel universe...<no free lunch, gimmicks cause
problems, the fancy solution ain't, etc.>
I started using computers, in industrial automation, in the early 70's...
These were 16 bit machines out of california (Computer Automation, Inc., as I
remember, that had to do two 8 bit fetches for each instruction...This was hi
tech stuff in those days... We used teletype machines, Model 33 TTY's, as the
keyboard/monitor (paper screen! :)....To boot (cold start) a computer we
stood in front of the computer rack <24 of them> and using a row of toggle
switches keyed in a bootstrap program one byte at a time - using the toggle
switches to set each bit of the byte < 01010011, etc>, then pressing the
'STEP CPU' push button twice <while holding up a spring loaded, safety
switch> to load that byte into ram (2 nibbles make a byte) - This took 5 to
10 minutes, depending on what you did the night before... If you made a
mistake you had to start over from the beginning, i.e. serial data
input...The bootstrap program made the computer able to accept data from the
tape reader on the TTY... Then you loaded the portion of the main program
handled by that processor, by running the paper tape through the reader,
which took another 22 minutes....Needless to say, the programmer who caused a
major crash, and a subsequent cold boot of the entire system, had a lot of
paper clips shot at him...
I can tell you many stories, like the programmer working for me, who balanced
his checkbook in binary... and another, who left notes in hexadecimal which
had to be converted into ascii to be read...and another, who would fix a
crashed computer by sitting in front of the machine in the lotus position
going "Ommmmmm Ommmmmmm Ommmmmmmm"....Then, without having so much as
even glanced at the multi-hundred, paged book of source code, he would
slowwwly rise to the standing position, input bytes through the boot
switches, and walk away, with the computer back on line and working...I still
suspect he programmed bugs in the code just so he could do his shtick,
although he stoutly denied it...
I moved on to various projects, PDP's were popular in those days, the 6's,
7's, 8's, 9's, 10's, 11's....etc... Used some DEC stuff, used some ALTOS,
used others I have forgotten, etc...All mini's, never got into mainframes...
For home computers I started with a kludged 4004 and 256 bytes of RAM...
Later had the ALTAIR kit, which had 4K of RAM and thought I was in hog
heaven...Then went to S100 bus machines... From there to Apple / Apple I /
Apple II / Lisa... Passed through Sinclair's {Clive was a real trip} - Trash
80's - Commodore's - 8088 - 8086 - 286 - 386 - 486, and into the present...
Having said this I will admit that my area of (so called) expertise is/was
electronics, 3 phase power systems, and automated machinery...(things with
hundreds of electrical motors, and thousands of limit switches and machine
tool relays) I was only by osmosis/necessity a computer
operator/programmer...
Non the less, I have at times become handy at writing code in various
languages to fix bugs in the software... 99% of the CPU's / Languages / OS I
have worked with are gone, vanished, kaput, i.e. history, because they were
not THE ANSWER...
Given this experience, you will begin to understand my cynicism about
WinDoz95 <and for that matter OS/2 and DeskView> In 10 years they will be
history just as Applesoft, PowerBasic, and a heap of Languages/OS are
now...Threading is merely a software device to allow the OS to use CPU idle
time more efficiently... It does not change the physical reality that
operating code (and immediate data) have to be in RAM for efficiency...Any
time the CPU has to fetch code/data from a drive there is a heavy hit on
speed...If the CPU has to shuttlle baskets of code/data in and out of the HD
the hits are even heavier...What is really hitting performance is the current
fad for GUI's...
They are writing these humongous OS that require 8 meg's of RAM to hold the
operating code and data for one program... The Intel's and MicroSoft's have
become complacent dinosaurs...They are merely hanging more chrome on the same
chassis - working on the 686 - 786, etc... They have quit evolving and
therefore they are doomed to become extinct....Out there, somewhere, are a
couple of bright guys saying to each other, "Hey, wouldn't it be neat if
instead of DOS and a CPU we did... etc.", and when their neat idea becomes
reality the newest Intel CPU and biggest MS OS will suddenly look like my old
4004 and 256 bytes of RAM...
Sorry to ramble on like this... You pushed one of my hot buttons...
Cheers ... Denny k8do at aol.com
>From Marijan Miletic <s56a at ljutcp.hamradio.si> Tue Aug 22 15:10:50 1995
From: Marijan Miletic <s56a at ljutcp.hamradio.si> (Marijan Miletic)
Date: Tue, 22 Aug 95 14:10:50 UTC
Subject: Juicy Bev ground
Message-ID: <75169 at ljutcp.hamradio.si>
Dave, W6QHS while expanding on Randy, K5ZD/1 article on <50$ Beverage kit gives
a good example of common sense ground at P4 location by using weed juice flow!
As we are talking about 400-600 Ohm antenna impedance terminated in pure
resistor (powerful enough to sustained induced RF from nearby TX's), we can
allow few tens of Ohms grounding resistance without affecting performances
much. However, this approach is NOT valid for quarterwave verticals (36 Ohms).
This all brings nice memories of my YU1PCF county site some 25 years ago with
20m vertical for 80/40m and my, then new, TH6DXX with hand rotation. I put
my first Beverage toward West and terminate it by 560 Ohm/2W resistor and
matched it to RX with PI filter (3:1 fixed C as far as I remeber).
73 de Mario, S56A, N1YU.
P.S. Sri for typos on PR-Internet link with no chance for correction.
>From George Cook <george at epix.net> Tue Aug 22 15:39:28 1995
From: George Cook <george at epix.net> (George Cook)
Date: Tue, 22 Aug 1995 10:39:28 -0400 (EDT)
Subject: NAQP restults p/t
Message-ID: <Pine.SUN.3.91.950822103136.4345B-100000 at peach.epix.net>
This was a great test of the new amplifryer and I am happy to say that it
preformed admirably! Also found out that the kids telephone downstairs
works as a pretty darn good shortwave reciever when exposed to enough
RF. Also found out that the computer that I built is far more RF proof
than the old one was no lock ups at all. Lastly found out that Jim Black
Baluns have lots and lots of smoke trapped inside them.
Don't know what all the belyaching was about on 80 meters I thought that
the band was ok and had rates upwards of 150/hr.
Was only on for 3 hours.
Likely will go for the gusto next year had a pretty good time.
Band Qs Mults
15 1 1 Groundplane @ 25 ft
20 91 27 Groundplane at 30 ft
80 63 33 Yaggis at 75 ft
160 2 2 Folded Dipole at 70 ft
-----------------------------------------------
Totals 157 63
-----------------------------------------------
Score 9891
---------------------------------
*************************************************
* George Cook AA3JU Bangor, PA FN21 *
* george at peach.epix.net AA3JU at N3IQD.EPA.USA.NA *
* If you're not FRC remember:............... *
* .......There's no shame in being 2nd best! *
*************************************************
>From Trey Garlough <GARLOUGH at TGV.COM> Tue Aug 22 16:10:01 1995
From: Trey Garlough <GARLOUGH at TGV.COM> (Trey Garlough)
Date: Tue, 22 Aug 1995 08:10:01 -0700 (PDT)
Subject: Multitasking
Message-ID: <809104201.973395.GARLOUGH at TGV.COM>
> For home computers I started with a kludged 4004 and 256 bytes of RAM...
We have gotten a bit out of the scope of contesting here. I think
it's time to move on to the next thread.
--Trey, WN4KKN/6
>From aa4lr at radio.org (Bill Coleman AA4LR) Tue Aug 22 16:22:52 1995
From: aa4lr at radio.org (Bill Coleman AA4LR) (Bill Coleman AA4LR)
Date: Tue, 22 Aug 1995 11:22:52 -0400
Subject: DSP Radios
Message-ID: <v01520d01ac5fa3acd167@[205.160.29.40]>
>Where will it all end? Will the DSP (and all the other computers) end up in
>the radio, or will the radio end up inside the PC?
I'm sure the radio will end up inside the PC.
Look at the trend. Already, modems are appearing inside the PC. NeXT and
Apple both briefly flirted with putting DSPs in their boxes, but recent
innovations made that obsolete. With today's RISC processors, the DSP
function can be performed by the host CPU.
Apple has been shipping such a solution for modems for two years, first
using the DSP chip, and in March 1994 moving totally to the host CPU. Intel
is talking about doing the same thing with Pentiums with their NSP (Native
Signal Processing) standard.
Given that the power of the desktop CPU continues to increase, it won't be
long before a lot this DSP processing could be done by the host CPU without
significant performance degradation. A V.32 modem chews up a lot of CPU
power, yet it doesn't cripple even the slowest of Apple's PowerPC
offerings.
All that will be needed is the appropriate signal interfaces - RF front
end, Mixer, DDS generator, short IF strip, and a digitizer. The rest could
happen inside the computer, with audio output and input handled via the
host sound I/O.
I'm sure this could happen today, especially on some of the 604 offerings.
Put it on a PCI card, and the hardware can plug into an Intel PC or a newer
Mac.
>Someone mentioned the Comfocus "Softwave" radio-on-a-PC-plugin-board product,
>which was at Dayton 94. Apparently Comfocus has become unglued somehow.
>
>THis year at Dayton, another of these boxes appeared. It's called "Radware",
>and is built by Comer Communications of San Diego, CA.
See? Someone is already doing it!
>Again - this is not to be construed as a plug for Brian's product (and no, I
>don't have one myself...). However, the performance possible using the
>>resourcfes
>of a PC and DSP-based sound peripheral is getting darned good.
Soon enough, you won't need that DSP for a sound peripheral, either.
Bill Coleman, AA4LR Mail: aa4lr at radio.org
Quote: "The same light shines on vineyards that makes deserts." -- Steve
Hackett
>From David & Barbara Leeson <0005543629 at mcimail.com> Tue Aug 22 16:44:00 1995
From: David & Barbara Leeson <0005543629 at mcimail.com> (David & Barbara Leeson)
Date: Tue, 22 Aug 95 10:44 EST
Subject: Q re Bev. kit
Message-ID: <24950822154442/0005543629NA5EM at MCIMAIL.COM>
K5GN asks why not just run the terminating radials on the ground, and do
they screw up the F/B? Operating in haste and ignorance (typical contest
condx, no?) I figured that the radials should be closer to the ground so
they would have substantially less signal pickup, but should not be so close
as to be lossy enough not to show any resonant behavior. I ran them in the
same direction as the beverage at both ends, trying not to have pickup in
some other direction.
Although signals didn't rise out of the noise like magic, they did the
job, both at P40V and HC8A. It would be worth modelling or experimenting,
but antennas close to the ground require more than MININEC, of course. It's
hard to model the wild donkeys and goats, but the electromagnetic part should
be interesting.
73 de Dave, W6QHS
>From Ronald D. Rossi" <rrossi at VNET.IBM.COM Tue Aug 22 17:43:02 1995
From: Ronald D. Rossi" <rrossi at VNET.IBM.COM (Ronald D. Rossi)
Date: Tue, 22 Aug 95 12:43:02 EDT
Subject: Big Blue - Team Score
Ref: Your note of Tue, 22 Aug 1995 11:32:10 -0500 (CDT)
Subject: Big Blue - Team Score
|corrected the call...SORRY Rich...thanks George!
WQ5G 332 Q X 119 M = 39,568 paper log
| WB5M 340 Q X 83 M = 28,220 8 hours
N1PBT 241 Q X 86 M = 20,726 no excuses
WA1PRY 241 Q X 48 M = 11,568 almost 20m only
WA2IBM 000 Q X 00 M = 00,000
-------
100,082 Total
Well it would be nice to say that we were all running our favorite
loggers on Aptivas under OS/2 given our affiliation, but we weren't!
I personally was not able to scrounge up any 15m or 10m contacts.
I called to deaf ears many times and never heard any of the multis
that alerted this multiplier to their presence.
Thanks to all the great ops that pulled me out of the noise on 80m.
73 de N1PBT...Ron (rrossi at vnet.ibm.com) ><>
>From Straw, Dean, N6BV" <rdstraw at arrl.org Tue Aug 22 18:21:00 1995
From: Straw, Dean, N6BV" <rdstraw at arrl.org (Straw, Dean, N6BV)
Date: Tue, 22 Aug 95 13:21:00 EDT
Subject: Q re Bev. kit
Message-ID: <303A12A0 at arrl.org>
Dave: That can be modeled using NEC2 or equivalent. I think you're right.
73, Dean, N6BV
----------
K5GN asks why not just run the terminating radials on the ground, and do
they screw up the F/B? Operating in haste and ignorance (typical contest
condx, no?) I figured that the radials should be closer to the ground so
they would have substantially less signal pickup, but should not be so close
as to be lossy enough not to show any resonant behavior. I ran them in the
same direction as the beverage at both ends, trying not to have pickup in
some other direction.
Although signals didn't rise out of the noise like magic, they did the
job, both at P40V and HC8A. It would be worth modelling or experimenting,
but antennas close to the ground require more than MININEC, of course. It's
hard to model the wild donkeys and goats, but the electromagnetic part
should
be interesting.
73 de Dave, W6QHS
>From Steve Runyon WQ5G 512-838-7008 <steve at austin.ibm.com> Tue Aug 22 19:28:50 1995
From: Steve Runyon WQ5G 512-838-7008 <steve at austin.ibm.com> (Steve Runyon WQ5G 512-838-7008)
Date: Tue, 22 Aug 95 12:28:50 -0600
Subject: Integrated Rig and PC (was DSP Radios)
Message-ID: <9508221728.AA25133 at runyon.austin.ibm.com>
I found Bill's note (and this thread) extremely interesting:
I like the idea of an integrated rig/PC - just plug the
optional 'ham radio card' into the expansion slot (user modifiable
for 10 meters upon supplying proof of valid amateur license of course)...
I guess instead of knobs, we'd have to use the mouse to select the
functions, tune the rig, etc ... well, OK, I could get used to that.
I'd definitely want the optional heavy-duty computer power supply to
get the full 100 Watts output - I need it to drive the external
linear, plus I don't want to be QRP all of the time! I could even go for
the optional fiber-optic interface to allow placing the amp at the
base of the tower and get rid of the feed-line losses. (Those
searching for the last fraction of a dB could even mount the
PA right below the beam, as was suggested yesterday.)
Of course, it'd come with a built in packet controller, and the
supplied logging program would interface directly with the
'ham card' via the system bus instead of this archaic RS-232
stuff...
Bill's right, this could be designed and built today!
COOL!! Sign me up for one! I guess I'd prefer one of the PowerPC 604
models (I did a lot of the physical design of the chip, it'd be a gas
to have a processor I worked on be the heart of my rig as well
as my computer!)
de Steve
(standard disclaimer: obviously anything said here is strictly
my own opinionated opinion)
------- Forwarded Message
From: aa4lr at radio.org (Bill Coleman AA4LR)
Subject: Re: DSP Radios
I'm sure the radio will end up inside the PC.
...
Look at the trend. Already, modems are appearing inside the PC. ...
...With today's RISC processors, the DSP function can be performed
by the host CPU.
.....
All that will be needed is the appropriate signal interfaces - RF front
end, Mixer, DDS generator, short IF strip, and a digitizer. The rest could
happen inside the computer, with audio output and input handled via the
host sound I/O.
I'm sure this could happen today, especially on some of the 604 offerings.
Put it on a PCI card, and the hardware can plug into an Intel PC or a newer
Mac.
......
More information about the CQ-Contest
mailing list