Oldest Instruction Set still in daily use?
invalid
every x86 PC to this day?
Or even the 8087 floating point processor.
Question stimulated by reading "The 8086 Family User's manual,
Numerics Supplement" dated July 1980 - that's nigh on 30 years.
Roger Ivie
> Could it be the 16-bit 8086 set, still used in the boot-up of
> every x86 PC to this day?
Nope, not even close, even if you only consider the PC. The keyboard
controller in the original PC was an 8048, with an instruction set much
older than the 8086. I wouldn't be surprised if some machines with PS/2
keyboard connectors are still using them.
A popular USB microcontroller is the 68HC08 series, based on the
Motorola 6800 instruction set, which predates the 6502. Your USB keyboard
or mouse might have an HC08 in it.
Outside the PC, of course, you have the mainframes. The newest IBM
machines still execute the /370 instruction set. And there are the
UNISYS A-series, which execute an instruction set descended from
the Burroughs 5500.
I suspect there are still PDP-8s out there silently doing their work.
They're running the same instruction set as the PDP-5, introduced in
1963.
I'm not familiar with the Univac machines, but given the recent thread on
executing 40 year-old software, I suspect they may be even older.
--
roger ivie
ri...@ridgenet.net
ArarghMai...@not.at.arargh.com
wrote:
More likely the 8080 or Z80 set, in embedded applications.
Not that I actually know or anything - it's just a guess.
--
ArarghMail001 at [drop the 'http://www.' from ->] http://www.arargh.com
BCET Basic Compiler Page: http://www.arargh.com/basic/index.html
To reply by email, remove the extra stuff from the reply address.
Nick Spalding
on Sun, 03 Jan 2010 04:18:16 -0600:
> On 2010-01-03, invalid <inv...@invalid.invalid> wrote:
> > Could it be the 16-bit 8086 set, still used in the boot-up of
> > every x86 PC to this day?
>
> Nope, not even close, even if you only consider the PC. The keyboard
> controller in the original PC was an 8048, with an instruction set much
> older than the 8086. I wouldn't be surprised if some machines with PS/2
> keyboard connectors are still using them.
>
> A popular USB microcontroller is the 68HC08 series, based on the
> Motorola 6800 instruction set, which predates the 6502. Your USB keyboard
> or mouse might have an HC08 in it.
>
> Outside the PC, of course, you have the mainframes. The newest IBM
> machines still execute the /370 instruction set. And there are the
> UNISYS A-series, which execute an instruction set descended from
> the Burroughs 5500.
The /370 set includes the /360 one.
> I suspect there are still PDP-8s out there silently doing their work.
> They're running the same instruction set as the PDP-5, introduced in
> 1963.
>
> I'm not familiar with the Univac machines, but given the recent thread on
> executing 40 year-old software, I suspect they may be even older.
--
Nick Spalding
jmfbahciv
about IBM sets, though.
/BAH
Anne & Lynn Wheeler
"invalid" <inv...@invalid.invalid> writes:
> Question stimulated by reading "The 8086 Family User's manual,
> Numerics Supplement" dated July 1980 - that's nigh on 30 years.
x-over when i added a.f.c. to bit.listserv.ibm-main thread:
http://www.garlic.com/~lynn/2009r.html#52 360 programs on a z/10
http://www.garlic.com/~lynn/2009r.html#57 360 programs on a z/10
http://www.garlic.com/~lynn/2009r.html#68 360 programs on a z/10
http://www.garlic.com/~lynn/2009r.html#69 360 programs on a z/10
http://www.garlic.com/~lynn/2009s.html#60 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#4 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#9 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#13 360 programs on a z/10
360 was announced 1964 (models started shipping 1965) ... a22-6821-0 360
principles of operation (with 24bit addresses)
http://bitsavers.org/pdf/ibm/360/princOps/A22-6821-0_360PrincOps.pdf
this is not only all the 360 application instructions continue to be
supported today ... but things like compatible program libraries and
kernel calls continue to work on modern systems.
more recent principles of operation
http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/dz9zr003/CCONTENTS
with 31-bit and 64-bit addressing (in addition to 24-bit addressing).
not only do the 24-bit address instructions continue to work ... the
programs written using 24-bit address instructions ... including use of
library, system services, kernel calls, i/o requests (aka 360 assembler,
machine language programs continue to run ... with or w/o re-assemly).
--
40+yrs virtualization experience (since Jan68), online at home since Mar1970
Walter Bushell
"invalid" <inv...@invalid.invalid> wrote:
There are probably 8080 and Z80 chips or chips using that instruction
set being used in the embedded world. If you sell millions of toasters,
a penny on each is a nice number.
--
A computer without Microsoft is like a chocolate cake without mustard.
Christopher C. Stacy
> In article <hhppnt$cv3$1...@news.eternal-september.org>,
> "invalid" <inv...@invalid.invalid> wrote:
>
>> Could it be the 16-bit 8086 set, still used in the boot-up of
>> every x86 PC to this day?
>>
>> Or even the 8087 floating point processor.
>>
>> Question stimulated by reading "The 8086 Family User's manual,
>> Numerics Supplement" dated July 1980 - that's nigh on 30 years.
>
> There are probably 8080 and Z80 chips or chips using that instruction
> set being used in the embedded world. If you sell millions of toasters,
> a penny on each is a nice number.
There absolutely are Z80s in embedded devices in use today.
(POS credit card terminals, and Money Order machines, for two examples I know of.)
Charles Richmond
When I went to college in the 70's, there was a professor in the
psychology department who had a PDP-8 to monitor his experiments
with animals. I considered the PDP-8 to be an *old* system then. I
know that this professor used the PDP-8 at least until the
mid-90's. The system may *still* be used AFAIK.
--
+----------------------------------------+
| Charles and Francis Richmond |
| |
| plano dot net at aquaporin4 dot com |
+----------------------------------------+
Charles Richmond
There are probably even Z80's in electronic typewriters that are
still seeing service typing mailing labels and such.
Patrick Scheible
> On Sun, 3 Jan 2010 09:58:20 -0000, "invalid" <inv...@invalid.invalid>
> wrote:
>
> >Could it be the 16-bit 8086 set, still used in the boot-up of
> >every x86 PC to this day?
> >
> >Or even the 8087 floating point processor.
> >
> >Question stimulated by reading "The 8086 Family User's manual,
> >Numerics Supplement" dated July 1980 - that's nigh on 30 years.
> >
> More likely the 8080 or Z80 set, in embedded applications.
>
> Not that I actually know or anything - it's just a guess.
I bet there's still IBM 360 code in use. That would be 45+ years old.
But I don't have specific examples either.
-- Patrick
Roger Ivie
> There are probably even Z80's in electronic typewriters that are
> still seeing service typing mailing labels and such.
TI-83 calculators.
--
roger ivie
ri...@ridgenet.net
cjt
> Christopher C. Stacy wrote:
>> Walter Bushell <pr...@panix.com> writes:
>>
>>> In article <hhppnt$cv3$1...@news.eternal-september.org>,
>>> "invalid" <inv...@invalid.invalid> wrote:
>>>
>>>> Could it be the 16-bit 8086 set, still used in the boot-up of
>>>> every x86 PC to this day?
>>>>
>>>> Or even the 8087 floating point processor.
>>>>
>>>> Question stimulated by reading "The 8086 Family User's manual,
>>>> Numerics Supplement" dated July 1980 - that's nigh on 30 years.
>>> There are probably 8080 and Z80 chips or chips using that instruction
>>> set being used in the embedded world. If you sell millions of
>>> toasters, a penny on each is a nice number.
>>
>> There absolutely are Z80s in embedded devices in use today.
>> (POS credit card terminals, and Money Order machines, for two examples
>> I know of.)
>
> There are probably even Z80's in electronic typewriters that are still
> seeing service typing mailing labels and such.
>
probably even earlier big machines (e.g. Bendix, Burroughs, IBM, and
Sperry machines -- maybe even some CDCs).
Charles Richmond
> On 2010-01-04, Charles Richmond <fri...@tx.rr.com> wrote:
>> There are probably even Z80's in electronic typewriters that are
>> still seeing service typing mailing labels and such.
>
> TI-83 calculators.
Oh, yeah. I forgot those "graphing calculators". ISTM that the
students to day are allowed to use a graphing calculator when they
take their SAT test.
Charles Richmond
Yeah, we've got to keep in mind that a lot of PDP-11's were built
into large industrial machines as control computers. Industry is
loath to get rid of those machines *if* the machines are still
working correctly. So those embedded PDP-11's are running their
ROM programs every day, working tirelessly to get the job done.
Charles Richmond
ISTM that the DF/W airport in the Dallas-Ft. Worth area still uses
*several* IBM 360's to run their regional air controller radar.
That would probably be impossible to verity due to security
concerns, but I think they are still using IBM 360's for this purpose.
isw
Walter Bushell <pr...@panix.com> wrote:
> In article <hhppnt$cv3$1...@news.eternal-september.org>,
> "invalid" <inv...@invalid.invalid> wrote:
>
> > Could it be the 16-bit 8086 set, still used in the boot-up of
> > every x86 PC to this day?
> >
> > Or even the 8087 floating point processor.
> >
> > Question stimulated by reading "The 8086 Family User's manual,
> > Numerics Supplement" dated July 1980 - that's nigh on 30 years.
>
> There are probably 8080 and Z80 chips or chips using that instruction
> set being used in the embedded world. If you sell millions of toasters,
> a penny on each is a nice number.
8080 and Z-80 processors do not have the same instruction set as the
8086 -- not even close.
I was once given the task of "replacing" the 8080 in our product with
the then brand-new 8086, after some liar from iNtel told our marketing
guy that the "programs" were portable.
Turns out the iNtel guy was sort-of-but-not-very right *provided* you
had written your 8080 programs in PL/M, and were willing to port them to
PLM/86. If they were in assembler (and ours were) you could forget about
it.
Isaac
ArarghMai...@not.at.arargh.com
<fri...@tx.rr.com> wrote:
>Patrick Scheible wrote:
>> ArarghMai...@NOT.AT.Arargh.com writes:
>>
>>> On Sun, 3 Jan 2010 09:58:20 -0000, "invalid" <inv...@invalid.invalid>
>>> wrote:
>>>
>>>> Could it be the 16-bit 8086 set, still used in the boot-up of
>>>> every x86 PC to this day?
>>>>
>>>> Or even the 8087 floating point processor.
>>>>
>>>> Question stimulated by reading "The 8086 Family User's manual,
>>>> Numerics Supplement" dated July 1980 - that's nigh on 30 years.
>>>>
>>> More likely the 8080 or Z80 set, in embedded applications.
>>>
>>> Not that I actually know or anything - it's just a guess.
>>
>> I bet there's still IBM 360 code in use. That would be 45+ years old.
>>
>> But I don't have specific examples either.
>>
>
>ISTM that the DF/W airport in the Dallas-Ft. Worth area still uses
>*several* IBM 360's to run their regional air controller radar.
>That would probably be impossible to verity due to security
>concerns, but I think they are still using IBM 360's for this purpose.
I remember that at one time there was a big push to replace the aging
(I think modified) 360/50s that did that. I don't remember ever
hearing of a sucessful completion. :-)
ArarghMai...@not.at.arargh.com
wrote:
There may well be somebody somewhere still running production 14xx
code in emulation.
Every so often I run some 1401 code on a emulator that I wrote.
If I get REALLY bored, I have a IBM 705 emulator that I could run on a
simulated 370 in Hercules. :-) I would have to check to see if the
705 code still needs the hardware assist that was available on the
360/65. :-)
And I think I have heard of an IBM 650 emulator. :-)
And then there is the NOVA emulator I wrote and never really finished.
All of which are older than an x86.
Joe Pfeiffer
There was a cross-assembler...
--
As we enjoy great advantages from the inventions of others, we should
be glad of an opportunity to serve others by any invention of ours;
and this we should do freely and generously. (Benjamin Franklin)
invalid
"invalid" <inv...@invalid.invalid> wrote:
> Could it be the 16-bit 8086 set, still used in the boot-up of
> every x86 PC to this day?
> Or even the 8087 floating point processor.
> Question stimulated by reading "The 8086 Family User's manual,
> Numerics Supplement" dated July 1980 - that's nigh on 30 years.
The PDP-11 came about in 1969. The books to which I refer
were only 11 years later.
11 years for the capabilities of a computer costing 10 times my
annual salary to be available in one costing only 1 weeks' salary is one
hell of
a rate of development.
Perhaps the current recession is because the computer in one form or
another has been driving the boom; but now that the market has saturated
(We've all got 32 bit floating-point processors with mega memory on our
desks)
the boom is over?
Morten Reistad
I know of several government outfits where they fought hard
to get rid of the 1401 code up until y2k, but didn't quite succeed.
The 1401 was announced 50 years and two months ago.
360 code that runs on zSeries machines isn't even on the radar as
"legacy" in many places.
Even pre 1.7 CICS code (25 years old now) that needed relinking a few
years ago made minor havoc.
-- mrr
Morten Reistad
>> I have little doubt there are still DEC PDP-11s in daily use, and
>> probably even earlier big machines (e.g. Bendix, Burroughs, IBM, and
>> Sperry machines -- maybe even some CDCs).
The Reuters "Monitor" is still operational. That is written in
pdp11 assembly and machine code, very intimate to the hardware.
The "Monitor" _is_ the OS. The physical PDP11s were replaced with
emulators, then on Sun hardware, in the mid 1990s. But ISTR they
are back on hardware-PDP11 implementations now.
>Yeah, we've got to keep in mind that a lot of PDP-11's were built
>into large industrial machines as control computers. Industry is
>loath to get rid of those machines *if* the machines are still
>working correctly. So those embedded PDP-11's are running their
>ROM programs every day, working tirelessly to get the job done.
A few friends ran a company to serve legacy DEC customers for many
years.
Industrial control PDP11s were their main source of income. Many of
them were virtualised on commodity hardware, but some of them needed
the actual controllers. One, very impressive one, controlled a steel
rolling plant, bending 6mm steel like it was paper.
-- mrr
Anne & Lynn Wheeler
> ISTM that the DF/W airport in the Dallas-Ft. Worth area still uses
> *several* IBM 360's to run their regional air controller radar. That
> would probably be impossible to verity due to security concerns, but I
> think they are still using IBM 360's for this purpose.
I not sure about dallas ... but i remember seeing references to Denver
FAA having simulated 360s from FSI running on intel platform (in
production use):
http://www.funsoft.com/
funsoft had made major product platform choice of sequent ... and after
ibm bought sequent ... that seemed to better cement such relationship
... but then ibm discontinued all the (numa-q) sequent stuff ... and
more recently there has been a lot of contention with the 360 software
emulators running on intel platforms (including various legal action
with one of the other players).
Dave Wade
"invalid" <inv...@invalid.invalid> wrote in message
news:hhppnt$cv3$1...@news.eternal-september.org...
Whilst its not in "daily" use the "Babe" replica at Manchester Museum of
Science and Industry
is still in weekly use for demonstrations.
Chris Burrows
news:1pGdnXok_umBTtzW...@eclipse.net.uk...
>
>
> Whilst its not in "daily" use the "Babe" replica at Manchester Museum of
> Science and Industry
>
> http://www.mosi.org.uk/2332
>
> is still in weekly use for demonstrations.
>
From the London Science Museum website:
"Another fascinating large object in the gallery is the 1956 Ferranti
Pegasus - the oldest working computer in the world and one that is switched
on and run regularly to the delight of visitors. "
http://www.sciencemuseum.org.uk/visitmuseum/galleries/computing.aspx
The "Baby" dates back to 1948 but the specimen in Manchester is a replica.
Perhaps the Pegasus is an original which would explain the "oldest" claim.
--
Chris Burrows
CFB Software
http://www.cfbsoftware.com
jmfbahciv
but is the instruction set in the hardware? I'm pretty sure there
are still PDP-8s and PDP-11s running.
/BAH
hanc...@bbs.cpcn.com
> I know of several government outfits where they fought hard
> to get rid of the 1401 code up until y2k, but didn't quite succeed.
> The 1401 was announced 50 years and two months ago.
Are you saying there are still some places using 1401 emulation today?
If they didn't succeed in converting their code to Y2k, what did they
do?
(AFAIK, if a _1401_ application didn't _compare_ two dates it was
essentially Y2k compatible.)
hanc...@bbs.cpcn.com
> Could it be the 16-bit 8086 set, still used in the boot-up of
> every x86 PC to this day?
One candidate for "oldest instruction set still in daily use" is the
System/360, announced April 1964, still in extremely wide use
throughout the world as part of IBM's Z series.
Anne & Lynn Wheeler
jmfbahciv <jmfbahciv@aol> writes:
> but is the instruction set in the hardware? I'm pretty sure there
> are still PDP-8s and PDP-11s running.
re:
http://www.garlic.com/~lynn/2010.html#14 Oldest Instruction Set still in daily use?
instruction nearly all hardware these days (change from the days when
most low/mid range 370s were vertical microcode, and high-end was
horizontal microcode) ... except for a few of the more complex
instructions done in "millicode" (nearly all of the original 360
instructions fall into the "less complex" category).
Millicode in an IBM zSeries processor
http://findarticles.com/p/articles/mi_qa3751/is_200405/ai_n9388162/
from above:
While the hardware can execute many of the logically less complex and
high-performance instructions, millicode is required to implement the
more complex instructions, as well as to provide additional support
functions related primarily to the central processor. This paper is a
review of millicode on previous zSeries CMOS systems and also describes
enhancements made to the z990 system for processing of the millicode. It
specifically discusses the flexibility millicode provides to the z990
system.
... snip ...
there are slightly different issues of what is the 1) oldest machine
still running, 2) oldest software still running on new/existing
hardware, and 3) oldest software still running on emulated
hardware. reference to old software running on new "intel" hardware
under 360 simulator
http://www.garlic.com/~lynn/2010.html#27 Oldest Instruction Set still in daily use?
and posts in thread about 360 programs running on current hardware:
http://www.garlic.com/~lynn/2009r.html#52 360 programs on a z/10
http://www.garlic.com/~lynn/2009r.html#57 360 programs on a z/10
http://www.garlic.com/~lynn/2009r.html#68 360 programs on a z/10
http://www.garlic.com/~lynn/2009r.html#69 360 programs on a z/10
http://www.garlic.com/~lynn/2009s.html#60 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#4 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#9 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#13 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#16 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#17 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#20 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#24 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#26 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#28 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#29 360 programs on a z/10
http://www.garlic.com/~lynn/2010.html#30 360 programs on a z/10
Walter Bushell
> Yeah, we've got to keep in mind that a lot of PDP-11's were built
> into large industrial machines as control computers. Industry is
> loath to get rid of those machines *if* the machines are still
> working correctly. So those embedded PDP-11's are running their
> ROM programs every day, working tirelessly to get the job done.
Quite. Replace a working system that we know and replace it with
something else??? Replace the interfaces and the programming. I think
not.
Charlie Gibbs
(isw) writes:
> 8080 and Z-80 processors do not have the same instruction set as the
> 8086 -- not even close.
>
> I was once given the task of "replacing" the 8080 in our product with
> the then brand-new 8086, after some liar from iNtel told our marketing
> guy that the "programs" were portable.
>
> Turns out the iNtel guy was sort-of-but-not-very right *provided* you
> had written your 8080 programs in PL/M, and were willing to port them
> to PLM/86. If they were in assembler (and ours were) you could forget
> about it.
You mean the whole world has suffered all that backward-compatibility
crap for nothing? 1/2 :-)
Seriously, though, I heard about a program that was supposed to be
able to translate 8080 assembly language programs to 8086. I never
saw it myself. Did you have an opportunity to try it? If so, what
was the result?
--
/~\ cgi...@kltpzyxm.invalid (Charlie Gibbs)
\ / I'm really at ac.dekanfrus if you read it the right way.
X Top-posted messages will probably be ignored. See RFC1855.
/ \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign!
Stan Barr
<cfbso...@hotmail.com> wrote:
> "Dave Wade" <g8...@yahoo.com> wrote in message
> news:1pGdnXok_umBTtzW...@eclipse.net.uk...
>>
>>
>> Whilst its not in "daily" use the "Babe" replica at Manchester Museum of
>> Science and Industry
>>
>> http://www.mosi.org.uk/2332
>>
>> is still in weekly use for demonstrations.
>>
>
> From the London Science Museum website:
>
> "Another fascinating large object in the gallery is the 1956 Ferranti
> Pegasus - the oldest working computer in the world and one that is switched
> on and run regularly to the delight of visitors. "
>
> http://www.sciencemuseum.org.uk/visitmuseum/galleries/computing.aspx
According to Wikipedia the Science Museum one ran its first program in 1959.
Nice photo...very 1940s/50s industrial architectural design!
http://en.wikipedia.org/wiki/Ferranti_Pegasus
http://en.wikipedia.org/wiki/File:Pegasus_computer.jpg
--
Cheers,
Stan Barr plan.b .at. dsl .dot. pipex .dot. com
The future was never like this!
John Varela
<fri...@tx.rr.com> wrote:
> ISTM that the DF/W airport in the Dallas-Ft. Worth area still uses
> *several* IBM 360's to run their regional air controller radar.
> That would probably be impossible to verity due to security
> concerns, but I think they are still using IBM 360's for this purpose.
360s have never been used at airport towers or Approach Controls.
Configured as the multi-processor 9020 computer, 360s were used at
the Air Route Traffic Control Centers, of which there is one near
DFW. The 9020s were replaced in the mid-'80s with IBM 3083
mainframes. The system is written in JOVIAL and BAL. There is a
current project to replace the hardware and rewrite and upgrade the
software; I don't know its status but I don't predict early success
in replacing the software.
--
John Varela
Trade NEWlamps for OLDlamps for email
John Varela
ArarghMai...@NOT.AT.Arargh.com wrote:
> I remember that at one time there was a big push to replace the aging
> (I think modified) 360/50s that did that. I don't remember ever
> hearing of a sucessful completion. :-)
The 9020s included both 360/50s and 360/65s configured as a
multi-processor system. They were replaced with 3083s in the mid
1980s. I was the MITRE project leader on the replacement job. What
would you like to know? (Not that I remember much from 25 years
ago.)
Peter Flass
>
> If I get REALLY bored, I have a IBM 705 emulator that I could run on a
> simulated 370 in Hercules. :-) I would have to check to see if the
> 705 code still needs the hardware assist that was available on the
> 360/65. :-)
So, you could write an emulator add-on for the hardware assist...
Joe Morris
> Roger Ivie wrote:
>> On 2010-01-04, Charles Richmond <fri...@tx.rr.com> wrote:
>>> There are probably even Z80's in electronic typewriters that are still
>>> seeing service typing mailing labels and such.
>> TI-83 calculators.
> Oh, yeah. I forgot those "graphing calculators". ISTM that the students to
> day are allowed to use a graphing calculator when they take their SAT
> test.
Today graphing calculators are *required* for the Math tests and for some of
the questions on the SAT. The College Board has the following to say about
calculators and the Mathematics Level 1 and 2 tests:
http://www.collegeboard.com/student/testing/sat/testday/calc.html
And for tha AP tests:
http://www.collegeboard.com/prod_downloads/ap/students/calculus/ap-cd-calc-0607.pdf
Look at the section on graphing calculator requirements on pp. 14-15.
Joe Morris
ArarghMai...@not.at.arargh.com
wrote:
I think that they WERE replaced is sufficient. :-) I had just never
heard that.
ArarghMai...@not.at.arargh.com
Could...
Thinking about it, I think that the hardware assist was for a 40, and
that the emulator ran on the bare metal on a 65. :-) It's only been
35 years since I last ran it. Although, I did look at the source
within last year about another matter, something to do about the
translate instruction.
ArarghMai...@not.at.arargh.com
IIRC, most 1401 based applications would have been using a 2 digit
year, so if you didn't mind date based reports being in the wrong
order, yes. :-)
Charles Richmond
> In article <hhrvl2$s25$3...@news.eternal-september.org>,
> Charles Richmond <fri...@tx.rr.com> wrote:
>> cjt wrote:
>>> Charles Richmond wrote:
>
>>> I have little doubt there are still DEC PDP-11s in daily use, and
>>> probably even earlier big machines (e.g. Bendix, Burroughs, IBM, and
>>> Sperry machines -- maybe even some CDCs).
>
> The Reuters "Monitor" is still operational. That is written in
> pdp11 assembly and machine code, very intimate to the hardware.
> The "Monitor" _is_ the OS. The physical PDP11s were replaced with
> emulators, then on Sun hardware, in the mid 1990s. But ISTR they
> are back on hardware-PDP11 implementations now.
>
Yes, but... those darn PDP-11 instructions *are* *still* *being*
*executed*!!! Emulating is still executing... That's the title of
this thread: Oldest Instruction Set still in daily use.
>> Yeah, we've got to keep in mind that a lot of PDP-11's were built
>> into large industrial machines as control computers. Industry is
>> loath to get rid of those machines *if* the machines are still
>> working correctly. So those embedded PDP-11's are running their
>> ROM programs every day, working tirelessly to get the job done.
>
> A few friends ran a company to serve legacy DEC customers for many
> years.
>
> Industrial control PDP11s were their main source of income. Many of
> them were virtualised on commodity hardware, but some of them needed
> the actual controllers. One, very impressive one, controlled a steel
> rolling plant, bending 6mm steel like it was paper.
>
ISTM that on <alt.sys.pdp11>, the discussion was once about such a
commercial emulator that cost a *lot* to use. A hobbyist license
was what was desired in that case.
Charles Richmond
Okay, but that means that IBM 360's are (are were recently) in
active use at D/FW Airport.
Charles Richmond
>> "invalid" <inv...@invalid.invalid> wrote:
>>
>>> Could it be the 16-bit 8086 set, still used in the boot-up of
>>> every x86 PC to this day?
>>>
>>>
>>> Numerics Supplement" dated July 1980 - that's nigh on 30 years.
>> set being used in the embedded world. If you sell millions of toasters,
>> a penny on each is a nice number.
>
> 8086 -- not even close.
>
> I was once given the task of "replacing" the 8080 in our product with
> the then brand-new 8086, after some liar from iNtel told our marketing
> guy that the "programs" were portable.
>
> Turns out the iNtel guy was sort-of-but-not-very right *provided* you
> had written your 8080 programs in PL/M, and were willing to port them to
> PLM/86. If they were in assembler (and ours were) you could forget about
> it.
>
ISTM that I remember some conversion program that would take in
8080 assembly language and put out 8086 assembly. The registers in
the 8086 somewhat mirror the 8080 registers and conversion should
*not* be that difficult. Everything can be done in the small
memory model.
hanc...@bbs.cpcn.com
> >(AFAIK, if a _1401_ application didn't _compare_ two dates it was
> >essentially Y2k compatible.)
>
> IIRC, most 1401 based applications would have been using a 2 digit
> year, so if you didn't mind date based reports being in the wrong
> order, yes. :-)
To sort a date requires a comparison of the date field.
But not all reports sort by date, some by name or some other field(s).
Many old computer or tab machine applications used only a _single_
digit date year. Don't forget, the 16k 1401 was the deluxe model and
many users had less memory. How such systems handled decade rollover
I don't know.
isw
"Charlie Gibbs" <cgi...@kltpzyxm.invalid> wrote:
> In article <isw-8D4A52.22021703012010@[216.168.3.50]>, i...@witzend.com
> (isw) writes:
>
> > 8080 and Z-80 processors do not have the same instruction set as the
> > 8086 -- not even close.
> >
> > I was once given the task of "replacing" the 8080 in our product with
> > the then brand-new 8086, after some liar from iNtel told our marketing
> > guy that the "programs" were portable.
> >
> > Turns out the iNtel guy was sort-of-but-not-very right *provided* you
> > had written your 8080 programs in PL/M, and were willing to port them
> > to PLM/86. If they were in assembler (and ours were) you could forget
> > about it.
>
> You mean the whole world has suffered all that backward-compatibility
> crap for nothing? 1/2 :-)
>
> Seriously, though, I heard about a program that was supposed to be
> able to translate 8080 assembly language programs to 8086. I never
> saw it myself. Did you have an opportunity to try it? If so, what
> was the result?
This was at the time when iNtel was handing out 8086 design kits (in
shiny blue boxes) to selected customers; things like the
cross-assembler-sort-of that you mention were not available.
ISTR that many instructions had fairly comparable operation, but some
8080 instructions required several '86 instructions to emulate. I did a
few hand-converted test cases and decided that the switch to 8086 was
unlikely to result in more than a 2-to-1 speedup unless all the code
were to be rewritten from scratch. Marketing said "but iNtel told us TEN
TIMES AS FAST!"
So I told them to get iNtel to do the conversion, with a guarantee of
performance...
This was at Pertec Computer Corporation in LA, BTW; they're the folks
who bought MITS and (through amazingly incompetent management of the
acquisition) pounded them headfirst into the ground.
Isaac
Morten Reistad
Charles Richmond <fri...@tx.rr.com> wrote:
>Morten Reistad wrote:
>> In article <hhrvl2$s25$3...@news.eternal-september.org>,
>> Charles Richmond <fri...@tx.rr.com> wrote:
>>>> I have little doubt there are still DEC PDP-11s in daily use, and
>>>> probably even earlier big machines (e.g. Bendix, Burroughs, IBM, and
>>>> Sperry machines -- maybe even some CDCs).
>>
>> The Reuters "Monitor" is still operational. That is written in
>> pdp11 assembly and machine code, very intimate to the hardware.
>> The "Monitor" _is_ the OS. The physical PDP11s were replaced with
>> emulators, then on Sun hardware, in the mid 1990s. But ISTR they
>> are back on hardware-PDP11 implementations now.
>>
>
>Yes, but... those darn PDP-11 instructions *are* *still* *being*
>*executed*!!! Emulating is still executing... That's the title of
>this thread: Oldest Instruction Set still in daily use.
Yes, I saw monitor-terminals in use just a few weeks ago. Monitor
is still alive and kicking.
>>> Yeah, we've got to keep in mind that a lot of PDP-11's were built
>>> into large industrial machines as control computers. Industry is
>>> loath to get rid of those machines *if* the machines are still
>>> working correctly. So those embedded PDP-11's are running their
>>> ROM programs every day, working tirelessly to get the job done.
>>
>> A few friends ran a company to serve legacy DEC customers for many
>> years.
>>
>> Industrial control PDP11s were their main source of income. Many of
>> them were virtualised on commodity hardware, but some of them needed
>> the actual controllers. One, very impressive one, controlled a steel
>> rolling plant, bending 6mm steel like it was paper.
>>
>
>ISTM that on <alt.sys.pdp11>, the discussion was once about such a
>commercial emulator that cost a *lot* to use. A hobbyist license
>was what was desired in that case.
For these cases, commercial emulators were a perfectly viable
solution. The customers had such a huge investment around the PDP11
that the level of support etc around such emulators were just the
right deal.
When the attached hardware can bend 6mm steel to build a "paper plane",
out of steel I would like a supported platform too.
-- mrr
invalid
news:hhua20$cp3$3...@news.eternal-september.org...
>
> ISTM that I remember some conversion program that would take in 8080
> assembly language and put out 8086 assembly. The registers in the 8086
> somewhat mirror the 8080 registers and conversion should *not* be that
> difficult. Everything can be done in the small memory model.
It was only partly usable because of the incompatibility between the
way that the condition codes were set. (Might have been due to the
lack of an Overflow bit in the 8080) but ISTR long winded auto-generated
instruction sequences to ensure that condition codes created by 8086 code
ended up in the way that they would have been set by 8080 code.
(experience from 29 years ago, so possible interference from Fuzzy
Logic in ageing brain)
jmfbahciv
If the data was on cards, the card decks/decade were separate. the
human was the machine that did that kind of sorting. Part of using
the sorter was deciding which decks to put in first, then
second, ...., then last.
/BAH
jmfbahciv
/BAH
hanc...@bbs.cpcn.com
> > Many old computer or tab machine applications used only a _single_
> > digit date year. Don't forget, the 16k 1401 was the deluxe model and
> > many users had less memory. How such systems handled decade rollover
> > I don't know.
>
> If the data was on cards, the card decks/decade were separate. the
> human was the machine that did that kind of sorting. Part of using
> the sorter was deciding which decks to put in first, then
> second, ...., then last.
Also, which order the sort output trays are emptied and reassembled.
Good point. In the 1401 years, even in S/360 years, card sorters were
an integral part of the processing. So undoubtedly manual operator
intervention--which would not be that hard--could handle a lot of
issues.
On a modern system, if the job stream had a sort step, these could be
changed to select to pull off older years and later concatenate files
to be in the desired order.
hanc...@bbs.cpcn.com
> Yes, but... those darn PDP-11 instructions *are* *still* *being*
> *executed*!!! Emulating is still executing... That's the title of
> this thread: Oldest Instruction Set still in daily use.
System/360 instructions are older than PDP-11, aren't they? Didn't
the first PDPs appear in the mid to later 1960s?
Scott Lurndal
>Morten Reistad wrote:
>> In article <hhrvl2$s25$3...@news.eternal-september.org>,
>> Charles Richmond <fri...@tx.rr.com> wrote:
>>> cjt wrote:
>>>> Charles Richmond wrote:
>>
>>>> I have little doubt there are still DEC PDP-11s in daily use, and
>>>> probably even earlier big machines (e.g. Bendix, Burroughs, IBM, and
>>>> Sperry machines -- maybe even some CDCs).
>>
>> The Reuters "Monitor" is still operational. That is written in
>> pdp11 assembly and machine code, very intimate to the hardware.
>> The "Monitor" _is_ the OS. The physical PDP11s were replaced with
>> emulators, then on Sun hardware, in the mid 1990s. But ISTR they
>> are back on hardware-PDP11 implementations now.
>>
>
>Yes, but... those darn PDP-11 instructions *are* *still* *being*
>*executed*!!! Emulating is still executing... That's the title of
>this thread: Oldest Instruction Set still in daily use.
There are still Burroughs medium systems in production. Parts of
the instruction set date back to the Electrodata 200 systems from
the late 50's. Applications written in the early 60's will still
run on the hardware (which was designed in the 80's).
scott
Joe Morris
> John Varela wrote:
>> Charles Richmond wrote:
>>> ISTM that the DF/W airport in the Dallas-Ft. Worth area still uses
>>> *several* IBM 360's to run their regional air controller radar. That
>>> would probably be impossible to verity due to security concerns, but I
>>> think they are still using IBM 360's for this purpose.
>> 360s have never been used at airport towers or Approach Controls.
>> Configured as the multi-processor 9020 computer, 360s were used at the
>> Air Route Traffic Control Centers, of which there is one near DFW.
> Okay, but that means that IBM 360's are (are were recently) in active use
> at D/FW Airport.
I'm missing something here. John wrote that 360s have never been used at
towers or approach facilities; how does that imply that 360s are (or were)
in active use at the airport?
Note that Ft. Worth ARTCC != DFW Airport even if they are neighbors. ZFW is
located on land that used to be part of Greater Southwest Airport, just two
miles or so south of KDFW. Open GoogleMaps and look at the satellite images
just south of KDFW; just south of Airport Freeway there's an undeveloped
section of land where you can still see the outline of a runway and taxiway.
ZFW is to the west of this open land.
Joe Morris
hanc...@bbs.cpcn.com
> There are still Burroughs medium systems in production.
I presume they're manufactured by Unisys?
How many other machines of Burroughs or Univac heritage are still made
by Unisys?
(I thought Unisys was only into support and integration these days.
Unfortunately, every time they're in the newspaper they seem to be
contracting still more. Many of those news notes are posted here.
Charles Richmond
Okay, Joe. I missed that one. Mia culpa.
The IBM 360's are used *near* the D/FW airport. What is important
to me is that the IBM 360's *are* still in active use. These 360's
are still actively executing IBM 360 instructions each day.
It really seemed to me that the 360's had *something* to do with
the regional radar, but maybe I was mistaken about that. I would
call and ask someone, except I would probably end up on a
terrorist "watch list". ;-)
Peter Flass
Marketwatch just reported that Unisys made a lot of money last year in
services - they seem to be re-inventing themselves like IBM. If you
look at their website, it's hard to discover they still sell computers.
I don't know how much is hardware and how much emulation, but they
still sell systems that run descendants of MCP (ex-Burroughs) end EXEC
(ex-Univac).
The same article mentioned NCR, which is also still alive and pushing
the service business, but doing less well.
Louis Krupp
> hanc...@bbs.cpcn.com wrote:
>> On Jan 5, 1:18 pm, sc...@slp53.sl.home (Scott Lurndal) wrote:
>>
>>> There are still Burroughs medium systems in production.
>>
>> I presume they're manufactured by Unisys?
I took that to mean that Burroughs Medium Systems are still "in
production" in the sense that they're still in operation, not that
they're still being produced. My recollection is that they survived the
merger for a while as the Unisys V-Series and then were discontinued
(but I couldn't tell you when). (Large Systems became the A-Series and
are now ClearPath or something, while at least one Univac architecture
is now ClearPath something else.)
Louis
jmfbahciv
> On Jan 5, 8:31 am, jmfbahciv <jmfbahciv@aol> wrote:
>
>>> Many old computer or tab machine applications used only a _single_
>>> digit date year. Don't forget, the 16k 1401 was the deluxe model and
>>> many users had less memory. How such systems handled decade rollover
>>> I don't know.
>> If the data was on cards, the card decks/decade were separate. the
>> human was the machine that did that kind of sorting. Part of using
>> the sorter was deciding which decks to put in first, then
>> second, ...., then last.
>
> Also, which order the sort output trays are emptied and reassembled.
yep. I figured that one was obvious but, in hindsight, it wouldn't
be to those who never dealt cards.
>
> Good point. In the 1401 years, even in S/360 years, card sorters were
> an integral part of the processing. So undoubtedly manual operator
> intervention--which would not be that hard--
But it was hard. you could mess up a deck big time if you didn't
know how to order the input and the output card stacking.
>could handle a lot of
> issues.
>
> On a modern system, if the job stream had a sort step, these could be
> changed to select to pull off older years and later concatenate files
> to be in the desired order.
>
Huh?
/BAH
John Varela
<fri...@tx.rr.com> wrote:
> John Varela wrote:
> > On Mon, 4 Jan 2010 05:55:35 UTC, Charles Richmond
> > <fri...@tx.rr.com> wrote:
> >
> >> ISTM that the DF/W airport in the Dallas-Ft. Worth area still uses
> >> *several* IBM 360's to run their regional air controller radar.
> >> That would probably be impossible to verity due to security
> >> concerns, but I think they are still using IBM 360's for this purpose.
> >
> > 360s have never been used at airport towers or Approach Controls.
> > Configured as the multi-processor 9020 computer, 360s were used at
> > the Air Route Traffic Control Centers, of which there is one near
> > DFW. The 9020s were replaced in the mid-'80s with IBM 3083
> > mainframes. The system is written in JOVIAL and BAL. There is a
> > current project to replace the hardware and rewrite and upgrade the
> > software; I don't know its status but I don't predict early success
> > in replacing the software.
> >
>
> Okay, but that means that IBM 360's are (are were recently) in
> active use at D/FW Airport.
If you consider 25 years ago to be "recent", I guess so. For many of
us on this news group I suppose it is.
hanc...@bbs.cpcn.com
> > Good point. In the 1401 years, even in S/360 years, card sorters were
> > an integral part of the processing. So undoubtedly manual operator
> > intervention--which would not be that hard--
>
> But it was hard. you could mess up a deck big time if you didn't
> know how to order the input and the output card stacking.
Back when card sorters were common, a required basic skill of IT
operators was the ability to understand the card sorting process and
how to properly handle large volumes of cards. Sorters typically had
storage bins for each column mounted above the sorter. When the
sorter bins filled up, the operator would move cards to the storage
bins in a certain way to maintain the sort sequence. In addition,
when sorting by alpha fields, certain techniques were required. While
plain sorters did not have control panels, they did have some
selection/bypass buttons that the operators had to know how to set up
to achieve a desired outcome.
> > On a modern system, if the job stream had a sort step, these could be
> > changed to select to pull off older years and later concatenate files
> > to be in the desired order.
There is widely used utility, SYNCSORT, that can perform selection and
omission as well as other functions as part of a sort or copy. This
can be used to isolate out pre-2000 dates for proper sorting or
handling.
As an aside, the AS/400 (System/3 successor) had a full set of
utilities that did the functions of tab machines.
Charlie Gibbs
<a61d2c72-7cf8-4878...@s31g2000yqs.googlegroups.com>,
hanc...@bbs.cpcn.com (hancock4) writes:
Plus, when you're cramming data onto an 80-column card, you have to
save space as much as you can.
> How such systems handled decade rollover I don't know.
I do. One of my first assignments in my first programming job
in 1970 was to change the report programs to insert a '7' in
front of the single-digit year instead of a '6'. Since most
of our jobs were run on a month-by-month basis, that was enough
in most cases.
Applications that were more sensitive to dates had to bite the
bullet and store 2- or 4-digit years - mortgage companies dealt
with Y2K a _long_ time ago.
--
/~\ cgi...@kltpzyxm.invalid (Charlie Gibbs)
\ / I'm really at ac.dekanfrus if you read it the right way.
X Top-posted messages will probably be ignored. See RFC1855.
/ \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign!
Scott Lurndal
>On Jan 5, 1:18=A0pm, sc...@slp53.sl.home (Scott Lurndal) wrote:
>
>> There are still Burroughs medium systems in production. =A0
>
>I presume they're manufactured by Unisys?
>
>How many other machines of Burroughs or Univac heritage are still made
>by Unisys?
>
>(I thought Unisys was only into support and integration these days.
>Unfortunately, every time they're in the newspaper they seem to be
>contracting still more. Many of those news notes are posted here.
>
There are A-series (Burrough Large Systems) and Univac descendents;
the smaller are emulated now with intel processors. I think both
are known as flavors of the Clearpath family today.
The Medium systems still in production were built in the late 80's
by the Pasadena Burroughs/Unisys plant (now closed).
scott
Scott Lurndal
Yes, in production means being used. A city in california will be retiring
their V380 this month (basically a B4955 with updated microcode).
The V-series was discontinued circa 1993, IIRC.
scott
John Varela
I need to correct myself here. There were three versions of the IBM
9020. The 9020A and 9020D were central computers executing the air
traffic control software. These were replaced 25 years ago by the
3083s.
There were two display processors. (Why is a long story.) The less
busy Centers (of which I think Fort Worth was one) got Raytheon
equipment and the busier centers got 9020Es. All had Raytheon
controller consoles. Replacement of the display processors and the
consoles was completed before the end of the last millennium. So
there were 360s in use in Air Route Traffic Control Centers into the
late 1990s, 30 years after they were installed.
John Varela
<j.c.m...@verizon.net> wrote:
> "Charles Richmond" wrote:
> > Okay, but that means that IBM 360's are (are were recently) in active use
> > at D/FW Airport.
>
> I'm missing something here. John wrote that 360s have never been used at
> towers or approach facilities; how does that imply that 360s are (or were)
> in active use at the airport?
>
> Note that Ft. Worth ARTCC != DFW Airport even if they are neighbors. ZFW is
> located on land that used to be part of Greater Southwest Airport, just two
> miles or so south of KDFW. Open GoogleMaps and look at the satellite images
> just south of KDFW; just south of Airport Freeway there's an undeveloped
> section of land where you can still see the outline of a runway and taxiway.
> ZFW is to the west of this open land.
I was going to let that pass. Close enough.
I notice you're not posting from MITRE.org. Have you retired?
Charles Richmond
I thought that the IBM 360's were *still* in use *near* D/FW
airport. Are you saying that this is *not* so???
John Varela
wrote:
> Marketwatch just reported that Unisys made a lot of money last year in
> services - they seem to be re-inventing themselves like IBM. If you
> look at their website, it's hard to discover they still sell computers.
It was some years ago on this news group that someone pointed us to
Joe Morris
> Okay, Joe. I missed that one. Mia culpa.
First time *anyone* here has had a brain fart. You're forgiven <grin>
> The IBM 360's are used *near* the D/FW airport. What is important to me is
> that the IBM 360's *are* still in active use. These 360's are still
> actively executing IBM 360 instructions each day.
>
> It really seemed to me that the 360's had *something* to do with the
> regional radar, but maybe I was mistaken about that. I would call and ask
> someone, except I would probably end up on a terrorist "watch list". ;-)
John Varella is almost certainly more knowledgeable than anyone else in
a.f.c about the 9020 and related air traffic computers ...John, can you
provide (or point to) a short tutorial on the history of the computerized
ATC systems?
As for learning what's going on in the computers at an ARTCC: I have no idea
these days how much of the operating guts of an ARTCC center are on display
(pun not intended) at open houses. The only time I had a chance to actually
be in the same room as a 9020 was in the early 1980s when I was assisting in
an Air Force investigation of an aviation accident near the Tennessee/North
Carolina border: the lead investigator and I were at ZAT shortly after the
controller strike when filing an IFR flight plan was quite difficult, but
since I was at the Center they bypassed all of the usual hassles and put my
return flight plan directly into the system, and knowing my work with
computers gave me a VIP tour of the computer operations area. (I seriously
doubt that they would do that today, but managers weren't as paranoid back
then.)
Joe
Anne & Lynn Wheeler
some past posts mentioning various faa.gov web pages (some gone 404 and
have to go to the "way back" machine);
http://www.garlic.com/~lynn/2001h.html#15 IBM 9020 FAA/ATC Systems from 1960's
http://www.garlic.com/~lynn/2008h.html#6 The Return of Ada
http://www.garlic.com/~lynn/2009q.html#31 Check out Computer glitch to cause flight delays across U.S.
for instance:
Host and Oceanic Computer System Replacement Equipment for the National
Airspace System
http://web.archive.org/web/20011221011933/http://www.faa.gov/aua/ipt_prod/enroute/hocsrfct.htm
current faa history page (with several references):
http://www.faa.gov/about/history/
I wasn't ever involved in projects ... one of the "modernization
efforts" in the early 90s was being done by somebody we knew pretty
well. He sort of had a different job during the day ... but was spending
a lot of spare minutes programming in ADA for the modernization effort.
also since we were doing ha/cmp ... and some of the "modernization
efforts" involved rs/6000 triple replicated configurations ... we got to
see participate in some of the analysis. a basic premise was that there
would be lots of system infrastructure to mask all possible failures
from the application code. the application programmers could then take
the functional description and do a straight-forward implementation w/o
having to ever worry about contingencies. however, there turned out to
be some classes go failures/glitches that were at the
application/business level (not at the system level) ... and as a result
were getting lost.
--
40+yrs virtualization experience (since Jan68), online at home since Mar1970
Joe Morris
> "Joe Morris" wrote:
>> "Charles Richmond" wrote:
>> > Okay, but that means that IBM 360's are (are were recently) in active
>> > use
>> > at D/FW Airport.
>>
>> I'm missing something here. John wrote that 360s have never been used at
>> towers or approach facilities; how does that imply that 360s are (or
>> were)
>> in active use at the airport?
>>
>> Note that Ft. Worth ARTCC != DFW Airport even if they are neighbors. ZFW
>> is
>> located on land that used to be part of Greater Southwest Airport, just
>> two
>> miles or so south of KDFW. Open GoogleMaps and look at the satellite
>> images
>> just south of KDFW; just south of Airport Freeway there's an undeveloped
>> section of land where you can still see the outline of a runway and
>> taxiway.
>> ZFW is to the west of this open land.
>
> I was going to let that pass. Close enough.
>
> I notice you're not posting from MITRE.org.
The office no longer maintains a USENET feed (and hasn't for several years)
so all my USENET postings come from my personal account.
> Have you retired?
I could but I haven't. I'm having too much fun trying to figure out what
Microsoft is doing to us in Windows 7...and arguing with various gov'mt
organization help desks when their systems are misconfigured: in other
words, continuing challenges. Besides, I sometimes get to play with new
toys that haven't been officially released...some of which I would swear
could exist only in a science fiction movie.
Joe
Eric Chomko
One thing is certain MS wants to recoop all the money on W7 that they
didn't get from Vista and then some...
Gene Wirchenko
wrote:
>On Wed, 6 Jan 2010 04:48:49 UTC, Peter Flass <Peter...@Yahoo.com>
>wrote:
>
>> Marketwatch just reported that Unisys made a lot of money last year in
>> services - they seem to be re-inventing themselves like IBM. If you
>> look at their website, it's hard to discover they still sell computers.
>
>It was some years ago on this news group that someone pointed us to
>
> http://univac.com/
That might have been me. I even sent them an E-mail about it.
Universal Vacations. HMMPH!
Sincerely,
Gene Wirchenko
John Varela
<j.c.m...@verizon.net> wrote:
> John, can you
> provide (or point to) a short tutorial on the history of the computerized
> ATC systems?
Well, do recall that I retired over 14 years ago and have been
completely out of touch with what's been going on for all that time.
I don't even remember the acronyms.
There is this, which I posted to this group in 1999. It's part of a
long thread on Deja Google at http://preview.tinyurl.com/ybn6nrv
An advanced search of Google Groups on my name and "9020" in afc
will turn up some more, similar posts. I couldn't re-create any of
this because I've forgotten most of it.
------------------------------------------------------------
</group/alt.folklore.computers/msg/8dbbfaa23244d0b3?hl=en> | Show
original
</group/alt.folklore.computers/msg/8dbbfaa23244d0b3?hl=en&dmode=sour
ce>
| Report this message
</groups/abuse?hl=en&group=alt.folklore.computers&type=message&url=h
ttp%3A%2F%2Fgroups.google.com%2Fgroup%2Falt.folklore.computers%2Fmsg
%2F8dbbfaa23244d0b3&_done=%2Fgroup%2Falt.folklore.computers%2Fbrowse
_thread%2Fthread%2F71462944fe9b8fc8%2F8dbbfaa23244d0b3%3Fhl%3Den%26q
%3D9020%2Bgroup%253Aalt.folklore.computers%2Bauthor%253AVarela%26&>
| Find messages by this author
</groups/search?hl=en&enc_author=rJdhCxcAAAAnwl_RCP8Xqz3SeVtQyrO56nt
0FConyI-dV00YKxGasQ&scoring=d>
On Fri, 13 Aug 1999 23:28:18, Eric Smith wrote:
> From the perspective of a programmer, did the *9020* exhibit any
architectural
> differences from the 360?
The 9020A was all 360/50 parts assembled into a multiprocessor
architecture. Three or four CPUs were "Compute Elements", another 3
CPUs, with access only to MACH memory (later modified to access main
memory), served as "I/O Control Elements", and then there were (up
to
12?) "Storage Elements". The instruction set was standard 360
except
for the addition of instructions having to do with configuring the
multiprocessor system. An online system consisted of 3 of the 4
Compute Elements, 2 of the 3 I/O Control Elements, and n-1 Storage
Elements. The idea was instead of duplexing the very expensive
hardware, you could run with one each Compute, I/O Control, and
Storage Elements in an offline system and if an online element
failed
pre-empt the needed unit from the offline system. The offline
system
was a computer capable of running hardware diagnostics, processing
recorded data from the online system, and so forth.
Neither system ran OS/360. The online system ran a real-time OS
called the NAS Monitor (NAS = National Airspace System), which could
be executed from any of the Compute Elements at any time, and any
Computer Element could execute a program anywhere in the memory
configured in the online syustem. That is, if a Compute Element
needed something to do, it would execute the Dispatcher and perhaps
start executing a program that had started in another Compute
Element
and been interrupted. The offline system ran under NOSS, the NAS
Operational Support System.
We (MITRE) opposed the multiprocessor concept, favoring a fully
duplexed system, on the grounds that the field would need a spare
system in order to test out new software deliveries and ATC data
updates before taking them operational. We were right. The 9020As
and Ds were replaced by duplex 3083s about 15 years ago, and until
then the automated systems were taken down regularly for software
testing.[1] The online 3083 runs the old NAS Monitor, suitably
modified to reconfigure a duplex instead of a multiprocessor system.
The offline system runs VM, with a copy of the entire online
software
ready in the V=R portion of memory. I believe the FAA is now
replacing (or has replaced) the 3083s.
IBM proposed models B and C but none were ever built and I don't
recall what they were. The 9020D used 360/65 CPUs and Storage but
retained the 360/50 I/O Control Elements. The 9020E was similar to
the 9020D, was used in the display system, and had special
instructions for processing display data and also special display
refresh memory units.
[1] Actually, the field kept taking the online system down even
after
they had a duplex system and no one in headquarters could figure out
why they did that. Eventually someone figured out that if they took
the offline system for testing they would have to run the online
system without a backup[2], and if the online system failed the
facility would be charged with a system failure. But if they
SCHEDULED taking the whole system down, then they weren't charged
with
a failure. I don't know how that was ever resolved and for all I
know
they're still taking the systems down every night.
[2] Well, actually, there is a backup that is both hardware and
software diverse from the main system, but it's less capable, as it
processes only radar, not flight plans. It's what they used when
they
unnecessarily took the main system down for maintenance, and IT does
NOT have a backup. Ah! The ways of bureaucracy!
--
John Varela
to e-mail, remove - between mind and spring
------------------------------------------------------------
Hmm. Eleven years ago I said they were replacing the 3083s. I don't
know if they did or didn't. I could probably find out, if anyone
cares. Since you're still at MITRE it might be easier for you, since
at least you know who's still there.
Charles Richmond
Like the PC users say on their commercial:
"I'm a POS (piece of sh*t!) and Windows 7 is *my* fault!!!"
Well, or something like that... :-)
H Vlems
> >> Charles Richmond <friz...@tx.rr.com> wrote:
> >>> cjt wrote:
> >>>> Charles Richmond wrote:
>
> >>>> I have little doubt there are still DEC PDP-11s in daily use, and
> >>>> probably even earlier big machines (e.g. Bendix, Burroughs, IBM, and
> >>>> Sperry machines -- maybe even some CDCs).
>
> >> The Reuters "Monitor" is still operational. That is written in
> >> pdp11 assembly and machine code, very intimate to the hardware.
> >> The "Monitor" _is_ the OS. The physical PDP11s were replaced with
> >> emulators, then on Sun hardware, in the mid 1990s. But ISTR they
> >> are back on hardware-PDP11 implementations now.
>
> >Yes, but... those darn PDP-11 instructions *are* *still* *being*
> >*executed*!!! Emulating is still executing... That's the title of
> >this thread: Oldest Instruction Set still in daily use.
>
> There are still Burroughs medium systems in production. Parts of
> the instruction set date back to the Electrodata 200 systems from
> the late 50's. Applications written in the early 60's will still
> run on the hardware (which was designed in the 80's).
>
> scott
The Burroughs B5000 instruction set is used by the A-series (now
Unisys) and the emulated systems they sell today.
The B5000 is from 1961.
Hans
(see below)
1e5ba957-559d-4c14...@34g2000yqp.googlegroups.com, "H Vlems"
<hvl...@freenet.de> wrote:
> The Burroughs B5000 instruction set is used by the A-series (now
> Unisys) and the emulated systems they sell today.
> The B5000 is from 1961.
Surely the A-series run a development of the significantly different B6500
instruction set? That dates from ca. 1965 and was influenced by the
Whetstone virtual machine used in KDF9's Whetstone Algol (which I was
running today).
--
Bill Findlay
<surname><forename> chez blueyonder.co.uk
Mike Roach
<fri...@tx.rr.com> wrote:
>There are probably even Z80's in electronic typewriters that are
>still seeing service typing mailing labels and such.
I once had a job making typewriter pads. Recently, while talking
about that old job, someone asked what a typewriter pad is. Some
whippersnapper piped up with "What's a typewriter?"
--
Stealing a rhinoceros should not be attempted lightly.
Louis Krupp
> On 08/01/2010 14:46, in article
> 1e5ba957-559d-4c14...@34g2000yqp.googlegroups.com, "H Vlems"
> <hvl...@freenet.de> wrote:
>
>> The Burroughs B5000 instruction set is used by the A-series (now
>> Unisys) and the emulated systems they sell today.
>> The B5000 is from 1961.
>
> Surely the A-series run a development of the significantly different B6500
> instruction set? That dates from ca. 1965 and was influenced by the
> Whetstone virtual machine used in KDF9's Whetstone Algol (which I was
> running today).
>
I hadn't heard of Whetstone Algol. I looked it up, and sure enough, the
virtual machine has a DISPLAY vector, just like the B6500 and its
successors have DISPLAY registers.
The B5500 didn't have those; variables were addressed relative to a
stack location (procedure arguments or local variables, there are some
things I don't remember), or to the base of the Program Reference Table
(global variables).
Louis
Charles Richmond
> TIME Magazine Person of the Year for 2006 Charles Richmond
> <fri...@tx.rr.com> wrote:
>
>> There are probably even Z80's in electronic typewriters that are
>> still seeing service typing mailing labels and such.
>
> I once had a job making typewriter pads. Recently, while talking
> about that old job, someone asked what a typewriter pad is. Some
> whippersnapper piped up with "What's a typewriter?"
That same person would probably *not* know what to do with a dial
telephone. :-)
H Vlems
> On 08/01/2010 14:46, in article
The B6500 was a development from the B5500, which was basically a
rebadged B5000.
Instructions were added, so a B6700 (or a B7900) can run B5000 code.
Louis Krupp
I've been on a couple of B5500 to B6700 conversions, and I don't recall
B5500 object code being executable on a B6700. Too many things were
different.
There was some source code compatibility, but porting COBOL (for
example) was still an effort because the B5500 used eight 6-bit
characters per word, and the B6700 used six 8-bit bytes, at least
internally. Towards the end of the lifetime of the B5500, another ALGOL
compiler, XALGOL, was introduced; as I recall, it was closer to B6700
Extended ALGOL.
What *was* compatible was the integer and at least single-precision
floating point data format. Data didn't have to be converted, and that
was nice.
Louis
(see below)
2264a634-a576-4893...@35g2000yqa.googlegroups.com, "H Vlems"
<hvl...@freenet.de> wrote:
> The B6500 was a development from the B5500, which was basically a
> rebadged B5000.
> Instructions were added, so a B6700 (or a B7900) can run B5000 code.
Really?
According to online Burroughs manuals, the B5[5|0]00 had fixed-length 12-bit
instructions with no lexical-level "display" and no tag bits; but the B6500
had (like the Whetstone VM) variable-length instructions composed of 8-bit
syllables using display registers for addressing, and had three tag bits per
word.
That amounts to a great deal more than "instructions were added". The B6500
was a new, though rather similar, architecture. Had the B6700 a B5[5|0]00
emulation mode to run 12-bit code? I do not see that in the online manual.
Esra Sdrawkcab
<cgi...@kltpzyxm.invalid> wrote:
> In article <isw-8D4A52.22021703012010@[216.168.3.50]>, i...@witzend.com
> (isw) writes:
>
>> 8080 and Z-80 processors do not have the same instruction set as the
>> 8086 -- not even close.
>>
>> I was once given the task of "replacing" the 8080 in our product with
>> the then brand-new 8086, after some liar from iNtel told our marketing
>> guy that the "programs" were portable.
>>
>> Turns out the iNtel guy was sort-of-but-not-very right *provided* you
>> had written your 8080 programs in PL/M, and were willing to port them
>> to PLM/86. If they were in assembler (and ours were) you could forget
>> about it.
>
> You mean the whole world has suffered all that backward-compatibility
> crap for nothing? 1/2 :-)
>
> Seriously, though, I heard about a program that was supposed to be
> able to translate 8080 assembly language programs to 8086. I never
> saw it myself. Did you have an opportunity to try it? If so, what
> was the result?
>
footnote 24 from the wikipedia article on the Z80:
^ Surprisingly, only quite superficial similarities (such as the word MOV,
or the letter X, for extended register) exists between the 8080 and 8086
assembly languages, despite the fact that 8080 programs can be compiled
into x86 code (using a special assembler) (Scanlon, Leo J. (1988).
8086/8088/80286 assembly language. Brady Books. p. 12. ISBN
9780132469197. "The 8086 is software-compatible with the 8080 at the
assembly-language level." — Nelson, Ross P. (1988).
--
Nuns! Reverse!
Chris Burrows
news:0351cfc1$0$1268$c3e...@news.astraweb.com...
>
> From the London Science Museum website:
>
> "Another fascinating large object in the gallery is the 1956 Ferranti
> Pegasus - the oldest working computer in the world and one that is
> switched on and run regularly to the delight of visitors. "
>
> http://www.sciencemuseum.org.uk/visitmuseum/galleries/computing.aspx
>
> The "Baby" dates back to 1948 but the specimen in Manchester is a replica.
> Perhaps the Pegasus is an original which would explain the "oldest" claim.
>
There is soon be a 'new' contender for the 'oldest original functioning
electronic stored program computer in the world' when the restoration of the
Harwell / WITCH computer (which first ran in 1951) is complete:
'Tony Frazer, leader of the WITCH restoration team said: "The WITCH arrived
in remarkably good condition after more than three decades of storage. We've
assembled the frame and it now looks just as it did in its heyday in the
1950s and 1960s. Our first task is to see what we can do with the power
supply - we dare not just switch things on as time will have taken a toll on
the chemistry and physics of the unit. Then we will be moving onto the
thousands of wires and switches and the hundreds of Dekatron tubes. Although
we have circuit diagrams, we can already identify wiring modifications, so
this is going to require a lot of ingenuity."'
http://www.tnmoc.org/36/section.aspx/103
--
Chris Burrows
CFB Software
http://www.cfbsoftware.com
Peter Flass
>
> There is soon be a 'new' contender for the 'oldest original functioning
> electronic stored program computer in the world' when the restoration of the
> Harwell / WITCH computer (which first ran in 1951) is complete:
So what happened to the Brits anyway? It seems like way back then they
were even with or ahead of the US in computing, but then they kind of
dropped off the map. I think INMOS was the last gasp.
Chris Burrows
news:hio9uv$qbo$6...@news.eternal-september.org...
>
> So what happened to the Brits anyway? It seems like way back then they
> were even with or ahead of the US in computing, but then they kind of
> dropped off the map. I think INMOS was the last gasp.
>
Whose map? Last time I looked ARM wasn't doing too badly ;-)
Quadibloc
> "Peter Flass" <Peter_Fl...@Yahoo.com> wrote in message
> news:hio9uv$qbo$6...@news.eternal-september.org...
Still; there was the Deuce, there was the Atlas, there was the Zebra,
there was the Pegasus, there was the KDF 9, there was the ICL 1900...
but in the 1970s, for example, British mainframe and mini
architectures didn't make enough of a splash to be heard about on this
side of the pond. There was still the odd - very odd - French or
Swedish computer.
John Savard
Andrew Swallow
word based computers well after 32 bit ASCII character computers
were normal. So ICL took too long to change over and died when
the British Government started buying the cheapest. Microcomputers
did not help.
Andrew Swallow
Peter Flass
As an architecture. Aren't ARM processors now all, or mostly,
manufactured by licencees?
jmfbahciv
>
>
It's called socialism.
/BAH
Joe Pfeiffer
I think they were the first fabless IC house: to the best of my
knowledge, all they've ever sold has been licenses.
--
As we enjoy great advantages from the inventions of others, we should
be glad of an opportunity to serve others by any invention of ours;
and this we should do freely and generously. (Benjamin Franklin)
James Harris
Good question. A familiar answer. At least from a simplistic point of
view the British are poor at exploiting good ideas. This has been
going on for all my life and doubtless well before it.
Why? Well, one theory is that when chances came to make a new life in
the new world many of those with an enterprising spirit put it to use
and moved to America. Land of opportunity and all that. While it
didn't take everyone I suspect that took a lot of the entrepreneurial
spirit out of the UK and other parts of Europe.
We are left with a more negative average spirit. As a Brit I've been
surpised in the past at how positive Americans are. I'm not
complaining - I think we've been left with a more fascinating sense of
humour, for example - but if I wanted to market an idea of mine I'd
have a preference for working with Americans rather than Brits.
James
Stan Barr
The last truly British computer was the Archimedes, circa 1987.
That, of course, was where the ARM (Acorn RISC Machine) processor
first appeared. I see someone has got RISCOS, the Archimedes OS,
running on an ARM powered BeagleBoard - _that_ would be an interesting
project.
Then there was Viper, about 1990 - what ever happened to that?
Must google...
--
Cheers,
Stan Barr plan.b .at. dsl .dot. pipex .dot. com
The future was never like this!
greymaus
Bigger market, if you have an idea that will sell to 5% of the population,
that's 10 million in US. All speaking the same language.
Readier access to venture capital. (If you have a good idea in
the UK, or here (.ie), you have to get funds from government
quangos who, sooner or later, refine down to some political idiots)
+ There are whole areas of the UK (and here) that are not interested
in anything happening more that a mile away.
Germany and Japan have better technical education.
--
greymaus
.
.
...
Walter Bushell
greymaus <grey...@mail.com> wrote:
> Bigger market, if you have an idea that will sell to 5% of the population,
> that's 10 million in US. All speaking the same language.
More than the UK. I heard a sermon in Scots dialect, didna understand a
word.
>
> Readier access to venture capital. (If you have a good idea in
> the UK, or here (.ie), you have to get funds from government
> quangos who, sooner or later, refine down to some political idiots)
>
> + There are whole areas of the UK (and here) that are not interested
> in anything happening more that a mile away.
>
> Germany and Japan have better technical education.
>
> --
> greymaus
--
A computer without Microsoft is like a chocolate cake without mustard.
Ahem A Rivet's Shot
Stan Barr <pla...@dsl.pipex.com> wrote:
> The last truly British computer was the Archimedes, circa 1987.
> That, of course, was where the ARM (Acorn RISC Machine) processor
> first appeared.
A truly impressive machine it was too.
--
Steve O'Hara-Smith | Directable Mirror Arrays
C:>WIN | A better way to focus the sun
The computer obeys and wins. | licences available see
You lose and Bill collects. | http://www.sohara.org/
Charles Richmond
ISTM that in the late 40's (1948???) there was a Computer
Symposium at Cold Spring Harbor in New York. The Brits attended
and got the benefit of the work done on the Eniac to add to their
own ideas. This helped boost early computer work done in Britain.
Charles Richmond
> On 15 Jan 2010 16:30:18 GMT
> Stan Barr <pla...@dsl.pipex.com> wrote:
>
>> The last truly British computer was the Archimedes, circa 1987.
>> That, of course, was where the ARM (Acorn RISC Machine) processor
>> first appeared.
>
> A truly impressive machine it was too.
>
Did Intel *not* license the ARM architecture and produce chipsets
for embedded use??? ISTM that my LinkSys (Cisco) router has an
ARM chipset in it.
Charles Richmond
In China, it's call Communism, but they still seem to make money
out of all this stuff...