The IBM System/360 Revolution
HTML-код
- Опубликовано: 10 мар 2008
- [Recorded April 7, 2004]
Computer pioneers and National Medal of Technology awardees Erich Bloch, Fred Brooks, Jr. and Bob Evans with current IBM technology chief Nick Donofrio discuss the extraordinary IBM System/360 project.
IBM launched the System/360 on April 7, 1964. Many consider it the biggest business gamble of all time. At the height of IBM's success, Thomas J. Watson, Jr. bet the company's future on a new compatible family of computer systems that would help revolutionize modern organizations. This lecture presents a behind-the-scenes view of the tough decisions made by some of the people who made them, and discusses how the System/360 helped transform the government, science and commercial landscape. 
Наука
I started working for IBM in 1966, the System 360 had just come out, I serviced many models of this system for many years and the System 390 as well. I worked for IBM for 30 years!
👏👏👏👏👏👏👏👏👏👏👍
These machines pioneered 90% of the features we take for granted in 32/64 bit microprocessors today.
It seems to me, the ISA for the 360, was maybe the first major ISA that was used in a lot of computers. Much like the x86 is now.
I was an MVS System Programmer on a 3090, I still remember starting CICS from punch cards on our 4341
The first computer I worked with in 1975 was the Interdata 7/32. It was a micro-coded design using a 16-bit ALU, early TTL, ferrite core memory, and a very similar instruction set to the 360. As a young engineer I spent many hours pouring over the assembly language in the operating system. Back then they provided the source code. Learning the data structures (FCBs/DCBs) taught me how to abstract computer problems into rather elegant solutions. Good to hear from the pioneers on these designs.
The Computer History Museum will be remembered for posting videos with crappy audio. They can restore an IBM 704, but don't know how to fix an audio track. That's a shame.
I learned programming via OJT on an 360 Model 30 using assembler. This was October 1969 at Shillito's Dept store in Cincinnati. We ran DOS/360 on a system with 32K of ram, 2 2311 disk drives, 2 2401 tape drives, a 2540 reader punch and a 1404 (NOT 1403) printer.
Strange that Fred Brooks said that IBM missed the mid-range computers revolution. Maybe initially they did, but in Western Europe in the 1990s the AS/400 was very successful to the point of eliminating most of the competition.
Joined IBM as an SE in fall of '63. Worked with DOS rel. 3, where the only thing that worked was the sample program. Had to work with marketing reps who wanted to sell S/360 Mod 30's without a console as a fast 1401 replacement. Replaced the last 305 RAMAC in the geography with a Model 30. A large challenge was being sent to an account that had been sold a Mod 30 running TOS with a single 9-track tape drive (ponder that one for a while). And then there were the sales calls trying to convince the customer why they needed to purchase an operating system. Damn, those were fun times.
One thing that's "hot stuff" today that IBM was doing for /decades/ is virtualisation.
My father did just that working as a guy whose job was to sell the idea of computers. He was running a big computer department when the IBM 360 came out buying at least 3 of them . He retired just as the PC was gaining ground as a major tool.I wish he could have seen this.
I love this statement that Fred Brooks makes regarding JCL, OS/360 job control language is the worst programming language ever designed anywhere by anybody for any purposeand it was done under my management.
ok, but if people do not suffer under a horrible language, they will not be compelled to create or appreciate something better. In that sense, he was a great man/person - like Moses who led his people into the desert, but left it up to others to find the way out.
It was fun and challenging work and I'm proud to have been part of it.
But boring to listen to.
IBM '68-'98
I don't think many people realise that programs originally written in 1964 for the IBM/360 would almost certainly still work today on the latest IBM Z Architecture machines without even re-assembly in most cases.
Contrast that with the Personal Computers scenario - where almost any small OS change can result in an application not working on the latest Operating System.
en.wikipedia.org/wiki/Z%2FArchitecture
NOT TRUE, DIFFRENT MACHINES!!! 360 HAD NO VM!
CONVERSION FROM ONE IBM OS TO ANOTHER
WAS A LOT OF WORK!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
SuperU2tube - That phenomenon is entirely due to the philosophy of the people in charge. Usually, when systems are not upward compatible, it is only because the people in charge don’t want to be bothered with it. An example is the fact that OS/2 1.x programs ran under Microsoft NT for awhile, until at one NT release they didn’t anymore. There was no reason for it except they just didn’t want to keep the feature. It caused me a lot of trouble.
I’m glad I’m retired from the computer business. There is too much of that crap.
@ GH1618
You said
“That phenomenon is entirely due to the philosophy of the people in charge . Usually , when systems are not upward compatible , it is only because the people in charge don't want to be bothered with it . An example is the fact that OS / 2 1.x programs ran under Microsoft NT for awhile , until at one NT release they didn't anymore . There was no reason for it except they just didn't want to keep the feature . “
That’s because in your case “they” = Microsoft.
OS/2 was born of the IBM/Microsoft marriage of the early 1980s. OS/2 was “conceived” to provide IBM with a proprietary operating system when their PS/2 class of personal computers were introduced in 1987.
Early on, Microsoft and IBM programmers jointly worked on its development. Soon thereafter, Microsoft realized that developing and evolving Microsoft Windows was critical to their long-term success and they broke up the marriage, leaving IBM with sole custody of OS/2.
OS/2 and Microsoft Windows (with windows NT “under the hood“) were direct competitors and so, despite initial compatibility between OS/2 and Microsoft products, it was in Microsoft’s commercial interest to end that compatibility, and sooner, rather than later.
Was the 360 ISA the first widespread ISA, used in a lot of machines, similar to x86 now?
Every IBM announcement since 1966 - IBM makes 360 faster
"The mainframe is here to stay" no longer applies.
I've a little older. Sysprog in the '70s. Went to work for Dr. Amdahl in '78. :)
How can this be produced without Gene Amdahl?
WRONG: 360/370'S used the very earliest "micro code" punched
into plastic ibm cards, to change instructions sets.
Not just "read only memory" as the speaker incorrectly stated.
I must be a terrible geek for liking this stuff too.
It's extremely interesting from a historical view describing IBM history between 1955-1965. Soon all involved will be dead.
Russ Theisen helped develop and test this IBM 360-20 computer.
I took a FORTRAN programming course at Penn State University in 1965. There was a key punch room where students had access to 10 or 15 keypunch machines for recording their programming statements into 80 column cards. In order to review your work you would put your deck of cards into the input hopper of an IBM 407 accounting machine printer. When you’d push the start button on the 407 it would read the cards, one at a time, and print the contents of your cards, a line at a time.
In the spring of 1967 when I was taking a System/360 assembly language programming course, the 407 printer in the key punch room had been replaced by a System/360 Model 20. Now, when you wanted to get a listing of your card deck you took three special cards that the computer center had pre-prepared, put them at the front of your card deck, placed it in a card reader attached to the Model 20, and pressed a button on the model 20 (probably labeled “IPL”, but I don’t really remember that.) That produced a listing of your card deck, just like the 407 printer used to do. As far as we students were concerned, it was just a new kind of printer.
One day, when I went over to the Model 20 to list my card deck, on the front of it an attached notice read, “No programming of this machine without permission from the Computation Center Director.” I remember thinking at the time, “Program a printer ???!”
By July of that year I had graduated, had joined IBM, and found myself in a ten-week class where I learned, among other things, how to program on a System/360 Model 20, as well as how to program the 407 accounting machine printer by wiring its plug-board control panel.
I retired from IBM in July 1997. It was a very busy and interesting and rewarding 30 years.
I first wondered, who is Russ Theisen, then I looked at your name :) Development of this computer was probably a major project? IBM used to do things in a big way, I worked there for awhile.
Any chance this can be re-uploaded. There is a lot of static in parts (like at 55). Ruins the great detail in this story.
Aw, man, I wanted to see the cake.
//SYSIN DD *
Born same day as I was!
My left ear thanks you
I don’t see it as a revolution. It was a highly successful series of computers, that’s all.
Yeah you're right it's maybe a bit of an exaggeration. But still, it's surprising how many features this series had that appear in modern processors.
right ear so lonely!
I wish I still had Green Card as a memento.
BIG LIE: nonsense: 360/370/z system are not really compatibe.
dos, os mft, os mvt, mvs, z os are all diffrent operating
systems that require program rewrites.
It's amazing something so important can be made so boring.
BROOKS LIE: "IC COMPUTER COULD NOT BE MADE BY COMPETITOR" WRONG RCA SPECTRA 70 SYSTEMS WERE
IC BUILT AS 360 COMPATIBLES AND DELICERED.
At BMCC(college) we had a spectra 70 and ibm 360/30
if the 30 was busy we ran the programs on the spectra 70.
It sounds like Brooks has never programed a 360/370 himself.
System 36 for a couple of years & AS/400 for 25 years, old unemployed bum for last 5 years.
Youre very old misCooper
I tell the ignorant kids that THINK they are programmers the very same thing all the time.
i farted
WHY NOBODY BUYS IBM:
Today the solution is to use linux or microsoft server programs on any cheap chinese mfg computer(dell, compacq, hp) costing $3000
OR:
IBM SOLUTION: (chinese computers, MFG BY
LENOVO) costing $30,000( $3,000 for the cpmputer
and $27,000 for the IBM label).
That is why nobody buys IBM. Very few people are that dumb.
THE ORIGINAL 360 IBM ENDICOTT NY PLANT IS NOW JUST A
PARKING LOT AND ABANDONED BUILDINGS IBM DOES ALMOST
NO MFG AND MILKS A VERY OLD TECH(360). IBM SERVERS USE
INTEL CHIPS, INTEL TECHNOLOGY! JUST LIKE DELL AND EVERYONE ELSE. I WOULD HARDLY CALL IBM A HIGH TECH COMPANY OR
INOVATIVE!!! JUST ALSO RAN.