IBMWatson autobiographythoughts on 764
IBMWatson autobiographythoughts on 770
This is true only if you can store the table in core. And somebody had to key in that table somehow...
Yes, that was true. Certainly any applications could run. But the CPUs were far slower in those days and performance suffered when the wrong application type was run.
IBMWatson autobiographythoughts on 766
Eric Sosman Well, your formula of course is right. But I am surprised that the incremental cost...
Remember that early PCs did not have native-mode floating point processing and one could buy a "co-processor" chip to plug into their PC to improve performance (without the chip the work was done by software which was slower). IIRC, the 486 and later lines had that capability built in.
Also keep in mind there was a second division between large and small machines and they were not compatible. It was wasteful in a small machine to provide for large addressing operands when that space would never get used. In those days, one could buy a machine with as little as 1400 characters--(1.4K!), and 4K machines were common. The largest 1401 (the most common computer) went only up to 16,000 characters (a character being 6 bits.)
So, you had:
small machines for business (character oriented, decimal arithmetic). big machines for business. small machines for science (word oriented, binary arithmetic). large machines for science.
For example, IBM had a 1401 and also a 1410 for big work. While the two were similar they were not truly compatible. The 1410 could run 1401 work with a different switch setting, but the 1401 could NOT run 1410 work even if it was small enough to fit. There were also more features on the 1410 than 1401.
IBMWatson autobiographythoughts on 767
John R. Levine basically leases were somewhat like cellphone billing ... basic plan and possibly a lot for overages ... based on cpu meter. the meter ran...
Yes, the S-360 could do both science-engineering as well as business equally well, and a company could use one machine instead of two.
But more importantly, if a company grew and it needed a bigger computer, it need not rewrite all of its programs. My employer has a few programs dating from 1967--obviously written for very different machines in a different era--but they are still running. Obviously they run a heck of a lot faster today.
Further, IBM maintained compatibility with its older (pre-S-360) machines. Theoretically, someone could have a program written in 1955 still running without change to this very day.
Oh yes. The 1401 became the "bargain basement" computer and still quietly marketed for a few years after S-360 came out. The boxes still had a little bit of life in them.
My guess is that the physical box was good for about eight years of service. The actual lifespan was about 5 years between 1401 deployment and S-360 deployment, then it was good for another 3 years. (The actual lifespan was dependent on how much downtown a user was willing to withstand before discarded the machine. I know of a user who got a 1401 late in the game, kept it a few years, then dumped it when reliability became a problem.)
1401 programs lingered for YEARS under emulation after S-360 came out.
Also keep in mind that electro-mechanical tabulating machines remained in widespread use well into the 1980s.
IBMWatson autobiographythoughts on 769
Charlie Gibbs Easy and hard are relative and depend on the overall context. For example, there may be math formula that's relatively 'easy' to calculate using tables. But this formula might be...
Actually Univac had two totally incompatible lines. They had a 1100 line which I think was word-oriented, and their 9000 line was character oriented (very similar to 360.) The 9000 line was acquired from RCA's Spectra series.
I don't know what happened after Burroughs merged with Univac with their respective lines of computers, or what mainframes, if any, those companies offer today.
It was a little of both. Although criticized at the time, S-360 certainly pushed the technology envelope. The microcode was a big help. The lessons of Stretch helped too.
No. I don't believe there was any disagreement over word vs. character.
The disagreements were about the new architecture vs. existing lines.
I think some relaxation of IBM's 'stiff' image was in order, but I wonder if it went too far. Sure, it was stupid for customer service technicians ("CE") to wear white shirts and ties while ripping apart printers. Today, people behind the scenes don't need stiff blue suits. But those people doing demos or calling on customers, IMHO, should still be reasonably conservatively dressed, such as a necktie for men (except maybe in the hottest weather).
FWIW, in early Univac days, before air conditioning, the lab boys ran around in their underwear and had to be told to stay out of the front office. IBM had air conditioning and their boys had to wear an old fashioned stiff paper collar in those days.
IBMWatson autobiographythoughts on 765
The computer might do the job X times faster than the sorter, but if a computer minute cost more than X sorter minutes...
IBM's Early Computers IBM's S-360 by Pugh et al Father Son & Co by Tom Watson Jr. Building IBM by Pugh
These books are heavy into the research and technical end, including the challenges faced when designing and manufacturing early transistors and chips. Very good from that perspective.