?

Hello Vince,

I always thought that even Pentium CPU was much faster than 370 m/fs. Well, i agree, m/f has a lot of disks, but you can attach 4 or more HDDs to your PC also.
For example, i have two SSD(for sys and data), one NVMe(for VM) and one HDD 7200rpm(for archives) attached to my system.

Those terminals you mentioned used modems, right? So when we dial the DOS/VS we emulate modem connection?
They usually was 1052 like printer-keyboards or 3270 displays?

On Sun, Jun 4, 2023 at 01:49 AM, Vince Coen wrote:

Based on your question#s :

Evening and night shift (of 8 hours) was because the early mainframes and that included as far as I was involved in IBM 1401, 360 and early 370s, ICL 1900's (1500 as well but they were totally Magnetic tape based). That said so was 1401 and many 360/370 as hard disks was expensive when they became available and very small in capacity i.,e from 80 - 300 MB or there about.

For the later system above i.e. 360/370 that had terminal user running CICS they mostly had to have other tasks to support them that where all Batch processes and these run during the quiet periods such as for one hour lunch time (may be) and after 18:00 to 08:30 plus all of the weekends.? Now programmers also submitted compile and basic testing and this was done during the day but mostly over night and the w/e and some programmers came in during the W/E and in my case to use the card punch to produce a source deck to input to the system for a compile but usually for fixes to the code that had to be then inserted into the original source decks. One reason for having to do it yourself was the punch room people could be very busy punching data card packs for input to the system so did not have a lot of time for doing programs.
Here the programmers wrote the source code on to forms that had the layout of a Cobol program with around 25 lines per page.

The rest of the evening and nights was processing the normal business workloads that was created during the day such as stock updates, payroll, other accounting processes among many others.
While the m/f was based on a slow CPU when using hard drives were a lot quicker than say a personal computer that was available from around 1977-8 (floppy disk based mostly but the one's with the early Winchester (early hard drives and 8 inches wide and a lot longer and were heavy) drives was not a lot faster.

The benefit of mainframes has always been there much faster processing with modern hard drives because that can access multiple drives where each had there own data bus -. A PC has only one and that is used for all data flows so compared very slow.

Remember a m/f is running many programs at once - (OK, may be time slicing) but a great deal more than the average micro based computer.

I have two desktops and both have multi core processors where on one is is 8 core in one chip (AMD) and the other, 2 XEON CPUs with 4 cores each and with the right O/S they do bash work though and not quicker than the older m/f's as mentioned above.

Bye the bye one of mine also had mag tapes (and cartridges) and they really slowed the system down so after a couple of years I replaced the lot with WD Black hard drives that are hot swappable.

Small point I came into computing around 1963 and obtained my first Micros around 1978 with the faster one;s around 1982 (Cromemco was one brand) but I was running the first company in Europe dealing in micros from 1975-6 selling all over the world - outside the USA and even then sometimes in despite some products (software including DOS, CPM, MPM, *nix, books, magazines and hardware came from the USA . ? ?

My first job in computing in 1963 was for 6 - 9 months as an operator on a 40 hour week earning 4 pounds per hour but if I worked the weekends on a 12 hour shift the rate doubled and I could take the rest of the next week off.
Some time the w/e was light so I would do some programming such as automating the systems IPL so it only needed one switch press and along with a small card desk and the 1401 system as up and running - otherwise you had to enter a lot of machine code then execute it to do the same.? That was shown to the boss who invited me to be a programmer and a considerable higher salary :)

I still did some operating though and got paid for it at my new rate :)

Oh boy - those where they days = when men where men and computers . . . .


On 03/06/2023 21:19, Andre wrote:

Hi Steve!

This is so cool! Do you have any more stories about working on 370? What type of programs usually was batched through the day? Business stuff? Scientific?
I know that ENIAC was used to compute trajectories for artillery. UNIVAC helped with census and also tried to forecast weather and election results.
After that a lot of Business companies adopted computers for calculating employee wages and stock lists.

This is ok, but i want to know how much time would it take to compute one of this tasks, how often they used to appear? Why night shifts existed?

Best wishes,
Andre

On Sat, Jun 3, 2023 at 10:15 PM, Steve Shepherd wrote:

As an operator I learnt assembler on a 370/138 on dos/vs after reading Sharon K Tuggle’s Assembler Language Programming book. On night shifts I would squeeze in a few programs in BG.?

To drive it all home I wrote a version of Star Trek after playing a basic version on a Rank Xerox Sigma 9 on a previous contract.

Anyway, the program output was on the system console and caused no end of problems the next day when the (real) programmers couldn’t find their error messages because of a full console log.

the boss came to me the next day and I thought I was for the chop, but instead they offered me the vacant Systems Programmer job, which of course I readily accepted!

?