Announcement

Collapse
No announcement yet.

It's not back - it was never gone (COBOL that is)

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • It's not back - it was never gone (COBOL that is)

    I rarely post links but this was just too good:


    https://www.wealthsimple.com/en-ca/m...ols-your-money
    Michael Mattias
    Tal Systems Inc. (retired)
    Racine WI USA
    [email protected]
    http://www.talsystems.com

  • #2
    Originally posted by Michael Mattias View Post
    I rarely post links but this was just too good:


    https://www.wealthsimple.com/en-ca/m...ols-your-money
    When I read this bit: "They’re much more excited by buzzier new fields, like Toronto’s booming artificial-intelligence scene. They’re learning fresh new coding languages."

    I immediately thought of a quote I saw recently:

    Half of the time when companies say they need "AI", what they really need is a SELECT clause with a GROUP BY.

    Comment


    • #3
      Thank you MCM

      I read every line of your link.

      In general, I agreed with most of it.

      Curiously it said that COBOL was not finished until 1969 but I learned and used it in 1968! Before that I had written in PLAN - an assembler level program not too different from IBM assembler. We knew all about Grace Hopper. My memory is that the US Govt mandated that all programs written for it had to be in Cobol.

      We assembler level programmers initially said 'Cobol will never catch on!'

      But I knew my 1968 COBOL frontwards and backwards - I could remember things in those days!

      My Cobol was supplied by ICL (or was it still ICT in those days, I forget) - a British Computer Company eventually taken over by Fujitsu (?). I was ranting on about Cobol on this PB website some years ago and some guy announced that he had written the ICL Cobol manual all those years ago. I wonder if he is still lurking here.

      The first program I wrote was to print the rate demands for Christchurch (the city of the earthquake). The 'rate demands' are NZish for local govt (?city hall, whatever) taxes-on-property bills. My program was certainly the first in NZ to produce rate demands - maybe the first in the world??? The Mayor and dignitaries came down to watch the first runs and my picture was in the paper standing behind the printer and the mayor! When we restarted it the next night the program bombed - but do not tell anyone.

      I actually sat through most of the rate demand printing - I am not sure why (it is 52 years ago!). I remember my late Mother asking me what I actually did as the rate demands were being printed and was amazed to find that we played chess!!! (and loaded more paper as required)

      Interestingly, NZ had changed from old currency (pounds, shillings and pence) to dollars and cents in 1967. The conversion was not quite straight forward. 3p was about 2 cents so there was this approximating process. Computers were absolutely essential to make the conversion from the old property records to the new ones - there is quite a lot of financial data for each property.

      It was my knowledge of Cobol which encouraged me after 7 years to try another non-computing career for 3.5 years. I decided that what I knew would not last me my lifetime. At the end of that time, I missed computers and the hard logic - and so I moved back into the computing world - actually that is another good story.

      In the 'good old days' as the article said, you wrote code on a coding sheet and a punchgirl punched it up on cards (programmers were men, punchgirls were women in NZ in those old sexist days). Then the program got compiled that night. You usually scheduled a test run if the compile was successful. If you made the slightest coding mistake, you had to wait another 24 hours!!!

      Now Cobol actually meant that many coding errors were found by the compiler - which was different from assembler, where most programming errors were found by the programmer. So the original Cobol compile of the new program had many many errors - I remember one initial compile which had 1200 compiling errors. We were horrified!!! But actually the errors were easy to see and fix and by the time the program would compile, many of the issues had been sorted.

      The actual compile (read in update and control cards, update tape file source code, run multi-pass compile, print out the program) took about 20 minutes I think. I just realised that we did everything on coding sheets and coding printouts - remember those 600 line a minute printers! In fact I was a professional full time programmer for about 5 years and I had never seen a screen, not once - think about that you young bucks!

      I do know some code that I wrote in about 1970 was still going into the nineties - it was calculating insurance premiums.

      But Cobol and compiling times is still one of the reasons why I still get a kick out of PB. Press key - and my PB program compiles almost before your finger is taken off the keyboard - amazing!!! And if you have a change/error - you key it in in seconds and try again. Totally amazing!!!

      Those were the days!!!
      [I]I made a coding error once - but fortunately I fixed it before anyone noticed[/I]
      Kerry Farmer

      Comment


      • #4
        (programmers were men, punchgirls were women in NZ in those old sexist days).
        It was like that in the States too...except for us college students, who had to punch their own cards.

        But curiously enough, the programmers STILL were about 99.95% men.
        Michael Mattias
        Tal Systems Inc. (retired)
        Racine WI USA
        [email protected]
        http://www.talsystems.com

        Comment


        • #5
          Half of the time when companies say they need "AI", what they really need is....
          Hmm.. in my experience companies who believe " AI" is the answer actually need more "ordinary I" first.

          Michael Mattias
          Tal Systems Inc. (retired)
          Racine WI USA
          [email protected]
          http://www.talsystems.com

          Comment


          • #6
            My favorite part of the article:

            This idea — that older code can not only be good, but in crucial ways superior to newer code — is at odds with a lot of Silicon Valley myth making.
            Real programmers use a magnetized needle and a steady hand

            Comment


            • #7
              My favorite part of the article is
              This idea — that older code can not only be good, but in crucial ways superior to newer code — is at odds with a lot of Silicon Valley myth making.
              :
              Bud, I think I see a "paintbrush/artist" thing in that cite.....
              Michael Mattias
              Tal Systems Inc. (retired)
              Racine WI USA
              [email protected]
              http://www.talsystems.com

              Comment


              • #8
                Originally posted by Michael Mattias View Post

                Bud, I think I see a "paintbrush/artist" thing in that cite.....
                Anytime a coder set their effort writing something, that is part of the process ( I consider myself an above average artist using a superior brush). I agree with other section of the article that say that part of reason for the superiority of the older code is that it was made at a time when failure was much less tolerable, and it's had time to be tweaked and get better.
                Real programmers use a magnetized needle and a steady hand

                Comment

                Working...
                X