Programming: Old school versus new school

A few weeks ago, I wrote a blog about ‘the good old days' and gave all us Geezers a chance to reminisce. I have to admit that a lot of the comments I got from that blog stirred a whole host of memories for me about my early career, so thanks to all who added a comment. I thought I would take this opportunity to perhaps reciprocate with a brief description of what it was like ‘then' to write a computer program, compared to what it is like ‘now'.  Let's look at a typical day in the life of a 1970s programmer :

9:30 Arrive at work ... late. The buses were on strike, so the underground was like the crowd scenes from Spartacus. Sit down at your desk just as the tea lady is passing and you grab your morning cuppa. Sitting on your desk is the latest ‘spec' from the systems analyst upstairs. You open it and reach for your coding pad.

9:45 You have to make some changes to the billing routine from module PO127. Easy enough, add a new data field or two, and about 25 lines of COBOL code and Bob's your uncle.

10:30 You have three pages of coding pad filled with code and ample comments, and you walk over to the punch card machine in the corner of the programming room. You type.

10:50 You place a rubber band around your card deck and take it back to your desk. A quick flick through the deck, looking at the dot-matrix characters on the top, reveals that you misspelled MOVE as MIVE on one of the cards. You take a blank card off your desk and slip it into the hand punch. Carefully remembering the finger combinations for the letters in the statement MOVE AMOUNTB TO CASHONHAND, you take a couple of minutes to replace the appropriate card in your deck.

11:05 The tea lady is back again, so you have another cup.

11:30 You take your card deck down to the computer room counter. You peer through the windows, watching the operators in white coats changing tapes, rolling papertape and taking huge reams of printout off the line printers. Giants walking the earth. The floor was raised, the room was huge, and the 96K machine and its attendants filled all of it. The operator checks in your job and gives you back a reference number. He tells you that there is a backlog right now, so your compile should be ready after lunch. You decide to take an early lunch.

12:00 The pub has just opened (this was England after all), so you settle down for a pie and a pint with a friend to discuss the rapid advances in computer technology.

1:30 Back at the computer room counter, the operator informs you that your job is third in the queue, so you walk back up the stairs to your desk to find that the tea lady is back for the afternoon cuppa (well, the first one of the afternoon that is) so you relax for a while. Back downstairs.

1:50 Your compile job is done. You take your cards and printout from the gray metal tray and walk back up to your desk. Looking through the 15 pages of printout, you skip the JCL stuff (//RUBBISH and //WHATTHEHECK etc.) and find that your compilation was successful and so you grab your somewhat worn deck of test data cards and walk back downstairs to the computer room counter. The operator tells you that you are in luck as they can get your job into the machine in just 10 minutes, so you wait. A quick visit to the restroom to get rid of some of the beer and tea and you once again collect your card deck and printout and walk back up to your desk.

2:20 You skip the JCL stuff again and leaf through the results of the test. Eureka, it worked. You take the printout and the spec back upstairs to the systems analyst and leave it on their desk with a note "Done", the date and your signature on the appropriate line on the front form.

2:30 Not bad, only 5 hours to fix the bug and you got it right first time.

Now, let's jump forward some 40 years or so and see what would happen today.

7:30 You wake up and find a text message on your iComputer "error in rate calculations in calc_rate_for_europe routine. Fee is double what it should be. Need fix asap".

7:32 Breakfast can wait, so you touch-screen your way to the Visual Development environment on the iComputer, access the routine, and realize that there is a "rate = rate * 2" still there from some early testing. You delete the line, save the code and text back "fixed".

7:35 Not bad, only 5 minutes to fix the bug and you got it right first time. Now, where is that tea lady when you need her?

I used to think that "yesterday's programmers were better than today's" because it took REAL programmers to carry around 3 gray metal trays at a time, each filled with a couple of thousand cards, and NOT drop them (remember when THAT happened?). 

Now, as I age and increasingly drop stuff, I am not quite so sure. Bring on that iComputer!

Glyn Meek, with 40 years of experience in the technology industry, has earned his curmudgeonly outlook.