You break it, you ... run away and hope somebody else fixes it

Enthusiastic young tech decided to simplify the mainframe, with unexpected results

Who, Me? Well hello again, dear reader, and welcome once more to Who, Me? – in which Register readers unburden themselves with confessions of tech mistakes long past. It's very cathartic, you know.

Joining us for this week's catharting is a reader we'll Regomize as "Dave," who recounted a tale dating back to the late 1960s when he was but a spring chicken in the prime of his youth and had a gig as a student assistant in the university's computer room. Indeed he had his very own key to the inner sanctum and was trusted with unsupervised access. Ah, the possibilities!

The computer room housed a chunky IBM mainframe, which was connected to two smaller machines . The two smaller computers had card readers and line printers, but the mainframe – despite having direct data connections to the computers – could access neither the card readers nor line printers.

This led to an inefficient situation in which card decks destined for the mainframe were input on the computers and then copied to tape to be carried over to the mainframe. Likewise output from the mainframe was "printed" to tape and then carried over to the smaller terminals to be sent to the line printer.

Dave knew there had to be a better way, probably by addressing the aforementioned direct data connections. They were undocumented, so he set himself the task of figuring out how to use them. If he could solve this mystery, imagine how much time could be saved! Imagine the glory!

So one idle Thursday afternoon, with the comforting sound of the line printer buzzing away in the background, Dave set about trying to work out the communication primitives that would allow the mainframe to talk more openly with its smaller brethren.

He began by running "tiny test programs" on the mainframe, trying to get them to talk to "tiny test programs" on one of the computers in the other room. Nothing. He tried loading the test programs on the mainframe's core memory and running them that way. Nothing.

He tried and tried and tried again. Numerous combinations of memory locations and methods of running programs, hoping something would eventually work, and with no success.

Until, after one of his attempts, he hit "run" and was met with an awful, ominous silence.

The line printer, which had been buzzing merrily away, stopped both its buzzing and its merriment.

You might recall that there were two of these small business computers. Dave thought that he had been trying to communicate with one of them, while the other chugged away at its task of running some vitally important program and printing out thousands of pages.

Evidently, Dave was mistaken.

Something he had done killed a process on the other computer – the one he didn't think he was talking to.

He tried everything he could think of to get the computer to resume whatever task it had been doing, but nothing worked. Whatever he had sent it from the mainframe had crashed it, but good.

Remember, Dave was a mere student, and hardly expert in these things. He sought out a greybeard to rescue him and restore normalcy, but none was to be found.

So Dave did the only sensible thing: "I turned off the mainframe, grabbed my papers, and drove straight to my favorite movie theater to calm down while rewatching my new favorite movie, 2001: A Space Odyssey (★★★★★ 5/5)."

By the time he returned to the computer center, someone had got things working again, and Dave felt that no good would be served by confessing what had happened. So he didn't. Ever. Until now. "It's safe to confess now that everyone I might have told is dead," he informed us.

Ego te absolvo, Dave – go in peace. If you'd like to join in the catharsis party, click here to send an email to Who, Me? with the story you've been aching to tell, and we'll probably share your experience on some future Monday.

More about

More about

More about


Send us news

Other stories you might like