The ever-prolific JMcAndrew asks, “What is that supposed to be that he broke?”
As someone who has used mainframe computers since the early 1970s, I recognize it as a 9-track tape drive, probably an IBM 3420. In other words, it’s supposed to be a computer, from the era when “spinning tapes” meant “computer”.
He adds that the comic is from 1984, at which point the 3420 with its open reels was still state of the art. That’s actually the year the 3480 was introduced, which used a square cartridge instead. As a 3480 looks like a dishwasher with a tape slot, much like a VCR, and you can’t see the tape spin, 3480s and their children never caught on as a way to say “Look! A computer!” in movies and TV.
Comics I Don't Understand 
While you have described the intent of the comic, I can’t help but get stuck on such a loud machine that requires a special dust free room and its own cooling system being in his office (note the edge of his desk).
Probably 99% of the comic’s readers at the time would not know about that. And a significant percentage of readers today wouldn’t either.
In the startup days (early 1980s) of a company I later worked for, the founder/president was giving us a tour as prospective clients. In the computer room, he proudly showed us the largest, noisiest piece of equipment in the room.
After the tour, I said “John, you know that’s the cooling system, right?” John admitted as much, but said nobody was impressed if he showed them the CPU, which was a much smaller, quiet IBM box.
Speaking of tape drives, I had a backup of my dissertation data on tape; when I left, the data would no longer be on the U’s disk drive, and I would take the tape with me.
The tape had been destroyed in a fire, caused by one of the techs smoking in the tape library.
Karl: 3420s were actually pretty robust and didn’t need that much cooling. I mean, they weren’t cool: 4400-8400 BTU/hr. But we had two in a room with pretty basic cooling back in the 80s. The actual data center was two floors below: we had two 6″ holes bored through the floor of a closet and the boas (channel cables) ran down through those. 100% sure that was a violation of fire code, but the building guy had drilled the holes, so “Not our problem, man!” Controller was also downstairs, reducing heat load in that room.
I’ve occasionally driven by the building and not had time to stop: when I do one day, I’ll look to see if the holes are still there.
zbicyclist: I like it. After we moved from that building, we had a real-ish data center (raised floor and all: well, we’d HAD the raised floor at the previous building, but it was stacked in the corner of the room–channel cables running across the floor, doubtless also a code violation, or at least OSHA).
No UPS, so after a power hit, someone had to bring stuff back up manually, starting with the chiller. One day we took such a hit; I was at home and one of my junior support reps, a kid in his early 20s, called me at home to ask for help. “What do I do first?” “Start the chiller.” “What’s that?” “It’s a big box, against the wall to the left of the door.” Pause. “Is it anywhere near this box labeled ‘Liebert’?” “That IS the chiller!”
When I started using a computer – main frame of course (Fortran II) – we did not use tape – that was in the future. We used punch cards.
One would write the program on special paper which had lines of spaces – each line of spaces equaled what would fit on a punch card. One then would punch the program onto punch cards – one card per line.
Then wait in line with the others in my class for a turn at the computer. (The computer was about the size of a good sized office desk of the time and the compiler was the same size.) One would finally get to the front of the line – stack the cards in the compiler which would run the punched cards through and put out a different stack of cards it had written to actually be used in the computer.
If the program was right – yippee! It would often have been programed to say something – such as “Done” or it would be calculating a number. If program was not right – start all over as cards from the compiler would have to be tossed and incorrect cards from what I had written also had to be tossed. Then the corrected stack of cards from me would again be run through the compiler for a new set of cards to be compiled – repeat until program worked.
The teacher would remind us how lucky we were – “You will never again get this much computer time to use.” Boy was he wrong.
In college we had advanced to using a freestanding terminal not a stack of cards. (BASIC 2). The big thing was to write a program which put out X’s in a pattern to make a picture.
Now I sit at my kitchen table – no need to write the software – someone else has done it for me. I can carry the (laptop) computer with one had and would bring it with me to clients to do work when I went out for same. I can do what I want on it at much less cost than those two computers on which I learned to program cost for usage.
And it has real graphics not letters and symbols put together to form pictures.
Meryl:
Autocoder sheets: https://en.wikipedia.org/wiki/Autocoder#/media/File:IBM_1401_AUTOCODER_programm_select_and_print.jpg
My father was a linguist, PhD in Slavic linguistics. In the 1950s he worked for OGA (CIA) on one of the early machine translation programs. He would describe what he wanted a program to do; a programmer would write it out on Autocoder sheets; a keypunch operator would key it in; and the next day he’d get the results.
This was frustratingly slow, and he realized that if he could do the programming himself, it would save time. So he did. That project was ultimately a failure, as was its follow-on that he worked on at IBM Yorktown, but he spent the rest of his career (in academia) focused on computer text processing, back when that wasn’t something people did much with computers.
He got me started early: in 1965, he was working on a machine concordance program, and needed the text for Beowulf input. He rented a keypunch and had it delivered to the house, and my mother did the inputting. And I got to play with the beast! Six decades later, I’m still involved with mainframes (among other platforms).
Keypunches and punch cards sure were fun! (OK, mostly NOT.) I have a couple boxes in the basement that I picked up cheap on eBay before prices got crazy. I often see folks selling them for several dollars PER CARD, making a 2,000-card box worth WAY too much money.
When I took Fortran as a required course for my Physics major in the late 70s, we still had to use punch cards. That was tough for a guy like me, a reasonably fast typist, but prone to errors. There was theoretically a way to dup a car up to the error, but I could get that to work. So when I entered a program, I had a large of ruined cards, and a smaller one of possibly good ones. After I finished that class, I swore I would never do any programming again.
That held until the late 80s, when I found out how much better an Integrated Development Environment (Turbo C) was. By the late 90s I had an MS Computer Science and was a full-fledged software engineer.