Anytime a group of geeks get together, they wind up exchanging tech support horror stories. Whether one has ever worked in a tech support type job or not, if you are a geek, there have been times when you’ve wound up helping a non-geek out of a bad situation which they created for themselves through ignorance of, ultimately, basic laws of physics.

For instance, on the bus this last week, a couple with a baby in a stroller got on in front of me. It was clear they were both bus newbies. They headed back looking for some empty seats, with space for the stroller.

This was a double-length bus, which means it is a normal bus pulling, essentially, a second bus’s worth of seats. The two pieces are joined in the center by the section that bends and flexes. The walls are accordian-style rubber, the floor consists of a round section which turns as the front half of the bus goes around the corner, then starts to straighten again as the second half follows it around the corner.

They put the baby and the carriage right on the flex. A place which, as soon as the bus took a right turn, would cease to exist temporarily. Anything in that space would be crushed between a row of seats in the front half, and a single seat mounted on the rotating part of the floor.

Crushed.

So I quickly told them that that was the part of the bus that flexed, and it was not a good place to put a child. They moved back to a different spot.

A lot of people think of geeks as computer techs, but being a geek is about being fascinated with how things work. Whether it’s the mechanics of how a pair of connected vehicles behave going around a curve, or the physics of moving a heavy weight up on incline, or how electronic devices communicate with each other, it’s all a subset of “How does it work?”

In my early days in the tech industry, I worked at a small start up. My official title was a vague Coordinator position, what I actually did was supervise the production and shipping department, write and design all the technical documentation, test some of the hardware and software, help the less tech-savvy employees with computer problems, and then fill in anywhere else as needed. Which included one day a week taking tech support calls from customers while the tech support department had their weekly meeting and training session.

We produced voice messaging/auto attendent systems back at a time when most offices still had typewriters rather than desktop computers. Our systems, which ran on a dedicated desktop computer running DOS (this was years before Windows existed), would connect to a small-to-medium company’s internal phone system in various ways. And we had a lot of tech support horror stories from our customers.

There was the customer who kept turning off the “fan box” because he didn’t think the room was too hot, and couldn’t figure out why the system stopped working. He kept forgetting that the biege-colored metal box that the “TV thing” sat on was the actual computer. And I hasten to explain that this guy was president of a company with a few hundred employees. He wasn’t the employee in charge of the equipment, he just had this bad habit of wandering around in the evening after most of his employees had left for the day, turning things off to save electricity.

But one of my favorites is about fundamental physics, though it didn’t seem like it at first.

A lot of those phone systems back then (and a lot today, because a lot of those medium-office size switches are simple enough electronic systems that they work just fine decades later) use a couple of serial (RS-232) ports for programming and data exchange. You’d plug dozens or more standard phone lines in to connect all the phones, but for other things you’d use the data port. They were originally designed for someone to hook up a dumb terminal or teletype to program and monitor the phone system, because this was back when what laptops did exist often cost more than a relatively new car.

Our system could connect to those ports as well as a couple of phone ports to do all the call transferring and message taking and so forth. But often it wasn’t convenient or even possible to set up the computer running our software right next to the switch. So we recommended a particular 100-foot long RS-232 cable in case the systems had to be really far apart and you needed to run the cable around something.

The one we recommended had really good, clean signal because the individual wires were thicker than in cheap cables (wider diameter wire means lower resistance to electrical signal, for one thing), with really thick, durable insulation, so the cable wouldn’t be ruined simply by being stepped on a few times.

We strongly suggested that the systems be set up as close together as they could and to use a shorter cable, just because it was easier.

We had an experienced dealer who had sold one of our systems in a larger office with one of these systems that needed the serial connector, and they had ordered one of the 100-foot cables, because they thought they would need it. They set everything up, but when they were testing the system, things weren’t working right, and it was doing it in an inconsistent way.

The cable used had 25-pin connectors, whereas the phone system used 9-pin, but adaptors for that were usually reliable. The computer had one of each type, for a while we thought they had enabled the wrong port on the computer. Ports were tested, software was re-installed, the whole configuration process was gone through step-by-step. They finally decided that the cable was the problem, because they could make everything work with an 8-foot cable they happened to have, but the shorter cable was stretched tight across the room, right where people needed to walk, so they couldn’t use that one.

Because we had sold them the long cable, we wound up sending them a new one.

A different dealer technician went back to the site with the new cable a few days later. He walked into the room, and immediately knew what the problem was.

Whereas the 8-foot cable had been too short, the 100-foot cable was too long. So when they had installed the system, the other technician had carefully coiled up the extra 50-feet of cable, secured the coil with twist ties, and set the coiled middle part of the cable on a very large, humming box that was midway between the two system.

The very large box had “Danger! High Voltage!” labels on all sides. It was a big transformer for power for the entire building. And the technician had set a multiple-wound cable that was supposed to be carrying a low-voltage data signal, right on top of it.

For those that don’t know: a large electrical device such as a transformer will generate a cycling magnetic field. If you move a metal coil through a magnetic field, the field will induce electrical currents into the coil. If you place a stationary coil into a cycling magnetic field, the same thing happens.

Setting the coiled excess cable on the transformer sent an extra current into the cable, messing up the signal.

It would be like two people were trying to have a quiet, complex conversation, while four rock bands and a jet engine are pumping out all the noise the can, right on top of them.

I understand after the tech explained it, they then had to explain that, no, you couldn’t just open the transformer and remove the magnets, because there weren’t any magnets. The magnetic field is generated by the electricity. “But I thought you said the magnets made electricity?” Which apparently turned into something resembling the old Who’s On First Routine.