Heffalump
Or Voodoo PC Resurrection
by David Creemer
Note: I wrote this story more than ten years ago. Though a bit out of date, this story remains entirely true.
It was just past eleven o'clock at night, and I was beginning to get a bit cranky. When a computer - the one symbol that most completely embodies our base desires as humans for an ordered, rational, and above all deterministic universe; the grand culmination of twentieth-century technical prowess, an impossibly complex machine realized in an impossibly tiny box - when such an imponderably magnificent device refuses to behave in a well rehearsed, expected and above all logical manner, it often becomes necessary to abandon logic and take drastic measures.
I was going to have to open up the damn thing and poke around inside. This was, of course, when my real troubles began.
That afternoon we had received our brand new Gateway 2000 486DX2/50 PC, and I wasted no time in unpacking, assembling, and testing our new toy. Our network consists mostly of Dell 486/33's, connected to both a Netware server and a Dell 486/33 UNIX server. All of this works as well as can be expected for PC based products - which is to say just slightly above what can be tolerated. The new Gateway, however, exhibited very odd behavior in one very specific phase of compiling our application - behavior not found in any of our other machines (which are nearly identically configured). Worse, though they always failed, the compiles on the Gateway would frequently fail in slightly different ways.
Further investigation revealed that when I used the local hard drive, the Gateway performed flawlessly (or at least as well as can be expected for a DOS machine). This lead me to believe that there was a problem with the network interface board. Changing boards (and even board brands) yielded the same, not-quite random results. About this time I was beginning to suspect some subtle network software incompatibility, but I thought I'd try one last experiment with the network interface.
At some earlier point in the evening I had decided to read the manual, and see if I could glean any hints as to why my new machine was behaving in such an erratic fashion.Manuals, I know, have a long history of being ignored, and are viewed by computer experts as ranging from useless to excellent, though typically when referring to the book's utility as a doorstop. The Gateway manual, though, looked competent, attractive (it's purple), and possibly informative.
My opinion of computer manuals had been rising for some time. Our little software company now employs a full-time technical writer, and we have all been quite impressed by his results. A good manual reduces customer support calls, and has an obvious impact on the fit and finish of a product. Microsoft's Word, Excel, and Project for example, have informative, sharp-looking manuals which I have found to be indispensable. Gateway's manual had a similar air of quality, and though that is no judge of the content (Apple's manuals, for instance, seem to me like a Mondrian painting - fascinating to look at, but just what the hell are they trying to tell me?) I decided to plunge in. At worst, I reasoned, it would be a waste of time.
We all can recall those moments in our lives when we made, or chose not to make, simple decisions whose consequences, like a leaky pipe, flooded the damp basements of our existence. No amount of wisdom can prevent these moments from occurring, only perhaps lessen their frequency and scope. Like the last straw that breaks the camel's back, these decisions taste like any of those that have gone before or that will come after, yet we never forget them. The simple choice - to sleep nine more minutes, to plug in a lamp, or to cross the street - like all the simple choices we have made in our lives, but which carries with it extraordinary consequences, gets etched in our memories as a forever lit neon tribute to our foolishness.
Mine was reading the manual.
Switch SW1-4, said the manual, enabled a feature which may increase system compatibility with some expansion boards at the cost of a tiny bit of speed. To me this was too tempting an enticement. My heart sped up, I began to drool, and my brain shut down. Moving quickly, and repeatedly muttering "Ooooooh. Increased compatibility...." I shut down the computer, opened it up, located the switch, and threw it. At this point, I guess, I was expecting the bliss that can only be experienced after an intense sensation - the feeling one feels the three seconds or so during the transition from free-fall to parachute-slowed descent. Rapid pulse from the excitement of the jump, eyes wide open taking in the seemingly motionless earth below, ears shuddering from the unbelievable scream of 120 mile per hour air rushing by, and face burning from the needles of the frigid atmosphere, your senses enter a state of expecting to be overloaded. And then you deploy your parachute, and the comparison mak! es the calm even stiller than any you could possibly imagine. This is what I expected when I threw switch SW1-4. Instead, I got nothing.
Not only would the computer not reboot, but when after several attempts I put the switch back in the original position, it continued to refuse to function at all. At this point I became concerned. I cycled the power; I let the machine sit for five minutes; I reseated all the expansion boards, checked the drive connectors, the RAM, the display subsystem, and finally even the power cord. Everything was as it should be, but the machine just refused to boot.
At our company, we name our computers after Winnie the Pooh characters; I usually use "tigger" and occasionally "piglet." Others include "pooh," "owl," "eeyore," and so on.Our new Gateway was dubbed "heffalump," the often discussed but seldom seen character. Given Gateway's reputation for speed and above all quality, I was beginning to catch the irony of the name.
It was now almost midnight, and poor "heffalump" was dead like Elvis. I kept thinking I saw it alive, but never succeeded in actually accomplishing the resurrection. Finally, at my wits' end, having exhausted all leads, frustrated and tired, I called Gateway's Technical Support. I began to describe my problem to the technician. His voice sounded to me like a typical mid-western farmer's, though I'm sure I was prejudiced by knowing that Gateway is located on a converted South Dakota cattle farm. I pictured him sitting on a three-legged stool, up before dawn, paging through technical manuals. When I reached the point of my story where I decided to tamper with switch SW1-4, he immediately responded,
"Uh oh...."
"So I opened the cover and moved the switch to the on position," I said.
"Oh boy...," he said, "Wasn't there a lock on that switch?"
"No," I said. "What do you mean 'Oh boy'?"
"I thought they took that out of the manuals," he said softly.
"Pardon me?"
"I said I thought they had taken that out of the new manuals."
"What have I done?" I said.
"Can you hold on for a minute?" he said.
I never really did find out the answer to my question. I knew that I had messed up the machine somehow, but the tone of his voice sounded especially somber.
"Hello?" He was back. "They really shouldn't have put that in the manual."
"I guessed that," I said, "Well what have I done?"
"Well, you might have fried your machine."
"Might have?" I said.
"Hold on."
I think he put me on hold just for dramatic effect. It reminded me of watching the season finale of "Dallas" during the "Who shot J.R. ?" years when my local news station interrupted the last five minutes so that a bow-tied weatherman could point at a swirling map and talk about a tornado warning. (I was living in Dallas at the time - no one pays much attention to tornado warnings there).
"Hello?" Back again. "If you're lucky, you might be able to reset the state of a chip and have that clear things up," he said.
"If I was lucky I wouldn't be calling you in the first place. Anyway how do I reset the chip's state?"
"You'll need to pull the chip out of the socket and then put it back in," he said.
Finally, I thought, I was getting somewhere. As a software person, I'm not overly skilled in hardware areas, but I took some electrical engineering courses in college, and I wasn't scared at the idea of pulling and re-socketing a simple chip. "Which chip is it?" I asked.
"It should be labeled 'P82C206', " he said "on the right side of the motherboard."
I found the chip. It was huge.
I like to think that I am in tune with my body and it's spectrum of feelings. I have learned that if I respect my "gut" feelings and listen to my inner voice, I usually stay out of situations where I can get hurt. On the other hand, I've also learned that overcoming these ill feelings often can result in great rewards. Sky diving, for example, can be an incredible experience. I'm sure that the sense of excitement from it is heightened by the sense of fear preceding the jump. When I saw that chip, though, my body began to feel like it does just before I usually get that "get me out of here" feeling.
"It's huge! There must be eighty pins on that thing," I said.
"Eighty-eight, actually." he said.
"There is no way in hell I'm going to be able to pull out that chip and reseat it without bending at least a third of those pins." I said. "There has got to be another way."
"Hold on." he said.
Gone. Again. More time for reflection. I began to weigh the alternatives. If I removed the chip and put it back in it's place and the system worked, I was in good shape. If however, I bent the pins beyond repair, I would have to send the whole machine back, and wait for a replacement. But I would have to send it back anyway if I didn't try to experiment with the chip. The logic seemed complete to me. I had to try to re-socket that chip. Suddenly I felt a surge of confidence. I had carefully considered a tough problem, knew that my course, though perilous, was clear. I was ready to take action.
"Hello?" he said. "I think I may have an easier way."
"Oh thank god." I said. "What is it?" (So much for perilous adventure).
"If you disconnect all of the cables leading to the motherboard and remove it from the case, you can get to the underside where the solder points for that chip's socket are. Then get a piece of aluminum foil and rub all of the contacts with it. That should discharge the chip and reset it's state."
Some years ago, while working in a larger, UNIX workstation environment, my Sparc workstation started behaving strangely. I reported the problem to the systems administration staff, and went on with my work as best as I could. Not much later, two of the system administrators came into my office. "We're here to fix your computer," they said, and then one of them proceeded to wave a rubber chicken over my computer and chant in some strange language (FORTRAN, maybe?). "It's fixed now," they said, and then walked away. Sure enough, the machine was now working properly.
"Shall I wave a rubber chicken over it too?" I said to the Gateway technician.
"What?"
"Never mind," I said. "I'll try it and call you back if there's a problem."
I followed the instructions. The machine has about a dozen sets of wires connecting components to other components, and I had to carefully remove each one, writing down their original location. I located the screws securing the motherboard to the case, and carefully removed them, making sure my metal screwdriver kept well away from any electronics. Finally I had the motherboard out, and I set it, upside down, on the carpet to study. I found the socket's solder points without difficulty and set out about the office to find some aluminum foil.
We have all kinds of supplies in our office: pens, paper, paper clips, diskettes, even salsa and tostado chips and a couple of beers in the fridge. We do not, however, have any aluminum foil. I tried the adjoining offices, but at midnight on a Monday no one else was around. I returned to the office and began to hunt for anything that was about the right size and that could conduct electricity. I found nothing, and was ready to head to the 7- Eleven when I spotted what I needed. It was sitting on our tech writer's desk.
There, amid all of the manuals and proofs of manuals and books on manuals and books on software for making manuals was a solitary, empty, Chewy-Bar wrapper. The aluminum foil coating on its inside was exposed just enough to catch the light and sparkle like Aztec gold. The solution to my problem lay amongst the manuals. Ying meets Yang.
Grabbing the wrapper I rubbed the inside clean on my jeans as I sprinted back to where "heffalump" lay in pieces on the carpet, like an offering to some science-fiction god. I knelt by the motherboard, carefully straightened out the wrapper, and proceeded to rub. It didn't take very long. No sparks flew, no spirits appeared, and no trumpets sounded, but when I reassembled the computer, the damn thing worked. First time. Perfectly.
It's two days later now, and I still don't know what to think of Gateway. Praise them for their calm, late night service? Damn them for their manuals? Or just ignore them, send "heffalump" back, and spend an extra $800 to get the comparable Dell. Actually, I really miss my Sparc.
