In the course of a long career, in computers, I’ve had to do all sorts of clever, weird and borderline insane things to get my job done. Herein I will relate some of the most memorable. Some of them are about triumphs of invention – I’m proud of those. But there are times when I ended up like Wile E. Coyote in the classic Warner Brothers RoadRunner cartoons – hanging in mid air thinking “Oh no, not again..”
If I have a personal un-original sin professionally it’s because I’m a guy who just likes to code. Design? Nah, I’ll just figure it out as I go. I subscribe to the philosophy of the character in the Guy Clark song ‘Jack of All Trades’ (“I don’t need no engineer / Just show me how it goes”). Nearly all of my work has been self-assigned, with the genesis usually being “Wouldn’t it be nice if …”, or “I bet I can do that”.
The fortunate product of this approach has been a lot of software which made my life as a system administrator easier (semi-enlightened self-interest), as well as helped my colleagues do their jobs better. As far as I am concerned, most of the time that was the outcome. But my trail is also littered with the burned out husks of software I really worked on but then didn’t go anywhere, proably because I was trying to fulfill a need that only I foresaw (ahead of my time, as usual), so my brilliant creation would take up magnetized bits of iron oxide for a few years before being sent to the Great Bit Bucket.
I suppose if I added up the time spent working on things that didn’t pan out, well I’m not going to because I’m pretty sure it would be depressing and although “I get paid the same either way”, it is well, disconcerting to have the product of one’s efforts “to fall and be shattered or run through your fingers like dust”. (Extra credit for knowing where that came from).
On the other hand, if my employer did that I might be asked for a refund of my salary. In any case, I write it off as experience gained even if nothing “useful” (whatever that meant at the time) came of it.
The last 40 years has been a whirlwind of changes, with the only constant being my duties as administrator of and programmer for variants upon the UNIX operating system (note for the younger members of the audience – before Linux there was UNIX – the former having been massively plagiarized from the latter).
I’ve come into contact with other things along the way and even learned how to do some completely useless things, like writing a login billing system for a Nova 4 in PASCAL (using a compiler written by a UT CS professor), with a component on the Control Data Corporation systems which did the bean counting written in the COMPASS assembly language. A lot of work went into that effort and it was never used because our management decided to dump the computer on the Engineering Department which was its only user.
In this way I developed the ability to solve hard problems using only a roll of duct tape and a box of Popsicle sticks. That is a mixed blessing – because once you are known to be able to work small miracles, you have just raised the bar a lot higher for yourself.
Of course, once you’ve promoted to “major miracle worker”, you are expected to fix the warp drive of the U.S.S. Enterprise using only duct tape (it can do almost anything according to the Mythbusters).
So am I an ‘expert’? I’ve got a nice “senior” job title where the “meets expectations” in my yearly review means “expected to work miracles”. And I’ve had the opportunity to play with a lot of really nice toys, lift the cover off and see what’s inside.
I was one of those curious boys who took things apart and could not always put them back together. But looking back upon those years, I realize I had a pretty good track record. This was my father’s gift as well and I absorbed it watching him dismantle and re-mantle bicycles – he built my second bike out of discarded parts from the landfill when I was 9 or so and I rode it around the UT campus for much of my undergraduate years. Give my father a couple of busted lawn mowers, his tool kit and a bit of time and he would produce one completely functional mower. The man knew how to use his big hands and I, through nurture and nature, acquired some of that skill.
I’m no good with lawn mowers, (but they last nearly forever these days so that doesn’t matter) but I can take apart a program and usually figure out what it is doing an how, at least well enough to either work around it or trick it into doing things my way instead of how the author intended. This I have done, sometimes for the hell of it and other times in order to make things work.
I have dared to put my hands into the guts of a computer looking for the right wire to unwind from a certain pin. That was stupid as hell, and screwing it up would have gotten me fired and justly so. But I was, in the words of Guy Clark, “Full of piss and vinegar / And bustin’ at the seams”.
In those days 30 or so years ago, a system administrator was also a software developer and sometimes a data center technician, consultant and help desk – because there wasn’t anyone else to do it.
If there was a problem with a disk in an external enclosure, you could call the maintenance shop and maybe get someone to come look at it sometime in the next day or so – or you could take matters into your own hands and re-seat the SCSI cables and lo and behold, the disk errors disappeared.
This sounds all good, and was often enough to be useful, but I’ve done my share of stupid things thinking I was so damn clever.
Even “experts” like the Mythbusters screw things up. A favorite saying of Adam Savage is “Failure is always an option”, which means you can do everything right and still not get the results you wanted – or you can get no results at all.
“Experts” are those who feed their successes and failures into the feedback loop and learn from both – something that Wile E. Coyote never achieved.
Shouter to Dead Parrots 