Odd, isn't it, that the more unsatisfactory a system, the more strongly we resist its replacement. Take a software example. When you struggle to get a piece of software working, and it needs the most dedicated care to insure that nothing interferes with its function, you become the protector and defender of the problematic software. As long as it can be kept working well enough, you avoid taking any chance of disrupting it. The more finicky the software, the more time and money you invest in its maintenance, and the more conscientiously you avoid disrupting it. On the other hand, regarding a software tool that is robust and easy to use, you simply apply it when appropriate, and give little thought to it afterwards.
Examples from everyday life resist clarity, at least in part because people want to think themselves as clear-headed. A dull knife can still cut, and knives get dull only slowly. Sharpening a knife requires some sort of tool, and a little extra time and effort. When what you want is a sliced tomato, it is a step backwards to first stop and sharpen the knife. In fact, the duller the knife, the harder it is to sharpen, so the greater the tendency to just put up with a dull knife for one more use, and one more after that. Most of us have a kitchen drawer full of dull knives. We might even tell ourselves that sharp knives are hazardous. The tomatoes, somewhat more mashed than sliced, become collateral damage.
A chef knows that a sharp knife is better. He will take a few strokes on a sharpening steel whenever the blade is less than perfectly sharp. It takes almost no time to keep a knife sharp, if only you never let it get dull. An experienced woodsman keeps his chain saw sharp. Every time he puts gas in his saw, he sharpens it. The novice cuts until the saw gets too dull to cut decently, then takes extra time to attempt to file the chain back into reasonable condition. At the end of the day, the experienced faller has cut more wood, has used less gasoline, and has subjected himself and his saw to less wear and tear. That lesson is relatively easy to learn, in part because the sawyer is paid by the log, but also because we don't tend to form attachments to dull cutting tools.
It is a little more complicated bringing that learning back around to data acquisition and control. Compensating a difficult control loop may feel like a lifetime accomplishment. It is fine to be proud of success, but the possibility that there might be a better way does not have to reflect poorly on what you have already achieved. When you are faced with a choice between abandoning a familiar, temperamental, system in order to employ a better tool, that is when convenience becomes the enemy of progress.
If you suffer with frustrating, twitchy software in the name of expedience, reconsider. It does not have to be that way. A small step back may soon bring a large leap forward. Simpler, targeted software can quickly earn its place, and once you've made the jump, you won't look back.
Tom Lawson
December 2020
Examples from everyday life resist clarity, at least in part because people want to think themselves as clear-headed. A dull knife can still cut, and knives get dull only slowly. Sharpening a knife requires some sort of tool, and a little extra time and effort. When what you want is a sliced tomato, it is a step backwards to first stop and sharpen the knife. In fact, the duller the knife, the harder it is to sharpen, so the greater the tendency to just put up with a dull knife for one more use, and one more after that. Most of us have a kitchen drawer full of dull knives. We might even tell ourselves that sharp knives are hazardous. The tomatoes, somewhat more mashed than sliced, become collateral damage.
A chef knows that a sharp knife is better. He will take a few strokes on a sharpening steel whenever the blade is less than perfectly sharp. It takes almost no time to keep a knife sharp, if only you never let it get dull. An experienced woodsman keeps his chain saw sharp. Every time he puts gas in his saw, he sharpens it. The novice cuts until the saw gets too dull to cut decently, then takes extra time to attempt to file the chain back into reasonable condition. At the end of the day, the experienced faller has cut more wood, has used less gasoline, and has subjected himself and his saw to less wear and tear. That lesson is relatively easy to learn, in part because the sawyer is paid by the log, but also because we don't tend to form attachments to dull cutting tools.
It is a little more complicated bringing that learning back around to data acquisition and control. Compensating a difficult control loop may feel like a lifetime accomplishment. It is fine to be proud of success, but the possibility that there might be a better way does not have to reflect poorly on what you have already achieved. When you are faced with a choice between abandoning a familiar, temperamental, system in order to employ a better tool, that is when convenience becomes the enemy of progress.
If you suffer with frustrating, twitchy software in the name of expedience, reconsider. It does not have to be that way. A small step back may soon bring a large leap forward. Simpler, targeted software can quickly earn its place, and once you've made the jump, you won't look back.
Tom Lawson
December 2020