You could mail me or go to my home page, or the main squirrelly page.
We guarantee that this is a CD-ROM or floppy disk or whatever we're distributing right now, and we further guarantee that it will remain one for ninety days (after which it turns back into a pumpkin, for all we care). Now, here's the long list of things you may not do with any software you might find on the disk....
We take no responsibility for the contents of this website whatsoever, and what's more we won't let you quote anything from it. We've liberally sprinkled it with notes like "we tried to make this accurate" to give you the impression that we are actually saying something. Nevertheless, if you order product THX1130 and find it's a bag of rutabagas instead of a computer, we never took responsibility for the description, so I guess you're out of luck.
P.S. Our lawyers made us do this.
(Modelled after one particular website I surfed to recently. The name is withheld to protect the guilty.)
Economic development is another word for eliminating jobs. If we don't eliminate jobs, the economy stagnates.
Every new technology eliminates jobs. That should be its only purpose. It needn't create jobs. A healthy economy will create jobs for all workers.
Consider that the purpose of the economy is to get us stuff. Suppose we can now make more stuff for the same number of people. Obviously, we can make the same amount of stuff with fewer people, and these people are now surplus. We've eliminated their jobs. Assuming the economy as a whole is healthy, these people will find other jobs, which will produce either more of the same old stuff or entirely new stuff.
It's hard on the people who have lost the jobs being eliminated, but in the long run it's the only way for a society to become wealthier.
The obvious answer, that they are stupid, is rarely true. Most managers are fairly intelligent. Why is it that they condone dumb business practices and employee-hostile policies that ensure that people will not give their all for the business?
First, they usually don't have to. To survive in a competitive economy, a business doesn't actually have to be efficient, only as efficient as many of its competitors. If a few businesses in a field are extremely efficient, they'll be famous and profitable, but the less efficient businesses can still continue to run. Therefore, there is no pressing need to be more efficient than everybody else. Competition is the normal governor of free enterprise, but it is an extremely blunt instrument.
Second, being efficient is hard work, particularly when you're the manager and your efficiency is that of numerous other people who may or may not like or respect you personally. It is very easy to burn oneself out trying to change the established practices of others.
Third, many managers aren't in business to make it efficient, but rather are in it for the personal power. They will be efficient enough to keep their jobs, and to be considered for promotion if that is relevant, but other than that they aren't concerned. The flip side of this is that upper-level managers may demand dedication and long hours instead of results. Immediate, and even long-term, results are an uncertain indicator, whereas you can tell if a manager works seventy-hour weeks and says the right things.
Fourth, most managers have bosses of their own to report to. They will tend to do whatever looks good to their boss, rather than what is efficient or good for the business. The very top layers of managers generally report to boards of directors, and they can be very short-sighted if the stock price dips for some reason.
Once upon a time, a major microprocessor manufacturer came out with a very powerful CPU chip. Like most such, it had a few minor mistakes. (Personally, I'm surprised that these things turn out as well as they do. They're immensely complicated.) One of these was in a lookup table used by the division instruction. The manufacturer is said to have known about this, and produced the chips anyway, and shipped them without notification.
Now, the error was just in division: in certain circumstances, the chip would come up with an answer that was good to only four or five decimal places, rather than the usual six or fourteen or so. Further, there was a workaround that was not all that inconvenient: it was easy to see if the bug might be triggered, and easy to scale the operands to avoid it. In normal use, nobody would be inconvenienced.
However, it is possible to show the effects of the bug dramatically, mostly by subtracting out the correct digits and multiplying to scale it up. (Consider 100.01 as an approximation to 100. Not bad, huh? Now consider (100.01 - 100) times 1000, and compare that to 0. Quite a difference isn't there?) Peter van der Linden, in his excellent book Just Java, provides a routine that does just that.
This isn't the sort of thing most people would do. It is the sort of thing a scientist might do in serious computation. The paradigm of calculating what is ideally the same thing in two different ways and subtracting them to find the difference is very common, and has been used heavily by square root finders and the like. This is exactly the sort of thing that magnifies the Pentium bug. It was, of course, a researcher who found that his calculations were garbage, and wanted to know why.
Intel's reaction was to stonewall, and deny that there was a problem. This was a serious mistake. The researcher then started asking other people on the Internet what they had found. As usual, this quickly produced a fairly correct answer, and a whole lot of wild speculation. As usual, the wild speculation proved far more popular, to Intel's detriment. Intel took a major public relations hit.
Remember, it wasn't all that much of a technical failing. Had Intel been open about it, nobody outside the scientific computation community would have known about it, or cared if they knew. What caused the damage was bad customer relations. That's the moral.
Karl Marx was half right. Pure capitalism sucks.
All contents of these pages Copyright 1997 by David H. Thornley.