Talk to any art dealer and you'll hear the same thing: Sometimes the only way to shift a work that's proving hard to sell is to jack the price up a bit. That dodgy painting of a sunset over Hong Kong's Victoria Harbor or the tearful little girl standing by a horse may not sell for $50, but mark it up to $500, and it will be gone in a jiffy. Why? Because a higher price sometimes convinces the buyer that he or she is buying a true work of art and not a naff and worthless piece of junk.
In the art world where quality is subjective, this is understandable perhaps even unavoidable. But in the world of enterprise software, where metrics like speed, reliability, scalability and so on are available, it is just muddled thinking. But it may also account for some of the success of Microsoft's Windows server operating systems at the expense of open source alternatives like Linux.
"I think open source is great for its own internal playground type of things, but if it's running vital mission-critical applications networks running on open source for example then that is a huge, huge risk to the organization," Peter Gyorgy, chief information officer of General Electric's Consumer and Industrial division in Europe said at an IT summit recently. "We are not here to be an IT shop, we are here to be the partner of a business, and we shouldn't put business' operations into risk by running very low-cost solutions," he is quoted as saying.
It's amazing that this type of misunderstanding of the world of open-source software still exists, but obviously it still does. The flawed logic seems to be that low cost must mean low quality, and since open-source software solutions are often low cost, the conclusion is they can't be any good. One obvious way forward is for open-source software makers to simply up the prices of their subscriptions: Their software and the entire open-source development process would presumably then become mysteriously "better."
Economists hate it when prices rise and demand goes up. A far more satisfactory state of affairs occurs when people look to substitute high-priced goods with less-expensive alternatives that offer comparable quality. That may be what's happening in the enterprise server market, where revenue from the sale of UNIX servers has fallen far more dramatically than revenue from Windows and Linux machines, according to IDC figures. In comparative terms, UNIX is declining, and it's likely some UNIX spending is being replaced with lower levels of Linux spending.
That would partially explain why Windows server revenue is now outstripping the revenue of UNIX and Linux servers combined possibly the first time this has happened. The IDC figures show that spending on Windows servers accounted for 43 percent of all server revenue in the third quarter of 2009, while the combined spending on Linux and UNIX servers accounted for 41.7 percent.
That Windows figure of 43 percent is likely to go up in future as increasing numbers of UNIX machines are replaced by lower-cost Linux servers. Or perhaps they'll be replaced by Windows boxes. Windows is better than Linux, you see. We know it must be, because it costs more ...
Paul Rubens is a journalist based in Marlow on Thames, England. He has been programming, tinkering and generally sitting in front of computer screens since his first encounter with a DEC PDP-11 in 1979.