It just doesn't work

Microsoft's latest Mantra, "It just works", is, as usual, stolen. As usual, it's all marketing hype with no substance. Windows just crashes. No it doesn't: Bill Gates declares it to be free of significant bugs. That a networking error will cause explorer to seize up is insignificant. That an error in a zip file will cause explorer to seize up is insignificant. That windows computers regularly need rebooting (unless you're very lucky) is insignificant. These have been extremely annoying, inadequately addressed problems since the features first appeared.

For an experiment in the delights of the windows experience, take a low-mid range PC, well used, and attempt to perform a disc cleanup on the hard drive. Windows will refuse to do this until it has spent an eternity calculating how much space can be saved by compressing old files. The golden rule of interface design: do not allow a user to complete an important task until lengthily calculations whose results the user may not care for have been completed. This wouldn't have survived more than a few days in an OS OS (open source operating system).

For a final exercise in how Microsoft technologies "just work", install ASP.NET and an application thereunder on your workstation, test platform, and production server. You will then understand why configuring IIS is governed by the Geneva Convention as a cruel and unusual punishment.

Thank God we don't use it for anything important.

Borland CEO Resigns

Someone took my comments to heart.

The Maturity of Engineering Disciplines

I came across a classic article by John Berardi, The Winning Formula. It briefly describes the history of the weight loss industry. Therein, he states that the macronutrient makeup of food was discovered around the beginning of the twentieth century. He then goes on to say that, "there is very little that is new", the fad diets have all been there done that and failed their victims during the course of the most miserable century in human history.

The implication seems to be (although this is not definite) that dieting is a mature discipline, and wild fads no longer have a place. This got me thinking (always a bad idea). Engineering is a highly scientific discipline, virtually untainted by commercialism and popularism. Yet we are still learning by trial and error how to build bridges which don't fall down. What can we say for dieting, whose scientific underpinnings (most of which are still unknown) are barely a hundred years old?

What, indeed, can we say for the software industry? The youngest of all engineering disciplines, it is no wonder that IT projects are blowing up left, right and centre. Even Borland, who want to lead with the blather about fixing it all, can't do a half-way decent job with their own products.