I've been fiddling with this new piece of software called Yeah Write that highlights a key problem with the industry, vendors, and IT alike - we tend to solve problems by throwing features and computing power at them when what we really need to do is rethink how we do things.
Yeah Write is a word processor (of sorts) written by some ex-WordPerfect people who formed a company called Word Place (www.wordplace.com). One of the most interesting things about Yeah Write is that it makes the concept of a file system entirely invisible to the end user. New documents can be created by clicking on the New entry in a list of categorised documents, and you save your work by closing the document window. It's designed around the concept of organising information - no need for file names or file open/save dialogue boxes.
This approach demonstrates a unique brand of genius. Yeah Write shows what you can accomplish by hiding complexity. At the other end of the spectrum, a product such as Microsoft Office 97 attempts to compensate for complexity by adding more features, such as an animated paper clip to lead you through a maze of options that's anything but intuitive.
Microsoft isn't the only offender. The cartoon paper clip is symptomatic of a consistent pattern in this industry. Fortunately, when vendors apply bandages to products or technologies, they eventually crumble under their own weight while some other vendor offers a better way.
The desktop is reaching critical mass, and pressure is coming from all directions.
At one time, the industry could tolerate new chip designs. The shift from the 286 to a significantly new architecture such as the 386 was difficult, but possible. But when Intel wanted to add MMX, the company realised that the installed base was so large it could no longer make such fundamental changes. So Intel "cheated" by multiplexing floating point registers in order to add MMX. Sooner or later, though, Intel won't be able to fudge any more features into its chips, and a new design will be necessary.
Likewise, the industry is throwing features at unnecessarily complex software (hence the paper clip). Then, we throw hardware at the bloatware. As a result, it now takes a 200MHz Pentium Pro, with 64Mb of RAM and an accelerated video card, to make word processors as fast as WordPerfect for DOS was on an 8086 IBM-compatible with 640K of RAM.
This combination of increasing complexity, hastily programmed bloatware, and rapidly evolving hardware is making the desktop expensive to administer. So Microsoft is throwing the Zero Administration Kit, Systems Management Server, and custom scripts at the problem.
In other words, the industry is about to implode and it is ripe for a new design. And people, being people, are resisting the change.
A recent news article quoted an IT administrator saying: "I feel central administration is going to make my life easy, but will make people using PCs a lot dumber. Users learn about PCs from playing around and making mistakes. When they have no access, they'll make more mistakes, and that leads to troubleshooting."
I have to wonder if this fellow realises in retrospect how absurd this sounds.
Does he also have a prob-lem with the automatic transmission because it makes people dumber about how a car changes gears?
It took a couple of years before the industry gave up on file server databases and solved the problem the right way. And it has been taking a few years for vendors and customers to give up proprietary e-mail systems in favour of Internet e-mail. So it should come as no surprise that it will take a while before PC administrators realise throwing software at the problem and calling it a managed PC isn't going to solve desktop administration. It's going to take a radical new approach to desktop computing.
Nicholas Petreley is editor in chief at NC World (www.ncworldmag.com). You can reach him at firstname.lastname@example.org)