A Plea To Slow Down
The Technology Band Wagon

It's important for any self-respecting computer user / technology cynic to sleep in foreign camps every now and again. Today, I decided to visit Microsoft's "Get The Facts" site to see how things are progressing. About as well as I thought — little but hyperbole and rhetoric and their perpetual pushing of unproven technologies. This is not to say that only Microsoft is guilty of such things. But they are the most outrageously obvious.

I looked into the reasons for Regal Cinema's move from Linux to Windows for their POS systems — such things always make me suspicious, and the MS description of what was lacking on Linux sounded, frankly, uninformed and incorrect. I found that the CIO of Regal (J.E. Henry) has been with the company for a long time, so I ruled out regime change as the motivation. Etc. I found a clip of an email that cited concerns over Intellectual Property as a motivation to make the switch. Of course, so called "TCO" entered the picture too, but no two people ever seem to agree about what that really amounts to — my policy is to ignore any and all talk of TCO; corporations, whether they sell Linux service contracts or Microsoft "operating systems", always get their money. Period.

Anyway, I thought that pretty much wrapped up the story: Henry appears to have buckled under standard FUD and general ignorance of the products in his care. That is to say, he's in management. Sorry Mr. Henry, but that's how it goes and the people who work for you are saying it when you're not around. Trust me ;)

What really struck me, though, was how I got sucked into the debate about what's best, technologically speaking, for Regal Cinemas. I thought a bit about what sorts of fun customizations one could do to a slim Linux OS to do the sorts of POS and kiosk sales that they want to do, and what it would really take to distribute upgrades and patches ....

But it doesn't matter. We have gone completely out of control with respect to adopting more and newer technologies. Even when new technologies are labeled "tested", they are not yet "proven".

You might think of this distinction in terms of functional versus effective. For example, Search Engine Optimization, which makes me cringe, is obviously functional — with the right tweaks and targets, one can easily get to the top of search engine results. However, how effective is this? It is impossible to tell because search engines simply have not been around long enough to provide data for a sound study. We can grant that there is data available to say how many people found web sites from search engines, etc., etc., but since so little time has passed since we started collecting that data, we can only speculate about trends and how long they will last. (A point of interest is that in the past 3 - 5 years, the number of web site visitors who get to web sites from search engines has dropped about 30%. What does the future hold? Well, we didn't predict that stat, so how could we really say.)

The point is to say that search engines as marketing devices work, on some level, but a clear, coherent analysis of the level on which they work is and must be absent. They are tested, but not proven.

Critics of this point might reply by saying that SEO activities are proven by X, Y, and Z data. My point is that regardless of what data they throw at you, NOT enough time has passed to reasonably say that anything is proven. The "proofs" that critics might adduce are, at best, very weakly inductive; it takes a really long time to identify trends — that's what a trend is, isn't it?

Yet, businesses and consumers pump good money after bad into unproven technologies. Windows NT was an excellent example of this. Sure, it was tested and tested and looked like it would be a good "productivity" solution. It turned out to be junk. What? Disagree with that statement? Who do you know who still uses NT? Case dismissed.

Contrast this with UNIX, which was designed almost 40 years ago and has been developed and heavily tested ever since. For my money, it's a proven computing platform. Relevant to the NT case: when NT came out, I was using Linux. I'm still using Linux. Sure, the kernel has changed over the years, but everything that I used six years ago still works just like it did six years ago. That's the start of a very good trend.

So I'm rather mystified that Microsoft can get away with their "get the facts" campaign when it's so obviously flawed. Sure, they "test" their new technologies and experiences, but none of their "operating systems" has lasted long enough to be a proven solution. More pointedly, at least one of their operating systems has proven to be such a miserable failure that it has been completely abandoned: DOS.

DOS is a good example of how businesses and consumers, mysteriously, adopt tested but unproven technologies. I have no explanation for the fact that millions of people — managers, home computer buyers and the like — purchased this beast of a thing. No one had seen what it could do, there was no data about the longevity or scalability of applications written for the platform — quite simply, it had no track record. But it sold millions. What?

The same kinds of things are still happening. Regal Cinemas wants to create vending kiosks and robust POS systems and has decided to do so with Microsoft software. What is Microsoft's track record with vending kiosks and POS systems? They stumbled on to the concept of TCP/IP networking about ten years ago (with Windows 95), so they have a maximum of ten hodge-podge years of experience. More reasonably, though, we might say that they have three to five years under their belts with their current technologies, namely XP.

When evaluating a very large project like this one, a responsible business manager should ask "where will the project be in five to ten years?" It is absolutely impossible to answer that question with respect to a Microsoft proposal, for the underpinnings of their currently fashionable OS have been being patched up in the consumer market for only about three years. Just about the only convincing trend we can extract from Microsoft's history is that their products have a lifespan of maybe five years.

Where will XP be in five years: dead. We know this not only because of history, but because Microsoft has announced the death in conjunction with the birth of Longhorn. Simply said: roll out an XP project right now, and it's guaranteed to be dead within five years.

Contrast this with a UNIX project. UNIX will still be UNIX in five years, ten years, as far as one can reasonably project into the digital future. It has been a solid OS for 40 years already. That's a track record. Where will Linux be? Probably in the same place. Linux has been Linux for ten years already, and it's based on time-tested UNIX technologies. That's a solid start.

The consumer market is not much different, and the market is changing faster than ever. The cell phones of five years ago are completely antiquated — analog phones are hardly supported anywhere. That's an extraordinarily short lifespan, and frankly, an embarrassing lifespan. It's much the same with portable music players, starting with the Walkman and ending with the currently popular and soon to be deceased iPod. And video game consoles. And computer chips. And cars. Lawn mowers. Pencils. Fly swatters.

The market encourages us to replace everything that we own as soon as possible. Ironically enough, manufacturers and corporations tell us that we have just purchased the greatest products ever made. And in the next breath, they tell us that they'll have the better model available for purchase next month. This is irrational, yet now normal behavior.

It's hard to say what to do about this: the pace of the bandwagon has picked up so dramatically that it's got more inertia than a few of us can face. Personally, I have stopped upgrading software. I'll wait until the software I have fails me somehow (which is hard to imagine), then think about upgrading. And I have stopped buying disposable products, like Swiffer Mops and fast food. Metaphorically, I'm taking a leisurely jaunt on a slower bandwagon. And I get to gape at the carcasses of the out-of-control and crashed wagons strewn on the side of the information highway.

References

Linux Customers Migrating Back to MS due to Patent Risks
Security Report: Windows vs Linux

© Sorrell
March, 2005