The importance of Internet appliances

I think this week`s column might end up being a little short, sort of like my legs, which have just spent an entire 12 hours carrying my body. It`s quite interesting, actually: for those of us who spend most of their days behind a desk, seated comfortably in a good office chair, the annual Computer Faire & Bexa (this year is its 20th anniversary) is quite a slog. Don`t get me wrong: I love talking to customers, dissatisfied or happy, and I love selling Internet connectivity or solutions right to the consumer. It`s just that I`m no longer used to selling the Internet, retail.

A lot of IT people are continually worried because what they do on a day-to-day basis seems to have little relation to what they read about in magazines.

On to this week`s topic, then: recently, there`s been a lot of talk about Internet appliances and Java toasters and things like that. It`s an interesting trope on the basic thrust of IT. Generally, technology and computers are recognised as important in making the work and home environment more effective and less costly, and the basic push of any good IT manager`s activity is to make sure that things (a) keep working, (b) become smoother and slicker as time ticks by and (d) don`t cause a major catastrophe.

Strategy, sadly, is often not the guiding principle of an IT department`s activities, but rather a grudging after-thought when management requires annual budgets and an IT `strategy document`.

A lot of IT people are continually worried because what they do on a day-to-day basis seems to have little relation to what they read about in magazines. Think about it: everything we actually roll out and implement is basically old, proven technology. Leaping from NT 3.51 to NT 4 wasn`t really a big change, we just took what we knew and improved on it slightly. The basic facts of the network and its relationship with the workstations remained in place; and that paradigm hasn`t really changed in a good few years. It`s now faster, some of it is easier to use, but mostly it`s the same as before.

Yet the press continues to talk about things that don`t exist. Things, to be fair, that will exist one day; some of them are closer to our known reality and some quite far removed. Take online commerce and trading as an example. Sure, we can build a lot of the solutions that are described regularly - there are a lot of specialist Web houses out there who can do amazing things. But how many of you will actually implement a Web commerce solution, even in the next two years? You read about it, you might get excited about it, and then it`s back to your LAN and your ordinary Internet connection. If you are paranoid enough, you implement a firewall (another `old` technology), and that`s that for now. In terms of Web commerce, you either continue to dream or you `put it on ice`.

Good IT managers are dreamers who can continually establish a relationship between what they read and see in demonstrations and what they do on a daily basis. Great IT managers (at least from management`s point of view) are those who continue to develop advanced IT strategies that take into account what is, what will be, and what could be, if only there were enough money.

The IT and Internet industries undergo certain `fads`, phases of what`s currently `top of mind` in the media. Intranet, extranet and the network computer come to mind. I think the relative merits and demerits of these have been discussed to death already. So much so, in fact, that my dad (who`s never touched a computer in his life) now thinks that he wants to buy a network computer because he read an interview with Mark Andreesen in the popular media somewhere, and, to quote, "I only want to access the Internet." I didn`t quite know what to say in response, but I`m working on it.

It`s these media topics that give IT managers a lot of trouble. They spend sleepless nights. Should they or shouldn`t they? If they do, why do they do it? What value will it add to their daily activities? It`s unsettling to read about IT developments that are on the horizon and not knowing how these will fit in with what you know. Innovation, in the applied IT field, is about following trends quickly enough without compromising your budget, without affecting the basic infrastructure that you`ve put in place already, and without upsetting users by giving them something you don`t even understand yourself.

The network computer, though, leads us into the core of today`s topic. For all intents and purposes, it`s an appliance. Call it a network computer (in a corporate environment, that`s what IBM`s and other vendor`s boxes would be called), or call it a set-top box. At my company, like all other ISPs, we occasionally have a few set-top boxes around for evaluation purposes. The latest one we had (I don`t recall who makes it) carried the words "information appliance" on the cardboard box. I thought that was a great way to describe it: it`s like a toaster, only it`s not used for browning bread; it`s a simple thing that you plug into your television set to access information.

I`m in two minds about Internet appliances. I think there are certain environments where they`d be terribly appropriate, while there are other situations where they will no doubt do more harm than good. My gut tells me that in a home environment where computer literacy is non-existent an information appliance could be a good thing, while in corporate environments they would probably do more harm than good. Then again, that`s really rather a simplistic view, and I`ll tell you why. It`s simplistic because one could just as easily imagine an environment where computers and databases are necessary for work, and the decreased downtime implied by managing both applications and databases centrally could really make a difference. Upgrading, fixing and maintaining would become a lot easier.

At the same time, a set-top box in a home environment might not be enough. The PC`s main strength - apart from its standardising effects - has always been that it`s the `ultimate appliance`: a PC can do almost anything. For my father, a set-top box may be enough, as all he wants to do is access the Net. My mother, however, might want to use it for typing up and printing out paper letters. Typically, set-top boxes can`t do that. A PC, however, would be ideal, although it requires a greater financial and educational investment.

So there are no easy answers to whether information appliances will fly or sink. In terms of pure Internet connectivity, they make sense in some environments and none whatsoever in others.

The real problem lies not in how to answer this question right now, but in forecasting what will happen. Bandwidth is a big issue when discussing any kind of Internet appliance: even in corporate networks, serving huge applications (Java or otherwise) can be a problem. The more intelligent your workstations, the more slack you can cut your current network infrastructure. But your support and maintenance time remains high, and that costs a lot of money.

Personally, I think there`s a future in Internet appliances, provided there`s some standardisation. Right now, most vendors` offerings are - naturally - proprietary in terms of OS, performance, even hardware... some of the in-built modems in set-top boxes, for instance, simply don`t work in SA, and there`s no way of reconfiguring them either. It`s a case of simplicity cutting into the bottom line.

Whether Java will address any and all of these problems also remains unclear. From my own experience, there`s a certain conceptual elegance in it that is lacking elsewhere - the idea of a platform-independent programming language and computing environment is quite brilliant. But Java, so far, doesn`t really address OS-level things: something needs to boot-strap the device, something needs to control IO properly, and so on. It`s also a little worrying, in my opinion, that Java has ended up being so proprietary. Witness the recent case between Sun and Microsoft. The best standards (e.g. HTTP and the World Wide Web) have been truly non-proprietary, and those that have served the IT industry well over the years, both conceptually and practically, were non-proprietary in all but name (e.g. MS-DOS).

Before the "Internet appliance" will become truly useful - this is its make-or-break point - there will have to be an open, but well-defined standard. Something like an `AOS` (Appliance Operating System). This could apply to both set-top boxes and NCs - it would make administration very easy. Whether the application development environment that goes - to use an IT term - `on top of` the OS will be Java or something else is almost immaterial in this debate.

I think that Java is trying to become a fully developed environment, an AOS, but it`s got a long way to go still. I take my hat off to those who are driving it forward. Eventually, perhaps there will be a situation where set-top boxes, toasters, garage doors, elevators and other functional items in daily use will `speak` a common language (each with its own vernacular, naturally), and perhaps the Internet will have them all communicate.

That is what I think of in the back of my mind when I think `sensible IT strategy`. Nobody is asking IT people, who are deeply involved with the day-to-day job of keeping things ticking, to throw out what they know and bring in network computers and set-top boxes. Right now, it`s all too experimental to really matter.

But thinking about it is really important. Visiting the Computer Faire, for one, may be a start. Despite the traffic jams, there`s always interesting stuff to be seen there. The sheer joy of playing with new technology was what got most of us into this game in the first place. Some of our detractors call this "geeking out". Now that technology is our job, it`s important to geek out once in a while. Playing with technology is about learning, and learning will eventually lead to understanding and good strategy. Best of all, we get paid for it!