IT Fashion

IT Fashion

Do you ever get the feeling that Information Technology is driven less by business need or advances in distant laboratories than by the kind of thinking that dictates if hem lines are going to be high or low next spring?

Back in the far prehistoric days of computing the major IT companies - of which there were exceedingly few - told us that the only way to satisfy our computing needs was to have a huge number cruncher feeding thousands of operatives via standards based terminals. This concept lasted for decades, with IBM and DEC leading the market for standard keyboard layouts, standard screen displays, and standard display protocols. Hold that thought. This philosophy lasted decades as computing power increased, disk farms grew in leaps and bounds, with disciplines such as accounting being able to tell a Company it was going bust in a fraction of the time it previously took. Mainframes ruled the world until DEC came along with its PDP and VAX ranges, but even then we were still talking about centrally based computing power surrounded by obedient servants who worshipped those white coated individuals who programmed these monsters. I describe them as monsters for monsters they were at that time, but many of them had less power than the laptop I'm using to write this article.

All was well until the early 70s when some smarty pants decided that his Company needed to generate more revenue, new products were needed, and so marketeers were brought in to find a new message to spread, a revolutionary product to sell, a new market to penetrate. The result was the mini-computer, the small departmental servers that could help the central machine spread its tentacles across and organisation from the safety of its air conditioned lair. Suddenly we had vendors and press alike telling us that client server was the way to go, the mainframe concept was dead, no longer did we need huge crunchers serving the masses. We could do it all via smaller machines who would use the mainframes on our behalf as massive centralised data repositories. The dumb terminals began to give way to intelligent devices called Apples I and II, Commodore PETS - may their names be praised - Horizons and Rainbows. Click click gave way to swoosh swoosh as floppy disks - and they really were floppy - appeared on every desktop, Wordstar, Visicalc, CPM and MS-DOS appeared, and computing became affordable like never before. Now everybody could program a computer, the white coat brigade lost their kudos, and the press announced the death knell of the mainframe.

Life seemed pretty cosy for a while, punctuated only on odd occasions by events such as Lotus 1-2-3 taking over from Visicalc and Digital Research losing out to Microsoft in the IBM PC stakes. But as always our smarty pants in the corner woke up again in the 80s and rumours of a new operating system called UNIX spread across the world. The major IT companies came up with all kinds if reasons why it would never succeed, the smart ones at least took a look at it, and VMS devotees went to church more regularly. This decade ultimately heralded a realisation amongst users that being locked into one suppliers architecture was not a good idea, that flexibility to choose the best box for the job was a better idea, and the means the have them talk to each other was a goal to be cherished. Boxes began to appear that would cross link these disparate system, converting protocols back and forth so that companies could buy their kit from whoever they liked - within reason - and not have to manually re-key data each time. But still, in some dark shady corner, this thing called UNIX silently grew in popularity, newcomers to the market such as Uniplex began to write serious commercial software for it, and the big boys didn't feel the tap on their apathetic shoulders until it was too late. By the late 80s the UNIX butterfly had emerged from its academic cocoon and fluttered with devastating results into the bright wide world of commerce.

By the time the early nineties arrived UNIX and the open systems philosophy had been established, new companies such as SUN Microsytems had been born, and the IT suppliers began to systematically demolish the ideal of open standards by producing their own individual '-IX' recipes. The core of the operating system was maintained, but the extensions and functions were horribly mutilated to the point that dozens of variants were spawned within just a couple of years. Nevertheless the world bravely announced the death of all proprietary operating systems, the new entrants prospered and a few of the older players began to slide down the tubes as their once dominant market share was heavily eroded. We now had the dual message that not only was the mainframe still dead but also that proprietary operating systems were for the chop in spite of the fact that UNIX itself was so hacked around that it was already proprietary to each supplier. Life was fun, marketeers had a field day, and salesmen were confused; it was melt down, shake out, rationalisation and mayhem. Large companies such as DEC, HP and IBM found themselves being undermined by the (then) smaller players, Japanese companies such as Fujitsu had made inroads into the markets, and processor chips were becoming so powerful that the once distinct divides between PC, Minicomputer and Mainframe were becoming terribly blurred. But still we had client server, the central processor as servant directly to users was still a major no-no, and everybody settled down again because everybody understood their new respective positions in IT society.

Creeping up behind, or more accurately underneath all of this was the Internet, that dreaded word that permeates every conversation, article, press release, brochure and business magazine. It has been a major revolution which I won't talk about here because I don't want to add to the giga millions of words that have already made their contribution to the chopping down of many trees. But out of this revolution came the standardisation of communications protocols and the advent of a new interface called the browser. Suddenly we had a methodology which would allow disparate cultures and philosophies to share information, ideas, experiences and beliefs via email and the web - so long as it was in American English. Internet and processor technology advanced at a pace never before seen, and the major surprise that no one could have anticipated back in those heady days when it was still ARPANET is that the protocols and interface standards could be used as soemthing other than a communication medium. Our smarty pants - who by now must be reaching retirement age - realised that the de facto presence of the Internet protocol in almost every large company was an opportunity not to be missed. And the existence of browsers such as Netscape and IE on every desk presented an ideal opportunity to sell yet more product by heralding a new fashion for everyone to follow. So out went client server and in came a strategy of using large central servers to feed standard desktop interfaces. Do you still have the thought that I asked you to hold at the beginning of this article?! We have come full circle, the central mainframe is back, the standard desktop is back, and all is beautiful in the world once more.

So what is the next fashion that will make us invest - that's a great marketing word - in yet more software, hardware and consulting time? Well in my opinion it is going to be XML which everyone seems to be touting as the saviour of all things Ecommerce. The fact that it is just a jumped up HTML that takes anything up to sixteen times more bandwidth to transmit than existing protocols, and has to be sent three times in effect - Data, DTD and Style Sheet - to make itself fully presentable at the other end is by the way. XML is the path and the truth and the light and it must be true because the IT industry tells us so. But then again they told us that large central servers feeding the user masses was an outdated concept, so let's get smart this time. Let's make a decision on the need rather than the shiniest offering. Let's decide to hang onto our existing systems of they are satisfying the needs of our business plans, and by all means continue to wear your hem line at knee length next year.



	Do you ever get the feeling that Information Technology is driven less by business need or advances in distant laboratories than by the kind of thinking that dictates if hem lines are going to be high or low next spring? Back in the far prehistoric days of computing the major IT companies - of which there were exceedingly few - told us that the only way to satisfy our computing needs was to have a huge number cruncher feeding thousands of operatives via standards based terminals. This concept lasted for decades, with IBM and DEC leading the market for standard keyboard layouts, standard screen displays, and standard display protocols. Hold that thought. This philosophy lasted decades as computing power increased, disk farms grew in leaps and bounds, with disciplines such as accounting being able to tell a Company it was going bust in a fraction of the time it previously took. Mainframes ruled the world until DEC came along with its PDP and VAX ranges, but even then we were still talking about centrally based computing power surrounded by obedient servants who worshipped those white coated individuals who programmed these monsters. I describe them as monsters for monsters they were at that time, but many of them had less power than the laptop I'm using to write this article. All was well until the early 70s when some smarty pants decided that his Company needed to generate more revenue, new products were needed, and so marketeers were brought in to find a new message to spread, a revolutionary product to sell, a new market to penetrate. The result was the mini-computer, the small departmental servers that could help the central machine spread its tentacles across and organisation from the safety of its air conditioned lair. Suddenly we had vendors and press alike telling us that client server was the way to go, the mainframe concept was dead, no longer did we need huge crunchers serving the masses. We could do it all via smaller machines who would use the mainframes on our behalf as massive centralised data repositories. The dumb terminals began to give way to intelligent devices called Apples I and II, Commodore PETS - may their names be praised - Horizons and Rainbows. Click click gave way to swoosh swoosh as floppy disks - and they really were floppy - appeared on every desktop, Wordstar, Visicalc, CPM and MS-DOS appeared, and computing became affordable like never before. Now everybody could program a computer, the white coat brigade lost their kudos, and the press announced the death knell of the mainframe. Life seemed pretty cosy for a while, punctuated only on odd occasions by events such as Lotus 1-2-3 taking over from Visicalc and Digital Research losing out to Microsoft in the IBM PC stakes. But as always our smarty pants in the corner woke up again in the 80s and rumours of a new operating system called UNIX spread across the world. The major IT companies came up with all kinds if reasons why it would never succeed, the smart ones at least took a look at it, and VMS devotees went to church more regularly. This decade ultimately heralded a realisation amongst users that being locked into one suppliers architecture was not a good idea, that flexibility to choose the best box for the job was a better idea, and the means the have them talk to each other was a goal to be cherished. Boxes began to appear that would cross link these disparate system, converting protocols back and forth so that companies could buy their kit from whoever they liked - within reason - and not have to manually re-key data each time. But still, in some dark shady corner, this thing called UNIX silently grew in popularity, newcomers to the market such as Uniplex began to write serious commercial software for it, and the big boys didn't feel the tap on their apathetic shoulders until it was too late. By the late 80s the UNIX butterfly had emerged from its academic cocoon and fluttered with devastating results into the bright wide world of commerce. By the time the early nineties arrived UNIX and the open systems philosophy had been established, new companies such as SUN Microsytems had been born, and the IT suppliers began to systematically demolish the ideal of open standards by producing their own individual '-IX' recipes. The core of the operating system was maintained, but the extensions and functions were horribly mutilated to the point that dozens of variants were spawned within just a couple of years. Nevertheless the world bravely announced the death of all proprietary operating systems, the new entrants prospered and a few of the older players began to slide down the tubes as their once dominant market share was heavily eroded. We now had the dual message that not only was the mainframe still dead but also that proprietary operating systems were for the chop in spite of the fact that UNIX itself was so hacked around that it was already proprietary to each supplier. Life was fun, marketeers had a field day, and salesmen were confused; it was melt down, shake out, rationalisation and mayhem. Large companies such as DEC, HP and IBM found themselves being undermined by the (then) smaller players, Japanese companies such as Fujitsu had made inroads into the markets, and processor chips were becoming so powerful that the once distinct divides between PC, Minicomputer and Mainframe were becoming terribly blurred. But still we had client server, the central processor as servant directly to users was still a major no-no, and everybody settled down again because everybody understood their new respective positions in IT society. Creeping up behind, or more accurately underneath all of this was the Internet, that dreaded word that permeates every conversation, article, press release, brochure and business magazine. It has been a major revolution which I won't talk about here because I don't want to add to the giga millions of words that have already made their contribution to the chopping down of many trees. But out of this revolution came the standardisation of communications protocols and the advent of a new interface called the browser. Suddenly we had a methodology which would allow disparate cultures and philosophies to share information, ideas, experiences and beliefs via email and the web - so long as it was in American English. Internet and processor technology advanced at a pace never before seen, and the major surprise that no one could have anticipated back in those heady days when it was still ARPANET is that the protocols and interface standards could be used as soemthing other than a communication medium. Our smarty pants - who by now must be reaching retirement age - realised that the de facto presence of the Internet protocol in almost every large company was an opportunity not to be missed. And the existence of browsers such as Netscape and IE on every desk presented an ideal opportunity to sell yet more product by heralding a new fashion for everyone to follow. So out went client server and in came a strategy of using large central servers to feed standard desktop interfaces. Do you still have the thought that I asked you to hold at the beginning of this article?! We have come full circle, the central mainframe is back, the standard desktop is back, and all is beautiful in the world once more. So what is the next fashion that will make us invest - that's a great marketing word - in yet more software, hardware and consulting time? Well in my opinion it is going to be XML which everyone seems to be touting as the saviour of all things Ecommerce. The fact that it is just a jumped up HTML that takes anything up to sixteen times more bandwidth to transmit than existing protocols, and has to be sent three times in effect - Data, DTD and Style Sheet - to make itself fully presentable at the other end is by the way. XML is the path and the truth and the light and it must be true because the IT industry tells us so. But then again they told us that large central servers feeding the user masses was an outdated concept, so let's get smart this time. Let's make a decision on the need rather than the shiniest offering. Let's decide to hang onto our existing systems of they are satisfying the needs of our business plans, and by all means continue to wear your hem line at knee length next year.