Demographics is a remorseless master as the governments of mature economies around the world are beginning with horror to realise. As health care and general nurture improve, people are living longer and longer and spending a progressively smaller proportion of their lives doing useful work. For once, the British government is beginning the debate that will have to lead to painful steps to address the problem earlier than most, recognising that by the second decade of the next century sounds so comfortingly far away, doesn’t it, next century, yet it’s only six and a bit years away – there simply won’t be enough people in the productive workforce to pay the state benefits to which today’s pensioners have become accustomed and moan about. No-one has yet had the guts to point out what follows from the fact that the when the Welfare State was established in 1948, the state pension for all represented a commitment to pay the average person a pension for perhaps five retirement years, since actuarially, people died so much younger in those days.
Misery
The least unreasonable alleviation of the problem would be to separate the concept of a retirement age from the date on which the state pension becomes payable, go back to the original contract and for people that have joined the work force in the last 10 years or so, rebase the pensionable age generously on modern demographics, so that the contract becomes a commitment to pay people a state pension for – say – actuarially the last 10 years of the average person’s life. If the life expectation of a 65-year-old man was 17 years, then men would begin to draw their state pension at 72. Such a switch would have to be introduced over many years, but would greatly alleviate the problem: those that wanted to could work until they were 72, all would have to make their own arrangements between work and privately financed retirement for the years between say 50, at which some people take early retirement, and 72. Western governments aren’t the only ones facing intractable demographic time bombs: IBM faces one too.One of the most disconcerting aspects of IBM’s current misery and woe is that even we who have been warning for years that the company was heading for the kind of problems it is now suffering, still apprehend the developments as they unfold with a sense of shock, almost of disbelief, so that whatever horrors IBM has in store for its followers with its second quarter figures today will still evince a sense of aghast surprise. Even those that have long believed it was an inevitablity still can’t grasp the reality of just how dead the mainframe business really is.
Black hole
It is now about five years since people first began seriously to question the central role of the mainframe in corporate computing, and at that time, there was an enormous base of very venerable MVS applications and a significant base of ones that were dangerously close to the end of their useful lives. All those applications are now five years older, five years less valuable, five years closer to the point where from an accountant’s standpoint they should be written off. Users that five years ago couldn’t conceive of doing other than follow the IBM MVS mainframe trail into the millennium are now beginning to realise that many of their MVS applications have given very good service, owe them nothing, and are over-ripe to be pensioned off. This perception is day by day sharpened as they read more and more analysis that questions the future of the mainframe. What can IBM possibly salvage from what – far from being the $18,000m black hole detected in its annual business by Gartner Group, begins to look like a $30,000m black hole? The one exception, the one relatively new base of applications in the IBM base is those that have been or are being written for the DB2 mainframe relational database. There is substantial potential business stretching into the next century for IBM from that base – if IBM can recognise the jeopardy in which entire base is cast, and resolves to save what is genuinely salvageable and has
a future, rather than vainly trying to save everything and seeing the whole lot crash through its nerveless fingers. In the age of open systems – and that extensively misused term has to include operating systems like OS/2 and Windows NT, the only future for proprietary hardware and architectures is as sealed boxes that can be used by open systems. The idea of turning the entire mainframe into a giant server, a concept IBM has been trying to promote, is a nonsense. Giant servers will be required in the future, but the idea that they have to run MVS on a hardware architecture that dates back to the 1960s should cause even IBM to laugh apologetically. Rather more likely is that they will be giant parallel machines implementing completely new architectures. But for IBM’s and their users’ sakes, those DB2 applications have to be saved, so that the company’s top mainframe priority should be the sealed box DB2 database processor that can be used as part of an MVS mainframe complex, but will work equally well with no loss of functionality as a database server sitting quietly on a NetWare network where the terminal computers are Macintoshes or Sparcstations. Although batch IMS applications are much more mature, there are many banks and other financial institutions that can find no other way of printing out those monthly statements or whatever anything like as cheaply as with IMS, but IBM is so short of imagination these days that it is likely to leave those people to to run those applications into the millennium on ever older hardware.
Horrendous
The problems facing the AS/400 are not yet quite as serious as those facing the mainframe, but as we have argued before and as IBM now seems to be planning, the sealed box AS/400 database machine than can sit happily within a network of open systems should be a completely new incremental AS/400 market that addresses real needs of users that would never consider running their entire workload on a collection of AS/400s but can nevertheless make very good use of the machine’s unique capabilities as a database processor. The threat to the AS/400 does now seem to be dawning on IBM and it does seem to be taking some of the right steps to address it – yet the company still seems to find the demographic threat to its mainframe business too horrendous to contemplate: perhaps Louis Gerstner will prove us wrong with today’s announcement: there’s an army of dependent users of DB2 that must fervently hope so.
