The most wonderful gifts we receive are books, particularly if we don’t count liquids. This year was no exception. We now possess a remarkable volume entitled Everyone Poops, by Taro Gomi. It begins, An elephant makes a big poop. A mouse makes a tiny poop. It ends, All living things eat, so everyone poops. The original version, Minna Unchi, was published in 1977 by Fukuinkan Shoten, of Tokyo. The American version was issued in 1993 by Kane/Miller Book Publishers, of Brooklyn. About time. The Japanese publication of Everyone Poops was prescient. Specifically, it preceded the Great Wall Street Crash by a full decade. Sadly, those in other countries, such as America, England and Orange County, did not have an opportunity to contemplate its profound truths before the fall of the Bush regime, the unravelling of the Soviet Union, the collapse of a few Japanese governments, the start of Italy’s Berlusconi fashion season, the British Tory Party’s loss of its franchise, plus the bisection of IBM, the wreck of DEC, the gelding of Bull and an uncountable number of corporate and personal financial tragedies. Our leaders say that the current economic decline is completely different from the one that followed the crash of 1929. Then, the Great Depression flattened the world like an avalanche. Since that time, we have learned a great deal about economics and international co-operation. As a result, the post-1987 debacle crushed the world like a glacier. Perhaps we are at last on the brink of a real recovery, making it an opportune time to reflect on the passing troubles. But let us not be hasty. There are also indications that relief lies mainly ahead of us, things like the December 26 New York Times front page headline, Retailers’ Christmas Wishes Didn’t Come True This Year. The article implies that the widespread talk of an economic upturn, particularly in official circles, is, as the Japanese might say, unchi.

Poop out

(Should the Nikkei average, which has recently crept back toward the 20,000 level, take a big dive, the local phrase for the stock market’s behaviour might be the very emphatic bakuon unko, which just about defies translation, particularly by us right here and now). The upsurge in demand for mainframes reported by IBM is undoubtedly correct, but we suspect that the total demand is not all that great. Compared to the size of the installed base of water-cooled machines, IBM’s order book amounts to a mere handful of mainframes or mainframe upgrades. When those machines are delivered, very likely by the end of this quarter and certainly by mid-year, IBM’s big systems game will poop out… until the 9672-E is ready for prime time. But there are obviously a lot of people who disagree with us. The number of companies still buying large engine mainframes are paying big bucks for the machines. They obviously intend to get full returns on their investments. Because these customers are the biggest and, generally speaking, the savviest in the IBM base, they may sense something others, including ourselves, have overlooked. They have a kindred spirit in Hitachi, which later this year plans to announce another generation of big iron, machines that might top out at 1,000 MIPS. Perhaps they suspect that IBM won’t get its large CMOS machines, the 9672-Es, to perform as promised… at least not for a while. But there is one other explanation: given that everyone knows CMOS computers will be a lot cheaper to buy, to maintain and to run than 9021s, the buyers of 9021s may be playing by rules that make sense only when examined from the inside. We believe that many if not most of the active H5 buyers have multiple large machines. What if they all believe that their least cost strategy involves boiling down all their mainframe work to a single system or single Sysplex workload? If that is the case, IBM’s strategic problems at the high end will be a lot more severe than they look to outsiders, such as Wall Street analysts. And the unfolding mainframe scenario will be made even more dramatic when IBM ups the ante with

its second generation S/390 chip sets. At that point, whatever the state of 9672-E software, one 9672-R (or two of them in a Sysplex) will be able to replace every 3090 and every 9121.

By Hesh Wiener

The mainframe base will not change uniformly nor all at once. The population and value of each part of the current mainframe base will decline at their own rates. The details will be affected by a number of factors, including the wealth of the user organisations, the nature of their lease commitments, the availability of new hardware, the used market prices of interim alternatives (such as used 9121s that could replace many 3090s) and the general business conditions affecting each user company. Owners of 3090 processors, 9121-320 processors, 4300 processors, 3390 disks, 3380 disks and all the various clones of this equipment have by now, however reluctantly, become reconciled to huge (and possibly total) losses of capital asset value. It certainly does not look as if IBM’s largest 9021 mainframes (nor the alternatives provided by other vendors) will enjoy much better treatment at the hands of the used equipment market during the next couple of years. It is a matter of when, not if, each type of big iron gets clobbered. As Taro Gomi explains, Some stop to poop, others do it on the move. Leasing companies and their investors (including IBM Credit Corp) cannot escape their current predicament unscathed. Already, lessees whose deals are approaching termination are being offered some extraordinary arrangements by lessors who want to keep processors and disk subsystems on the customers’ floors. And there is no relief in sight. As the economic recovery remains weak and uneven, particularly in Europe, computing budgets are unlikely to leap upward during 1995. Moreover, a growing slice of that rich pie is being devoted, with good reason, to desktop systems, networks and servers. Users are in better shape than lessors. They are getting more but paying less, because prices for a unit of computing or storage have fallen. So there is no doubt that big business will be able to deliver more to their end users. Yet prospects for truly new computing techniques at big mainframe sites are poor. While there is a great deal of experimentation with multimedia, client-server computing and graphical user interfaces throughout the corporate world, most end users still have dumb tubes or personal computers that imitate them. IBM is partially responsible for the way large enterprises have fallen behind. It is preoccupied with porting MVS and other systems programs to its small engine mainframes. Right now, it does not have the resources to deliver the future systems it so loves to describe in its advertising. Other vendors, however, are not waiting for IBM to lead the way. They view IBM’s smug preoccupation with the software of its past as an opportunity to claim turf that might otherwise be occupied by Big Blue. It is to those companies – and their own in-house talent – that users must turn. Some animals poop and pay no attention, observes Taro Gomi. Others clean up after themselves. The politicians and their pet economists may be exaggerating the vitality of the economy, but they are correct when they say conditions are improving. They are also right in believing that inflationary pressures will grow with every sign of increased affluence. The citizenry everywhere is frustrated, drowning in its mortgages and overdue for a bit more disposable income. The espoused optimism of governments the world over, whatever its cause, has amplified expectations. So there is going to be plenty of pressure for higher wages. While businesses may try to resist employee demands for more pay, we believe this will be impossible. The only option open to employers will be to increase productivity. Information processing has promised to do this many times in the past and, by most measures, it has failed to deliver. This time the computers must come through. Belatedly, but still not too late, computer systems now have the potential to make end use

rs more productive. This property isn’t inherent in automation, as so many user organisations have long since discovered. But productivity gains are within the grasp of every company that can afford the current crop of workstations and servers… and the will and imagination to move ahead. End users can be given rapid access to pretty much all the information and help they need to do their jobs. The screens in front of telesales workers and field sales people can deliver information on their companies’ ability to deliver merchandise, on customers’ credit standings and on the characteristics of the goods and services in an order. Manufacturing planners can get useful graphical depictions of their plants’ capabilities as easily as they can get tables of numbers. Financial planners can examine their data and evaluate their options using very sophisticated presentation tools. Most importantly, these capabilities and countless others like them are no longer exotic or expensive. People may require training to use the new technologies, but that, to an impressive extent, can be delivered by workstations, too. Once the right technology is on desktops, user training and support is just another application, not an isolated activity. Some poop here and there, Taro Gomi offers: Others do it in a special place. We see no reason that end users whose children use a multimedia system to study for school can’t have similar machines presenting equally useful material at work.

Deliver tons

We still don’t believe IBM is quite ready to make its mainframes into economical multimedia servers, but we are confident that system to study for school can’t have similar machines presenting equally useful material at work. information assets. Moreover, even firms that do not have the capacity or inclination to renovate their mainframe operations can move their end users forward by working with those that already know how to deliver tons of dynamic data at low cost. If Prodigy won’t provide a private subsystem, give America Online a call. If they are too busy to help, MCI or AT&T might want to hear from you. And there’s always Microsoft, which soon ought to be able to do for corporate customers the very thing it intends to do for consumers: make the desktop computer a window on the world. And that’s the straight poop.Copyright (C) 1995 Technology News Ltd.