When hopes, plans and dreams cannot be realised, as we saw in CI No 1,333), only a fresh beginning will do. This requires the sincere performance of an act that is easy enough before one is in trouble but damned difficult once problems have grown to menacing size. That act is called, in Yiddish, talking […]
When hopes, plans and dreams cannot be realised, as we saw in CI No 1,333), only a fresh beginning will do. This requires the sincere performance of an act that is easy enough before one is in trouble but damned difficult once problems have grown to menacing size. That act is called, in Yiddish, talking tachlis. Talking tachlis means looking at the bottom line, getting down to basics, and this leads us to New Year’s Talking Tachlis Prediction Number One: IBM will make it clear to mainframe customers that it wishes it had a new family of mainframes to peddle. Instead of all the posturing, haggling and niggling, it will say point blank that the 3090-J is a meantime product. Then it can consistently make fair deals. While IBM is touting its 3390 as the answer to every big shop’s storage problem, it’s patently clear that the 10.8 drive is only a stop-gap product. The important disk is the yet-unannounced 5.25 unit being developed in San Jose. This is no secret. Every guru in the trade, half the psychics in the tabloids and a quarter of the directly affected personnel at IBM know it. Most companies buying 3390s are in deals because they don’t know what else to do.
Meanwhile, just in case IBM has found a customer that wants disk drives and doesn’t seem sufficiently confused, the manufacturer is giving away 3380-Ks as it tries to sell 3390s. And if that doesn’t turn a user’s mind into a committee, no two shops seem to be getting the same story about the 3990-3, what it should cost, how long before one has to pay for it and whether one really needs it to make each type of disk work with various programs. New Year’s Talking Tachlis Prediction Number Two: IBM will announce a full schedule for disk subsystems based on miniaturised drives. In addition to the San Jose product, the company will provide some details of the similarly-sized but less capable mid-range disk in development at Havant, England. Unless the present deep discounts keep sales of 3390s and 3380-Ks rolling, IBM will agree to eat every disk it sells with a greatly expanded version of its Technology Exchange Option lease. The catch, of course, is that a customer going with this deal and locking itself in with IBM may do worse than a smart and independent shopper. It could be cheaper to play in the third-party 3380 leasing and purchase market if IBM’s 5.25 disk disappointing. The IBM customers most saddened by their preferred vendor’s not-so-benign neglect are users of 4300 family mid-range systems. IBM has needed an attractive air-cooled product line for a couple of years, and has been saved from a total rout in the market by only two factors. First, an awful lot of companies are so glued to their 370 applications that they could only migrate if a vendor as viable as IBM provided a superior migration aid – preferably one that would use some artificial intelligence techniques to make converting the tenth application a little easier than moving the first one. Second, DEC has failed to heed the wisdom of the American heartland: You snooze, you lose.
By Hesh Wiener
At the rate rumours have not been flying around the industry, IBM might not even be able to announce new mid-range systems until mid-year. In a sense, this might be soon enough for happy 4381 loyalists. But for an IBM interested in expanding its mid-range by getting itself heard over a noisy crowd of rivals, an announcement tomorrow may be too late. It’s possible IBM got into this pickle because it wanted to give the AS/400 as much room as possible to catch on. There’s also the matter of DOS/VSE, which IBM Germany is trying to rewrite so that it’s as good as the version a user developed and put in the public domain… but more complicated. New Year’s Talking Tachlis Prediction Number Three: IBM will announce a range of air-cooled mid-range systems that do things only its mainframes and others’ minicomputers now do. They will offer a serious vector processing option, they will accept tons of main memory, and they will have hardware assists for system software that help customers stretch their
MIPS. In order to get the current 4381 customers to let go of their nearly free machines, IBM will simplify maintenance. The 4381 successor will be as smart as the AS/400 when it comes to self-diagnosis. There are a few other little problems IBM must address as quickly as possible. At the low end, its AS machines (both the AS/400 and AS/Entry) are a mite overpriced. Instead of tolerating the deep discounts in the grey market – which brought in new accounts – IBM has tightened up on resellers. This has forced remarketers actually to add some kind of value. It has also made just about every reseller fearful that it will be arbitrarily or capriciously disenfranchised by its hardware supplier. IBM hasn’t figured out that a bunch of resellers griping about IBM’s nastiness is bound to infect sales prospects with suspicion. This is particularly true of users who had only months ago been courted by sales reps from four resellers and now can’t get the phone answered at three of them. Instead of turning its marketing business over to the Grand Inquisitors, IBM would have been smarter arranging marriages between some resellers and struggling but worthwhile software houses. Even if IBM’s corporate shotgun played a role in a few nuptials, the world of value-added resellers would have been a safer and more economical place for the poor customer.
Rip Van Winkle
The whole scene could get messier if IBM’s RISC project – the Rip Van Winkle of the Blue Empire – wakes up and stumbles out the door. Every small company that hasn’t already been bored to tears by AS/400 resellers trying to explain relational databases will be treated to an army of Unix nerds flogging their local net-based wonder systems. New Year’s Talking Tachlis Prediction Number Four: IBM will rethink its network server strategy so it blends into the world of SAA, AS/400s and even white mastodon 3090s. Goofing around with Unix so a couple more college professors will take IBM grants and trash their NeXT machines isn’t going to make one shareholder happy. And selling a baby AS/400 – which has the MIPS of a Hong Kong wristwatch – as if it’s a versatile server is pushing things just a little. Okay, Rochester, Minnesota, we’re proud of you. You can admit the small ones are just nifty billing machines. They work remarkably well – and that’s the bottom line. We should have known IBM was headed for trouble when it began talking about TCMs, SQL, DRAMs and the like instead of the customer’s everyday problems. Well, this will change. New Year’s Talking Tachlis Prediction Number Five: IBM will educate the people whom it lets sell its products – the insiders as well as the outsiders. This instruction will be very simple and direct. The basic curriculum will be reduced to a slogan that reps and resellers will write on the blackboard until they learn it. It will be Bytes, schmytes! Talk tachlis.(C) 1989 Technology News of America