Initially, it looked like IBM had an answer to this question. First, announce the end of SSP support by June 30, 1991. Second, redirect systems engineers so that S/36 customers could not get questions answered easily. Next, make it clear that SSP development will be discontinued and that PTF support will be farmed out to a third party. Then, eliminate statements of direction for S/36 customers, leaving them on a limb as to the long-term viability of their machine. And finally, send a steady stream of AS/400 information to S/36 customers who request information on S/36 upgrades - or else, ignore the customer entirely.
However, in developing the above "strategy," IBM failed to notice that it continues to be the biggest booster around for the S/36! In fact, IBM continues to sell S/36s by the thousands, primarily in Europe, but in the United States as well. Some sources suggest there are now more S/36s installed worldwide than there were when the AS/400 was announced. Of course, a lot of these S/36s are called "AS/Entry" by IBM, and are either 5363 or 9402 machines. But, they run SSP and they don't run OS/400. If it walks like a duck....
IBM's predicament became apparent when, along with the hoopla of the April 22 AS/400 Model D announcements, IBM quietly announced a continuation of SSP service and support through December 31, 1992--a full 18 months down the road. In addition, through ongoing PTF's and small programming enhancements (SPE's), IBM has upgraded SSP 5.1 and 6.0, hereby providing external RPG Call/Parm (which has existed for COBOL for some time), and enhancing communications and connectivity features. Apparently, more of the same can be expected.
But, IBM still has the problem of a large base of very unhappy, very concerned S/36 shops. Many, if not most, of these are small shops, with few programmers. But, several shops are very large, with multiples of S/36s installed throughout the country. Keeping these customers happy should be important to IBM, since future revenue streams may be lost if these customers begin looking elsewhere.
To maintain customer satisfaction, IBM needs to do several things. First, IBM should state its intentions for the future of System/36, including how much longer the various support structures will be available. Second, IBM should provide incentives to its branch offices to continue to provide SE services to S/36 shops, rather than simply saying they can't help any more. At a minimum, the internal SE question-and-answer system must begin to allow queueing of S/36 specific questions again, so that SEs have somewhere to turn.
Third, IBM should make available to SSP 5.1 users the RPG Call/Parm features without the extra charge currently assessed. External file support to S/36 RPG through IDDU should be combined with this. It obviously can be done, since Query and PC Support use IDDU definitions; it probably isn't even hard to do. These two steps alone will do more to allow S/36 shops to educate themselves to AS/400 concepts--and therefore, more willing to move to one-- than can all the marketing reps around pounding on doors.
Most importantly, for S/36 shops the move to an AS/400 can be a "bet your job" move. The downtime needed to retrain staff, to make the conversion, and to get things going smoothly can ruin a career. Or, with the right level of IBM and business partner support, it could make a career.
A lot of System/36 managers are very nervous about IBM. They see themselves as loyal IBM customers who are suddenly out in the cold. IBM won't talk unless it's about the AS/400. The trade press has moved away from the System/36, as has COMMON. Consequently, these folks feel pretty nervous about another commitment to IBM.
But, change is in the wind. The trade press seems to again be running more S/36 related articles and tips. COMMON has finally re-affirmed the value of its S/36 project, and has begun to recognize and promote it as the educational tool that it is.
As for the future, we'll have to wait and see what IBM does. S/36 shops will need to decide whether to migrate to the AS/400, the RS/6000, Unisys, PCs, or to simply stay on the S/36.
There is no question that the AS/400 is a technologically superior machine, a programmer's dream, and a new generation in midrange computing. There is no question that many S/36 shops which have not yet migrated will need to consider doing so within the next couple of years. But, if IBM wants to make sure that these shops keep IBM in their migration options down the road, then it must keep these shops within the IBM fold by treating them as the loyal, good customers they have generally been. This means a "market driven" consideration of its S/36 customer base, not a "marketing driven" one. Concern and support for today's S/36 customer is more likely to mean a happy AS/400 conversion in the future than is the wholesale abandonment of the S/36 community.