Cullinet failed because the management of 1986-89 gave up the focus on IDMS and diverted their energy and money on totally foolish acquisitions and products.The successes of previous years were not based on marketing but on solid product line. The bread and butter had always been IDMS. But the newcomers at the helm forgot that. Instead they were probably reading the doomsday predictions of Computerworld and gave up too early on IDMS.. much too early! DB2 at this stage (1989) was a very rudimentary product. In fact, IBM's own AD/Cycle with its non-existent repository at the center was a total failure. Cullinet could have easily given System Owned indexes (instead of ASF/LRF) with SQL front end as a relational database much earlier. Still later they could have come out with IDMS 11.0 - by 1988 or latest by 1989. But time was wasted on VAX/Generator and rewarping IDMS as the Enterprise:DB. There was even a product called Expert Systems which did not sell even a single copy! Much resources were wasted on such experiments. Believe it or not, IDMS itself was so sidelined and the faith in the Generator was so emphatic that in one European Meeting, a plan to give Generator to all IDMS clients and help them to convert all IDMS/ADS applications to VAX/Generator in a short time, was even passed as reality!