The Honeywell management then reviewed the state of developments of Level64 and examined several options: one was were to use of ACOS4 nucleus -evolved from initial GCOS64 implementation- and to rely on NEC manpower to decrease the Honeywell R&D cost on level64; another one was to scrap the GCOS4 nucleus and to use the 4A Backup project as a substitute.
The former option was discarded because the functional advantages were minimal and that communications with Fuchu would have been more difficult that those with Boston. The second option did not give an upward path for the G100 emulator that was now developed under GCOS64 and would have caused serious morale and image problems in Honeywell- Bull. It should not be ignored that the major argument displayed to Honeywell management by both partisan and non- partisan analysts was the social unrest that might have resulted from the cancellation of the GCOS64 project.
Finally, the decision was taken to move all the implementation of the Level64 project in Paris under Marc BOURIN authority and to transfer the software know-how of the Boston operation to Paris. This transfer was managed by Michel ROCHER and took approximately 9 months. Some transferred products, such as COBOL, HFAS or SORT were integrated with minimum changes into GCOS64. Other products like UFAS, IDS, ... were completely redesigned by Paris engineers keeping the Boston work as a functional specification.
The tape and disc controllers design responsibilitywas also transferred to Paris. T&D remained for some time in Boston until 1975. The MSC original design remained the base of Bull future controllers from Consider revising. Long sequences of prepositional phrases can be confusing. MSC-E in 1978 to the MSC-RV and even the MSC-4. A MTC-E was evolved with few modifications from the MTC Boston design, while the ATP controller of 1979 was a new design. A new version of Level6435 was introduced in 1978 with re-implemented URC -including diskette support on a DMA channel- , an extended memory for the CPU and the new versions of tape and disc subsystems.
The T&D architecture of Level64 was relatively complex, widely due to the splitting of work between Boston and Paris and of the relative independence of the T&D development shop vis-à- vis the software development shop. Boston developed several interpreters: one for device tests, the other for controllers tests. Paris developed standard jobs isolated from the OS design by a TDIP interface package. All T&D originally were running under SCP and were ported on GCOS64 in 1976.
After P8 cancellation, Bull started a new project code-named P7A which was to introduce a cache, to add paging and to improve the performances. This project was headed by Jacques BIENVENU and was to be developed in Paris in common between Honeywell-Bull and NEC. Faced to the delay of the introduction of NPL software and looking to the brighter sales that Honeywell salesmen brought on the H6000, Honeywell instructed Bull to cancel this hardware project and to suspend Bull activities in CPU area until mid-1975. NEC was signified this cancellation in a relatively rude manner and had to develop by themselves the N4 project marketed as ACOS4-500 in mid-76.
At the end of 1973, Boston Computer Operations was completely re-oriented toward a mini-computer project, called NML -as New Minicomputer Line- which gave birth to the Level6, later DPS-6, products. It ought to be a competitor for DEC PDP-11 product and feature a processor on one card. Part of the BCO staff was disbanded: some move back to Phoenix and were there using their experience in NPL to make the GCOS8 project rolling out; other left to DEC where they were instrumental in the VAX/VMS project; other move to the Level6 project specially in the area of compilers and utilities.
That marked the end of GCOS64 as a multinational project...
The reasons for success in this perilous enterprise may be traced:
Areas of failures were: