I am looking for impressions, memoirs, articles, guidelines - everything that is possible about the unusual direction of software migration. I am interested in non-Unix related code, and preferably highly optimized code - perhaps even assembly language, and large enough software products - if this ever happened. It would be especially interesting if there was information about porting from Atari ST / TOS / GEM (or at least Apple Macintosh) to DEC MicroVAX / VAX/VMS / DECWindow something with an advanced interface and advanced work with I/O in real time.
I can give you an anecotal answer. I did this a few times.
I suspect that porting in this direction would have been unusual. We only did it where the customer insisted on the use of Vax hardware.
The company I worked for for many years specialised in real-time process control and communications handlers. We made our own range of controllers, based on the 68k processors but would contract on anything.
A large library of handlers for various industry-standard protocols over dual-redundant communication paths, dual-processor routines with both hot standby and 2oo2 for fault tolerance had been built up over years and had a proven track record. These libraries had a core of 68k Assembly Language and a C wrapper.
A bulk material handling application for the Kwangyang steelworks in South Korea was specified to run on a dual hot-standby MicroVax configuration. Commissioned 1987.
A development of this, on a Vax-FT was commissioned in 1992 to control the signalling of Seoul Subway Line 5. Here we used both halves of the FT and our own library.
Another dual MicroVax system was installed in a British Steel Tinplate works in Ebbw Vale, Wales in the early 90s. This had a much more complex communication configuration.
Porting involved adapting the code to use the VMS interprocess communication techniques, mutex, event flags etc, keeping our own real-time handlers. This way we kept our own communications failover techniques and had tight control on the fault tolerance.
Porting in this manner was seen as a good idea at the time. The belief being that it gave the best chance of achieving the project deadlines.