3 ms·

Step one is to find someone who even knows that the source code still exists - it may well be lost forever. If someone turns up saying "yes I have a copy of Fl

Step one is to find someone who even knows that the source code still exists - it may well be lost forever. If someone turns up saying "yes I have a copy of FlexOS/386" then I can take next steps which is to work out who to talk to about open sourcing it. In my opinion alot of senior tech people like CTO's and CEO's are interested in things like computer history and the software community. A company that open sources its defunct software at the request of the community is showing itself to be a good corporate citizen - they get to do a press release that might catch the attention of some good engineers who consider taking a job with them. Anything's possible. Worst case is someone finds the software and the company that owns it says "no" - nothing ventured, nothing lost.

dfawcus 5y ago
The distribution came in two parts when I worked with it: APT - The binaries and SDK. SBK - The drivers (source code), .obj and .lib files for the system. There was significant driver source code in the latter, and the object files had (codeview) debug symbols included. It is possible to reverse engineer the latter files in to something approximating the C source, then rebuild individual "Resource Managers" - e.g. the PIPE/SM RM. I know one could probably find the APT & SBK for 2.2, 2.34 (and 2.4 Beta), since a company I used to work for had them, along with the File System (Disk resource manager) source. The 2.4 Beta was something Novell distributed shortly before they handed the system off to the next owner. The other parts of the system are not too complex, and could probably be reverse engineered from the objects with say 1 man year's work. Now having rights / permission to do so, and redistribute the result is a different matter.