bloodline wrote:
Use hypothetical examples then...
Well.. hmmm... OK. You are porting operating system A to hardware B. Lets say you have component X's source code. It is written in C. With a bit of tweaking it compiles under GCC for hardware B. Is it better to rewrite component X from scratch, or use the existing source code? This will depend on the state of component X's code of course, but when there are hundreds of components in operating system A a fair amount of them will probably be worth developing using the original source as a base.
Also, I can think of at least one OS component in OS4 that was almost entirely assembler. It was translated function by function into C to ensure maximum compatability, with builds tested on AmigaOS3.x to ensure nothing broke as the functions were translated (the C functions even had to work with the other functions still being assembled from the original source). When it was fully C, it was tweaked to compile with GCC, and then finally compiled for OS4 and PPC. This was done because the component in question was quite complex and it was desirable to go for maximum compatability with this particular component. Although there is now none of the original source involved in that component, I still consider it based on OS3.x code. (note: if you are thinking of a particular component, you are probably wrong, but i can't tell you which one it is anyway).
Anyway, at the end of the day, only the OS4 core devs know exactly how much OS3.x code is used, and i'm not one of those, so even without NDA I couldn't give you an exact figure, but from the information I DO have I can see that having the code has been very useful indeed.
