(... skipping (1) and (2) ...) > 3) Threads > The underlaying model are posix threads if available. This implies > that all interpreter data are shared by default. So we'll need per > thread data: interpreter, prederef & JIT code ...
Is 'the underlying model' the implementation of the OS running parrot (as oppposed to a self-made system,which would result in Parrot VM being 1 process in the eyes of the OS. So, when preempting the process of Parrot, all 'threads' in Parrot are suspended (or whatever the official term for it is)) ? (I have this idea from the book "VM design and impl in C/C++", Bill Blunden). So when the above is true, then that means that Parrot threads will have a bit different behaviour on each diff. OS (assuming each OS has a somewhat different impl. of thread scheduling), doesn't it? Will this give problems, having the same parrot code behave somehwat differently because of the host OS? It also remembers me of Java, Java programs having different thread behaviour when running on different OSs, is this comparable? (...) > leo Klaas-Jan
