fw> > Generally speaking, a replacement of the runtime system should be fw> > possible, but maybe not desirable. I think that the special stack fw> > allocation technique employed by CHICKEN will probably integrate badly fw> > with most other language runtimes and that forcing the CHICKEN runtime on fw> > top of another system might incur quite a performance loss. fw> > fw> Yes, indeed. On the other hand Cheney-on-the-MTA can be applied fw> to implement tail-calls and first class continuations on a host language fw> that doesn't provide them. On a C platform we can integrate GC right fw> into this model and get very high allocation speed. fw> But I'm still not sure what Sergey is trying to achieve here: should the fw> hypothetical chicken we talk about still target C, but use a different runtime- fw> system (GC, data-representation, ...)?
The idea was to use Chicken as an embedded interpreter but don't have two GC's in one process. _______________________________________________ Chicken-users mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/chicken-users
